decision making in geriatric oncology · between overtreatment and undertreatment and must be...

Decision m

aking in geriatric oncology

M

arije Emilie H

amaker

Decision making in geriatric oncology

Marije Emilie Hamaker


PhD thesis, University of Amsterdam, The Netherlands

© M.E. Hamaker, Amsterdam, The Netherlands 2012

All rights reserved. No part of this thesis may be reproduced or transmitted in any form or

by any means without prior permission of the author. A digital version of this thesis can be

found at www.dare.uva.nl

Lay‐out: J.W. Broek

Printed by: Buijten & Schipperheijn

ISBN 978‐90‐9027018‐0

The printing of this thesis is financially supported by: Stichting Oncocare Alkmaar, Servier Nederland Farma B.V., Amgen B.V., Vifor Pharma Nederland B.V., Janssen‐Cilag B.V., Celgene B.V., Nutricia Advanced Medical Nutrition and Diakonessenhuis.

http://www.dare.uva.nl/


ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

Prof. dr. D.C. van den Boom

ten overstaan van een door het college voor promoties ingestelde

commissie, in het openbaar te verdedigen in de Agnietenkapel

op vrijdag 30 november 2012, te 14.00 uur

door Marije Emilie Hamaker

geboren te Assen

Promotiecommissie Promotor: Prof. dr. M.M. Levi

Prof. dr. S.E.J.A. de Rooij Co‐promotores: Dr. B.C. van Munster

Dr. C.H. Smorenburg Overige leden: Prof. dr. A.H. Zwinderman

Prof. dr. C. Koning

Prof. dr. D.J. Richel

Prof. dr. J.W.R. Nortier

Prof. dr. M.G.M. Olde‐Rikkert

Dr. H.A.A.M. Maas

Faculteit der Geneeskunde

Table of contents

Introduction 7

Part I. Breast cancer treatment in the elderly

Chapter 1 Breast cancer in the elderly: a retrospective study on diagnosis and

treatment according to national guidelines 13

Chapter 2 Trends in breast cancer treatment in the elderly at a breast cancer out‐

patient clinic (translated from Dutch) 29

Chapter 3 Omission of surgery in elderly patient with early stage breast cancer 39

Chapter 4 Non‐referral of nursing home patients with suspected breast cancer 53

Chapter 5

Slow accrual of elderly patients with metastatic breast cancer in the

OMEGA study of the Dutch Breast cancer Trialists’ Group (BOOG) 65

Part II. Treatment of elderly patients with other cancer types

Chapter 6 Colon cancer in the elderly: does age still matter? 79

Chapter 7 Diagnostic choices and clinical outcome in octogenarians with iron‐

deficiency anaemia 93

Chapter 8 Age‐related differences in guideline adherence for head and neck

cancer 107

Part III. Decision making in geriatric oncology

Chapter 9 The value of a comprehensive geriatric assessment in patient care and

prognostication for elderly cancer patients acutely admitted to hospital 125

Chapter 10 Baseline comprehensive geriatric assessments predicts toxicity and

survival in elderly patients with metastatic breast cancer receiving

palliative chemotherapy

141

Chapter 11 The value of geriatric assessments in predicting treatment tolerance

and prognosis in older cancer patients – a systematic review 157

Chapter 12 Frailty screening tools for predicting outcome of a comprehensive

geriatric assessment in older cancer patients – a systematic review 183

Part IV. General discussion, summary and acknowledgements

Chapter 13 General discussion 205

Chapter 14 Summary / Samenvatting 213

Author Affiliations 227

Curriculum Vitae 231

Acknowledgements (Dankwoord) 235

Introduction

Introduction

Demographic changes In Western societies, the population is ageing. In the Netherlands, the proportion of

inhabitants aged 75 years and older has increased from 5% in 1990 to 7% in 2012, and will

continue to increase rapidly to 14% by 2030 before plateauing at around 15% by 2040.1

These changes are caused by two separate demographic developments. First, the post‐

war baby boom generation ‐ clearly visible as a wave passing through the demographic

pyramid (Figure 1) ‐ is soon reaching their seventies. Second, due to medical advances and

increased prosperity, the elderly are becoming older. In the course of the past 50 years,

life‐expectancy has risen by 10 years for women and by 8.5 years for men.1 Another 5

years will be added in the coming thirty years, meaning that by 2040, the average life‐

expectancy will be 87.4 years for women and 84.5 years for men.1

Ageing is very much an individual process, in the course of which differences in genetic

predisposition and life‐style will become apparent, intercurrent and chronic diseases will

leave their mark and the speed with which physiologic reserves decline will vary from

person to person. Essentially, in the course of a life‐time an individual will become more

and more unique and as a result, the elderly form a highly heterogeneous population.

Figure 1: Age structure of the Dutch population in 2012, 2025 and 2040

age age age

-----------------------------------------------------------------------------------------

2012 2025 2040

men x 1000 women x 1000 men x 1000 women x 1000 men x 1000 women x 1000

Source: Statistics Netherlands (cbs.nl)

8

Introduction

Ageing and cancer As the population ages, many diseases that predominantly affect the elderly will become

more prevalent. This also holds true for cancer: in the past 20 years, the number of newly

diagnosed cancer patients rose by 67% in the Netherlands, from 57,000 to over 95,000.2

Although malignant disease can occur at all ages, it disproportionally strikes those aged 75

years and older. While only 7% of Dutch inhabitants currently falls into this age group,

30% of malignancies occur in individuals aged 75 years and older as well as 46% of cancer‐

related deaths,2 and these percentages are expected to rise even further over the coming

decades. Optimal cancer care for the older patient ideally strikes the golden mean

between overtreatment and undertreatment and must be tailored to the individual

patient as well as the customary disease‐related factors incorporated in treatment

guidelines. Cancer specialists are now collaborating with geriatricians to optimize the

decision‐making process. However, such initiatives are complicated by the lack of elderly‐

specific data from clinical trials or even observational cohort studies on which to base such

ecisions. d

Aim and outline of this thesis The aim of this thesis is two‐fold: first, to study current treatment practice and decision

making in older cancer patients and second, to investigate the potential value of a

comprehensive geriatric assessment in this decision‐making process.

Part I consists of four cohort studies, focusing on current patterns of care for breast

cancer in elderly patients. Chapter 1 is a retrospective analysis of the diagnostic process

and treatment in a single centre breast cancer outpatient clinic and compares these to

national guidelines. In Chapter 2 the trends in these treatment practices over a six‐year

period is studied, as well as the impact of the introduction of a multidisciplinary breast

cancer team. Chapter 3 examines survival and cause of death in a historical cohort of 187

patients with resectable breast cancer for whom surgery was omitted. Chapter 4

addresses decision making and referral practices for elderly patients residing in nursing

homes who are suspected of breast cancer. As Chapter 5 demonstrates, accrual of elderly

patients in clinical trials is not straight‐forward, emphasizing the value of observational

studies in understanding and optimizing cancer treatment for these patients.

In Part II, current treatment practices for older patients with two other cancer types are

studied. Chapter 6 compares treatment choices and outcome for older colon cancer

patients with their younger counterparts. Iron‐deficiency anaemia can be a first sign of

colon cancer, and in Chapter 7, we assessed diagnostic decision making and clinical

outcome for a cohort of 112 patients with this condition. Chapter 8 addresses age‐related

differences in the treatment of head and neck cancer, and the consequences of guideline

discordance for cancer‐specific survival.

9

Introduction

10

Part III focuses on the potential value of the comprehensive geriatric assessment (CGA) in

decision making for older cancer patients. In Chapter 9, the value of CGA for patient care

and prognostication is studied in a cohort of elderly cancer patients acutely admitted to

hospital. Chapter 10 addresses the ability of a CGA to predict toxicity and survival in

patients undergoing chemotherapy for metastatic breast cancer. This issue is addressed

further in a systematic review in Chapter 11, focusing on the value of geriatric

assessments in predicting treatment tolerance and prognosis in older cancer patients. As a

CGA is a time‐consuming process, Chapter 12 addresses the value of frailty screening tools

for selecting patients likely to benefit from further geriatric assessment.

Finally, Part IV contains a summary of the main findings of this thesis, and a discussion of

e potential consequences of these finding for clinical practice and future research. th

References 1. www.cbs.nl 2. www.cijfersoverkanker.nl

http://www.cbs.nl/

http://www.cijfersoverkanker.nl/

Part I

Breast cancer treatment in the elderly

Chapter 1

Breast cancer in the elderly: retrospective study on diagnosis and treatment according

to national guidelines

M.E. Hamaker, W.H. Schreurs, H.J. van Slooten, J.M. Uppelschoten, C.H. Smorenburg

The Breast Journal 2009;15:26‐33

Chapter 1

Abstract Introduction: We set out to investigate the level of accordance of diagnosis and treatment

of elderly breast cancer patients with national guidelines and to study predictors of

deviation.

Methods: Data on patient and tumour variables were collected from the charts of 166

patients aged 70 years and older, diagnosed at our hospital in 2002‐2004. Diagnostic

work‐up and treatment were compared to guidelines and reasons for deviation were

recorded.

Results: In all, 122 (74%) patients were diagnosed and treated in accordance with

guidelines. Diagnosis was incomplete in 19 patients (11%). Surgery, radiotherapy and

hormonal therapy were withheld in 19 (11%), 11 (7%) and 9 (5%) patients, respectively.

Guideline deviation was motivated in 18 patients (11%) (comorbidity n=11, patients’

preferences n=5, age n=2), unmotivated in 18 (11%), and undeliberate in 8 (5%).

Conclusion: Our study demonstrates that deviation from guidelines in elderly breast

cancer patients mainly occurs due to a deliberate adjustment to patient’s comorbidity and

preference.

14

Breast cancer in the elderly

Introduction In the Netherlands, breast cancer is diagnosed in over 11,500 women each year, of which

30% is aged 70 or older. It is the most frequently diagnosed cancer in women and its

incidence increases with age. In western societies, due to increasing life expectancy and

ageing of the population, the number of elderly patients with breast cancer will increase

substantially over the next decades. It is expected that by the year 2035, 60% of all new

breast cancer patients are 70 years or older.1

Unfortunately, insufficient data are available on the optimal treatment of elderly patients

with breast cancer. Various studies have demonstrated that treatment for women with

breast cancer differs substantially with age.2‐13 If patients are older, they are treated less

extensively, and treatment guidelines are less likely to be followed.6,14

However, this does not automatically imply that elderly patients are undertreated.

Treatment guidelines are based on clinical trials in which patients 65 years of age or older

are greatly underrepresented or even excluded.15‐17 Increasing comorbidity with age

results in an increasingly heterogeneous patient population.5,6,10 Both the physiological

process of ageing as well as comorbidity result in a decreased physical reserve, and the

question whether or not older patients are able to tolerate treatment as well as younger

patients has not been answered conclusively.18 Comorbidity in elderly patients may also

increase the risk of dying of other causes, thereby decreasing the relative impact of breast

cancer on overall survival.19‐23 In addition, studies have demonstrated that elderly women

are often diagnosed at a later stage of the disease,4,19 and that the biology of breast

cancer changes with age.19 Furthermore, patient’s preferences in treatment modalities

need to be taken into account, as older patients may be less willing to trade current

quality of life for survival.24

These differences suggest that guidelines based on studies examining mainly non‐elderly

patients without comorbidity cannot automatically be extended to elderly patients. It may

be possible that observed differences in treatment between older and younger patients

are not a reflection of age but are adequate adjustments to altered biology, physiology,

comorbidity and preferences of older patients.

We set out to investigate the level of adherence to Dutch guidelines in newly diagnosed

breast cancer patients aged 70 years or older and to study various predictors of deviation

t our hospital. a

Methods This study is a retrospective cohort study of women aged 70 years and older diagnosed

with breast cancer at the Medical Centre Alkmaar in the Netherlands between January

2002 and December 2004. Patients with non‐invasive breast cancer or a second primary

breast tumour were excluded from this study.

15

Chapter 1

Using patient’s charts, surgical records, and pathology reports, we collected data on the

following variables: patient’s age, comorbidities, date of diagnosis, tumour histology,

tumour grade, mitotic activity index (MAI), stage of disease (TNM), oestrogen receptor

status, progesterone receptor status, as well as the initial treatment with surgery,

chemotherapy, radiotherapy and/or hormonal therapy. Breast cancer histology was

classified as ductal carcinoma, lobular carcinoma, other, or “not determined” if no

histological or cytological examination was performed. Steroid receptors were determined

by immunohistochemistry and classified as positive (10% or more cells stained positive),

negative (<10%) or unknown. Stage of disease was classified in five groups: stage I (T1 N0

M0), stage II (T0‐2 N1 M0, or T2 N0‐1 M0, T3 N0 M0), stage III (T3 N1 M0,Tany N2‐3 M0), stage

IV (Tany Nany M1) and unknown. Information on functional status was lacking in most

patient charts and subsequently was not collected.

Patients were followed until March 1st 2007. Data on survival was collected from the

charts; if this was insufficient, the patient’s general practitioner was contacted for

additional information.

In addition, diagnostic work‐up and treatment modalities were compared with work‐up

and treatment advised by national guidelines. In the Netherlands, guidelines on diagnosis

and treatment of breast cancer are frequently updated by the Dutch National Breast

Cancer Platform (NABON) and the Dutch Society for Medical Oncology (NVMO). Table 1

summarizes these guidelines for patients aged 70 and older, as used in the period 2002‐

2004.25 This guideline is similar to that of younger post‐menopausal women for diagnostic

work‐up, surgery, hormonal therapy and radiotherapy. The guideline provides no strict

indication for adjuvant chemotherapy, stating that this can be considered in patients with

high‐risk hormone receptor negative disease.

For patients not receiving the standard treatment, reasons for deviating from guidelines

were collected from the charts. These reasons were classified in three categories:

“motivated” if the reason for deviation was documented, “deliberate, reason not

documented” if the treating physician had noted that the guideline was not followed,

without motivating this decision in the patient’s chart; and “undeliberate” if the chart

owed no remark on any deviation from guidelines. sh

Statistical analysis To compare different groups in terms of comorbidity, disease stage, and different

treatment modalities, a chi‐square test was used. A p‐value of lower than 0.05 was

considered to be statistically significant. Statistical analyses were performed using the

tatistical program SPSS for Windows version 14.0. s

16


Table 1: Diagnosis and treatment guidelines for patients aged 70 and older in accordance with the Dutch multidisciplinary guidelines for treatment of breast cancer 2002

25

I. Diagnosis Histology In all patients, histological confirmation of breast cancer should be sought. Lymph node status In all patients with invasive tumours receiving surgery, a diagnostic procedure to determine axillary node status is required. Distant metastases In all patients with T3‐4 and/or N2‐3 tumours, imaging studies for distant metastases is recommended. II. Treatment Surgery In all tumours except T4 and/or M1, surgery is required. Radiotherapy 1. After breast conserving therapy a. radiation of the breast is always required b. regional lymph nodes: in case of pN2‐N3 tumours 2. After modified radical mastectomy, locoregional radiation is required if: a. mastectomy was irradical b. pN2‐N3 c. positive top axillary node d. cT4 e. to be considered in T3 tumours Hormonal therapy 1. If tumour is hormone receptor negative, no hormonal treatment is necessary. 2. In case of lymph node N0 and receptor positive tumour, hormonal therapy is required if : a. tumour size is > 3 cm b. tumour size is 1‐3 cm and mitotic activity index > 10 3. Lymph nodes N1‐2 and receptor positive tumour always require treatment. 4. M1 disease if hormone receptor positive Chemotherapy For patients over age 70, there is no strict guideline with indications for (neo)adjuvant chemotherapy.

17

Chapter 1

Results

Patient and tumour characteristics Between January 2002 and December 2004, 791 patients were diagnosed with breast

cancer in our hospital, of which 205 patients (26%) were aged 70 years or older. Thirty‐

nine patients were excluded from our analysis because of a second primary breast tumour

(n=23), non‐invasive breast cancer (n=14) and treatment elsewhere (n=2). As a result, 166

patients were included in this study (165 females and one male). Patient characteristics

are listed in Table 2. The median age at diagnosis was 78 years (range 70‐96 years).

In our series, 51 patients presented with stage I disease (31%), 57 with stage II (34%,

node‐negative n=29, node‐positive n=28), 29 with stage III (17%) and 17 with metastatic

disease (stage IV, 10%). Overall, older patients presented with more advanced disease

(p=0.003). Tumour stage was unknown in 12 patients (7%) because of insufficient data on

lymph node involvement or tumour size. The percentage of patients with unknown stage

correlated with increase in age.

Table 2: Patient characteristics per age group

70 ‐ 79 yrs 80 ‐ 89 yrs 90+ yrs p‐value*

Number of patients 100 56 10

Chronic and/or current disease 0 1 2 3+

42 (42%) 29 (29%) 16 (16)% 13 (13)%

10 (19%) 14 (26%) 17 (31%) 13 (24%)

1 (11%) 2 (22%) 3 (33%) 3 (33%)

0.04

Stage of disease I II III IV unknown

40 (40%) 34 (34%) 14 (14%) 8 (8%) 3 (3%)

10 (19%) 22 (39%) 10 (19%) 7 (12%) 7 (12%)

1 (10%) 1 (10%) 5 (50%) 1 (10%) 2 (20%)

0.002

0.003

Tumour morphology Not determined Ductal Lobular Other

1 (1%)

77 (77%) 15 (15%) 7 (7%)

3 (5%)

32 (57%) 14 (25%) 7 (12%)

2 (20%) 4 (40%) 1 (10%) 3 (30%)

Hormone receptors Oestrogen receptor positive Progesterone receptor positive

81 (84%) 67 (69%)

40 (82%) 34 (72%)

4 (80%) 4 (80%)

0.95 0.82

Tumour grade Grade I Grade II Grade III Not determined

12 (12%) 35 (35%) 14 (14%) 39 (39%)

3 (5%)

13 (23%) 10 (18%) 30 (54%)

1 (10%) 0 (0%) 2 (20%) 7 (70%)

ns ns ns 0.02

* p‐value in bold signifies a significant differences between age groups (p<0.05); ns = not significant

18


In the majority of patients (n=160; 96%), a histological examination of tumour was

performed. Tumour morphology and grade were less frequently determined in older age

groups (p=0.023). Hormone receptor (HR) status was determined in 90% of tumours.

There were no significant differences in HR status between age groups.

Data on comorbidity were available for 159 patients (96%). Only 53 patients (32%) had no

current or chronic disease. Twenty‐nine patients (18%) had three or more current diseases

at the time of diagnosis. Both presence and number of comorbidities increased with age

(p=0.044, Table 2). Cardiovascular disease was most frequently observed (53%), including

hypertension (29%), coronary artery disease (16%), and TIA or stroke (7%). Other

comorbidities were diabetes mellitus (13%), COPD (5%), thyroid disease (5%), dementia

(3%), and depression (3%). Apart from a prior diagnosis of breast cancer (excluded),

nother 10 patients (6%) had a prior or concurrent malignancy. a

Treatment Early breast cancer (Stage I‐II) In accordance with guidelines, 99 out of 108 patients (92%) with early breast cancer (stage

I and II disease) had surgery; 66 patients had a modified radical mastectomy (MRM), and

33 patients had breast conserving therapy (BCT). A sentinel lymph node procedure (SNP)

was done in a majority of 62 patients (63%), of whom 14 were tumour positive and

followed by an axillary lymph node dissection (ALND). A primary ALND was performed in

34 patients (34%). In three out of 99 patients (3%), no axillary lymph node surgery was

done. Nine patients did not have any surgery.

Twenty‐nine out of 33 patients (88%) with BCT received radiotherapy as indicated by

guidelines. According to the guidelines, radiotherapy was indicated in eight patients after

MRM, and seven patients (88%) were treated accordingly.

Adjuvant hormonal therapy was indicated in 47 patients and 40 patients (85%) were

treated accordingly. Four patients with stage II disease received adjuvant chemotherapy

due to irradical surgery or lymph node metastases.

Figure 1 lists treatment modalities per age group for patients with early stage breast

cancer. Older patients were less likely to receive surgery (97% in patients aged 70‐79 years

vs. 50% in the patients aged 90 and over, p=0.002). If surgery was performed, older

patients more often had a MRM than BCT (84% in patients aged 80‐89 years vs. 52 % in

patients aged 70‐79 years, p<0.05 after correction for stage of disease). Differences in the

type of lymph node staging procedure were not significant after correction for stage of

disease. Older patients were less likely to receive radiotherapy as compared to younger

patients (0% vs. 49%, p<0.001). The use of hormonal therapy increased with age: 38% in

atients aged 70‐79 compared to 100% of patients aged >90 (p=0.024). p

19

Chapter 1

Figure 1: Treatment modalities per age group for patients with early stage breast cancer

Locally advanced disease (stage III) Twenty‐nine patients were diagnosed with stage III disease. In 19 patients with locally

advanced disease, diagnostic work‐up consisted of a chest X‐ray, abdominal ultrasound,

and skeletal scintigraphy. In five patients, a chest X‐ray and ultrasound were done and in

another three patients a chest X‐ray only. In the remaining two patients, both with T4

disease, no further work‐up was done despite guidelines.

Eleven patients (38%) with locally advanced disease had primary surgery; eight patients

had a MRM and three patients had BCT. Another eight patients had an MRM after

neoadjuvant treatment (hormonal therapy in five patients and chemotherapy in three

patients). Ten patients (35%) did not have any surgical therapy. Radiotherapy was given to

24 out of 25 patients in accordance with guidelines. Hormonal treatment was not

indicated in four patients due to HR negative tumours. In HR positive breast cancer, 24 out

f 25 patients received adjuvant hormonal therapy. o

Metastatic disease (stage IV) Sixteen out of 17 patients (94%) with metastatic disease received hormone therapy. In

three patients chemotherapy was part of palliative therapy. The oldest patient to receive

chemotherapy was 78 years old. One patient received no form of systemic therapy, and

none had surgery or radiotherapy as primary treatment.

20


Unknown stage of disease Two out of 12 patients (17%) with unknown stage of disease had surgery; one BCT and

one MRM. No axillary lymph node surgery was performed. One patient, with HR negative

disease, received no adjuvant treatment. Eleven patients, with either HR positive disease

r in which receptor status was unknown, received primary or adjuvant hormonal therapy. o

Adherence to guidelines The accordance with guidelines and reasons for deviation are listed in Table 3. Overall,

122 out of 166 patients (74%) were adequately staged and given surgical and adjuvant

treatment in accordance with guidelines.

Diagnosis and staging were incomplete in 19 patients (11%). In six patients (4%), diagnosis

was made on clinical data only, without cytological or histological confirmation;

furthermore, for three of these patients, stage was not determined. In six patients (4%),

no lymph node staging procedure was performed during surgery, and in another seven

patients (4%), the diagnostic strategy supplied insufficient information for staging.

In older patients, treatment guidelines were followed less frequently than in younger

patients (88% of patients aged 70‐79 was treated in accordance to guidelines vs. 40% in

patients aged 90+, p<0.001).

Surgery was withheld in nineteen patients (11%), of whom nine with stage I‐II disease and

ten with stage III disease. The reason stated most frequently was the presence of

comorbidity (n=10), varying from moderate chronic obstructive pulmonary disease to

severe cardiovascular disease, dementia or other metastatic malignancy. Other reasons

stated were patient’s request (n=4), age (n=2) and in two patients the reason was not

clear. One patient died of myocardial infarction before any treatment could be started.

Eleven patients (7%) did not receive radiotherapy despite guidelines. One patient refused

radiotherapy, and for the other patient, the reason stated was her age and comorbidity. In

the remaining patients, it remains unclear why the guidelines were not followed.

Nine patients (5%) did not receive hormonal therapy according to guidelines; reasons for

this were not stated in patients’ charts.

Overall, deviation from diagnostic and treatment guidelines was deliberate in 36 patients

(22%); in 18 of them the decision was motivated in patient’s charts. In only eight patients

(5%), the treating physician seemed unaware of the fact that guidelines were not

followed. In the majority of these patients, this consisted of omitting adjuvant hormonal

or radiotherapy. All eight had early stage breast cancer.

Patients not treated according to guidelines were significantly older (83.0 vs. 76.9 years,

p<0.001), and had more comorbidity (1.7 vs. 1.2 comorbid conditions, p=0.024). There

were no differences in hormone receptor status or stage of disease. In a multivariate

analysis, no single factor was an independent predictor of adherence to guidelines, with

21

Chapter 1

the exception of patients with stage II disease, in which case both advanced age (p=0.028)

and increasing comorbidity (p=0.016) were significantly correlated with deviation from

guidelines.

Table 3: Adherence to national guidelines

70‐79

yrs

n=100

80‐89

yrs

n=56

90+

yrs

n=10

Total

n=166

Treatment and diagnosis in accordance with

guidelines***

86

(86%)

34

(61%)

4

(40%)

122

(74%)

Incomplete diagnosis

No histological/cytological confirmation

No axillary lymph node procedure during surgery

Insufficient information for staging

1

2

2

3

4

6

2

0

2

6

6

10

A.*

Insufficient treatment

No surgery

No radiotherapy

No hormonal therapy

4

4

5

12

3

4

3

4

0

19

11

9

B. ** Motivated deviation***

Comorbidity

Malignancy of another origin

Dementia

Moderate to severe cardiovascular disease

Multiple sclerosis

Parkinson’s disease

Moderately severe COPD

Age

At patient’s request

Deliberate deviation, reason not documented

Undeliberate deviation

3

0

1

4

4

4

2

4

9

4

4

0

0

3

0

11

2

2

4

1

1

1

2

5

18

8

* Panel A lists the number of patients in which guidelines were followed and the ways in which treatment or staging deviated from guidelines. ** Panel B lists the reasons for deviating from guidelines as stated in the patients’ chart. *** In some patients, both staging and treatment was not according to guidelines, or more than one treatment modality was withheld. Similarly, in some patients, multiple reasons for deviating from guidelines were given Therefore, patients may be listed more than once.

22


Figure 2: Kaplan Meier survival plots, stratified per stage of disease

Survival After a median follow‐up of 38 months, 70.5% of patients were still alive. Figure 2 shows

Kaplan Meier plots, stratified per stage of disease.

Patients who died were significantly older (p<0.001), had more comorbidity (p=0.001), had

more advanced disease (p=0.034) and were less likely to have been treated in accordance

with guidelines (p=0.003). In a multivariate analysis, only comorbidity and stage of disease

mained independent predictors of mortality. re

Discussion In this study, we examined the diagnosis and treatment patterns of women with newly

diagnosed breast cancer, aged 70 and older, in relation to age, disease stage and

comorbid conditions. Using Dutch national guidelines, which are stage‐specific and

modified for age, we found that in 74% of patients guidelines were followed, and that

adherence to guidelines decreased with advancing age.

Several other studies have examined adherence to treatment guidelines in elderly breast

cancer patients. Giordano et al noticed that in patients aged >75 years, accordance with a

stage‐specific guideline decreased with age, varying from 80% for surgical therapy to 29%

23

Chapter 1

for adjuvant chemotherapy.14 Yancik et al found that only 7% of 448 patients with stage I

and II breast cancer aged >70 were given less than minimally expected treatment, but this

study only examined surgical treatment.6 Bouchardy et al observed that nearly 50% of

patients aged >80 had sub‐optimal treatment strategies;7 however, optimal treatment

was not clearly defined and was not stage‐specific. As breast cancer treatment guidelines

depend on age and stage of disease, both factors need to be taken into account to

examine accordance to these guidelines. One of the strengths of our study is that national

guidelines we used are stage‐specific and have been modified for age.

Several studies have shown that patients with comorbidity are less likely to be treated

according to guidelines than patients without comorbidity.5‐7,9,20,24 Similarly, we found that

comorbidity was the most frequently stated reason for deviating from guidelines (11 out

of 18 motivated deviations, Table 3), also being an independent predictor of mortality.

Other studies have demonstrated that the impact of breast cancer on prognosis decreases

as the risk of dying from comorbid disease increases.20‐23 For example, Coebergh et al

found that, although overall 10‐year survival is much lower in women aged >70 as

compared to younger patients (16% vs. 41%, respectively), the cancer‐specific survival is

quite similar (49% vs. 53%).27 Therefore, serious comorbidity can be a legitimate reason

for deviating from treatment guidelines if the life‐expectancy of a patient is significantly

reduced by it. In our study, the adjustment of treatment in ten patients because of

comorbidity seems adequate, considering the severity of these comorbid conditions

(Table 3).

Age was not a major factor in determining whether guidelines were followed, being a

reason for deviating from treatment guidelines in only two patients in our study. Although

older women were less likely to be treated according to guidelines, age was not an

independent predictor of guideline deviation after correcting for comorbidity and stage of

disease. Bickell et al recently published a study on reasons for omitting adjuvant therapy

in both elderly and non‐elderly breast cancer patients.26 Physicians did not recommend

therapy because of older age, comorbidity or unawareness of treatment benefit in 14%,

11% and 3% of treatment omissions, respectively, while 31% of treatment omissions was

due to patient preference. In our study, deviation from guidelines was undeliberate in

eight out of 44 patients with any guideline deviation; this occurred primarily in patients

with early stage breast cancer and frequently involved hormonal therapy. Given the

relatively low treatment burden for the patient, this possibly implies undertreatment. Of

notice, other studies have observed that older women with a relatively low risk of

recurrence (i.e. early stage disease) are less likely to receive treatment according to

guidelines. For example, Giardano et al found that in post‐menopausal women, deviations

from guidelines were found primarily in early stage breast cancer, while advanced disease

24


was highly predictive of accordant therapy.14 This suggests that in early stage disease the

need for adjuvant therapy is not as obvious as it is in advanced disease.

At present, diagnostic and therapeutic decisions for patients with breast cancer in the

Netherlands are made in multidisciplinary meetings with dedicated breast surgeons,

radiographers, pathologists, oncologists, radiation oncologists and breast care nurses. In

our hospital, these meetings were initiated in 2006, thus after the time period of this

study. We intend to evaluate the effect of these multidisciplinary meetings on the

adherence to national guidelines in the elderly.

Most studies on the effect of age and comorbidity in elderly breast cancer patients have

used national databases, with central data collection by national cancer institutes.6

However, patient’s medical records are generally regarded as the most complete source of

information on the patient’s past and current health status.20 Thus, one of the strengths of

this study is that the data we accumulated using patients’ charts may be more clinically

precise than those obtained from self‐reports or administrative databases using discharge

data.

However, in a retrospective analysis, it is impossible to determine whether discordance

with guidelines is justifiable based on patient’s health status or should be considered as

undertreatment. Furthermore, it is still not certain how to measure health status or

functional reserves in elderly cancer patients.7,9,19,28‐29 Therefore, prospective studies are

needed to validate comprehensive geriatric assessments as a decision‐making tool in

geriatric oncology.28 Until such a tool is developed and incorporated into guidelines,

treating physicians will have to determine for each individual patient whether or not

uidelines provide adequate treatment. g

In conclusion, this study provides new insight into the accordance with diagnosis and

treatment guidelines in elderly breast cancer patients. Our study demonstrates that

deviation from treatment guidelines occurs in a fourth of patients, mainly due to

deliberate deviation from guidelines as an adjustment to the patient’s comorbidity and

preference.

25

Chapter 1

References 1. Hillen HF, Hupperets PS. [Breast cancer in patients 70 years and older]. Ned Tijdschr Geneesk

2000;144:1099‐104. 2. Vulto AJ, Lemmens VE, Louwman MW et al. The influence of age and comorbidity on receiving

radiotherapy as part of primary treatment for cancer in South Netherlands, 1995‐2002. Cancer 2006;106:2734‐42.

3. Wanebo HJ, Cole B, Chung M et al. Is surgical management compromised in elderly patients with breast cancer? Ann Surg 1997;225:579‐89.

4. Bergman L, Kluck HM, Leeuwen FE van et al. The influence of age on treatment choice and survival of elderly breast cancer patients in South‐Eastern Netherlands: a population‐based study. Eur J Cancer 1992;28A:1475‐80.

5. Janssen‐Heijnen ML, Maas HA, Lemmens VE et al. Samenhang van leeftijd en comorbiditeit met therapie en overleving bij patiënten met kanker in Noord‐Brabant en Noord‐Limburg 1955‐2001. Ned Tijschr Geneeskd 2005;149:1686‐90.

6. Yancik R, Wesley MN, Ries LAG et al.. Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. JAMA 2001;285:885‐892.

7. Bouchardy C, Raptiti E, Rioretta G et al. Undertreatment strongly decreases prognosis of breast cancer in elderly women. J Clin Oncol 2003;21:3580‐7.

8. Livi L, Paiar F, Saieva C et al. Breast cancer in the elderly: treatment of 1500 patients. The Breast Journal 2006;4:353‐9.

9. Louwman WJ, Janssen‐Heijnen ML, Houterman S et al. Less extensive treatment and inferior prognosis for breast cancer patients with comorbidity: a population‐based study. Eur J Cancer 2005;41:779‐85.

10. Hurria A, Leung D, Trainor K et al. Factors influencing treatment patterns of breast cancer patients age 75 and older. Crit Rev Oncol Hematol 2003;46:121‐6.

11. Gajdos C, Tartter PI, Bleiweiss IJ et al. The consequence of undertreating breast cancer in the elderly. J Am Coll Surg 2001;192:598‐707.

12. Enger SM, Thwin SS, Buist DS et al. Breast cancer treatment of older women in integrated health care settings. J Clin Oncol 2006;24:4377‐83.

13. Ballard‐Barbash R, Potosky AL, Harlan LC et al. Factors associated with surgical and radiation therapy for early stage breast cancer in older women. J Natl Cancer Inst 1996;88:716‐26.

14. Giordano SH, Horobagyi GN, Kau SW et al. Breast cancer treatment guidelines in older women. J Clin Oncol 2005;23:783‐91.

15. Townsley CA, Selby R, Siu LL. Systematic review of barriers to the recruitment of older patients with cancer onto clinical trials. J Clin Oncol 2005;23:3112‐24.

16. Hutchins LF, Unger JM, Crowley JJ et al. Underrepresentation of patients 65 years of age or older in cancer‐treatment trials. N Engl J Med 1999:341:2061‐7.

17. Early Breast Cancer Trialist Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15‐year survival: and overview of the randomised trials. Lancet 2005;365:1687‐717.

18. Silliman RA. What constitutes optimal care for older women with breast cancer? J Clin Oncol 2003;21:3554‐6

19. Diab SG, Elledge RM, Clark GM. Tumour characteristics and clinical outcome of elderly women with breast cancer. J Natl Cancer Inst 2000;92:550‐6.

20. Houterman S, Janssen‐Heijnen ML, Verheij CD et al. Comorbidity has negligible impact on treatment and complications but influences survival in breast cancer patients. Br J Cancer 2004;90:2332‐7.

21. Rijke JM de, Schouten LJ, Hillen HF et al. Cancer in the very elderly Dutch population. Cancer 2000;89:1121‐33

22. Extermann M, Balducci L, Lyman GH. What threshold for adjuvant therapy in older breast cancer patients. J Clin Oncol 2000;18:1709‐17.

23. Read WL, Tierney RM, Page NC et al. Differential prognostic impact of comorbidity. J Clin Oncol 2004;22:3099‐103.

24. Yellen SB, Cella DF, Leslie WT. Age and clinical decision making in oncology patients. J Natl Cancer Inst 1994;86:1766–1770.

26


27

25. Multidisciplinaire richtlijn voor de behandeling van borstkanker. Kwaliteitsinstituut voor gezondheidszorg CBO 2002. Revised 2004.

26. Bickell NA, LePar F, Wang JJ, Leventhal H. Lost opportunities: physician’s reasons and disparities in breast cancer treatment. J Clin Oncol 2007;23:2516‐21.

27. Coebergh JW, Heyden LH van der, Jansen‐Heijnen ML. Cancer incidence and survival in the South‐east of the Netherlands. Eindhoven: integraal kankercentrum Zuid 1995.

28. Terret C, Zulian G, Droz JP. Statements on the interdependence between the oncologist and the geriatrician in geriatric oncology. Crit Rev Oncol Hematol 2004;52:127‐22.

29. Extermann M, Aarpo M, Barnabei R et al. Use of comprehensive geriatric assessment in older cancer patients. Crit Rev Oncol Hematol 2005;55:241‐52.

Chapter 2

Trends in breast cancer treatment in the elderly at a breast cancer outpatient clinic

M.E. Hamaker,

W. H. Schreurs, H.J. van Slooten, J.M. Uppelschoten, C.H. Smorenburg

Nederlands Tijdschrift voor Geneeskunde 2009;153:A562

A translation of: “Trends in de behandeling van borstkanker bij ouderen op een mammapolikliniek: minder vaak operatieve behandeling”

Chapter 2

Abstract Aim: To assess the treatment of elderly patients with resectable breast cancer treated at

the Medical Centre Alkmaar before and after the implementation of a multidisciplinary

breast cancer team in February 2006.

Methods: Retrospective chart review of all patients aged 70 years and older with a newly

diagnosed resectable breast tumour (stage I and II) treated between 2002 and 2007.

Results: Of 232 patients, 84% received surgical treatment. Adjuvant treatment with

radiotherapy, endocrine therapy and chemotherapy was given to 88, 91 and 5 patients

respectively. Over time, there was a significant decrease in the percentage of patients that

were treated surgically, particularly for patients aged 80 years and older. The number of

patients receiving radiotherapy in accordance with guidelines improved over time, while

the guideline adherence for endocrine treatment remained more or less stable. The

reporting of decisions regarding treatment and guideline discordance improved.

Conclusion: Between 2002 and 2007, treating physicians seemed more aware of guidelines

and guideline discordance was more clearly motivated in the patient’s chart. Guideline

adherence for radiotherapy improved but there was a significant decrease in surgical

treatment over time, particularly in patients aged 80 years and older.

30

Trends in breast cancer treatment for the elderly

Introduction In the coming decades, increasing life expectancy and ageing of the population will result

in a substantial increase in the number of older cancer patients;1‐4 it is expected that by

2035, 60% of all new breast cancer patients will be over 70 years of age.2 Although

attempts have been made to actively include older patients in oncologic trials, most

treatment guidelines are still based on research done in relatively fit patients.3 In addition,

changes in tumour characteristics with increasing age have important prognostic and

therapeutic implications.5‐7 Furthermore, comorbidity, polypharmacy, decreased

functional capacity and decreased physiological reserve result in increasing heterogeneity

in the elderly population;8,9 these factors must be taken into account in the decision‐

making process.10,11

Previous studies have shown that case review by a multidisciplinary breast cancer team at

a breast cancer clinic resulted in an alteration of the treatment for over 50% of patients.12

Furthermore, pre‐operative consultation with an oncologist has been shown to result in

better guideline adherence for breast cancer patients.13

In the Netherlands, decision making for breast cancer treatment is increasingly done in a

multidisciplinary breast cancer team (MBCT),14 which includes treating physicians

(surgeons, oncologists, radiotherapists), supporting specialists (pathologists, radiologists)

and specialised nurses. These MBCTs have been developed by the Integral Cancer Centres

(IKC) and their procedures have been recorded in the “Organization of Breast Cancer

Care” written by the National Breast Cancer Council Netherlands (NABON).

In February 2006, a weekly MBCT meeting was initiated at the Medical Centre Alkmaar in

the Netherlands. In this study, we examined if the implementation of this MBCT

influenced treatment decisions and guideline adherence for older patients with a

sectable breast cancer treated at the breast cancer out‐patient clinic of this hospital. re

Methods We performed a retrospective cohort study of patients aged 70 years and older, treated at

the Medical Centre Alkmaar, in the Netherlands, for a newly diagnosed, primary

resectable breast cancer (stage I or II) between January 2002 and December 2007.

Patients with a second primary breast cancer, involvement of a fixed parasternal (N2),

subclavicular or supraclavicular lymph node (N3) or distant metastases (M1) were

excluded.

The following variables were collected from the patient’s medical and surgical charts, and

pathology reports: age, date of diagnosis, histological data, tumour grade, mitotic activity,

stage (TNM), oestrogen and progesterone receptor status, expression of human epidermal

growth factor type 2 (Her2Neu), as well as data on the initial treatment. The stage of

31

Chapter 2

disease was subdivided into two categories: stage I (T1N0M0) and stage II (T0‐2N1M0 or

T3N0M0).

The presence of the following comorbid diseases was recorded: valvular disease, angina

pectoris, myocardial infarction, heart failure, arrhythmias, left ventricular hypertrophy,

hypertension, diabetes, stroke, Parkinson’s disease, pulmonary embolism, chronic

obstructive lung disease, rheumatoid arthritis, renal insufficiency, prior malignancies,

osteoporosis, thyroid disease, dementia, mood disorders, psychotic disorders and

personality disorders.

The actual diagnostic process and initial treatment were compared to the national

guidelines, as developed by the NABON.15,16 Table 1 lists an overview of the

recommendations for diagnosis and treatment for patients aged 70 years and over during

the study period.

In case of discordance with guidelines, the patient’s chart was consulted to find the

motivation behind these treatment choices. Table 1: Diagnostic and treatment guidelines for stage I and II breast cancer in patients aged 70 years and older

15,16

Indication

Histology Breast cancer should be histologically confirmed in all patients. Diagnostic procedures Lymph node

status In case of surgical treatment, a diagnostic procedure for assessing lymph node status is required.

Surgery Surgery is indicated in all patients with stage I and II disease.

Radiotherapy 1. Radiotherapy is always required after breast conserving treatment. 2. Locoregional radiotherapy is also required in case of irradical mastectomy.

Endocrine therapy Jan 2002 to Sept 2005

1. In case of hormone receptor negative breast cancer, endocrine treatment is not required. 2. In the absence of lymph node metastases and in case of hormone receptor positive disease, endocrine treatment is recommended if: a. tumour is larger than 3 cm b. tumour size is between 1 and 3 cm and mitotic activity is greater than 10 3. In case of lymph node metastases and hormone receptor positive disease, endocrine treatment is always required.

Endocrine therapy Oct 2005 to Dec 2007

1. As in Jan 2002 to Sept 2005 2. In the absence of lymph node metastases and in case of hormone receptor positive disease, endocrine treatment is recommended if: a. tumour is larger than 1 cm and grade III disease b. tumour is larger than 2 cm and grade II‐III c. tumour is larger than 3 cm 3. As in Jan 2002 to Sept 2005

Treatment

Chemotherapy

There are no strict recommendations for chemotherapy in patients aged 70 years or older. Chemotherapy should be considered for patients with hormone receptor negative disease and a high risk of recurrence.

32


Statistical analysis To assess differences between age groups and differences in guideline adherence

between various subgroups, the SPSS (Statistical Package for the Social Sciences) version

14.0 was used. The chi‐square test was used for nominal and ordinal variables. For

continuous variables with a normal distribution, the Student t‐test was used, and for

ontinuous variables with a non‐normal distribution the Mann‐Whitney test. c

Results

Patient and tumour characteristics Between January 2002 and December 2007, 285 patients aged 70 years and older

presented with a resectable breast cancer at our clinic. After exclusion of 53 patients with

a second primary breast cancer, 232 patients (229 women, 3 men) remained for further

analysis. Median age of these patients was 77.9 years (range 70‐94 years; Table 2). For 102

patients, disease was at stage I (44%), and for 120 at stage II (56%).

For the great majority of patients, the presence of breast cancer was histologically

confirmed (n=227, 98%). Hormone receptor status was assessed in 220 patients (95%);

86% (189/220) was oestrogen receptor positive and 74% (163/220) was progesterone

receptor positive. Her2Neu status was assessed in 84 patients, and an overexpression of

these receptors was found in seven patients (8.3%).

Over time, no significant changes were seen in the stage of disease at presentation or the

tumour characteristics.

Data on comorbidity was available for 230 patients; 68 had no concurrent disease (30%).

The most frequent comorbid conditions were cardiac arrhythmias (n=33), angina pectoris

or myocardial infarction (n=40), hypertension (n=81), diabetes (n=32), heart failure (n=11),

stroke (n=22), pulmonary disease (n=13), psychiatric disorders and dementia (n=19) and a

prior malignancy (n=18). Comorbidity increased over time; in 2002, 19% of patients had

three or more comorbid conditions, which increased to 33% by 2007 (p=0.059).

Table 2: Patient characteristics

Total 2002‐3 2004‐5 2006‐7 p‐value*

Number of patients 232 71 75 86

Median age 75.9 years 76.8 years 78.4 years 78.3 years 0.178

Stage I Stage II

102 (44%) 130 (56%)

33 (46%) 38 (54%)

31 (41%) 44 (59%)

38 (44%) 48 (56%)

ns

Oestrogen receptor positive Progesterone receptor positive Her2neu overexpression

189/220 (86%)163/220 (74%)

7/84 (8%)

55/67 (82%)48/67 (71%)

0/0 (‐)

65/74 (88%)55/73 (75%)4/25 (16%)

76/86 (88%) 65/86 (75%) 3/59 (5%)

ns ns ns

Number of comorbid disease 1.5 1.3 1.4 1.8 0.059 *ns = not significant

33

Chapter 2

Treatment Overall, 195 patients received surgical treatment (84%); 109 had a modified radical

mastectomy and 86 patients had breast conserving surgery. In 192 of the 195 patients

treated surgically, an additional procedure for staging of the axillary nodes was

performed; for 128 of these, a sentinel node procedure was used.

Of the 86 patients with breast conserving surgery, 80 received adjuvant radiotherapy

(93%) in accordance with guidelines. Radiation after a modified radical mastectomy was

recommended by guidelines for nine patients, and eight were treated accordingly (88%).

Adjuvant hormonal treatment was recommended for 103 patients, and was given to 91 of

these (88%). In addition, five patients received adjuvant chemotherapy due to irradical

surgery, high risk disease or lymph node metastases.

Despite resectable disease, surgery was withheld in 37 patients (13 with stage I disease, 24

with stage II). These patients were significantly older (median age 86.1 years vs. 76.3 years

in patients treated surgically, p<0.001) and had significantly more comorbidity (median

number of disease 2.43 vs. 1.33 in patients treated surgically, p<0.001). For one patient,

no treatment was given and for two patients, treatment was unclear. The remaining 34

patients received primary endocrine treatment (tamoxifen n=21, anastrozole n=13). Two

patients eventually did receive a modified radical mastectomy, one due to disease

progression after 10 months of endocrine treatment, and one due to an alteration in the

patient’s wishes 11 months after diagnosis. At the end of the study period, after a median

follow‐up of 1.9 years, 14 of these 37 patients had stable disease, three had progressive or

metastatic disease and 13 patients had died. Cause of death was generally not available. A

further seven patients were lost to follow‐up.

Figure 1 shows the changes in guideline adherence over time. The use of endocrine

treatment and chemotherapy remained more or less the same, while the use of

radiotherapy increased (2002‐2003 85%, 2004‐2005 97%, 2006‐2007 100%, p=0.03). There

was a significant decrease in guideline adherence for surgical treatment: in the first two

years 94% of patients with stage I and II were treated surgically, but this decreased to 84%

in 2004‐2005 and was as low as 76% in 2006‐2007 (p=0.002). This decrease was

particularly evident in patients aged 80 years and over, for whom guideline adherence

decreased from 80% in 2002‐2003 to 33% in 2006‐2007 (p=0.01); in the same time period,

the percentage of patients aged 70‐80 years that received surgical treatment, as

commended by guidelines, remained over 95%. re

Guideline adherence and the MBCT Overall, 170 of the 232 patients were diagnosed and treated in accordance with national

guidelines (77%). Over time, a non‐significant decrease in guideline adherence was seen:

83% of patients (59/71) in 2002‐2003, 77% (58/75) in 2004‐2005, and 74% (62/85) in

34


2006‐2007 (p=0.23). As demonstrated by Figure 1, this decrease was primarily due to the

decrease in the percentage of patients treated surgically.

Reporting of reasons behind guideline discordance improved over time: in 2002‐2003 this

decision was motivated in 22% of patients (2/9), in 2004‐2005 in 93% (12/13) and in 2006‐

2007 in 100% (21/21, p<0.001). Of the 35 documented decisions not to follow guidelines,

19 were due to comorbidity (42%) including previous malignancy (n=3), cardiac disease

(n=5), dementia (n=1), depression (n=3) and stroke (n=4). Other documented reasons for

guideline discordance were the patient’s overall condition (n=4, 9%), age (n=5, 13%) or the

patient’s preference (n=14, 35%).

The weekly MBCT conference at the Medical Centre Alkmaar was initiated in February

2006. In the first year, 82% of newly diagnosed patients were discussed at this conference;

this percentage increased to 98% in 2007. Although a direct comparison of treatment

before and after the start of the MBCT conference does show significant differences in

guideline adherence, Figure 1 shows that these are due to trends that were already visible

in the years before the MBCT meetings were initiated. This makes causal relationship

between the start of the MBCT conferences and these developments unlikely.

Figure 1: Treatment patterns per two year interval

Surg=surgery, Ax=surgical procedure for axillary lymph node staging, RT=radiotherapy, ET=endocrine therapy, CT=chemotherapy Percentages represent the amount of patients for whom a specific treatment was recommended that actually receiving this treatment. For chemotherapy, this percentage represents the number of patients for whom chemotherapy should have been considered that actually received chemotherapy. The multidisciplinary breast cancer team meetings started in February 2006.

35

Chapter 2

Discussion In this study, we assessed changes in the treatment of resectable breast cancer in patients

aged 70 years and older in a large peripheral Dutch hospital between 2002 and 2007. Over

time, the use of endocrine treatment and chemotherapy remained stable, and an increase

was seen in the use of adjuvant radiotherapy. However, we found a significant decrease in

the number of patients treated surgically, particularly in patients aged 80 years and older,

for whom adherence to surgical guidelines decreased from 80 to 33% over time. These

patients were older and had more comorbidity than surgically treated patients, and often

received primary endocrine therapy instead. Interestingly, we also found an increase in

the number of comorbid conditions for the entire patient group. This trend has been seen

in other studies.17,18

Breast cancer surgery in itself has a low mortality risk (0.0‐0.3%).1,10 However, previous

studies have demonstrated that comorbid conditions are the primary cause of post‐

operative complications and mortality.19 Furthermore, 10% of older patients experience a

post‐operative decrease in cognitive function.20 In addition, multiple studies have shown

that comorbidity plays a greater role in overall survival than cancer‐specific mortality,

particularly in case of early stage disease.1,10,11,21 This is also demonstrated by a recent

Cochrane review, on breast cancer in women aged 70 years and older,22 which found no

difference in overall survival between patients receiving surgical treatment only when

compared to endocrine monotherapy. Limited life‐expectancy due to comorbidity could

therefore be a legitimate reason to withhold surgery. However, progressive disease can

have a significant impact on quality of life; the same Cochrane review demonstrated that

primary endocrine treatment results in a significant decrease in progression‐free

survival.22

We found that between 2002 and 2007 treating physicians seemed to be more aware of

guidelines as guideline discordance was more frequently motivated in the patient’s chart:

in 2002‐2003 only 22% of these decisions were motivated in the chart compared to 100%

in 2006‐2007 (p<0.001). Comorbidity (in 42%) and patient’s preference (in 35%) were the

primary reasons to adjust treatment.

Although assessing for differences between treatment before and after the

implementation of the weekly MBCT conference reveals some differences, these are

based on trends already visible before the start of the MBCT meetings (Figure 1). The

implementation of these conferences thus seems to be a part of a greater awareness of

the importance of adequate treatment for older patient with breast cancer, and a

recognition of the complexity of treatment decisions, requiring a multidisciplinary

approach. This fact has also been stressed in a recent report by the Dutch Health Care

Inspection (IGZ) on coordinated care trajectories for cancer patients, addressing the

36


importance of multidisciplinary treatment and collaboration of the different specialists

volved in the treatment process.23 in

In conclusion, between 2002 and 2007, we found an increased awareness of treatment

guidelines among the treating physicians involved in the treatment of breast cancer;

treatment decisions were better motivated. For older patients, guidelines on adjuvant

radiotherapy were followed more frequently, but we found a significant decrease in the

use of surgical treatment, particularly in patients aged 80 years and older. Comorbidity

and patient’s preference were of primary importance for guideline discordance. Well‐

coordinated multidisciplinary treatment remains of the utmost importance, particularly in

older cancer patients.

37

Chapter 2

38

References 1. Wildiers H, Kunkler I, Biganzoli L et al. Management of breast cancer in elderly individuals:

recommendations of the International Society of Geriatric Oncology. Lancet Oncol 2007; 8:1101‐1115. 2. Hillen HF, Hupperets PS. [Breast cancer in patients, 70 years or older]. Ned Tijdschr Geneeskd

2000;144:1099‐1104. 3. Smorenburg CH, Sijp JR. [Breast cancer in the elderly]. Ned Tijdschr Oncol 2006;3:247‐252. 4. Janssen‐Heijnen ML, Maas HA, Lemmens VE et al. [The correlation of age and comorbidity with therapy and

survival in cancer patients in North‐Brabant and North‐Limburg, 1995‐2001]. Ned Tijdschr Geneeskd 2005;149:1686‐1690.

5. Molino A, Giovannini M, Auriemma A et al. Pathological, biological and clinical characteristics, and surgical management, of elderly women with breast cancer. Crit Rev Oncol Hematol 2006;59:226‐233.

6. Gennari R, Curigliano G, Rotmensz N et al. Breast carcinoma in elderly women: features of disease presentation, choice of local and systemic treatments compared with younger postmenopasual patients. Cancer 2004;101:1302‐1310.

7. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical outcome of elderly women with breast cancer. J Natl Cancer Inst 2000;92:550‐556.

8. Extermann M, Overcash J, Lyman GH et al. Comorbidity and functional status are independent in older cancer patients. J Clin Oncol 1998;16:1582‐1587.

9. Louwman WJ, Vulto JC, Verhoeven RH et al. Clinical epidemiology of breast cancer in the elderly. Eur J Cancer 2007;43:2242‐2252.

10. Wyld L, Reed M. The role of surgery in the management of older women with breast cancer. Eur J Cancer 2007; 43:2253‐2263.

11. Janssen‐Heijnen ML, Maas HA, Houterman S et al. Comorbidity in older surgical cancer patients: influence on patient care and outcome. Eur J Cancer 2007;43:2179‐2193.

12. Newman EA, Guest AB, Helvie MA et al. Changes in surgical management resulting from case review at a breast cancer multidisciplinary tumor board. Cancer 2006;107:2346‐2351.

13. Keating NL, Landrum MB, Ayanian JZ et al. Consultation with a medical oncologist before surgery and type of surgery among elderly women with early‐stage breast cancer. J Clin Oncol 2003;21:4532‐4539.

14. van Nes JG, van de Velde CJ. [The multidisciplinary breast cancer care team: promoting better care]. Ned Tijdschr Geneeskd 2005;149:1929‐1931.

15. Multidisciplinaire richtlijn voor de behandeling van borstkanker. Kwaliteitsinstituut voor gezondheidszorg CBO 2002. Revisie 2004.

16. Multidiciplinaire richtlijn voor de behandeling van borstkanker. Kwaliteitsinstituut voor gezondheidszorg CBO 2005.

17. Jagger C, Matthews RJ, Matthews FE et al. Cohort differences in disease and disability in the young‐old: findings from the MRC Cognitive Function and Ageing Study (MRC‐CFAS). BMC Public Health 2007;7:156.

18. Deeg D. Gezond ouder worden. Geriatrie dagen; 2008. 19. Tiret L, Desmonts JM, Hatton F, Vourc'h G. Complications associated with anaesthesia‐‐a prospective survey

in France. Can Anaesth Soc J 1986;33:336‐344. 20. Samain E, Schauvliege F, Deval B, Marty J. Anesthesia for breast cancer surgery in the elderly. Crit Rev Oncol

Hematol 2003;46:115‐120. 21. Satariano WA, Ragland DR. The effect of comorbidity on 3‐year survival of women with primary breast

cancer. Ann Intern Med 1994;120:104‐110. 22. Hind D, Wyld L, Beverley CB, Reed MW. Surgery versus primary endocrine therapy for operable primary

breast cancer in elderly women (70 years plus). Cochrane Database Syst Rev 2006;(1):CD004272. .

Chapter 3

Omission of surgery in elderly patients with early stage breast cancer

M.E. Hamaker, E. Bastiaannet, D. Evers, W. van de Water, C.H. Smorenburg, E. Maartense,

A.M. Zeilemaker, G.J. Liefers, L. van der Geest, S.E. de Rooij, B.C. van Munster, J.E.A. Portielje

European Journal of Cancer 2012 Sept 6 [Epub ahead of print]

Chapter 3

Abstract Aim: To assess national trends over time in surgery for elderly patients with resectable

breast cancer and to evaluate clinical outcome and cause of death after omission of

surgery in a regional cohort of elderly patients.

Methods: National trends in 1995‐2005 were calculated using cancer registry data. In

addition, a chart review was performed in a cohort of patients aged ≥75 years, with early

stage breast cancer but no primary surgery, diagnosed at five Dutch hospitals in 1990‐

2005. Patient characteristics, comorbidity and reason for omission of surgery were

collected from the chart. Cause of death was retrieved from death certificate data

registered at Statistics Netherlands.

Results: Omission of surgery increased significantly over time for patients aged 80 years

and older (p<0.05). Of the 187 patients in the regional cohort (median age 85.9 years

(range75.0‐97.7), 174 (92%) received hormonal therapy. Omission of surgery was at the

patient's request in 59 patients (32%). Of the 178 patients that died during follow‐up, 60

patients (34%) died of breast cancer. For 81 patients (45%), breast cancer was not

clinically relevant at the time of death. Median overall survival was 2.3 years (range 0.2‐

10.7) and did not differ between breast cancer and other causes of death (p=0.9).

Conclusion: Omission of surgery for elderly patients with resectable breast cancer has

increased significantly over the past decade; instead patients often received primary

endocrine treatment. Although this may appear an effective alternative to surgery, the

potential for a longer term negative impact on disease control and quality of life deserves

further investigation.

40

Omission of breast cancer surgery in the elderly

Introduction Western societies are ageing; for example, in the Netherlands, the proportion of

inhabitants aged 75 years and over will rise from its current 7% to as much as 15% by

2045.1 As breast cancer incidence increases with age,2 these demographic changes mean

that the number of elderly breast cancer patients will also increase considerably over the

coming decades.3

Cancer specialists are faced with the challenge of determining the appropriate treatment

for these elderly patients. National guidelines are based on trials from which older

patients and those with comorbidity were often excluded.4,5 It is uncertain if the results of

these trials can be extrapolated to the elderly population, with their somatic and

psychogeriatric comorbidity and decreased physiological reserves.6 These conditions can

alter the ability of a patient to tolerate treatment and represent competing causes of

death.7,8 Limited life‐expectancy of frail elderly patients may indicate that they do not live

long enough to benefit from treatment aimed at decreasing the risk of cancer recurrence,8

while they may suffer the side‐effects and complications of this treatment.

As a result, elderly cancer patients do not always receive standard treatment. For

example, surgery may be replaced by primary endocrine therapy and adjuvant treatment

may be omitted.6,9 In a recently published study on treatment patterns for older breast

cancer patients in a single cancer clinic, we concluded that the number of patients aged 80

years and older that received surgery for resectable breast cancer decreased greatly over

the last decade, from 80% to 33%.10 However, the number of patients and duration of

follow‐up was insufficient to ascertain the clinical outcome for these patients.

Therefore, the aim of this study was two‐fold: first, to determine whether this local trend

towards less surgery for elderly patients with resectable breast cancer could also be found

on a national level and second, to evaluate the effect of omission of surgery by examining

clinical outcome and cause of death for a multicentre cohort of elderly patients (aged 75

ears and older) with resectable breast cancer that did not undergo surgery. y

Methods In the Netherlands, all new cancer diagnoses are collected in the National Cancer Registry,

based on data from the central pathology database and the ICD‐9 (international

classification of disease‐9) codes as registered by the hospital. Along with the cancer

diagnosis, patient and tumour characteristics are recorded, as well as the primary cancer

treatment. Using this national database, we examined treatment choices for all patients

newly diagnosed with resectable breast cancer over a period of 10 years (1995‐2005).

Resectable breast cancer was defined as stage T1‐3N0‐2M0. Primary cancer treatment was

dichotomized as surgery vs. no surgery, irrespective of (neo)adjuvant or alternative cancer

41

Chapter 3

treatment. For each incidence year and per five‐year age group, the percentage of

patients receiving surgery was calculated.

To ascertain clinical outcome after omission of surgery for elderly patients with resectable

breast cancer, we examined a regional cohort of such patients in greater detail. All

patients over the age of 75 years, who were diagnosed with a resectable primary breast

cancer and not receiving primary surgery at five different centres in the Netherlands

(Leiden University Medical Centre, Leiden; Haga Hospital, The Hague; Medical Centre

Alkmaar, Alkmaar; Rijnland Hospital, Leiderdorp; and Reinier de Graaf Hospital, Delft)

between 1990 and 2005 were included. For these patients, the following data were

collected from their medical charts: date of birth, date of diagnosis, prior medical history,

tumour stage, and the oestrogen and progesterone receptor status. The human epidermal

growth factor receptor 2 (Her2) was not available in the Netherlands until 2003 and was

therefore not included. Comorbidity burden was assessed using the Charlson comorbidity

index.11 In addition, data on the primary treatment and the reason for omission of surgery

were also collected. Date and cause of death were obtained from the national database at Statistics

Netherlands, based on death certificate data. In the Netherlands, when a patient has died,

it is mandatory that a physician fills out a death certificate including date of death and the

primary cause of death, as well as a maximum of three underlying diseases or secondary

causes of death. Based on primary cause of death as registered on the death certificate,

deceased patients were classified as death due to breast cancer or due to other causes.

Statistical analysis Stata version 11.0 was used for all analyses. For the assessment of national time‐trends,

three‐year moving means were calculated; changes over time were assessed using a

logistic regression analysis. To examine differences in clinical parameters between those

patients that died of breast cancer and those that died of other causes, we used the chi‐

square test. For cancer‐specific survival, breast cancer as primary cause of death was

counted as an event. A univariate and multivariable Cox regression analysis was

performed to assess which of the following baseline characteristics were associated with

breast cancer specific survival: time period (1990‐1999 vs. 2000‐2005), age, T‐stage, lymph

node status, comorbidity using the Charlson score, hormone receptor status and

prescription of endocrine treatment. For each multivariable model, the Cox proportional

hazards assumption was tested. Factors with a p‐value <0.10 in the univariate analysis

were included in the multivariable analysis, as were clinically relevant factors. Kaplan

Meier survival plots, with a log‐rank analysis, were used to determine differences in

survival stratified by cause of death. Relative mortality was calculated as the number of

observed deaths divided by the number of expected deaths based on the corresponding

42


(sex, age, year) general population, based on data from Statistics Netherlands.1 For all

nalyses, a p‐value p <0.05 was considered statistically significant. a

Results Trends over time Between 1995 and 2005, 108,651 patients were newly diagnosed with a resectable breast

cancer in the Netherlands. Of these, 95.6% received primary surgical treatment: the

proportion of patients was 98.7% for those aged 75 years and under, but decreased in

older patients, reaching only 48.9% for those aged 90 years and older (p<0.001). Figure 1

shows the trends over time per age group. For patients aged 75 years or less, the

proportion of patients undergoing surgery remained stable at around 99%. For patients

aged 75 to 79 years, these was a trend towards decrease (98.4% in 1995 to 95.6% in 2005,

p=0.08) while for patients aged 80 years and older this decrease was significant,

particularly in the last five years (p=0.001, Figure 1).

Figure 1: National time‐trends in surgical treatment for resectable breast cancer per age‐group

*For age groups marked with an asterisk, p‐value for change of time was <0.05.

43

Chapter 3

Table 1: Baseline characteristics

Characteristic n %

Period 1990‐1995 1996‐2000 2001‐2005

38 55 94

20 29 50

Age (median + range) 85.9 (75‐97.7)

Charlson 0 1 2 3 4+ Missing

39 43 28 19 16 42

21 23 15 10 9 23

Stage

T1 T2 T3 Unknown

19 44 8 29

19 44 8 29

Lymph node status Negative Positive Unknown

91 24 72

49 13 39

Oestrogen receptor status

Positive Negative Unknown

87 7 93

29 4 50

Progesterone receptor status Positive* Negative Unknown

50 16 121

27 9 65

*All progesterone receptor positive tumours were also oestrogen receptor positive.

Clinical characteristics of older patients not receiving surgery Between 1990 and 2005, 187 patients with resectable breast cancer and not receiving

surgery were identified at the five participating centres; baseline characteristics of these

patients can be found in Table 1. Median age was 85.9 years (range75.0‐97.7). The median

Charlson comorbidity index score was 1 (range 0‐7) and 18 patients had a score of 3 or

more (12%); data on comorbidity were not available for 33 patients. Tumour stage was T1

in 35 patients (16%), T2 in 83 (44%), T3 in 15 (8%) and unknown in 54 (29%). Lymph node

status was negative in 91 patients (49%), positive in 24 patients (13%) and not determined

in 72 patients (39%). All patients were registered as having no metastases. Oestrogen

receptor (ER) status was known in 94 patients (50%) while progesterone receptor (PR)

status was available for 66 patients (35%). The number of patients with unknown receptor

status was approximately 70% until 2000 (1990‐1995 71%, 1996‐2000 73%) but decreased

to 28% between 2001 and 2005. Of patients with a known receptor status, 93% had a

positive ER and/or PR status.

Over half of included patients were diagnosed between 2001 and 2005. Median age

increased in the course of the study period (1990‐1995 83.6 years, 1996‐2000 85.9 years,

2001‐2005 86.4 years, p=0.02) as did the mean Charlson comorbidity index (1990‐1995 0,

1996‐2000 0.5, 2001‐2005 0.9, p=0.02).

44


Figure 2: Kaplan Meier survival curve for (A) observed all‐cause survival compared to expected survival and (B) breast‐cancer related mortality compared to all other causes of death

A.

B.

45

Chapter 3

Treatment For 59 patients (32%), the reason for omitting surgery was stated to be patient's choice.

Other reasons mentioned were age (8 patients, 4%), comorbidity (11 patients, 6%) and

overall health status (10 patients, 5%). For the remaining patients (68%), the reason for

omission of surgery could not be retrieved. The proportion of patients for which omission

of surgery was done at patient’s request did not change over time (p=0.72).

Of the 169 patients for which data on treatment were available, 156 (92%) received

endocrine therapy and 13 (8%) did not receive any treatment. All patients with hormone

receptor positive disease received endocrine treatment, as did 43% of patients with

receptor negative disease and 89% of those with unknown receptor status. The use of

endocrine treatment did not change in the course of the study period (p=0.87). No

patients received radiotherapy or chemotherapy as primary treatment. Seven patients

(4%) required secondary surgical intervention due to disease progression (median time to

surgery 1 year); two of these patients had hormone‐receptor negative disease. Of note,

ne patient died during surgery due to a cardiac arrest. o

Outcome Patients were followed until December 31st 2010. At the end of follow‐up, 178 patients

had died (95.2%); for three patients, survival status was unknown. Median time to death

was 2.3 years (range 0.2‐10.7 years, Figure 2a). Observed survival was significantly lower

than the expected survival based on the matched general population, with a relative

mortality ratio of 2.3 (p<0.001, Figure 2a).

Breast cancer was the primary cause of death for 60 of the 178 patients that died (34%,

Table 2); for 32 of these, breast cancer was the only registered cause of death without

additional contributing diseases. Cardiac disease was the most frequent non‐malignant

cause of death (36 patients, 20%), followed by dementia (17 patients, 10%), pulmonary

disease (13 patients, 7%) and cerebrovascular disease (10 patients, 6%). Ten patients died

of another type of cancer (6%). In addition, breast cancer was stated to be an underlying

cause of death for 37 patients (21%) whose primary cause of death was not breast cancer

(Table 2). For the remaining 81 patients (45%), breast cancer was not mentioned on the

death certificate. Survival did not differ between patients dying of breast cancer vs. those

dying of all other causes (p=0.9, Figure 2b).

Table 3 shows the multivariable analysis for the breast cancer‐specific mortality. Positive

and unknown lymph node status were both significantly associated with mortality (hazard

ratio 3.1 (95% confidence interval (CI) 1.3‐7.1,) and 2.0 (95%CI 1.0‐4.0) respectively,

p=0.02). Increasing comorbidity demonstrated a trend towards a significant association

(p=0.1), as did the time period (p=0.1), with increased hazard of dying of breast cancer in

later years.

46


Table 2. Primary causes of death and underlying diseases in breast cancer patients not receiving surgery

Primary cause of death Relevant underlying disease

Breast cancer

60 (34%) None Cardiovascular disease Dementia Other

32 (53%) 8 (13%) 9 (15%) 11 (18%)

Non‐breast cancer cause of death Cardiac disease (Cerebro)vascular Dementia Pulmonary disease Second malignancy Other

118 (66%) 36 (20%) 10 (6%) 17 (10%) 13 (7%) 10 (6%) 32 (18%)

Breast cancer 37 (31%)

Table 3: Multivariable model for breast cancer‐specific survival

Variable Univariate Hazard ratio p‐value

Multivariable Hazard ratio p‐value

Age Continuous 1.0 (0.9‐1.0) 0.8 1.0 (0.9‐1.0) 0.6

Stage T1 T2 T3 Unknown

Reference 0.7 (0.3‐1.4) 1.3 (0.5‐3.4) 0.8 (0.4‐1.7)

0.5 Reference 0.8 (0.3‐1.7) 0.9 (0.3‐2.4) 1.1 (0.5‐2.8)

0.7

Lymph nodes Negative Positive Unknown

Reference 3.1 (1.5‐6.6) 1.6 (0.9‐2.9)

0.01 Reference 3.1 (1.3‐7.1) 2.0 (1.0‐4.0)

0.02

Charlson 0 1 2 3+

Reference 1.9 (1.0‐3.7) 1.0 (0.4‐2.2) 0.7 (0.3‐1.6)

0.08 Reference 1.7 (0.9‐3.3) 1.1 (0.4‐2.4) 0.6 (0.2‐1.3)

0.1

Period 1990‐1999 2000‐2008

Reference 1.8 (1.0‐3.1)

0.03 Reference 1.6 (0.9‐2.8)

0.1

Oestrogen receptor Positive Negative Unknown

Reference 1.0 (0.3‐3.4) 0.9 (0.5‐1.5)

0.8

Endocrine treatment Yes No Unknown

Reference 1.0 (0.3‐2.7) 1.7 (0.7‐4.3)

0.5

Factors with p<0.1 and clinically relevant factors (age and stage) were entered in the multivariable model.

47

Chapter 3

Discussion This study addresses the incidence and effects of omission of surgery in patients aged 75

years and older with resectable breast cancer. We found that the proportion of these

patients for whom surgery is omitted has increased significantly in the past decade,

particularly in those aged 80 years and older. Data from the regional cohort of 187

patients show that omission of surgery was frequently at patient’s request (32%). When

surgery is omitted, primary endocrine therapy (PET) is generally the alternative of choice

(92% of patients). Median survival was 2.3 years, significantly worse than expected when

compared to an age‐ and gender‐matched cohort of the general population (relative

mortality ratio 2.3). One‐third of patients died of breast cancer, while breast cancer was

not clinically relevant at the time of death in 45%.

This study confirms the findings of our prior single‐centre study that surgery is increasingly

omitted for elderly breast cancer patients.10 Several factors could explain this finding. On

the one hand, therapeutic choices of breast cancer specialists may have changed, based

on clinical trials on primary endocrine treatment for resectable breast cancer in patients

unfit or unwilling to undergo surgery.12 Another explanation could be an alteration in

referral patterns. Not all patients with suspected breast cancer are referred to a breast

cancer clinic, for instance due to significant comorbidity or because the patient and/or

family prefer an expectative approach.13 However, in the past decade, public awareness

on the relevance of breast cancer has increased, as has the general practitioners’

awareness of alternatives to surgical treatment. This could mean that the decrease in the

number of patients undergoing surgery is due to an increased referral of frail patients who

are unsuitable or unwilling to receive surgery for diagnosis and possible hormone therapy.

A third option is a change in patient preferences over time; however, although omission of

surgery was at patient’s request for one‐third of patients in our cohort, this proportion did

not increase during the study period.

The question rises whether omission of surgery for an elderly patient with breast cancer is

justified. Prior studies have demonstrated that concurrent diseases will have a greater

impact on the overall survival and quality of life of older breast cancer patients than the

cancer itself.14‐18 Therefore, withholding treatment in the presence of serious comorbidity

that significantly limits the remaining life‐expectancy seems legitimate, provided that

adequate follow‐up is in place to allow timely intervention if complications do occur. It

must be kept in mind, however, that estimation of life‐expectancy is notoriously difficult,19

particularly in the very elderly who can experience a persistent level of disability and

frailty over an extended period of time, before succumbing to a minor illness due to a lack

of physical reserves.20

In our cohort, it appears that the availability of primary endocrine treatment (PET) may

have facilitated the decision to omit surgery, as 92% of patients received this alternative.

48


However, although studies in unselected elderly patients with resectable breast cancer

have demonstrated that the survival benefit of surgery compared to PET does not become

clinically relevant until three years after diagnosis,21 PET has been proven to result in a

significantly decreased locoregional control compared to primary surgery.12 This could

seriously affect a patient’s quality of life and for many older patients, quality of life and

functional independence are considered to be of greater importance than survival.22,23 In

addition, although increasing age, poor general health and comorbidity – which are

frequently stated as reasons for choosing PET – do affect the ability of a patient to tolerate

surgery,14,15,24 these factors will also increase the risk of side‐effects of endocrine

treatment.25‐27 These include constitutional symptoms such as fatigue, dizziness, nausea,

headaches, muscle weakness and pains, as well as an increased risk of thrombo‐embolic

events.25‐27 Furthermore, exposure to these adverse effects will be much longer in case of

PET. These factors suggest that although endocrine treatment may appear an easy, safe

and effective alternative to surgery, the potential for a negative impact on the quality of

the patient’s remaining lifetime should not be underestimated.

Patient’s refusal of treatment was another important reason for omission of surgery in our

cohort. Older patients tend to rely much more on their physician’s opinion than younger

patients and generally prefer a more passive role when it comes to treatment

decisions.28,29 As a result, part of what is stated to be the patient’s preference could in fact

be a reflection of the physician’s preference. In addition, the decision that a patient makes

will be greatly influenced by the information that he or she has been offered by the cancer

specialist,30 particularly in case of older patients who are less likely to access alternative

sources of information,29 such as the internet.

The relatively poor survival of the patients in our regional cohort is most likely due to the

fact that this represents a selected cohort of unfit patients, with a poor a priori prognosis

irrespective of the breast cancer. However, the patient charts generally did not provide an

extensive description of the patient’s health status and rarely included data on functional

limitations or the presence of geriatric syndromes that can significantly diminish overall

prognosis.31 Therefore, we were unable to quantify the fitness of the patients in our

cohort. Another factor contributing to the poor survival rate could be a possible

underestimated or incomplete staging of disease. This is suggested by the fact that 17 of

the 60 patients (28%) that died of breast cancer died within the first year (Figure 2b),

which is very unlikely in genuine early stage disease. Finally, it is possible that breast‐

cancer related therapeutic decision affected survival, but in an observational cohort study

with limited clinical data, it is not possible to provide an accurate estimate of this effect.

This study has several weaknesses. Due to the retrospective nature of this study, we could

only draw on what was recorded in the patient’s charts, resulting in missing data. In

addition, we were not always able to retrieve the medical records, particularly for patients

49

Chapter 3

diagnosed in the earliest years of this study. Furthermore, patients were frequently

referred back to their primary care physician for follow‐up and thus, we had to rely on the

death certificate to ascertain the clinical relevance of the breast cancer at the time of

death. In addition, establishing the cause of death in elderly patients with multiple

interacting comorbidities can be susceptible to errors.32,33 Despite these issues, we believe

that this real‐life patient cohort provides an interesting and relevant supplement to

linical trial data. c

In conclusion, omission of surgery for elderly patients with resectable breast cancer and

the use of primary endocrine therapy have increased significantly over the past decade.

Breast cancer was not considered clinically relevant at the time of death for a significant

number of patients, suggesting that PET was successful in suppressing the cancer for the

patient’s remaining lifetime. However, although PET may appear an easy, safe and

effective alternative to surgery, the longer term negative impact on disease control and

quality of life should not be underestimated. For the very frail, it may be preferable and

legitimate to withhold treatment altogether. As clinical trials addressing this issue in such

a vulnerable patient population are unfeasible, prospective observational cohort studies

incorporating a broad scope of patient characteristics including frailty data and geriatric

syndromes, may aid in identifying which baseline factors could guide decision making

garding the initiation or withholding of the available treatment options. re

Acknowledgements: the authors would like to thank the Netherlands Cancer Registry, as

well as the Zoleon foundation and Statistics Netherlands.

50


References 1. www.cbs.nl. 3‐2‐2012. 2. www.cijfersoverkanker.nl. 2‐3‐2012. 3. Alberg AJ, Singh S. Epidemiology of breast cancer in older women: implications for future healthcare.

Drugs Aging 2001;18:761‐772. 4. Hutchins LF, Unger JM, Crowley JJ et al. Underrepresentation of patients 65 years of age or older in

cancer‐treatment trials. N Engl J Med 1999;341:2061‐2067. 5. Peto R, Davies C, Godwin J et al. Comparisons between different polychemotherapy regimens for early

breast cancer: meta‐analyses of long‐term outcome among 100,000 women in 123 randomised trials. Lancet 2012;379:432‐444.

6. Hamaker ME, Schreurs WH, Uppelschoten JM, Smorenburg CH. Breast cancer in the elderly: retrospective study on diagnosis and treatment according to national guidelines. Breast J 2009;15:26‐33.

7. Bastiaannet E, Liefers GJ, de Craen AJ et al. Breast cancer in elderly compared to younger patients in the Netherlands: stage at diagnosis, treatment and survival in 127,805 unselected patients. Breast Cancer Res Treat 2010;124:801‐807.

8. Ring A, Sestak I, Baum M et al. Influence of comorbidities and age on risk of death without recurrence: a retrospective analysis of the Arimidex, Tamoxifen Alone or in Combination trial. J Clin Oncol 2011;29:4266‐4272.

9. Lavelle K, Moran A, Howell A et al. Older women with operable breast cancer are less likely to have surgery. Br J Surg 2007;94:1209‐1215.

10. Hamaker ME, Schreurs WH, van Slooten HJ et al. [Trends in breast cancer treatment in the elderly at a breast cancer outpatient clinic: guidelines followed better]. Ned Tijdschr Geneeskd 2009;153:A562.

11. Charlson ME, Pompei P, Ales KL, Mackenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373‐383.

12. Hind D, Wyld L, Beverley CB, Reed MW. Surgery versus primary endocrine therapy for operable primary breast cancer in elderly women (70 years plus). Cochrane Database Syst Rev 2006;(1):CD004272.

13. Hamaker ME, Hamelinck VC, van Munster BC et al. Nonreferral of Nursing Home Patients With Suspected Breast Cancer. J Am Med Dir Assoc 2012;13:464‐9.

14. Wildiers H, Kunkler I, Biganzoli L et al. Management of breast cancer in elderly individuals: recommendations of the International Society of Geriatric Oncology. Lancet Oncol 2007; 8:1101‐1115.

15. Wyld L, Reed M. The role of surgery in the management of older women with breast cancer. Eur J Cancer 2007; 43:2253‐2263.

16. Janssen‐Heijnen ML, Maas HA, Houterman S et al. Comorbidity in older surgical cancer patients: influence on patient care and outcome. Eur J Cancer 2007;43:2179‐2193.

17. Satariano WA, Ragland DR. The effect of comorbidity on 3‐year survival of women with primary breast cancer. Ann Intern Med 1994;120:104‐110.

18. Patnaik JL, Byers T, Diguiseppi C et al. The influence of comorbidities on overall survival among older women diagnosed with breast cancer. J Natl Cancer Inst 2011;103:1101‐1111.

19. Christakis NA, Lamont EB. Extent and determinants of error in physicians' prognoses in terminally ill patients: prospective cohort study. West J Med 2000;172:310‐313.

20. Gill TM, Gahbauer EA, Han L, Allore HG. Trajectories of disability in the last year of life. N Engl J Med 2010; 362:1173‐1180.

21. Fennessy M, Bates T, MacRae K et al. Late follow‐up of a randomized trial of surgery plus tamoxifen versus tamoxifen alone in women aged over 70 years with operable breast cancer. Br J Surg 2004;91:699‐704.

22. Yellen SB, Cella DF, Leslie WT. Age and clinical decision making in oncology patients. J Natl Cancer Inst 1994; 86:1766‐1770.

23. Steinhauser KE, Christakis NA, Clipp EC et al. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 2000;284:2476‐2482.

24. Tiret L, Desmonts JM, Hatton F, Vourc'h G. Complications associated with anaesthesia‐‐a prospective survey in France. Can Anaesth Soc J 1986;33:336‐344.

25. Hernandez RK, Sorensen HT, Pedersen L et al. Tamoxifen treatment and risk of deep venous thrombosis and pulmonary embolism: a Danish population‐based cohort study. Cancer 2009;115:4442‐4449.

51

Chapter 3

52

26. Biganzoli L, Wildiers H, Oakman C et al. Management of elderly patients with breast cancer: updated recommendations of the International Society of Geriatric Oncology (SIOG) and European Society of Breast Cancer Specialists (EUSOMA). Lancet Oncol 2012;13:e148‐e160.

27. Aapro M, Monfardini S, Jirillo A, Basso U. Management of primary and advanced breast cancer in older unfit patients (medical treatment). Cancer Treat Rev 2009;35:503‐508.

28. Fallowfield L. Participation of patients in decisions about treatment for cancer. BMJ 2001;323:1144. 29. Pinquart M, Duberstein PR. Information needs and decision‐making processes in older cancer patients. Crit

Rev Oncol Hematol 2004;51:69‐80. 30. Pierce PF. Deciding on breast cancer treatment: a description of decision behavior. Nurs Res 1993;42:22‐

28. 31. Ferrucci L, Guralnik JM, Studenski S et al. Designing randomized, controlled trials aimed at preventing or

delaying functional decline and disability in frail, older persons: a consensus report. J Am Geriatr Soc 2004;52:625‐634.

32. Ederer F, Geisser MS, Mongin SJ et al. Colorectal cancer deaths as determined by expert committee and from death certificate: a comparison. The Minnesota Study. J Clin Epidemiol 1999;52:447‐452.

33. Lu TH, Shih TP, Lee MC et al. Diversity in death certification: a case vignette approach. J Clin Epidemiol 2001;54:1086‐1093.

Chapter 4

Non‐referral of nursing home patients with suspected breast cancer

M.E. Hamaker, V.C. Hamelinck, B.C. van Munster, E. Bastiaannet, C.H. Smorenburg,

W.P. Achterberg, G.J. Liefers, S.E. de Rooij

Journal of the American Medical Directors Association 2012;13:464‐9

Chapter 4

Abstract Introduction: People with suspected breast cancer who are not referred for diagnostic

testing remain unregistered and are not included in cancer statistics. Little is known about

the extent of and motivation for non‐referral of these patients.

Methods: A web‐based survey was sent to all elderly care physicians (ECPs) registered at

the National Association of Elderly Care Physicians and Social Geriatricians in the

Netherlands, inquiring about the number of patients with suspected breast cancer they

encountered and subsequent choices regarding referral.

Results: Surveys were completed by 419 out of 1239 ECPs (34%); 249 of these had

encountered one or more patients with suspected breast cancer in the past year (60%).

Seventy‐four ECPS reported not referring the last patient (33%). Reasons for non‐referral

were end‐stage dementia (57%), patient/family preference (29%) and limited life‐

expectancy (23%). Referral was frequently thought to be too burdensome (13%). For 16%

of non‐referred patients, hormonal treatment was started by the ECP without diagnostic

confirmation of cancer.

Conclusion: In this survey, over 33% of nursing home patients with suspected breast

cancer were not referred for further testing, in particular those with advanced dementia,

limited life‐expectancy and poor functional status. As the combination of dementia and

suspected breast cancer is expected to double in the coming decades, now is the time to

optimize cancer care for these vulnerable patients.

54


Introduction Cancer statistics show that in 2009, a total of 13,177 women were diagnosed with breast

cancer in the Netherlands.1 These data are based on the Netherlands Cancer Registry

(NCR),1 a nationwide network which collects histo‐ and cytopathology data from all Dutch

hospitals, supplemented by data from the national hospital discharge databank. After

cancer cases are identified, trained personnel from regional cancer registries gather

additional data on diagnosis, staging and treatment.

As all oncologic treatment in the Netherlands is provided by hospital‐based specialists, the

registry can provide a comprehensive overview of current cancer treatment. It also allows

for a comparison of actual treatment with treatment as recommended by guidelines. For

example, using registry data, studies have demonstrated that older breast cancer patients

are often treated less extensively than their younger counterparts and that they are at risk

for being undertreated.2‐5

In the Netherlands, primary care physicians form an important first link in the cancer

treatment pathway (Figure 1), as they are generally responsible for referral to hospital

specialists – although some alternative routes are possible. For patients residing in nursing

homes, either permanently or temporarily in case of rehabilitation, this tasks falls on

specially trained physicians, called elderly care physicians (ECPs), for whom nursing homes

are the primary place of work.6 This differentiation between primary medical care and

hospital‐based care in the Netherlands, results in an important limitation of the cancer

registry: patients with a clinical suspicion of cancer that are not referred to hospital for

further diagnostic testing will remain unregistered and will not be included in Dutch

cancer statistics.

Surprisingly, little is known about the issue of non‐referral. Studies using Medicare‐data

Figure 1: Global overview of breast cancer care pathway in the Netherlands

55

Chapter 4

in the United States show that little cancer care is claimed for patients living in a nursing

home setting,7 and that Alzheimer patients receive less treatment for breast cancer than

comparable female Medicare beneficiaries,8 but the authors could not determine whether

this was due to less cancer vigilance resulting in missed cancer diagnoses or to omission of

referral for specialised cancer care. Even less is known about the motivation behind non‐

referral or the consequences for the patient.

For this study, we sent a survey to all members of the National Association of Elderly Care

Physicians and Socials Geriatricians, to determine a) the extent of and reasons for non‐

referral of patients suspected of breast cancer by ECP, and b) the motivations behind this

hoice. c

Method We developed a web‐based survey using the SurveyMethods, Inc. software.9 The survey

contained questions relating to the incidence of suspected breast cancer in nursing

homes, whether or not these patients were referred and the motivation behind referral

choices. The content of the survey is depicted in Figure 2. After a concept of this survey

was successfully tested in 19 ECPs, it was subsequently sent to all ECPs registered at

Verenso, the National Association of Elderly Care Physicians and Social Geriatricians in

August 2011. Of the 1525 ECPs active in the Netherlands, 1238 are registered at Verenso;

onsequently 81% of all Dutch ECPs were invited to participate in the survey. c

Statistical analysis To compare differences between referred and non‐referred patients, the SPSS (Statistical

Package for the Social Sciences) version 19.0 was used. The chi‐square test was used for

nominal and ordinal variables. For continuous variables with a normal distribution, the

Student t‐test was used, and for continuous variables with a non‐Gaussian distribution the

ann‐Whitney test. M

Results

Response rate Surveys were completed by 419 of the 1239 ECPs (response rate was 34%, Figure 2).

Characteristics of respondents are listed in Table 1. The median age of respondents was

47 years (range 25‐66 years) and 66% were female. Responses came from all over the

country, covering over 90% of the 90 primary zip code‐areas in the Netherlands. Almost

60% of respondents stated they had encountered at least one patient with suspected

breast cancer in the past year; of these patients, 33% were not referred for further

diagnostic testing (Figure 3).

56


Figure 2: Content of survey

Question 1: How old are you?

Question 2: What sex are you? Question 3: What are the four numbers of the zip code where your practice is located? Question 4: How many patients did you care for in the past year?

Question 5: What percentage of these patients is female?

Do you have any additional comments relating to this survey?

IF YES IF NO

IF NONE IF ONE IF MORE THAN ONE

Question 6: In the past year, did you suspect one or more of your patients of having breast cancer?

Question 7: how many of these patients did you refer for further testing and/or treatment?

The following questions concern the last patient you suspected of having breast cancer

Question 8: What was this patient’s age? Question 9: Did you refer this patient for further testing and/or treatment?

Question 10: With whom did you discuss this decision?

Question 10: With whom did you discuss this decision?

Question 11: What were your reasons for referring this patient?

Question 11: What were your reasons for not referring this patient?

Question 12: What treatment did the patient receive?

Question 12: Did you yourself start any anti‐cancer treatment?

Question 13: How is the patient doing now?

Question 13: How is the patient doing now?

Question 14: Have you recently had any patients who you did

Question 14: Have you recently had any patient who you not choose to refer,

and if so, what were your reasons for this?did choose to refer, and if

so, what were your reasons for this?

END OF SURVEY

57

Chapter 4

Referral vs. non‐referral Table 2 lists a comparison of patients that were or were not referred. Patients not referred

were older (median age 86 vs. 82 years, p<0.001), although some non‐referred patients

were as young as 60 years. Over 99% of physicians discussed their decision on referral

with at least one other party: in 54% of cases, it was discussed with the patient, while in

87% a family member was consulted; in 9% it was only discussed with another physician.

Of note, of the patients that were not referred, less than half were personally involved in

making this decision.

The motivations for choosing to refer patients to hospital (Table 3) were primarily the

desire to confirm the diagnosis (28%), the fear of future ulceration or metastases (21%),

good general health and life‐expectancy (19%), and patient’s or family’s preference for

referral (18%). Current or imminent ulceration was stated in 9% of cases, while

maintaining quality of life or optimizing palliative care were stated in 7% and 4%

respectively. For 11%, the main reason for referral was to assess the suitability of primary

hormonal treatment, as the ECP felt that due to cognitive or functional status, the patient

was not a candidate for more invasive treatment.

The primary reason stated for not referring was end‐stage dementia (57%, Table 4). Other

reasons were the preferences of patient and/or family (29%), limited life‐expectancy

(23%), poor functional status or somatic comorbidity (18% and 16%, respectively) and

advanced age in 8%. The expected burden of the hospital visits and subsequent diagnostic

process and treatment for the patient was stated in 13%, particularly for patients with

dvanced dementia. a

Treatment and outcome Of the patients that were referred to hospital, seven were found to have a benign tumour

(5%); 16% received no treatment and 24% received hormonal treatment only. Surgery was

performed in 28% of patients, radiotherapy given to 8% and chemotherapy to one patient.

For 18%, the diagnostic process was still on‐going. In addition, twelve non‐referred

patients (16%) were prescribed primary hormonal treatment by the ECP without

confirmation of breast cancer.

The current health status of referred and non‐referred patients is listed in Table 5. Thirty‐

four patients were lost to follow‐up. Three referred patients died of breast cancer or

breast cancer treatment, and three patients suffered from locally advanced or metastatic

disease (two referred and one non‐referred patient). Forty‐four patients had died of

causes other than breast cancer (17%).

58


Figure 3: Flow chart of response rate and referrals

Number of survey returned: n=419/1238 (34%)

One or more patient with suspected breast cancer No patients with suspected breast cancer n=249 (59.4%) n=170 (40.6%)

Table 1: Characteristics of respondents

Elderly care physicians

Response rate 419/1238 (34%) Median age of respondents (range) 47 years (25‐66) % female of respondents 66.1% % with ≥1 patients suspected of breast cancer 59.4% Number of patients suspected of breast cancer in past year 0 1 2 3 4 5 more than 5

170 140 81 20 4 3 1

% of patients referred to hospital 67.1%

Table 2: Comparison of patients that were and were not referred

Patients referred (n=151) Patients not referred (n=74) p=

Median age of patients (range) 82 (45‐99) years 86 (60‐102) years <0.001

(non‐) Referral discussed with* No one

Patient Family member Colleague Clinical geriatrician Oncologist Surgeon Radiotherapist Others Nursing staff

0%*61%85%14%4%

13%29%3%7%6%

1%* 41% 91% 23% 1% 5% 5% 0%

12% 7%

ns <0.001 <0.001 <0.001 0.02

<0.001 <0.001 0.01

0.002

* n=146 of these responses originated from question 9 and n=17 from question 14

Last patient referredn=151 (67.1%)

Last patient not referred Missingn=74 (32.9%) n=24

59

Chapter 4

Table 3: Reasons for referral

Reason Frequency

(out of 121 responses)* %

Dementia/cognitive function 69 57 Preference of patient and/or family 35 29 Limited life expectancy 28 23 Functional status 22 18 Somatic comorbidity 19 16 Burden of referral too high for specific patient 16 13 Tumour characteristics 10 9 Advanced age 10 8 Lack of subjective burden of tumour 6 5 No expected benefit of referral for patient’s quality of life 3 2 * n=146 of these responses originated from question 9 and n=17 from question 14

Table 4: Reasons for non‐referral

Reason

Frequency (out of 121 responses)*

%

Dementia/cognitive function 69 57 Preference of patient and/or family 35 29 Limited life expectancy 28 23 Functional status 22 18 Somatic comorbidity 19 16 Burden of referral too high for specific patient 16 13 Tumour characteristics 10 9 Advanced age 10 8 Lack of subjective burden of tumour 6 5 No expected benefit of referral for patient’s quality of life 3 2 * n=80 of these responses originated from question 9 and n=41 from question 14

Table 5: Current status of patients

Referred patients n=151

Non‐referred patients n=74

Lost to follow‐up 32 21% 2 3% Stable/asymptomatic disease or disease‐free

97 64% 46 62%

Locally advanced/metastatic disease 2 1% 1 1% Died of other causes 19 13% 25 34% Died of breast cancer or breast cancer treatment

3 2% 0 0%

60


Discussion We found that 60% of the responding ECPs had encountered one or more patients whom

they suspected of having breast cancer in the past year, and 33% of these patients were

not referred. The primary reasons for non‐referral were dementia, poor functional status,

comorbid diseases and limited life‐expectancy, as well as the expected burden of a visit to

a clinic or the subsequent treatment. Of referred patients, only 28% received surgical

treatment, while 40% received no oncologic treatment or primary hormonal therapy only.

To our knowledge, this is the first study to address the issue of non‐referral of nursing

home residents with suspected breast cancer. We believe it provides valuable information

on a vulnerable population that has thus far remained outside the scope of cancer

research and national cancer statistics.

This study has some weaknesses. First of all, the response rate was 34%. This is an issue

frequently encountered in survey‐based studies.10 For this survey, it is not unlikely that

those ECPs who had recently dealt with the issue of suspected breast cancer were more

prone to respond to the survey than those who had not. This makes it difficult to know to

what extent the incidence of suspected breast cancer in nursing home patients can be

extrapolated from these results. Furthermore, as this survey requires ECPs to recollect

their last patient, data may be somewhat influenced by recall bias. Another limitation is

that this study was done in a single country only; as the organisation of care and of cancer

registries will differ from country to country, it remains unclear whether our findings can

be extrapolated to other countries.

This study highlights an important limitation of the current cancer registration in the

Netherlands and consequently of cancer statistics, particularly for the very elderly where

non‐referral is likely to be more prevalent. Although there is a mandatory registration of

confirmed cancer cases, there is no obligation to report suspected but unconfirmed cases;

what is more, a procedure for reporting such cases is currently lacking. As the prevalence

of dementia is expected to double in the coming decades,11 and the proportion of newly

diagnosed patients with breast cancer aged 85 years and over will rise from 9% to 17% by

2030,12 the combination of patients with advanced dementia and suspected breast cancer

will also increase greatly. If no procedure is developed for their registration, the number

of very elderly or frail cancer patients that remain unregistered is likely to increase,

making the cancer statistics for these patients increasingly unreliable. Addressing this

issue in the registry will be challenging, however, as suspected cancer is not confirmed

cancer, and these additional patients cannot automatically be added to what is currently

recorded.

The increasing number of patients suffering from both dementia and suspected breast

cancer asks for a careful evaluation of the current care process. Although the diagnostic

process for breast cancer is not very invasive, and breast cancer surgery has a low risk of

61

Chapter 4

perioperative complications,13 for a patient with advanced dementia, even the process of

going to an out‐patient clinic or undergoing physical examination can be of great burden.

However, this needs to be weighed against the risks of leaving a suspected malignancy

unaddressed. Uncontrolled breast cancer, particularly when ulceration occurs, may have

serious impact on a patient’s comfort and quality of life.

Of course, as this study demonstrates, many patients who were thought to be too frail to

refer for further testing have a life‐expectancy that is limited, leaving little time to suffer

the potential consequences of untreated breast cancer or the potential benefits of

treatment. Watchful waiting with regular physical examination to determine rate of local

progression and symptomatic treatment of cancer‐related complaints such as pain, can be

a useful strategy in such patients. However, estimating life expectancy is not always

easy,14 particularly in those with advanced dementia who can experience a persistent level

of severe disability and frailty over an extended period of time, before succumbing to a

minor illness due to lack of physical reserves.15 Therefore, if the extend of remaining life‐

years is not clear, and there is a desire to start oncologic treatment, but burden of a visit

to clinic is considered too great, what options are left?

One possibility is to start treatment with endocrine therapy without actual confirmation of

breast cancer diagnosis or assessing hormone receptor status. In our study, this option

was chosen for 16% of patients that were not referred. As over 75% of patients aged 80

years or older have oestrogen receptor positive disease,16 and partial remission and loco‐

regional control can often be attained17 – albeit temporarily – this is not an unreasonable

option. However, there will be a proportion of patients who either have hormone

receptor negative disease, or who have no breast cancer at all, and therefore will not

profit from treatment but will be exposed to side‐ effects of treatment. These side‐effects

are limited, but even in fit subjects have been shown to affect their feeling of well‐being,

particularly in the first months of treatment.18,19 For example, all types of hormonal

treatment can cause hot flushes, dizziness, gastro‐intestinal complaints such as nausea

and anorexia, as well as psychological effects such as depression or agitation.20

Furthermore, the very frail are more likely to experience adverse effects,21 and what is

seen as a minor side‐effect for a fit subject can have great impact on the quality of life,

functional status and behaviour of the very frail.

Another option is to alter the diagnostic testing process in a way that minimizes the

burden for these vulnerable patients. For example, one ECP explained that the pathologist

came to their nursing home, to take biopsies of palpable tumours, offering the possibility

of confirming the diagnosis and assessing receptor status. Although for some patients

even this may be too burdensome, for many, a consultation in their own care setting – by

a pathologist, surgeon or oncologist – may be a solution.

62


The results of this study can form a starting point for a range of future clinical studies. First

of all, as this is the first study on non‐referral of nursing home patients, from a single

country, similar studies should be done in other countries to confirm our findings. In

addition, a more in‐depth case review of non‐referred patients may provide additional

information to supplement the survey data. Second, studies could look at non‐referral of

other patient groups, such as frail elderly patients living at home, or nursing home

residents suspected of having other types of cancer. Third, studies on guideline adherence

– particularly in older patients – should take the possibility of non‐referral of patients into

account and address in what way this could influence the outcome of their results.

However, more importantly, studies should focus on the potential of non‐oncologic non‐

pharmacologic interventions to optimize quality of life and minimize cancer‐related

symptom burden, and on developing new treatment pathways, such as a specialist

consultation in the patient’s place of residence, suitable for these vulnerable patients.

Possibly, the option of initiating endocrine treatment without histological confirmation of

breast cancer – as is sometimes chosen already – could be evaluated in a placebo‐

controlled study weighing the benefit in disease control against the potential harmfulness

f side‐effects. o

In conclusion, our survey shows that suspicion of breast cancer is not uncommon in a

nursing home setting. Over 33% of patients were not referred for further testing, in

particular those with advanced dementia and poor functional status, because the burden

of referral is expected to be greater than the benefit for the patient. With the expected

increase in the occurrence of both dementia and breast cancer, now is the time to start

thinking about how best to provide them with the care they need in a way that is suitable

their overall condition. to

Acknowledgements: the authors would like to thank all participating elderly care

physicians, and in particular the chair of the Dutch organisation of the Elderly Care

Physicians, Mieke Draaijer and the director, Franz Roos, for their willingness to collaborate

with us.

63

Chapter 4

64

References 1. www.cijfersoverkanker.nl 2. Hamaker ME, Schreurs WH, Slooten HJ van et al. Trends in breast cancer treatment in the elderly at a

breast cancer outpatient clinic: guidelines followed better. Ned Tijdsch Geneeskd 2009;153:A562. 3. Hamaker ME, Schreurs WH, Uppelschoten JM, Smorenburg CH. Breast cancer in the elderly: retrospective

study on diagnosis and treatment according to national guidelines. Breast J 2009;15:26‐33. 4. Bastiaannet E, Liefers GJ, Craen AJM de et al. Breast cancer in elderly compared to younger patients in the

Netherlands: stage at diagnosis, treatment and survival in 127,805 unselected patients. Breast Cancer Res Treat 2010;124:801–807.

5. Bastiaannet E, Portielje JEA, Velde JH van de et al. Lack of survival gain for elderly women with breast cancer. Oncologist 2011;16:415‐23.

6. Koopmans RT, Lavrijsen JC, Hoek JF et al. Dutch elderly care physician: a new generation of nursing home physician specialists. J Am Geriatr Soc 2010;58:1807‐9.

7. Bradley CJ, Clement JP, Lin C. Absence of cancer diagnosis and treatment in elderly Medicaid‐insured nursing home residents. J Natl Cancer Inst 2008;100:21‐31.

8. Sheinfeld Gorin S, Heck JE, Albert S, Hershman D. Treatment for breast cancer in patients with Alzheimer’s disease. J Am Geriatr Soc 2005;53:1897‐1904.

9. www.surveymethods.com 10. Dillman DA. Internet, mail, and mixed‐mode surveys: the tailored design method. Mail and internet

surveys. Hoboken, N.J.: Wiley & Sons; 2009. 1941. 11. www.alzheimer‐nederland.nl 12. Hayat MJ, Howlader N, Reichman ME, Edwards BK. Cancer statistics, trends and multiple primary cancer

analysis from the surveillance, epidemiology and end results (SEER) program. Oncologist 2007;12:20‐37. 13. Audisio RA, Bozzetti F, Gennari F et al. The surgical management of elderly cancer patients:

recommendations of the SIOG surgical task force. Eur J Cancer 2004;40:926‐38. 14. Christakis NA, Lamont EB. Extent and determinants of error in doctors’ prognoses in terminally ill patients:

prospective cohort study. BMJ 2000;320:469‐73. 15. Gill TM, Gahbauer EA, Han L, Allore HG. Trajectories of disability in the last year of life. N Engl J Med 2010;

362:1173‐80. 16. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical outcome of elderly women with breast

cancer. J Natl Cancer Inst 2000;92:550‐617. 17. Hind D, Wyld L, Beverley C, Reed MW. Surgery versus primary endocrine therapy for operable primary

breast cancer in elderly women (70 years plus). The Cochrane Library: DOI: 10.1002/14651858.CD004272.pub2.

18. Cella D, Fallowfield L, Baker P et al on behalf of the ATAC trialists’ group. Quality of life of postmenopausal women in the ATAC trial after completion of 5 years’ adjuvant treatment for early breast cancer. Breast Cancer Res Treat 2006;100:273‐84.

19. Fallowfield L, Cella D, Cuzick J et al. Quality of life of postmenopausal women in the arimidex, tamoxifen, alone or in combination adjuvant breast cancer trial. J Clin Oncol 2004;22:4261‐71.

20. www.fk.cvz.nl 21. Routledge PA, O’Mahoney MS, Woodhouse KW. Adverse drug reactions in elderly patient. Br J Clin

Pharmacol 2004;52:121‐6.

Chapter 5

Slow accrual of elderly patients with metastatic breast cancer in the Dutch

multicentre OMEGA study

M.E. Hamaker, C. Seynaeve, J.W.R. Nortier, M. Wymenga, E. Maartense, E. Boven, A.E. van Leeuwen‐Stok, S.E. de Rooij, B.C. van Munster, C.H. Smorenburg

[Submitted]

Chapter 5

Abstract Background: In a Dutch multicentre study, elderly (65+ year) metastatic breast cancer

patients, eligible for first‐line chemotherapy, were randomised between two types of

single‐agent chemotherapy. As accrual was slow, with 78 randomised patients between

April 2007 and September 2011, we explored potential barriers in the accrual process and

their consequences for characteristics of included patients.

Methods: We sent surveys on the reasons for non‐inclusion to all coordinating

investigators. We also examined inclusion in a concurrent, non‐elderly breast cancer study

of the trialists’ group and analysed baseline geriatric characteristics of included patients.

Results: Investigators from fifteen participating centres returned the survey. Most

commonly reported barriers to inclusion were: patient’s refusal of chemotherapy (n=8) or

of randomisation (n=9), impaired cognition (n=3) and insufficient cardiac function (n=2).

Oncologists’ preference for combination regimens over single‐agent chemotherapy was

reported twice. Twenty‐eight potentially eligible patients, aged 65 to 71 years, were

included in a concurrent, study investigating combination chemotherapy in fit non‐elderly

patients with metastatic breast cancer. However, baseline characteristics of the included

patients showed that the OMEGA study succeeded in including frail and older patients,

with a performance status of 2 in 22% of patients and 54% of patients aged 75 years or

older.

Conclusion: Accrual in this study was mainly hampered by patient’s refusal or preference

for a particular type of treatment, and an overall condition considered as too fit or too

frail for inclusion. Future trials in elderly metastatic breast cancer patients should focus on

non‐restrictive inclusion criteria as well as on education of physicians and elderly patients

on the advantages of trial participation.

66

Slow accrual in the OMEGA study

Introduction In the Netherlands, breast cancer is diagnosed in over 13,000 women each year, of which

53% is aged 65 or older. It is the most frequently diagnosed cancer in women and its

incidence increases with age. Due to increasing life expectancy and ageing of the

population, the number of elderly patients with breast cancer will increase substantially

over the next decades. It is expected that by the year 2035, 60% of all new breast cancer

patients will be 70 years or older.1

Breast cancer treatment in older patients is complicated by the lack of elderly‐specific

evidence on which to base treatment decisions, especially for chemotherapy. Older

patients and those with comorbidity have historically been excluded from clinical trials; for

example, in a study published in 1999, only 9% of patients in breast cancer trials was aged

65 years and over, while at that time 49% of breast cancer patients in the United States

fell into this age group.1 Although these findings have led researchers to develop trials

without upper age limits, more recent publications still report an underrepresentation of

elderly patients and underline the need for trials focusing on this patient population.2

The Dutch Breast Cancer Trialists Group (BOOG) has performed a multicentre, randomised

clinical trial regarding the tolerance and efficacy of first‐line, single‐agent palliative

chemotherapy for metastatic breast cancer patients aged 65 years or older, with

incorporation of a comprehensive geriatric assessment, the OMEGA trial (trial number

NTR897, BOOG 2006‐02). This study was specifically designed to allow inclusion of frail,

elderly patients and those with comorbidity: selection criteria were limited to acceptable

bone marrow, kidney and liver function, an European Cooperative Oncology Group (ECOG)

performance status ≤ 2 (or ≤ 3 if limitations are caused by pain or a pre‐existent disabling

disease), a left ventricular ejection fraction above the lower limit of normal, and the

absence of clinically relevant cardiac disease or a previous malignancy in the past five

years. Patients were randomised between pegylated liposomal doxorubicin (Caelyx®)

administered intravenously once every four weeks for a maximum of six cycles, or

capecitabine, given orally on days 1 to 4, every three weeks for a maximum of eight cycles.

In addition to standard analyses of toxicity and efficacy, this study performed an

evaluation of quality of life and a comprehensive geriatric assessment3‐7 at baseline,

twelve weeks, end‐of‐treatment (24 weeks if all cycles were completed) and after twelve

months.

Despite a total of 25 participating Dutch institutions, accrual for the OMEGA study was

slow with only 78 patients randomised between April 2007 and September 2011 (end of

study); this is just over half of the intended 154 patients based on power calculations and

a pre‐study audit on potentially eligible patients (Figure 1). To determine the implications

of this difficult accrual process for future clinical studies in frail elderly cancer patients, we

explored potentials barriers in the accrual process.

67

Chapter 5

Figure 1: Expected versus actual accrual

Methods In the summer of 2008, a survey was sent to all coordinating investigators of the

participating centres (19 at that time, six centres joined this study after 2008) to assess

potential barriers to accrual (Appendix 1), focusing on study‐related barriers such as in‐

and exclusion criteria of the study protocol, patient‐related factors such as patient’s

refusal to participate or preference for a specific type of treatment, and physician‐related

factors like the preference for a particular study arm and their attitude towards treatment

with chemotherapy in older patients.

To assess if any of these factors ultimately influenced the characteristics of enrolled

patients or only affected the number of patients, we analysed the baseline data of the 78

randomised patients. These data were collected by the research nurses, and included:

social status and living situation, comorbidity as assessed by the Charlson comorbidity

68


index (CCI),3 Eastern Cooperative Oncology Group performance status (PS), functional

status assessed by the Lawton & Brody scale for instrumental activities of daily living

(IADL),4 polypharmacy (defined as 5 or more types of medication), undernutrition (defined

by a BMI <20 kg/m2), cognitive dysfunction assessed by the mini‐mental state examination

(MMSE),5,6 depressive symptoms assessed by the geriatric depression scale (GDS)6 and

frailty assessed by the Groningen Frailty Index (GFI).7

In addition, data were obtained on the accrual of patients aged 65 years and older in the

ATX study, a concurrent BOOG study for metastatic breast cancer patients, in which

patients were randomised to first‐line combination chemotherapy with either

bevacizumab and paclitaxel or bevacizumab, paclitaxel and capecitabine. This study

cruited fit patients aged ≤ 75 years with an ECOG performance status of 0 or 1. re

Statistical analysis O

nly descriptive data were used.

Results

Potential barriers to accrual Surveys regarding barriers to accrual were sent to the coordinating investigators of the 19

centres that were participating in the OMEGA study in 2008; surveys were returned from

15 centres. Regarding potential patient‐related barriers, eight respondents stated that

potentially eligible patients had refused chemotherapy (53%), and nine respondents

mentioned that potential patients had refused randomisation (60%). The most important

physician‐related factors were their preference for a particular type of chemotherapy

(7%), their hesitance in offering chemotherapy to older patients (13%) and a lack of man‐

power (7%). In response to questions on study‐related barriers to accrual, three

investigators stated that cognitive impairment in patients limited their ability to

participate in the trial (20%) and two mentioned that the exclusion of patients with an

impaired left ventricular ejection fraction posed a barrier (13%), while one respondent

mentioned that many of the patients were too frail to receive any form of chemotherapy

(7%) and another found the patient information form as too complicated (7%). The

comprehensive geriatric assessment was considered to be too time‐consuming by the

patient (one respondent, 7%) or the investigators (two respondents, 13%). The use of

single‐agent chemotherapy as opposed to combination treatment was stated by one

respondent (7%), while three respondents reported that concurrent trials in this particular

patient group limited accrual (20%).

69

Chapter 5

Table 1: Baseline characteristics of patients included in the OMEGA study

Number of patients 78

Median age (range) 75.5 years (65.8–86.8)

Marital status Married Widowed Single Missing

43 23 8 4

Current living situation Independent Residential care Missing

71 2 5

Additional persons in household 0 32 1 37 2 5 Missing 4

Offspring Yes 60 No 14 Missing 4

Education level College/university 6 High school/vocational education 51 Primary school 15 None 1 Missing 5

Performance status 0 23 1 37 2 16 3 1 Missing 1

Body Mass index Undernutrition (BMI <20 kg/m2)

Normal (BMI 20‐30 kg/m2)

Overweight (BMI >30 kg/m2)

5 50 23

IADL4 Fully independent (28‐30 points) 8

Partially dependent (14‐27 points) 65 Fully dependent (≤13 points) 0 Missing 5

Polypharmacy 0‐4 types of medication 39 5+ types of medication 37 Missing 2

Charlson comorbidity index3 0 45

1‐2 22 3‐4 2 5+ 1 Missing 8

MMSE5 No cognitive dysfunction (24 points or more) 69

Cognitive dysfunction (≤23 points) 5 Missing 4

GDS6 No depressive symptoms (0‐4 points) 49

Mild depressive symptoms (5‐9 points) 21 Moderate/severe depressive symptoms (10 points or more) 3 Missing 5

GFI7 Not frail (0‐3 points) 32

Frail (4 or more points) 40 Missing 6

70


Inclusion in a concurrent non‐elderly study A total of 28 potentially eligible patients for the OMEGA study, aged 65 to 71 years, were

included in the concurrent ATX‐study with combination chemotherapy in metastatic

breast cancer by centres participating in the OMEGA trial during the time period that the

tter trial was open to inclusion. la

Baseline characteristics of patients included in the OMEGA study Baseline characteristics of included patients are shown in Table 1. Median age was 75.5

years (range 65.8 – 86.8 years) and 32 patients (54%) were aged 75 years or older.

Patients primarily lived independently (n=71, 91%) and 54% of patients (n=42) lived with

at least one other person. Although most patients had an ECOG performance status (PS) of

0 or 1, seventeen patients (22%) had a PS of 2 or 3. Five patients had a BMI below 20

kg/m2 (6%). Over 80% of patients (n=65) were partially dependent in their instrumental

activities of daily living (IADL), but no patients were fully IADL dependent. The median

number of different medications being taken at baseline was five (range 0‐14). For 45

patients (58%), no significant comorbidity was present (Charlson comorbidity index score

0). Twenty‐two patients had a Charlson score of 1 or 2 (28%), while only three patients

scored 3 or higher (4%). Few patients had cognitive dysfunction (n=5, 6%) but mild to

moderate depressive symptoms were present in 24 patients (31%). According to the

roningen frailty index, more than half of patients were considered frail (n=40, 51%). G

Discussion In the current report on accrual problems in the OMEGA study, a first‐line chemotherapy

study in frail older patients with metastatic breast cancer, we identified three major

barriers: 1) the patient’s refusal of chemotherapy, or their preference for a particular

treatment arm, 2) patient’s being considered too frail for chemotherapy, and therefore

more suitable for supportive care only, or alternatively, fit enough to be suitable for more

aggressive (combination) treatment such as in the ATX study, and 3) study characteristics

such as the time investment required for the geriatric assessments resulting in patient’s

refusal to participate or physician’s reticence to discuss this study with their patients.

However, despite these barriers and the subsequent lower than intended number of

patients included in this study, the baseline characteristics suggest that we did succeed in

including frail, elderly patients with geriatric syndromes.

This analysis of barriers to accrual may be hampered by the fact that the survey

retrospectively inquired about the barriers as experienced by the investigators which may

not necessarily reflect the actual process of accrual. Unfortunately, screening and

enrolment logs were not available for most participating centres, and no data were

71

Chapter 5

available on the number of potentially eligible patients who were not even considered for

enrolment.

The results of our survey demonstrate the dilemmas in designing a chemotherapy study

that allows for inclusion of sufficient numbers of older patients, while at the same time

ensuring sufficient homogeneity in the baseline characteristics to permit drawing

conclusions on the treatment outcome. Due to the great heterogeneity in the elderly

patient population, cancer treatment for these patients typically requires tailored care,2

while cancer trials by definition offer predefined and standardized care. Still, the baseline

characteristics of the included patients show that it is possible to enrol an older, frail

population, and future studies should maintain non‐restrictive inclusion criteria regarding

comorbidity and performance status.8 In addition, trials could use a form of stratification

to allow for inclusion of greater number of patients, or allow for dose‐adjustment (for

example, a primary dose reduction with the option of dose escalation if treatment is well

tolerated); Although this strategy was successfully implemented in previous studies,9 the

resultant heterogeneity in dosing schedules could affect the generalizability of study

outcomes. In addition, there is no consensus on what kind of instruments should be used

to guide stratification.10

The OMEGA trial is not the first study in older breast cancer patients to encounter accrual

issues. Recently, the CASA and ACTION trial, both focusing on adjuvant chemotherapy in

older women with breast cancer, were closed prematurely due to poor patient

recruitment.11 In the ACTION trial, patients were randomised between four cycles of

anthracycline/cyclophosphamide or alternatively observation only.12 In their analysis of

potential causes, the ACTION trial investigators stated that in contrast to their pre‐trial

audits, many patients were not eligible, and eligible ones often refused participation

because they did not want to be randomised to the chemotherapy arm. The investigators

stated that this is a particularly important issue in trials that use a no‐treatment arm.

Although the OMEGA study did not have a no‐treatment arm, patients were randomised

between two distinct types of chemotherapy, one given orally and one intravenously.

Previous studies have demonstrated that cancer patients receiving chemotherapy often

have a preference for a particular administration method,13 and in our survey, preference

for a particular treatment arm was indeed an important reason for patient’s refusal of

participation. It is likely that studies with more comparable treatment arms will be able to

circumvent this issue, at least partially.

A final issue is the time component: our survey showed that the time‐consuming nature of

the geriatric assessments, considered necessary for a meaningful characterisation of the

overall health status of elderly patients in clinical trials,14 was an issue for some patients

(and/or their family members) and time restraints can also cause reticence in treating

physicians to discuss the possibility of trial participation with the patient. Explaining a

72


clinical trial to an older patient often requires a greater time investment than to younger

patients. However, a study on the attitudes of older patients towards enrolment in clinical

trials demonstrated that, despite initial hesitation due to randomisation, over 75% of

patients were willing to participate after additional explanation of the trial process.15 In

another study, the amount of time spent with the patient (by the treating physician or a

clinical research assistant) was identified as an important factor associated with the

likelihood of participation of older patients in clinical trials.16 Another prime factor was the

ndorsement of the clinical trial by the treating physician.15,16 e

In conclusion, our analysis of the accrual process in the OMEGA study demonstrates that

accrual of older breast cancer patients in this randomised clinical trial on chemotherapy

was mainly hampered by either patient’s refusal of chemotherapy or randomisation or a

medical condition considered as being too fit or too frail for inclusion. In line with previous

randomised studies on adjuvant chemotherapy in elderly patients, preference for a

specific treatment was an important reason for patient’s refusal of trial participation.

Ultimately, we did manage to include frail, elderly patients, and the results on toxicity,

efficacy and the impact of comprehensive geriatric assessments are expected this year.

Future trials in elderly breast cancer patients should focus on non‐restrictive inclusion

criteria regarding co‐morbidity and performance status, avoid concurrent trials with

overlapping eligibility criteria, as well as on education of physicians and elderly patients on

e advantages of trial participation. th

Financial support: The OMEGA study received unrestricted financial support from Amgen

BV the Netherlands, Janssen‐Cilag BV the Netherlands, MSD the Netherlands and the

Dutch Cancer Society.

73

Chapter 5

Appendix 1: Survey sent to participating centres

Could you state which of the following factors could have affected the accrual of patients to the

OMEGA study in your centre:

1. Inclusion and exclusion criteria

O insufficient cognitive function

O poor cardiac function

O other ……

2. Patients

O patient information too complicated

O patient refusal of chemotherapy

O patient refusal of randomisation because of preference for treatment arm

O other ……

3. Time restraints

O geriatric assessments too time consuming

O insufficient staff (research nurse/oncology nurse)

O other … …

4. Physician

O as a physician, I am hesitant to offer chemotherapy to older patients

O I feel that combination chemotherapy is preferable to monotherapy

O this study has a lower priority than other current studies I am participating in

because ……

O other ……

5. Any other comments? …

74


75

References 1. Hutchins LF, Unger JM, Crowley JJ et al. Underrepresentation of patients 65 years of age or older in

cancer‐treatment trials. N Engl J Med 1999;341:2061‐2067. 2. Wildiers H, Kunkler I, Biganzoli L et al. Management of breast cancer in elderly individuals:

recommendations of the International Society of Geriatric Oncology. Lancet Oncol 2007;8:1101‐1115. 3. Charlson ME, Pompei P, Ales KL, Mackenzie CR. A new method of classifying prognostic comorbidity in

longitudinal studies: development and validation. J Chronic Dis 1987;40:373‐383. 4. Lawton MP, Brody EM. Assessment of older people: self‐maintaining and instrumental activities of daily

living. Gerontologist 1969;9:179‐186. 5. Folstein MF, Folstein SE, McHugh PR. "Mini‐mental state". A practical method for grading the cognitive

state of patients for the clinician. J Psychiatr Res 1975;12:189‐198. 6. Sheikh JI, Yesavage JA, Brooks JO et al. Proposed factor structure of the Geriatric Depression Scale. Int

Psychogeriatr 1991;3:23‐28. 7. Slaets JP. Vulnerability in the elderly: frailty. Med Clin North Am 2006;90:593‐601. 8. Ferrucci L, Guralnik JM, Studenski S et al. Designing randomised, controlled trials aimed at preventing or

delaying functional decline and disability in frail, older persons: a consensus report. J Am Geriatr Soc 2004; 52:625‐634.

9. ten Tije AJ, Smorenburg CH, Seynaeve C et al. Weekly paclitaxel as first‐line chemotherapy for elderly patients with metastatic breast cancer. A multicentre phase II trial. Eur J Cancer 2004;40:352‐357.

10. Maas HA, Janssen‐Heijnen ML, Olde Rikkert MG, Machteld Wymenga AN. Comprehensive geriatric assessment and its clinical impact in oncology. Eur J Cancer 2007;43:2161‐2169.

11. Reed MW, Wyld L, Ellis P et al. Breast cancer in older women: trials and tribulations. Clin Oncol (R Coll Radiol ) 2009;21:99‐102.

12. Leonard R, Ballinger R, Cameron D et al. Adjuvant chemotherapy in older women (ACTION) study ‐ what did we learn from the pilot phase? Br J Cancer 2011;105:1260‐1266.

13. Liu G, Franssen E, Fitch MI, Warner E. Patient preferences for oral versus intravenous palliative chemotherapy. J Clin Oncol 1997;15:110‐115.

14. Audisio RA, Van LB. When reporting on older patients with cancer, frailty information is needed. Annals of Surgical Oncology 2011;18: 4‐5.

15. Townsley CA, Chan KK, Pond GR et al. Understanding the attitudes of the elderly towards enrolment into cancer clinical trials. BMC Cancer 2006;6:34.

16. Wright JR, Whelan TJ, Schiff S et al. Why cancer patients enter randomised clinical trials: exploring the factors that influence their decision. J Clin Oncol 2004;22:4312‐4318

.

Part II

Treatment in elderly patients with other cancer types

Chapter 6

Colon cancer treatment and adherence to national guidelines: does age still matter?

J.M. Jonker, M.E. Hamaker, M. Soesan, C.R. Tulner, I.M.J.A. Kuper

Journal of Geriatric Oncology 2012;3:131‐137

Chapter 6

Abstract Introduction: In the past decades, much attention has been given to the risks of

undertreatment of cancer in older patients. We set out to determine whether current

treatment of colon cancer in older patients still differs from younger patients and to

identify risk factors and physician’s reasons for deviation from Dutch treatment guidelines.

Patients and methods: Retrospective cohort study of all consecutive patients newly

diagnosed with colon cancer at the Slotervaart Hospital in Amsterdam between January

2002 and December 2007. Data were collected using clinical charts.

Results: Of 286 newly diagnosed colon cancer patients, 183 were 70 years or older.

Ninety‐one percent of older patients received curative surgery and 32% received adjuvant

chemotherapy in accordance with guidelines compared to 100% and 85% in the younger

group (p=0.002 and p<0.001 respectively). The primary reasons stated by the treating

physician for withholding surgery were comorbidity and poor general health. For adjuvant

chemotherapy, the main reason stated was age. For both surgery and chemotherapy,

multivariate analysis revealed that deviation from guidelines was most strongly associated

with age (surgery OR 1.18 (CI 1.06‐1.30); p=0.002 and chemotherapy OR 1.19 (CI 1.08‐

1.31); p<0.001 respectively). Despite this selection, older patients experienced more

postoperative morbidity and mortality than younger patients. Chemotherapy toxicity was

equal in both groups, despite less aggressive regimens for older patients.

Conclusion: At our centre, guideline adherence for surgery in older patients was high, and

deviations were well motivated. Age still seems to be the most important factor for

withholding chemotherapy, despite evidence suggesting benefit in selected older patients.

80

Colon cancer in the elderly

Introduction Colon cancer is one of the most common cancers worldwide, particularly in developed

countries.1 In the Netherlands, it ranks third in the incidence of all cancers, with 37.8 new

cases per 100.000 inhabitants in 2008. In 2008 the mortality rate of colon cancer in the

Netherlands was 16.8/100.000. The occurrence increases with age, with 58% being 70

years or older.2 As a result of increasing life expectancy, the incidence of colon cancer in

the older population is expected to increase even further in the coming decades.

Much is yet unknown about how to treat these older cancer patients. Ageing is an

individual process, making this group heterogeneous in terms of comorbidity, general

health, functional status and social network. In view of this heterogeneity, it is incorrect to

automatically assume that older patients not receiving standard oncologic treatment are

being undertreated, as good reasons may exist for deviation from treatment guidelines.

Recent studies have demonstrated that surgery and chemotherapy are safe and

efficacious in selected older patients with colon cancer,3‐7 but as clinical trials have often

excluded older patients and those with severe comorbidity, it is unclear how these results

are to be interpreted when treating this older, heterogeneous population.8

Many experts agree that chronological age itself should not be the reason for withholding

cancer treatment. However, several studies on colon cancer treatment in older patients,

based primarily on patient cohorts diagnosed between 1990 and 2000, have

demonstrated that age alone was the prime risk factor for deviation from standard

treatment.9‐15 Several of these studies did not address the motivation behind discordance

with recommended cancer treatment.

In this study, we set out to determine whether current treatment of colon cancer in older

patients still differs from that in younger patients in terms of adherence to national

treatment guidelines, to describe physician’s reasons for withholding treatment in the

older patients group and to identify risk factors for deviations from recommended

eatment. tr

Methods This is a retrospective cohort study of all consecutive patients newly diagnosed with colon

cancer in a middle‐sized teaching hospital in Amsterdam, the Netherlands, between

January 2002 and December 2007.

Patients were identified from the regional cancer registry, the IKA (Integraal Kanker

Centrum Amsterdam). To ensure that no patients were omitted, the registry data were

compared with the hospital’s pathology database. This study was reviewed by the Medical

Research Ethics Committee at Slotervaart Hospital and found it exempt from IRB review.

For all patients, the following data were collected from their charts: age, sex, Charlson

Comorbidity Index16, polypharmacy (the use of 4 or more types of medication), TNM

81

Chapter 6

stage, histological grade, initial treatment regimens (surgery and/or chemotherapy), and if

applicable, post‐operative complications and toxicity of chemotherapy.

Initial treatment was compared to treatment as recommended by national treatment

guidelines.17 For stage I and II colon cancer, the recommended treatment is surgery only.

Chemotherapy for patients with stage II disease is not recommended as standard care, but

might be offered to patients with high risk disease, for example T4 or imminent blow‐out.

For stage III colon cancer, standard treatment consists of surgery and subsequent adjuvant

chemotherapy. In patients with metastatic disease, palliative chemotherapy is advised,

with surgical intervention when appropriate. The latter is not considered standard care.

For patients not receiving standard treatment, the physician’s correspondence to the

general physician of the patient was consulted to collect the reasons for deviation from

the guidelines. When the reason was not found, the patient’s medical chart was

consulted. Duration of follow‐up was 2 years. Data on mortality was collected from the

Municipal Data Registry. Post‐operative mortality was defined as death within 30 days of

rgery. su

Statistical analysis To compare baseline characteristics and guideline adherence, patients were categorized

into two age groups, younger than 70 years and 70 years and older. For these

comparisons, the chi‐square test was used.

To determine which factors were associated with guideline discordance a univariable

logistic‐regression analysis was performed. For surgical treatment, five factors were

included in the analysis: age, sex, comorbidity scored by the Charlson index, the presence

of polypharmacy and anaemia. For chemotherapy, the occurrence of post‐operative

complications was added to these five factors. All factors with a p‐value <0.2 and possible

confounders (sex, age and comorbid disease) were subsequently analysed in a

multivariable logistic‐regression analysis. A backward selection procedure, accepting a p‐

value of p<0.05, was used. Kaplan Meier survival plots, with a log‐rank analysis, were used

to determine survival.

The SPSS (Statistical Package for the Social Sciences) version 16.0 was used for the

nalyses. a

Results

Patient characteristics From January 2002 to December 2007, 286 new cases of colon cancer patients were

identified, of which 183 were 70 years or older (64%). Patient characteristics are listed in

Table 1. Mean Charlson comorbidity score was 1.2 (SD 1.4) in the older adults compared

82


to 0.6 (SD 1.0) in the younger group (p=0.001). Tumour stage did not differ between age

groups. However, 9% of the older patients could not be staged due to omission of surgery

the absence of clinically evident metastases. in

Guideline adherence Six patients died before treatment could be initiated. These patients were not included in

further analyses. Guideline adherence per age group is presented in Figure 1. All patients

under 70 years of age with stage I to III disease received surgery in accordance with

national guidelines. Of the 147 patients aged 70 and over, 133 received surgical treatment

(91%, p=0.002), and this percentage remained over 80% even in patients aged 85 years

and older. Resection margins were tumour free in 99% of younger patients and 98% in

older patients. Only 12 of 38 patients older than 70 years with stage III colon cancer

received adjuvant chemotherapy in accordance with guidelines (32%), compared to 29 of

34 younger patients (85%, p<0.001). Treatment with chemotherapy decreased steadily

Table 1: Patient characteristics per age group

Total < 70 years ≥ 70 years p‐value

No. of patients 286 103 (36.0%) 183 (64.0%)

Median age (range) 75 (21‐97) 62 (21‐69) 80 (70‐97)

Female, n (%) 165 (57.7) 50 (48.5) 115 (62.8) 0.013

Charlson comorbidity score, n (%)

0

1

2

≥ 3

134 (46.9)

76 (26.6)

50 (17.5)

26 (9.1)

65 (63.1)

21 (20.4)

12 (11.7)

5 (4.9)

69 (37.7)

55 (30.1)

38 (20.8)

21 (11.5)

<0.001

0.045

Tumour stage, n (%)

I

II

III

IV

Unknown

41 (14.3)

99 (34.6)

72 (25.2)

58 (20.3)

16 (5.6)

18 (17.5)

29 (28.2)

34 (33.0)

22 (21.4)

0 (0.0)

23 (12.6)

70 (38.3)

38 (20.8)

36 (19.7)

16 (8.7)

0.001

Histology grade n (%)

Well differentiated

Mildly differentiated

Poorly differentiated

Not determined

4 (1.4)

213 (74.5)

45 (15.7)

24 (8.4)

2 (1.9)

71 (68.9)

24 (23.3)

6 (5.8)

2 (1.1)

142 (77.6)

21 (11.5)

18 (9.8)

0.007

83

Chapter 6

with increasing age (Figure 1). Similar results were seen for patients with stage II disease:

chemotherapy was prescribed in 7% of the older patients with stage II colon cancer versus

38% of the younger group. Eight patients had high risk disease; three younger patients

with T4 and four older patients with T4 and one older patient with imminent blowout.

None of them received adjuvant chemotherapy. Interestingly, there was a trend toward

more palliative surgery, such as tumour debulking or bypass surgery, in older patients with

stage IV disease (81 vs. 59% in younger patients; p=0.07), although the procedures were

often less extensive in the older age group. In contrast, older patients received less

palliative chemotherapy (19% vs. 64%; p<0.001).

For both adjuvant and palliative chemotherapy, older patients mostly received single‐

agent chemotherapy (74%). In comparison, this regimen was chosen in only 43% of

younger patients (p=0.01), who generally received more extensive combination therapy

(p=0.01). Dose reduction was prescribed in only one older patient.

Reasons for deviation from guidelines in the older age group, as stated in patient’s charts,

are shown in Table 2. The primary reasons mentioned for withholding surgery were

comorbidity and poor general health (8 out of 14 patients). In only one case, a 90‐year old

woman, age itself was stated as the reason for not performing surgery. In contrast, age

was the most frequently stated reason for not considering chemotherapy in stage III colon

cancer (8 out of 26 patients); three of these patients did not have any comorbid disease.

Cognitive decline was stated in three older patients as the reason for withholding

Figure 1: Guideline adherence for surgery and chemotherapy in per age group

84


chemotherapy, while for seven patients the reason was not stated. In the younger group

reasons for withholding chemotherapy were prolonged post‐operative course (n=2),

comorbidity (n=1) and poor general health (n=1). One patient declined chemotherapy.

Univariate and multivariate analysis, performed to determine baseline factors associated

with guidelines discordance are shown in Table 3. Advancing age was shown to be the

most important factor for both surgery (odds ratio (OR) 1.18 (95% confidence interval (CI)

1.06‐1.30); p=0.002) and chemotherapy (OR 1.19 (95%CI 1.08‐1.31); p<0.001).

Comorbidity was borderline significant for surgery (OR 1.39 (95%CI 0.99‐1.95); p=0.06) but

was not associated with chemotherapy use (p=0.44). However, withholding of

chemotherapy was associated with post‐operative complications (OR 8.81 (95%CI 1.74‐

44.50); p=0.01) while polypharmacy showed a borderline significant association (OR 4.62

5%CI 0.98‐21.85); p=0.05). (9

Complications of surgery and chemotherapy Complications of treatment are shown in Table 4. Nearly 50% of the older patients (75 out

of 157) experienced some kind of complication after surgery; in 30 patients (19%), these

complications were life threatening (requiring intensive care treatment and/or resulting in

death). This is significantly higher than in the younger group, for which these percentages

were 30% (p=0.003) and 9% (p=0.02) respectively. The two groups did not differ in the

occurrence of wound infection (9 versus 12% in younger and older patients respectively)

or anastomotic leak (7 versus 10% respectively) but older patients experienced

significantly more cardiopulmonary complications (13% versus 27%; p= 0.01). Post‐

operative mortality was 12% for older patients, compared to 3% in the younger age group

(p= 0.03).

Table 2: Reason stated in clinical charts for deviation of guidelines older age group

Surgery Chemotherapy

Number of patients not treated in accordance with guidelines

14 26

Reason stated Death during treatment Age Comorbidity Patient refusal Cognitive decline Poor general health Prolonged post‐operative course Unknown

3 1 5 2 0 3 ‐ 0

3 8 2 1 3 0 2 7

85

Chapter 6

Table 3: Univariate and multivariate analysis

Univariate analysis Multivariate analysis *

Odds ratio

95% confidence interval

p= * Odds ratio


p= *

Surgery Age 1.16 1.05‐1.29 0.005 1.18 1.06‐1.30 0.002 Sex 0.74 0.18‐3.12 0.68 0.75 0.18‐3.17 0.70 Comorbidity 1.42 0.96‐2.10 0.08 1.42 1.00‐1.95 0.06 Polypharmacy 0.69 0.16‐3.02 0.62 Anaemia 0.35 0.036‐3.37 0.36

Chemotherapy Age 1.17 1.06‐1.29 0.002 1.19 1.08‐1.31 <0.001 Sex 1.10 0.22‐5.48 0.91 1.14 0.23‐5.63 0.87 Comorbidity 1.37 0.61‐3.07 0.44 1.37 0.62‐3.05 0.44 Polypharmacy 3.74 0.76‐18.47 0.11 4.62 0.98‐21.85 0.05 Anaemia 0.58 0.13‐2.64 0.48 Postoperative complications

7.97 1.54‐41.25 0.013 8.81 1.74‐44.5 0.008

*Variables with a p‐value <0.20 in the univariate analysis were included in the multivariate analysis. Age, sex and

comorbidity were also included as potential confounders. A backward selection procedure was applied; a

forward analysis selected the same variables.

Table 4: Treatment complications

< 70 years ≥ 70 years p‐value

Surgery*

All complications Wound infection Anastomotic leak Cardiopulmonary complications Life‐threatening complications Postoperative mortality (30‐days)

26/88 (30%)8/88( 9%)6/88 ( 7%)

11/88 (13%)8/88 ( 9%)3/88 ( 3%)

75/157 (48%)18/157(12%)15/157(10%)42/157(27%)30/157 (19%)18/157 (12%)

0.004 0.36 0.32

0.006 0.03 0.02

Chemotherapy**

Toxicity all grades

Toxicity grade III‐IV

Hospital admission due to toxicity Termination therapy due to toxicity

44/50 (88%)16/50 (32%)9/50 (18%)

10/50 (20%)

18/22 (82%)6/22 (27%)4/22 (18%)5/22 (23%)

* Polipectomie not included ** 6 patients lost to follow up (4 young, 2 old)

86


Of the 78 patients receiving chemotherapy, four younger patients and two older patients

no information on toxicity was available, as they were treated elsewhere. No statistically

significant differences in toxicity or toxicity‐related sequalae were seen between age

groups, but numbers are small and treatment regimens were diverse. In both age groups,

18% required admission to hospital due to toxicity of chemotherapy and around 20%

rminated their initial chemotherapy regimens before receiving all planned cycles. te

Survival Figure 2 shows two‐year survival plots. Patients not receiving standard treatment had a

significantly poorer prognosis: one‐year survival was 84% for those patients receiving

standard treatment, and 68% in case of a deviation from treatment guidelines while two‐

year survival was 74% and 56% respectively (p=0.04).

Figure 2: Kaplan Meier survival plots

87

Chapter 6

Discussion In the past decades, much attention has been paid to the risks of undertreatment of

cancer in older patients. In this study, we set out to determine whether current treatment

of colon cancer in older patients still differs from standard treatment and to explore

treating physicians’ reasons for deviating from guidelines. We found that 91% of patients

aged 70 years and older received surgery in accordance with guidelines, compared to

100% of younger patients. Only 32% of older patients receiving adjuvant chemotherapy,

compared to 85% of younger patients. With advancing age, use of adjuvant chemotherapy

declines drastically. Age was the factor most strongly associated with withholding

chemotherapy, followed by the occurrence of post‐operative complications, while for

surgery, age and comorbidity were the most important factors.

Although various prior studies have addressed guideline adherence in older patients with

colon cancer,3‐7 most did not address the reasons behind discordance with treatment

recommendations. We believe that these data are necessary to differentiate between

undertreatment and adequate adjustment of treatment to the individual patient. Our

study is unique in that we have collected these reasons, in combination with information

on surgical complications and toxicity of chemotherapy. Another strength of this study is

that it represents every day practice rather than a selected population. An important

limitation of this study, however, is its retrospective character, as not all parameters were

systematically documented and some data could not be retrieved from patient’s charts.

For example, we were not able to include relevant factors such as functional and

nutritional status, cognitive function and social support. The data we collected were from

a single‐centre, and therefore, it is unclear if our findings can be extrapolated to

treatment patterns at other centres.

The rate of guideline adherence for surgical treatment in our study is higher than those

found in previous studies,9,10,18 which state rates of 67 to 88% for older patients. As these

studies addressed patients diagnosed between 1990 and 2000, perhaps the higher rate

found at our centre is a reflection of an increased awareness of the risks of

undertreatment.

Post‐operative morbidity and mortality were high for our older patients (all complications

48% and postoperative mortality 12%). This finding is confirmed by various previous

studies, which, despite substantial differences in complication rates and mortality, all

found these to be significantly higher in older patients.3 In our study, comorbidity and

poor general health were the primary reasons stated by the treating physician for

withholding surgery. In light of the higher surgical risk for older patients, these may be

good reasons for deviating from guidelines.

We found that older patients received chemotherapy less often than younger patients,

and when administered, less extensive treatment regimens were chosen. Earlier European

88


studies showed similar results, with 13‐24% of patients aged 75 years and older receiving

chemotherapy; in the United States treatment rates are considerable higher (≥75 years

range 35% to 52%).18 In our study, chronological age was the most important reason

mentioned by the treating physician for not following chemotherapy guidelines. Despite

selection of older patients and less extensive treatment regimes in our cohort, one‐third

of patients experienced grade III‐IV toxicity.

Prior studies addressing the reasons for not offering adjuvant chemotherapy in older

patients found that patient’s refusal was the primary reason in the patients aged 71 to 80

years and age alone or age in combination with comorbidity for those aged 80 years and

older.11,14,19 Thus, age is still an important factor in treatment decisions regarding

chemotherapy.

Treating physicians may consider the benefit of adjuvant chemotherapy in preventing

cancer recurrence to be too small or toxicity too great. Eighty percent of recurrences of

colon cancer occur within 3 years and 90% of patients with recurrence die within 24

months.20 In 1990 the first study with levamisole and fluorouracil in stage III colon cancer

showed a 15% absolute risk reduction for recurrence and 16% absolute risk reduction of

mortality compared to patients without this adjuvant chemotherapy.21 Later studies,

replacing levamisole with leucovorin, have shown similar benefit of adjuvant

chemotherapy in selected older patients.22 Furthermore, various studies have

demonstrated that fit older patients with no significant comorbidity are able to tolerate

standard colon cancer treatment as well as younger patients.3‐7,23,24 In summary, any

patient with a life expectancy of 3 to 5 years, in reasonable general health, and a high risk

of recurring disease, could benefit from the addition of chemotherapy. The average life‐

expectancy for an 80 year old woman is another 9.3 years and for a man 7.5 years so

curative treatment should be considered in this age group.

Therefore, the challenge surgeons and oncologists are facing is how to differentiate the fit

from the vulnerable, and how to predict life expectancy. Although chronological age

obviously influences life expectancy and a recent study showed that older patients may

experience more chemotherapy toxicity, ageing is a heterogeneous process and age alone

should not be the reason for withholding a potentially curative treatment.25

Elements of a comprehensive geriatric assessment (CGA) can predict morbidity and

mortality in older patients with cancer, which are both important factors in determining if

cancer treatment is appropriate.26 A CGA provides information on the functional ability,

comorbid medical conditions, cognition, nutritional status, psychological state and social

support of older cancer patients.27‐29 Screening instruments such as American Society of

Anaesthesiologists (ASA)‐score and Karnofsky performance status (PS) score were not

developed to detect these geriatric conditions.27,30‐31 However, it may well be that these

additional factors determine if an older patient is vulnerable or has a good chance to

89

Chapter 6

undergo cancer treatment without serious adverse effects.32 For example, a recent study

showed that the results of a CGA were predictive of complications after surgery for

colorectal cancer in 171 older patients (median age 80 years), while age, ASA class, and

stage of disease were not.33 In addition, preoperative assessment in elderly cancer

patients (PACE) showed that IADL, fatigue and performance status were associated with

50% increase in relative risk of post‐operative complications.34 Two recent studies focused

on predicting chemotherapy toxicity in older adults with cancer. One study found that

hearing impairment, one or more falls in the last 6 months, limitations in walking one

block, the need for assistance in taking medications, and decreased social activities were

risk factors for toxicity.25 In the other study IADL was a risk factor of hematologic toxicity,

while self‐rated health, mini‐mental status scores and mini‐nutritional assessment score

were predictors for non‐hematologic toxicity.35 Further research into the optimal form of

CGA and its role in cancer treatment decision making is much needed and is slowly

ecoming available.36‐38 b

In conclusion, colon cancer treatment in the older patients still differs greatly from their

younger counterparts in our centre. Curative surgery in the older patient is not withheld

based on chronological age alone but mostly because of comorbidity and physical fitness.

As chronological age alone is the most important reason mentioned for omitting adjuvant

chemotherapy, it is possible that a potentially curative treatment was withheld in these

patients. Further research is needed on the role of CGA in differentiating frail older

patients from those that will benefit from standard treatment despite advanced age.

90


References 1. Jemal A, Bray F, Centre MM et al. Global cancer statistics. CA Cancer J Clin 2011;61:134 . 2. IKCnet (Integraal kankerregistratie: Dutch cancer registry); Available from URL: http//www.ikcnet.nl 3. Colorectal Cancer Collaborative Group. Surgery for colorectal cancer in elderly patients: a systemic review.

Lancet 2000;356:968‐74. 4. Sargent DJ, Goldberg RM, Jacobson SD et al. A pooled analysis of adjuvant chemotherapy for resected

colon cancer in elderly patients. N Engl J Med 2001;345:1091‐7. 5. Sundararajan V, Mitra N, Jacobson JS et al. Survival associated with 5‐fluorouracil‐based adjuvant

chemotherapy among elderly patients with node‐positive colon cancer. Ann Intern Med 2002;136:349‐57 6. Jessup, JM, Steward A, Green FL, Minsky BD. Adjuvant chemotherapy for stage III colon cancer: implication

of race/ethnicity, age, and differentiation. JAMA 2005;294:2703‐11. 7. Goldberg RM, Tabah‐Fisch I, Bleiberg H et al. Pooled analysis of safety and efficacy of oxaliplatin plus

fluorouracil/leucovorin administered bimonthly in elderly patients with colorectal cancer. J Clin Oncol 2006;24;4085‐91.

8. Hutchins LF, Unger JM, Crowley JJ et al. Underrepresentation of patients 65 years of age or older in cancer‐treatment trials. N Engl J Med 1999;341:2061‐7.

9. Damhuis RAM, Wereldsma JCJ, Wiggers T. The influence of age on resection rates and postoperative mortality in 6457 patients with colorectal cancer. Int Colorect Dis 1996;11:45‐48.

10. Aparicio T, Navazesh A, Boutron et al. Half of elderly patients routinely treated for colorectal cancer receive a sub‐standard treatment. Crit Rev Oncol Hematol 2009;71:249‐57.

11. Mahoney T, Kuo YH, Topilow A, Davis JM. Stage III colon cancers; why adjuvant chemotherapy is not offered to elderly patients. Arch Surg 2000;135:182‐5.

12. Schrag D, Cramer LD, Bach PB, Begg CB. Age and adjuvant chemotherapy use after surgery for stage III colon cancer. J Natl Cancer Inst 2001;93:850‐7.

13. Potosky AL, Harlan LC, Kaplan RS et al. Age, sex, and racial differences in the use of standard adjuvant therapy for colorectal cancer. J Clin Oncol 2002;20:1192–202.

14. Ayanian, JZ, Zaslavsky, AM, Fuchs CS et al. Use of adjuvant chemotherapy and radiation therapy for colorectal cancer in a population‐based cohort. J Clin Oncol 2003;21:1293‐300.

15. Lemmers VE, van Halteren AH, Janssen‐Heijnen ML et al. Adjuvant treatment for elderly patients with stage III colon cancer in the southern Netherlands is affected by socioeconomic status, gender, and comorbidity. Ann Oncol 2005;16:767‐72.

16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83.

17. Dutch national guidelines: Coloncarcinoom versie 1.2;www.oncoline.nl 18. Faivre J, Lemmens VE, Quipourt V, Bouvier AM. Management and survival of colorectal cancer in the

elderly in population‐based studies. Eur J Cancer 2007;43:2279‐84. 19. Ananda S, Field KM, Kosmider S et al. Patient age and comorbidity are major determinants of adjuvant

chemotherapy use for stage III colon cancer in routine clinical practice. J Clin Oncol 2008;26:2532‐7. 20. Sargent DJ, Wieand HS, Haller DG et al. Disease‐free survival versus overall survival as a primary end point

for adjuvant colon cancer studies: individual patient data from 20,898 patients on 18 randomized trials. J Clin Oncol 2007;25:4569‐74.

21. Moertel CG, Fleming TR, Macdonald JS et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med 1990;322:352‐8

22. Haller DG, Catalano PJ, Macdonald JS et al. Phase III study of fluorouracil, leucovorin, and levamisole in high‐risk stage II and III colon cancer: final report of Intergroup 0089. J Clin Oncol 2005;23:8671‐8.

23. Scheithauer W, McKendrick J, Begbie S et al on behalf of the X‐ACT Study Group. Oral capecitabine as an alternative to i.v. 5‐fluorouracil‐based adjuvant therapy for colon cancer: safety results of a randomized, phase III trial. Ann Oncol 2003;14:1735‐43.

24. Popescu RA, Norman A, Ross PJ et al. Adjuvant or palliative chemotherapy for colorectal cancer in patients 70 years or older. J Clin Oncol 1999;17:2412‐8.

25. Hurria A, Togawa K, Mohile SG et al. Predicting Chemotherapy Toxicity in Older Adults With Cancer: A Prospective Multicentre Study. J Clin Oncol 2011;25:3457‐65.

26. Extermann M, Hurria, A. Comprehensive Geriatric Assessment for older patients with cancer. J Clin Oncol 2007;25:1824‐1831.

91

http://www.ncbi.nlm.nih.gov/pubmed?term=%22Ananda%20S%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=%22Field%20KM%22%5BAuthor%5D

http://www.ncbi.nlm.nih.gov/pubmed?term=%22Kosmider%20S%22%5BAuthor%5D

Chapter 6

92

27. Repetto L, Fratino L, Audisio RA et al. Comprehensive geriatric assessment adds information to Eastern Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study. J Clin Oncol 2002;20:494‐502.

28. Monfardini S. Balducci L. A comprehensive geriatric assessment (CGA) is necessary for the study and the management of cancer in the elderly. Eur J Cancer 1999;35:1771‐72.

29. Extermann M, Aapro M, Bernabei R et al. Use of comprehensive geriatric assessment in older cancer patients: recommendations from the task force on CGA of the International Society of Geriatric Oncology (SIOG). Crit Rev Oncol Hematol 2005;55:241‐52.

30. Extermann M, Meyer J, McGinnis M et al. A comprehensive geriatric intervention detects multiple problems in older breast cancer patients. Crit Rev Oncol Hematol 2004;49:69‐75.

31. Gosney MA. Clinical assessment of elderly people with cancer. Lancet Oncol 2005;6:790‐797. 32. Aaldriks AA, Maartense E, le Cessie S et al. Predictive value of geriatric assessment for patients older than

70 years, treated with chemotherapy. Crit Rev Oncol Hematol 2011;79:205‐12. 33. Kristjansson SR, Nesbakken A, Jordhøy MS et al. Comprehensive geriatric assessment can predict

complications in elderly patients after elective surgery for colorectal cancer: a prospective observational cohort study. Crit Rev Oncol Hematol 2010;76:208‐17.

34. Audisio RA, Pope D, Ramesh HS et al. Shall we operate? Preoperative assessment in elderly cancer patients (PACE) can help: A SIOG surgical task force prospective study. Crit Rev Oncol Hematol 2008;65:156‐63.

35. Extermann, M, Boler I, Reich RR et al. Predicting the risk of chemotherapy toxicity in older: the chemotherapy risk assesment scale for high‐age patients (CRASH) score. Cancer 2012;118:3377‐86.

36. Hurria A, Gupta S, Zauderer M et al. Developing a cancer‐specific geriatric assessment. Cancer 2005;9:1998‐2005.

37. Hurria A, Cirrincione CT, Muss HB et al. Implementing a geriatric assessment in cooperative group clinical cancer trials: CALGB 360401. J Clin Oncol 2011;29:1290‐6.

38. Brunello A, Sandri R, Extermann M. Multidimentional geriatric evalution for cancer patients as a clinical and research tool. Canc Treat Rev 2009;35:487‐92.

Chapter 7

Diagnostic choices and clinical outcome in octogenarians with iron‐deficiency anaemia

M.E. Hamaker, T. Acampo, J.A. Remijn, S.A.C. van Tuyl, A. Pronk, H.A. Paling,

C.H. Smorenburg, S.E. de Rooij, B.C. van Munster

[submitted]

Chapter 7

Abstract Aim: To evaluate current clinical practice for octogenarians with iron‐deficiency anaemia

(IDA) by assessing referral patters and diagnostic choices, the clinical consequences of

omission of endoscopy, and the risks and benefits of IDA‐related surgery.

Methods: Chart review of all patients aged 80 years and older diagnosed with IDA (Hb ≤6.9

mmol/l and ferritin ≤25 ug/l) at a regional hospital‐based laboratory between January

2008 and December 2010.

Results: During the study period, 471 patients were newly diagnosed with IDA (median age

85.4 years). Of these, 276 (59%) did not undergo any diagnostic procedures for IDA. A

cause of anaemia was identified in 50% of patients undergoing initial work‐up. Of the 61

gastrointestinal malignancies that were identified, sixteen were found during follow‐up,

primarily in patients for whom initial diagnostic work‐up was limited or omitted. In case of

colon cancer, perioperative mortality was 15%, and survival benefit of surgery compared

to supportive care was not apparent until 1.3 years after ascertainment of iron‐deficiency

anaemia.

Conclusion: Omission of endoscopy for iron‐deficiency anaemia and omission of surgery

for colon cancer occur frequently in octogenarians, and appears legitimate in the presence

of relevant comorbidity and in case of limited remaining life‐expectancy. Further research

is needed to determine which baseline factors should guide decision making to optimize

outcome and quality of life.

94

Octogenarians with iron‐deficiency anaemia

Introduction Iron‐deficiency anaemia (IDA) occurs in 2‐5 % of men and post‐menopausal women in the

western world.1 Many benign causes are possible, such as insufficient iron intake or

absorption, arteriovenous malformations and erosive lesions, but 2‐15% of IDA patients

are diagnosed with colorectal cancer and 2‐6% with cancer of the upper gastro‐intestinal

tract.2‐5

In the care for an elderly patient with IDA, physicians are faced with a number of

dilemmas on how to proceed with diagnosis and treatment. First, although endoscopic

procedures are frequently performed in older patients, both the procedure itself – and in

case of colonoscopy the required bowel preparation – are burdensome and not entirely

without risk,6‐10 particularly in frail older patients or in case of significant comorbidity.

Second, omission of endoscopy will mean that the cause of IDA remains unclear. A lack of

diagnosis is most relevant in case of a potentially curable disease, such as colon cancer,

provided that a patient is fit enough and willing to undergo treatment. However, many

causes of IDA cannot be easily eliminated even when identified, in which case treatment

will focus on alleviation of symptoms with blood transfusions and/or iron suppletion; for

these patients, the clinical consequences of not identifying an exact cause of IDA due to

omission of endoscopy will be very limited.

In this study, we performed a population‐based cohort study of octogenarian patients

with iron‐deficiency anaemia to determine: a) current diagnostic practice in case of IDA,

by assessing the proportions of elderly patients receiving a full, partial or no endoscopic

work‐up; b) the clinical consequences of omission of endoscopy by assessing how

frequently a potentially curable cause of iron‐deficiency was diagnosed during follow‐up in

case of an incomplete initial work‐up; and c) to assess risks and benefit of cancer

treatment in case a colon malignancy is identified, by comparing survival in patients

ceiving surgical treatment with those for whom such treatment was omitted. re

Methods This is a cohort study of all patients aged 80 years and older for whom haemoglobin level

(Hb) and ferritin was determined at the Laboratory for Clinical Chemistry and

Haematology of the Gelre hospitals in Apeldoorn, the Netherlands, between January 2008

and December 2010. The Gelre hospital is the only laboratory in the Apeldoorn region and

serves as a regional laboratory for both hospital specialists and primary care physicians.

Patients were included in this study if they had a newly ascertained Hb‐level ≤ 6.9 mmol/l,

with a ferritin level below 25 ug/l. Patients were excluded if they had a previous diagnosis

of iron‐deficiency in the past five years.

For included patients, the following data were collected from the hospital database, which

contains all correspondence, data on diagnostic procedures and laboratory results: age at

95

Chapter 7

diagnosis, sex, medical history, medication use at diagnosis, laboratory findings

(haemoglobin, MCV, ferritin), the reason for blood sampling, symptoms or complaints

suggestive of anaemia or malignancy, the initial diagnostic procedures for anaemia

(imaging studies, endoscopic studies) and the underlying cause of anaemia. Comorbidity

was scored using the Charlson comorbidity index.11 All diagnostic procedures that were

part of the initial work‐up were classified as “initial diagnostic procedures”. In addition,

subsequent treatment was recorded, including the initiation or cessation of medication as

well as invasive treatment such as surgery.

Patients were followed until May 15th, 2012. All blood transfusions, secondary anaemia‐

related diagnoses, complications related to gastrointestinal disease and all‐cause deaths

ere recorded. w

Statistical analysis Descriptive statistics were used for most outcome data. For comparisons between patient

groups, the chi‐square test was used for nominal and ordinal variables. For continuous

variables with a normal distribution, the Student t‐test was used, and for continuous

variables with a non‐normal distribution the Mann‐Whitney test. All analysis was

performed using the SPSS (Statistical Package for the Social Sciences) version 19.0.

To determine which factors were associated with diagnostic choices, a univariate logistic

regression analysis was performed. Factors with a p‐value ≤0.20 were subsequently

entered into a multivariable model. To compare survival, Kaplan Meier survival plots with

log‐rank analysis were used. a

Results

Baseline characteristics Between 2008 and 2010, iron‐deficiency anaemia was ascertained in 471 patients aged 80

years and older living in the Apeldoorn region. Baseline characteristics can be found in

Table 1. Median age was 85.4 years (range 80‐101 years); 345 were women (73%).

Median haemoglobin level at diagnosis was 6.0 mmol/l (range 2.4‐6.9 mmol/l) and median

ferritin was 13 ug/l (range 2.0‐25.0 ug/l; for reference values, see Table 1). MCV was

below 80 fl in 289 patients (61%). Median Charlson comorbidity index was 1 (range 0‐7).

Laboratory testing was initiated by a hospital‐based specialist in 128 patients (27%) and by

the primary care physician in 343 (73%). Of the latter group, 138 were not referred to

hospital for analysis and therefore no further anaemia‐related information was available

(29% of all patients). For the remaining 333 patients, reason for testing haemoglobin

levels was suspected anaemia in 157 (47%), routine laboratory testing in 97 patients (29%)

and not retrievable in 79 (26%). Only 39 patients (12%) did not have any anaemia‐related

96


symptoms or complaints (Table 1); 51% of patients had constitutional symptoms, while

symptoms suggestive of upper and lower digestive tract pathology were present in 12%

nd 17% of patients respectively. a

Diagnostic procedures and diagnostic yield In addition to the 138 patients that were not referred to hospital, another 138 referred

patients did not have any diagnostic testing of the gastrointestinal tract (276 patients in

total, 59%). Ninety‐two patients had both a colonoscopy and a gastroduodenoscopy

(20%), 41 (9%) had a gastroduodenoscopy only, 43 (9%) a colonoscopy only and for 19

patients (4%), the diagnostic procedures were limited to imaging studies (ultrasound

and/or CT‐scan). Two of the 135 patients undergoing colonoscopy suffered from a colon

perforation (1.5%), one of which had a fatal outcome.

Table 2 shows factors associated with the likelihood of endoscopic procedures during the

initial diagnostic work‐up. In the multivariable analysis, female sex (odds ratio (OR) 0.57,

95% confidence interval (CI) 0.32‐0.91, p=0.02) and being 90 years of age or older (OR

0.34, 95%CI 0.17‐0.69, p=0.003) were independently associated with a decreased

likelihood that endoscopic procedures were performed while the odds of endoscopy

increased in the presence of gastrointestinal symptoms (OR 2.34, 95%CI 1.38‐3.96,

p=0.002).

The yield of the initial diagnostic procedures is listed in Table 3. A benign cause of anaemia

was identified in 46 of the 133 patients undergoing gastroduodenoscopy (34%) and a

malignancy in nine patients (7%). Colonoscopy identified a benign cause in 31 of 135

patients under this procedure (23%) and a malignancy in 31 patients (23%), of which five

were located in the rectum and 26 in the colon. When diagnostic procedures were limited

to imaging studies, a malignancy was identified in six out of 19 patients (32%), of which

four were located in the gastrointestinal tract. Two additional colon cancers were

identified on imaging studies performed in patients for whom colonoscopy could not be

completed. Overall, a cause of anaemia was identified in 97 of the 195 patients (50%)

during the initial IDA work‐up.

In addition to these nine upper and 37 lower gastrointestinal tract cancers identified in the

initial work‐up, another 24 malignancies were identified during follow‐up, of which six

were found outside the digestive tract, two in the pancreas, 15 in the colon and one in the

rectum. Of these latter 16 patients, three had received a complete initial work‐up; three

had received only imaging studies while for the remaining ten, no initial diagnostic

procedures were performed.

97

Chapter 7

Table 1: Baseline characteristics

Characteristics n = (%)

Age Median 85.4 years (range 80‐101)

Sex Female 345 (73%)

Laboratory tests (median (range)* Haemoglobin MCV Ferritin

6.0 mmol/l (2.4‐6.9)** 78.5 fl (48‐128) 13 ug/l (2‐25)

Physician responsible for initial laboratory testing Hospital specialist Primary care physician

128 (27%) 343 (73%)

Reason for initial laboratory testing Unknown Suspicion of anaemia Routine screening

217 (46%) 157 (33%) 97 (21%)

Concurrent conditions Charlson comorbidity index

11

Cardiac disease (Cerebro)vascular disease Dementia Pulmonary disease Prior malignancy 0‐1 2‐3 4+ Missing

109 (23%) 130 (28%) 47 (10%) 57 (12%) 44 ( 9%)

246 (52%) 140 (30%) 48 (10%) 37 ( 8%)

Medication use Median 5 types (range 1‐18) Anticoagulant use Acetylsalicylic acid

Coumarin derivatives Clopidrogel

136 (40%) 79 (23%) 31 ( 9%)

Other relevant drugs NSAID SSRI Prednisone Proton pump inhibitor Erythropoietin

33 (10%) 22 ( 7%) 24 ( 7%) 124 (37%) 5 ( 2%)

Symptoms*** Unknown 87 (34%) No symptoms or complaints 39 (12%) Constitutional symptoms

Fatigue Dyspnoea Cardiac complaints Dizziness or syncope Weight loss

170 (51%) 82 (25%) 73 (22%) 38 (11%) 49 (15%) 23 (7%)

Upper digestive tract symptoms

Heart burn Diminished appetite Melaena Nausea and/or vomiting

39 (12%) 14 (4%) 13 (4%) 11 (3%) 5 (2%)

Lower digestive tract symptoms

Visible rectal blood loss Altered bowel habits Abdominal pain

57 (17%) 17 (5%) 23 (7%) 18 (5%)

MCV mean corpuscular volume, NSAID non‐steroidal anti‐inflammatory drugs, SSRI selective serotonin reuptake inhibitors // *Reference values: Haemoglobin male 8.5‐11/0 mmol/l, female 7.5‐10.0 mmol/l; MCV 80‐100 fl; Ferritin 45‐250 ug/l. // **This is equivalent to 9.7 g/dL (range 3.8‐11.1 g/dL) //*** The 138 patients for whom anaemia was ascertained by the primary care physician and not referred to hospital were excluded, leaving 333 evaluable patients. Multiple complaints per patient were possible.

98


Table 2: Multivariable analysis of factors associated with performing of endoscopic procedures

Univariate Multivariable

Factor OR 95% CI p= OR 95% CI p=

Female sex 0.41 0.27‐0.62 <0.001 0.57 0.32‐0.91 0.02 Age >90 years 0.25 0.14‐0.46 <0.001 0.34 0.17‐0.69 0.003 Charlson 1.06 0.89‐1.27 0.5 Polypharmacy 1.30 0.84‐2.02 0.2 Use of anticoagulants 0.94 0.60‐1.48 0.8 Gastro‐intestinal symptoms 2.28 1.39‐3.74 0.001 2.34 1.38‐3.96 0.002 Haemoglobin level <6.0 mmol/l 1.63 1.12‐2.37 0.01 ‐ Ferritin level ≤12 mmol/l 1.36 0.94‐1.98 0.11 ‐ Median Corpuscular Volume <80 fl 1.35 0.92‐2.00 0.13 ‐

Table 3. Initial diagnostic procedures for anaemia and their outcomes

Procedure Outcome n= (%)

Gastroduodenoscopy Performed in 133 patients (28%)

No abnormalities Benign cause of anaemia Malignancy

78 (59%) 46 (34%) 9 ( 7%)

Colonoscopy/Sigmoidoscopy Performed in 135 patients (29%)

No abnormalities Benign cause of anaemia Malignancy*

83 (61%) 31 (23%) 31 (23%)

Imaging studies only Performed in 19 patients (4%)

No abnormalities Gastrointestinal malignancy Other malignancy

13 (68%) 4 (21%) 2 (11%)

Overall n= 471

No referral No initial work‐up No cause identified Benign cause of anaemia Gastrointestinal malignancy

138 (29%) 138 (29%) 95 (20%) 55 (12%) 45 (10%)

*

An additional 2 malignancies were identified with CT‐scan after incomplete colonoscopy

Treatment and outcome Pharmaceutical treatment of anaemia was initiated in 160 patients (34%), including iron

supplements in 146 patients (31%), Helicobacter eradication treatment in six patients (1%)

and proton pump inhibitors (PPIs) in 44 patients (9%). Of these 44, only eight had been

diagnosed with ulcerative lesions in the upper digestive tract; 31 had had no diagnostic

work‐up and five received PPI despite an unremarkable gastroduodenoscopy. Blood

transfusion was required for 219 patients (46%), of which 101 required repeated

transfusions (21%). Surgery was performed in 25 of the 32 patients initially diagnosed with

colon cancer (78%) and for nine of the 15 patients diagnosed during follow‐up (60%). Four

of the six patients with rectal cancer received radiotherapy (66%) and three of the nine

patients with upper gastro‐intestinal tract cancer were treated with chemotherapy (33%).

The remaining 20 patients diagnosed with gastro‐intestinal cancer and all patients

diagnosed with cancer outside the digestive tract received supportive care only.

99

Chapter 7

Table 4: Outcome per diagnosis

Diagnosis % recovery of anaemia

% requiring blood transfusion

Median survival

No diagnosis 49% 40% 3.2 years Benign cause upper gastrointestinal tract 70% 53% not reached

Benign cause lower gastrointestinal tract 38% 80% 3.3 years Malignancy lower gastrointestinal tract 47% 75% 3.0 years Malignancy upper gastrointestinal tract 22% 67% 1.1 years Malignancy outside of gastrointestinal tract 50% 70% 1.3 years

p‐value <0.001 0.10 <0.001

Figure 1a: Kaplan Meier survival plots per initial diagnosis

Log rank p<0.001

1. Benign cause upper gastrointestinal tract 2. Benign cause lower gastrointestinal tract 3. No diagnosis 4. Malignancy lower gastrointestinal tract 5. Malignancy upper gastrointestinal tract 6. Malignancy outside of gastrointestinal tract

100


Figure 1b: Kaplan Meier survival plots comparing patients with colon cancer who undergo surgery with those who do not

Median survival Surgery: 2.2 years No surgery: 1.5 years Log rank p=0.05

Prognosis varied greatly depending on the initial diagnosis (p<0.001, Table 4).

Haemoglobin levels normalized for 232 patients (49%); this was most likely in patients

with a benign upper digestive tract cause of anaemia (recovery rate 70%), and least likely

in case of a malignancy of the upper digestive tract (22%, Table 4).

During a median follow‐up of 3.0 years (range 1.4‐4.4 years), 220 patients died (47%).

Survival per diagnosis is shown in Figure 1a. Patients with a benign upper gastrointestinal

tract cause of the iron‐deficiency anaemia demonstrated the best survival (median not

reached, Table 4); patients with a benign cause located in the colon, a colon malignancy or

those without a diagnosis demonstrated an intermediate prognosis (median survival 3.0‐

3.3 years); and patients with an upper gastrointestinal tract malignancy or a malignancy

outside the gastrointestinal tract that caused iron‐deficiency anaemia had a poor

prognosis (median survival 1.1‐1.3 years, p<0.001).

101

Chapter 7

For the 34 patients receiving surgery for colon cancer, thirty‐day mortality was 15% (5 out

of 34) and one‐year mortality was 26% (9 of 34). One‐year mortality for patients with

colon cancer who did not undergo surgery was 17%. However, after 1.3 years the survival

benefit of patients undergoing surgery became apparent (Figure 1b), and median survival

for these patients was 1.0 years longer than for colon cancer patients those who did not

ceive surgery (p=0.05). re

Discussion The aim of this study was to evaluate current clinical practice for octogenarians with iron‐

deficiency anaemia by assessing referral patterns and diagnostic choices, the clinical

consequences of omission of endoscopy, and the risks and benefits of IDA‐related surgery.

We found that in 59% of patients no initial diagnostic procedures were performed. A

cause of anaemia was identified in 50% of patients undergoing an initial work‐up. Of the

61 gastrointestinal malignancies that were identified in this cohort, sixteen were found

during follow‐up, often in patients for whom initial diagnostic work‐up was limited or

omitted. In case of colon cancer, perioperative mortality was 15%, and benefit of surgery

compared to supportive care was not apparent until 1.3 years after ascertainment of iron‐

deficiency anaemia.

Few prior studies have assessed diagnostic choices in IDA in a population‐based cohort of

elderly patients. One study addressed adult patients of all ages diagnosed with iron‐

deficiency by the general practitioner and found that 69% of those patients did not

undergo any anaemia‐related work‐up,12 while another study found that of patients aged

40 years and older with IDA, also diagnosed by the general practitioner, only 35% received

a relevant referral for further analysis.13 These percentages are comparable to what we

found, suggesting that the lack of diagnostic procedures and referral in our patient cohort

is not entirely an age‐related issue. However, we did find that for patients aged 90 years

and older, the likelihood of endoscopic assessment in case of iron‐deficiency decreased

greatly.

In geriatric medicine, making a diagnosis is not a goal in itself, but rather a means towards

optimizing quality of life and survival; therefore, an undiagnosed cause of IDA is

acceptable if its ascertainment has no therapeutic consequences. In this study, the main

diagnosis‐related treatment choices were surgery for colon cancer and proton pump

inhibitors (PPIs). However, many patients were given PPIs even if no such erosive lesions

were found or no work‐up was done, and therefore the question remains whether these

patients would not also have received this treatment if no gastroduodenoscopy was

performed. For this reason, it seems that the primary benefit of a work‐up for IDA in

octogenarian patients would be to identify colon cancer as early as possible, so that

surgery can be performed before disease progression limits the resectability of the

102


tumour, potentially resulting in a poorer prognosis.14 However, one could question

whether the risks of delayed diagnosis of colon cancer would have justified performing a

colonoscopy in the 336 patients in our cohort that did not undergo this procedure during

the initial work‐up. Colonoscopy does not exclude a delayed diagnosis and is not without

risk: in addition to the 1.5% of patients that suffered an iatrogenic bowel perforation

during endoscopy in our cohort, 2% of patients that did have a colonoscopy were

nonetheless diagnosed with colon cancer during follow‐up.

In addition, if the patient is considered not fit enough or unwilling to undergo surgery

should a colon cancer be found, the benefit of making an early diagnosis is nullified.

Perioperative mortality for elderly patients undergoing colon surgery is high (10‐15%),15‐17

and these patients remain at an excess risk for mortality during the first year after

surgery.16 In our cohort, surgical benefit compared to best supportive care did not become

apparent until approximately 1.3 years after initial ascertainment of iron‐deficiency

anaemia. As omission of surgery is frequently related to comorbidity and poor general

health,15 the remaining life expectancy of these patients was likely to be shorter

irrespective of cancer and cancer treatment, and the surgical risks would generally have

been higher. Therefore, omission of surgery in case of poor general health and in the

presence of comorbidity that limits the remaining life‐expectancy seems legitimate.

Similarly, if the outcome of endoscopy will have no clinical consequences, the risks and

burden of endoscopy mean that withholding such diagnostic procedures is acceptable. On

the other hand, it is important to realize that estimation of life‐expectancy is notoriously

difficult,18 particularly in the very elderly who can experience a persistent level of disability

and frailty over an extended period of time, before succumbing to a minor illness due to

lack of physical reserves.19

Future research should focus on developing predictive models that will aid in estimating

potential benefits and risks of further examination of IDA and/or surgery for colon cancer

in elderly patients. Recently published data on the predictive value of the presence of

geriatric syndromes, such as functional limitations, polypharmacy, cognitive impairment,

and depressive symptoms for predicting prognosis and treatment‐related risks for older

cancer patients seem promising,22‐27 but more research is needed to clarify the exact role

of these geriatric conditions in the decision‐making process. Studies should also include

patient‐centred outcome measures, such as quality of life and functional independence, as

these are often considered of greater importance by elderly patients than survival

itself.20,21 Development of such models will allow for more precise tailoring of care to

prevent unnecessary or non‐beneficial procedures as well avoidable delays in diagnosis.

This study has several limitations. First, we have assumed that all diagnostic procedures in

these patients were performed at the Gelre hospitals. Over 95% of the patients in this

cohort had had at least one or more appointments with medical specialists based in this

103

Chapter 7

hospital, but this does not exclude the possibility that some patients were referred to

other centres for the diagnostic work‐up. A second limitation is that many different

definitions of iron‐deficiency anaemia exist, with different cut‐off values and laboratory

measurements.28,29 The composition of the patient cohort will change depending on the

definition that is used, which could alter the prevalence of the various causes of iron‐

deficiency anaemia and thus clinical outcome. This could limit the possibility to generalise

our results to other patient cohorts. A third limitation is that cause of death was

frequently unknown and we were therefore unable to calculate a cause‐specific mortality.

Finally, as this was a retrospective study, available data were limited to what was recorded

the patient’s charts. in

In conclusion, omission of endoscopy for iron‐deficiency anaemia and omission of surgery

for colon cancer occur frequently in octogenarians, and appears legitimate in the presence

of relevant comorbidity and in case of limited remaining life‐expectancy. Further research

is needed to determine which baseline factors should guide decision making to optimize

outcome and quality of life.

104


References 1. Joosten E, Pelemans W, Hiele M et al. Prevalence and causes of anaemia in a geriatric hospitalized

population. Gerontology 1992;38:111‐117. 2. Luman W, Ng KL. Audit of investigations in patients with iron deficiency anaemia. Singapore Med J

2003;44:504‐510. 3. Lindsay JO, Robinson SD, Jackson JE, Walters JR. The investigation of iron deficiency anaemia‐‐a hospital

based audit. Hepatogastroenterology 1999;46:2887‐2890. 4. Stephens MR, Hopper AN, White SR et al. Colonoscopy first for iron‐deficiency anaemia: a Numbers Needed

to Investigate approach. QJM 2006;99:389‐395. 5. Majid S, Salih M, Wasaya R, Jafri W. Predictors of gastrointestinal lesions on endoscopy in iron deficiency

anaemia without gastrointestinal symptoms. BMC Gastroenterol 2008;8:52. 6. Chatrenet P, Friocourt P, Ramain JP et al. Colonoscopy in the elderly: a study of 200 cases. Eur J Med

1993;2:411‐413. 7. Lukens FJ, Loeb DS, Machicao VI et al. Colonoscopy in octogenarians: a prospective outpatient study. Am J

Gastroenterol 2002;97:1722‐1725. 8. Avgerinos DV, Llaguna OH, Lo AY, Leitman IM. Evolving management of colonoscopic perforations. J

Gastrointest Surg 2008;12:1783‐1789. 9. Jafri SM, Arora A. Silent perforation: an iatrogenic complication of colonoscopy. Surg Laparosc Endosc

Percutan Tech 2007;17:452‐454. 10. Jafri SM, Monkemuller K, Lukens FJ. Endoscopy in the elderly: a review of the efficacy and safety of

colonoscopy, esophagogastroduodenoscopy, and endoscopic retrograde cholangiopancreatography. J Clin Gastroenterol 2010;44:161‐166.


12. Droogendijk J, Beukers R, Berendes PB et al. Screening for gastrointestinal malignancy in patients with iron deficiency anaemia by general practitioners: an observational study. Scand J Gastroenterol 2011;46:1105‐1110.

13. Damery S, Ryan R, Wilson S et al. Iron deficiency anaemia and delayed diagnosis of colorectal cancer: a retrospective cohort study. Colorectal Dis 2011;13:e53‐e60.

14. Weiser MR, Landmann RG, Kattan MW et al. Individualized prediction of colon cancer recurrence using a nomogram. J Clin Oncol 2008;26:380‐385.

15. Jonker JM, Hamaker ME, Soesan M et al. Colon cancer treatment and adherence to national guidelines: does age still matter? J Geriatr Oncol 2012;3:131‐137.

16. Gooiker GA, Dekker JW, Bastiaannet E et al. Risk Factors for Excess Mortality in the First Year After Curative Surgery for Colorectal Cancer. Ann Surg Oncol 2012.

17. Cull JD, Kukreja S, Hieken TJ. Surgical treatment of colorectal cancer in the advanced elderly: does outcome justify treatment? J.Clin.Oncol. ASCO meeting abstracts 2012:3629.

18. Christakis NA, Lamont EB. Extent and determinants of error in physicians' prognoses in terminally ill patients: prospective cohort study. West J Med 2000;172:310‐313.

19. Gill TM, Gahbauer EA, Han L, Allore HG. Trajectories of disability in the last year of life. N Engl J Med 2010;362:1173‐1180.

20. Yellen SB, Cella DF, Leslie WT. Age and clinical decision making in oncology patients. J Natl Cancer Inst 1994;86:1766‐1770.

21. Steinhauser KE, Christakis NA, Clipp EC et al. Factors considered important at the end of life by patients, family, physicians, and other care providers. JAMA 2000;284:2476‐2482.

22. Arnoldi E, Dieli M, Mangia M et al. Comprehensive geriatric assessment in elderly cancer patients: an experience in an outpatient population. Tumori 2007;93:23‐25.

23. Bamias A, Lainakis G, Kastritis E et al. Biweekly carboplatin/gemcitabine in patients with advanced urothelial cancer who are unfit for cisplatin‐based chemotherapy: report of efficacy, quality of life and geriatric assessment. Oncology 2007;73:290‐297.

24. Basso U, Tonti S, Bassi C et al. Management of Frail and Not‐Frail elderly cancer patients in a hospital‐based geriatric oncology program. Crit Rev Oncol Hematol 2008;66:163‐170.

105

Chapter 7

106

25. Tucci A, Ferrari S, Bottelli C et al. A comprehensive geriatric assessment is more effective than clinical judgment to identify elderly diffuse large cell lymphoma patients who benefit from aggressive therapy. Cancer 2009;115:4547‐4553.

26. Kristjansson SR, Nesbakken A, Jordhoy MS et al. Comprehensive geriatric assessment can predict complications in elderly patients after elective surgery for colorectal cancer: a prospective observational cohort study. Crit Rev Oncol Hematol 2010;76:208‐217.

27. Cheema FN, Abraham NS, Berger DH et al. Novel approaches to perioperative assessment and intervention may improve long‐term outcomes after colorectal cancer resection in older adults. Ann Surg 2011;253:867‐874.

28. Choi CW, Cho WR, Park KH et al. The cutoff value of serum ferritin for the diagnosis of iron deficiency in community‐residing older persons. Ann Hematol 2005;84:358‐361.

29. Beutler E, Waalen J. The definition of anaemia: what is the lower limit of normal of the blood hemoglobin concentration? Blood 2006;107:1747‐1750.

Chapter 8

Age‐related differences in guideline adherence for head and neck cancer

M.E. Hamaker, C.H. Smorenburg, R.J. Bun, G.T. de Kuyper, B.C. van Munster, S.E. de Rooij,

B.M. Wiarda, N.G. Breeuwsma, J.M. Uppelschoten

In print: Journal of Geriatric Oncology

Chapter 8

Abstract Background: The aim of this study was to examine which factors influence guideline

adherence and to determine the impact of non‐adherence on survival.

Patients and methods: Cohort of 606 patients (median age 65.3 years) newly diagnosed

with head and neck cancer at Medical Centre Alkmaar between 1997 and 2009. Treatment

was compared to guideline recommendations. Multivariable analyses were performed to

determine factors associated with non‐adherence and associated outcome.

Results: Ninety‐one percent of patients were treated in accordance with guidelines for

head and neck cancer. Reasons for discordant treatment were comorbidity, lack of

cooperation in patients aged <70 years, and patients’ refusal to undergo recommended

treatment. Age (OR 1.40 95%CI 1.04‐1.87), comorbidity (OR 1.68, 95%CI 1.32‐2.13) and

advanced disease (OR 1.83, 95%CI 1.46‐2.28) were independently associated with non‐

adherence. Five‐year overall survival was 64% for accordant treatment and 16% for

discordant treatment (p<0.001). Higher age (HR 1.03, 95%CI 1.01‐1.04), advanced stage

(HR 1.36, 95%CI 1.21‐1.53), recurrent disease (HR 3.29, 95%CI 1.97‐5.52) and treatment

discordant with guidelines (HR 3.22, 95%CI 2.15‐4.85) were independently associated with

cancer‐specific mortality.

Conclusion: Discordance with guidelines occurred in less than 10% of patients at our

centre and was associated with age, tumour stage, comorbidity and a significantly poorer

cancer‐specific survival.

108

Age‐related differences in head and neck cancer treatment

Introduction Head and neck cancers are the sixth most common malignancy and account for 6% of

cancer deaths.1 Worldwide, over 650,000 new cases of head and neck cancer are

diagnosed and 350,000 patients die of these types of malignancies every year.1 In the

Netherlands, over 3,000 new patients are diagnosed every year.2 The majority of new

cases occur in patients aged 50 years and older, primarily because of its relationship with

chronic tobacco and alcohol exposure.3,4 Approximately 25% of these patients are aged 70

and over.3,5 However, in the coming decades, the number of older patients with cancer

will increase substantially, as a result of increasing life expectancy and ageing of the

population.6

Due to the anatomical complexity of the head and neck region, and the high impact of

head and neck cancer and its treatment on quality of life, treatment decisions are

complex. Radical surgical treatment is often mutilating, while the effect of uncontrolled

disease on the patient’s speech, swallowing and breathing is devastating and often fatal.5

Although more aggressive treatment provides gains in terms of locoregional control and

disease recurrence,7 it can also have a negative impact on vital organ functions and quality

of life.8‐10 Moreover, patients with head and neck cancer often suffer from other alcohol‐

and tobacco related diseases, such as cardiac and respiratory illnesses, second primary

tumours, hepatic or metabolic diseases, as well as cognitive disorders,3,5,11 that may

decrease their resilience, quality of life and treatment tolerance.4,11‐15

As a result, older patients and those with comorbidity are often given a modified

treatment in an attempt to avoid treatment‐related complications and mortality.16‐19

Although this increases the risk of recurrent or metastatic disease in time,20 limited life

expectancy and competing causes of death could also mean that a patient does not live

long enough to suffer from this increased risk. In this case, more radical treatment would

be unnecessarily aggressive. On the other hand, there are concerns that elderly patients

are undertreated purely because of their age.21 Finding the right balance between

undertreatment and overtreatment for this vulnerable patient group presents a major

challenge and requires thorough knowledge of the potential benefit and risks of the

available treatment options.

The aim of this study was to examine which factors influence guideline adherence in

patients with head and neck cancer and to determine the impact of non‐adherence on

verall and cancer‐specific survival. o

Methods At the Medical Centre Alkmaar, a large teaching hospital, all medical charts of patients

treated for newly diagnosed head and neck cancer between 1997 and 2009 were

reviewed. Patients with a first presentation of a primary squamous cell carcinoma (SCC) of

109

Chapter 8

the oral cavity, lip, larynx, hypopharynx or oropharynx were included. Rare types of head

and neck cancer, such as salivary gland tumours, lymphomas and sarcomas were excluded

from our analysis.

The following data were collected: age at diagnosis, tumour localisation, stage of disease

(TNM), as well as the initial treatment with radiotherapy surgery and/or chemotherapy.

Comorbidity was scored using the Charlson comorbidity index;22 the current tumour was

not included in the score. Stage of disease was classified in seven groups: in situ, stage I

(T1 N0 M0), stage II (T2 N0 M0), stage III (T3 N0‐1 M0 or T1‐2 N1 M0), stage IVab (T4Nany M0,

TanyN2‐3 M0), stage IVc (TanyNany M1), and unknown. Staging was standardized for all

try. Cause of death was extracted from the data of the Central Bureau of

atistics.

for

the Cox regression

nalysis, adherence to guidelines and disease recurrence were added.

patients.

Delivered treatment modalities were compared with treatment advised by national

guidelines. In the Netherlands, treatment guidelines have been developed by the Dutch

Head and Neck Cancer Cooperative Working Group (NWHHT). A summary of these

guidelines per tumour localisation can be found in Appendix 1. Evaluation of guideline

adherence was done without knowledge of the patient’s survival status. For patients not

receiving the standard treatment, reasons stated by the treating physician for non‐

adherence to guidelines were subsequently collected from the charts; these were

subsequently classified as comorbidity, poor general condition, tumour characteristics,

patient’s refusal (if the patient chose not to have recommended treatment), and lack of

cooperation (if planned treatment could not be carried out due to patient factors, such as

failure to show up for treatment). Follow‐up data were retrieved after 1, 2 and 5 years

from the patient’s chart. For each time point, mortality was assessed, as was the presence

of disease recurrence in patients still alive. Mortality was determined using the Municipal

ata RegisD

St

Statistical analysis To assess differences between age groups and differences in guideline adherence

between various subgroups, the SPSS (Statistical Package for the Social Sciences) version

16.0 was used. The chi‐square test was used for nominal and ordinal variables. For

continuous variables with a normal distribution, the Student t‐test was used; and

continuous variables with a non‐normal distribution the Mann‐Whitney test was used.

A logistic regression analysis was performed to determine factors associated with

treatment discordance. To determine which factors were associated with mortality, a Cox

regression analysis was performed. For both, a backward selection procedure was applied,

accepting a p‐value <0.05. The following factors were included: age at diagnosis, stage of

disease, tumour localisation, Charlson comorbidity score, and sex; for

a

110


Results

65.3 years (range 23.6‐95.9 years); 220 patients

e frequently in

or prior malignancy (within

e years prior to inclusion) was observed in 9% of patients.

ent.

guidelines more often suffered from heart failure (12% vs. 2% in patients receiving

Patient and tumour characteristics Between January 1997 and December 2009, 757 new patients were seen by the

multidisciplinary head and neck cancer team. After exclusion of 153 patients (recurrent

disease or second primary tumour n=41, tumour location outside predefined area n=112),

a total of 604 patients were analysed. Patients and tumour characteristics are listed in

Table 1. Median age at diagnosis was

were aged 70 years or older (36%).

Tumour localisation varied with age; younger patients more often had oropharyngeal or

hypopharyngeal cancer, while cancer of the larynx and lip occurred mor

older patients (p<0.001). Stage of disease did not differ with age (p=0.47).

Data on comorbidity were available for 548 patients (91%). Mean Charlson index score

was 0.77 (SD 1.07) and increased over the age of 70 years. Table 2 lists the conditions

included in the Charlson index. Myocardial infarction was the most common comorbid

condition (11% of patients), followed by diabetes (9.5%), chronic pulmonary disease

.5%) and peripheral vascular disease (6.7%). A concurrent(7

fiv

Treatment choices Data on treatment was incomplete for three patients. Of the remaining 601 patients, 545

were treated in accordance with treatment guidelines (91%). Surgery was recommended

for 302 patients, and was performed in 276 (90%), but 3 patients only received local

treatment where locoregional treatment was recommended (1%). For radiotherapy, 310

out of 342 patients for whom it was recommended received full radiation treatment

(91%); 17 patients received only part of recommended dose (5%) and 20 did not receive

any radiotherapy (6%). Guidelines recommended either surgery or radiotherapy in 168

patients, of whom 60 received surgical therapy (36%), 72 radiotherapy (43%), 30 received

both (18%) and 6 had no therapy (4%). Chemotherapy was required in 97 patients

(concomitant n=73, induction n=24), and was given to 89 (92%), all of which received all

planned courses. Only 3% of patients (n=19) did not receive any form of treatm

Compliance to guidelines did not change in the course of the study period (p=0.763).

Increased age (odds ratio (OR) 1.40 95% confidence interval (CI) 1.04‐1.87 p=0.025), more

comorbidity (OR 1.68, 95%CI 1.32‐2.13 p<0.001) and more advanced disease (OR 1.83,

95%CI 1.46‐2.28, p<0.001) were independently associated with non‐adherence. This trend

was seen in all treatment modalities (Figure 1). Patients not treated in accordance with

111

Chapter 8

Table 1: Baseline characteristics per age group

All <70 years 70‐79 years 80+ years p‐value*

Number of patients 604 384 (63%) 148 (25%) 72 (12%)

% male 72.9% 74.5% 80.3% 50.0% 0.001

Mean Charlson score (SD) % without comorbidity % missing

0.77 (1.07) 55.8% 9.4%

0.60 (1.01)64.9%9.1%

1.16 (1.21)40.1%9.0%

1.03 (1.21) 40.3% 11.4%

0.001 <0.001

Tumour localisation Oropharynx Hypopharynx Lip Oral cavity Larynx

14.6% 5.8%

15.6% 34.8% 29.3%

18.8%7.6%

11.5%36.5%25.8%

10.1%2.0%

18.2%29.1%40.5%

2.8% 4.2%

31.9% 36.1% 25.0%

<0.001

Tumour stage In situ Stage I Stage II Stage III Stage IVab Stage IVc Stage unknown

1.7%

33.8% 18.8% 14.5% 27.3% 2.5% 1.3%

1.8%29.9%19.5%15.4%28.9%3.1%1.3%

2.0%40.5%14.9%14.2%24.3%2.7%1.4%

0%

40.3% 22.2% 11.1% 25.0%

0% 1.4%

0.47

* A ‐value < 0.05 is considered significant Table 2: Charlson comorbidity index

22, not including current cancer

Condition

Index factor

All patients

n=548

Treatment accordant with guidelines n=499

Treatment discordant with

guidelines n=49

p‐value**

Myocardial infarction 1 58 11% 55 11% 3 6% 0.21 Heart failure 1 14 3% 8 2% 6 12% 0.001 Peripheral vascular disease 1 36 7% 34 7% 2 4% 0.34 Cerebrovascular accident 1 23 4% 17 3% 6 12% 0.01 Dementia 1 14 3% 7 1% 7 14% <0.001 Chronic pulmonary disease 1 41 8% 37 7% 4 8% 0.51 Connective tissue disease 1 10 2% 9 2% 1 2% 0.61 Ulcer 1 10 2% 10 2% 0 0% 0.39 Chronic liver diseases or cirrhosis 1 6 1% 5 1% 1 2% 0.43 Diabetes mellitus 1 38 7% 35 7% 3 6% 0.55 Diabetes with organ complications 2 14 3% 12 2% 2 4% 0.36 Hemiplegia 2 3 1% 3 1% 0 0% 0.76 Kidney disease 2 10 2% 10 2% 0 0% 0.39 Cancer* 2 48 9% 39 8% 8 16% 0.05 Leukaemia or lymphoma 2 4 1% 3 1% 1 2% 0.31 Moderate to severe liver disease 3 6 1% 2 1% 4 8% 0.001 Metastases* 6 1 0% 1 0% 0 0% 0.91 AIDS 6 0 0% 0 0% 0 0% ‐ * not including current cancer ; ** a p‐value < 0.05 was considered significant (printed in bold)

112


Figure 1: Association between baseline characteristics and guideline adherence a. Age

*In the treatment guidelines, chemotherapy is stated as “not mandatory” for patients aged 70 and over. b. Stage of disease

c. Comorbidity

113

Chapter 8

recommended treatment, p=0.001, Table 2), a cerebrovascular accident (12% vs. 3%,

p=0.01), dementia (14% vs. 1%, p<0.001), a previous malignancy (18% vs. 9%, p=0.05) and

moderate to severe liver dysfunction (8% vs. 0%, p=0.001).

For 44 out of the 62 patients not treated in accordance with guidelines, the motivation

behind this decision was stated in the patient’s chart. Reasons mentioned were

comorbidity in 20 patients (45%), poor general condition in 12 patients (27%), lack of

cooperation in two patients (5%), tumour characteristics in one patient (2%), while 14

patients stated they did not wish to undergo the recommended treatment (32%). In

patients aged < 70 years, comorbidity and lack of cooperation were the most important

reasons for non‐adherence, while in older patients, particularly in those over 80 years old,

discordance was more often due to patient’s refusal of treatment (two out of 386 patients

aged < 70 years (1%), four of 148 patients aged 70‐79 years (3%) and seven of 72 patients

ged 80 and over (10%); p=0.022). a

Outcome Median follow‐up in this study was 3.1 years (range 0‐5 years); five year follow‐up was

available for 64% of patients. Overall survival and cancer specific survival are listed in

Table 3. For patients not receiving standard treatment, death occurred most frequently

during the first year after diagnosis (one‐year overall survival 42%) while for patients that

did receive recommended treatment, mortality was more spread out (Figure 2).

Of patients still alive, the number of patients without disease recurrence was 90% at one

year (397 of 442 patients), 87% at two years (294 of 338) and 74% at five years (120 of

161), resulting in a disease‐free survival of 74%, 65% and 51% respectively. Disease

recurrence was only associated with stage of disease (OR 1.39, 95% CI 1.14‐1.50; p=0.001).

It was not associated with guideline discordance (p=0.511).

Data on cause of death was not retrievable for five patients (2%). Cause of death was

directly related to the head and neck cancer in 122 of the remaining 228 deceased

patients (54%). Other malignancies were cause of death in 52 patients (23%) and

cardiovascular disease was the major non‐oncological contributor (27 patients, 12%).

Cancer‐specific survival, stratified for guideline adherence, is listed in Table 3. Cause of

death did not differ for patients treated in accordance with guidelines compared to those

receiving discordant treatment (p=0.787).

In the multivariable Cox regression analysis, which included sex, age at diagnosis, stage of

disease, tumour localisation, Charlson comorbidity score, adherence to guidelines and

disease recurrence, higher age (hazard ratio (HR) 1.03, 95%CI 1.01‐1.04; p=0.009), more

advanced stage (HR 1.36, 95%CI 1.21‐1.53; p<0.001), recurrent disease (HR3.29, 95%CI

1.97‐5.52; p<0.001) and treatment discordant with guidelines (HR 3.22, 95%CI 2.15‐4.85:

p<0.001) were independently associated with cancer‐specific mortality.

114


Table 3: Overall survival and cancer‐specific survival stratified, for guideline adherence

End point Overall survival*

Died of head and neck cancer

Died of other causes

Cancer specific survival*

One year 89% 7% 4% 93%

Two years 80% 13% 7% 87% Accordant treatment

Five years 64% 19% 17% 81%

One year 39% 41% 20% 59%

Two years 30% 46% 23% 54% Discordant treatment

Five years 16% 51% 33% 49% * comparison accordant vs. discordant treatment p<0.001 at all end points

Figure 2: Kaplan Meier survival plots for patients that were and those that were not treated in accordance with guidelines

115

Chapter 8

Discussion In this study, we examined adherence to treatment guidelines in a cohort of 604

consecutive patients newly diagnosed with head and neck cancer and found that

guidelines were followed in over 90% of patients. Guideline adherence decreased in

patients aged > 70 years. Non‐adherence was also associated with stage and comorbidity.

After correction for potential confounders, non‐adherence resulted in significantly poorer

overall and cancer specific survival.

This study has several limitations. Because of its retrospective nature, the usual limitations

associated with data collection and missing values apply. In particular, data on

comorbidity were limited to what was recorded in patient’s medical charts; for 10% of

patients, no data were available. For the same reason, we were unable to include data

such as functional or socio‐economic status, which could potentially have affected

guideline compliance. As patient’s charts were used as a data source, data extraction was

not entirely blinded to outcome. In addition, although the received treatment was

compared to site‐ and stage‐specific treatment guidelines, there are various details of

treatment decision making that were not incorporated in the assessment of guideline‐

compliant treatment. Another potential limitation is that the patient sample was accrued

over an extended period of time; however, no major changes were made in the Dutch

treatment guidelines in the study period. Furthermore, it is difficult to determine

retrospectively whether the decision to modify treatment in a particular patient is

justifiable.

We are aware of only two previous studies that have compared actual treatment to

guidelines with stage‐specific treatment recommendations in newly diagnosed head and

neck cancer patients. Sanabria et al. reported guideline adherence in 312 patients with

head and neck cancer aged 70 and older and found 80% was treated in accordance to

guidelines.19 Derks et al. observed a guideline adherence of 62% in 78 patients aged 70

and older and 89% in 105 patients aged 45‐60 years of age.17 Likewise, we found age to be

an independent predictor of guideline discordance, but overall adherence rates were

better (80% for patients aged > 70 years versus 94% in younger patients).

A possible explanation for decreased guideline adherence in older patients is the

increased presence of comorbidity. This is particularly true in head and neck cancer,

because its association with smoking and alcohol consumption puts patients at risk for

other lifestyle‐related diseases. As has been shown in other studies,4,17,19 comorbidity

increased with age (Table 1) and was a major factor associated with treatment

discordance, both in the multivariate analysis as well as in the reasons for deviating from

the guideline stated in the patient’s chart. In previous studies, the presence of comorbid

diseases has been found to result in more complications and decreased tolerance of

surgery,15,23‐25 radiation therapy,14 and chemotherapy.26 Therefore, adjusting treatment in

116


patients with significant comorbidity seems appropriate. Furthermore, in the first year

after diagnosis, we found that 20% of patients not treated in accordance with guidelines

died of causes other than the head and neck cancer, increasing to 33% of patients after

five years. In these patients, standard treatment could have been unnecessarily

aggressive. Despite this, cancer‐specific survival was severely compromised in patients

receiving discordant treatment (Table 3), suggesting the need for stricter guideline

adherence.

However, we also found that discordance with guidelines was often at patients’ request,

particularly in older patients: over 10% of those aged > 80 years refused the offered

treatment. Similarly, Derks et al. found that 9% of patients aged 70‐79 years and 18% of

those aged 80 years refused treatment.17 In another study, refusal increased from 2% in

patients younger than 75 years to 8% in those over 75 years.27 Refusal of treatment in

newly diagnosed older patients with cancer has been shown to be associated with

baseline functional impairment,28 or living alone.17,28 Furthermore, elderly patients often

prefer quality of life to length of life,17,29 and treatment‐related functional impairment is

feared more than death.30 As studies have shown that older patients experience poorer

outcomes for chewing,10 swallowing,31 speech10,31 and physical functioning,7,10,31 compared

to younger patients, and 32% of older patients experience incomplete recovery after one

year, this fear is not unrealistic.31

However, no studies are available comparing the quality of life and quality of death in

patients receiving discordant treatment to those receiving standard care. Future research

should focus not only on optimizing cancer‐specific survival through increasing guideline

adherence but more importantly on mapping out the short‐ and long‐term effects of head

and neck cancer and its treatment on quality of life. This information is much needed to

uide medical specialists and patients in their decision making. g

In conclusion, discordance with treatment guidelines was less than 10% at our centre, but

increased to 29% in patients aged 80 and over. It was independently associated with a

higher stage of disease and comorbidity and resulted in a significantly worse cancer‐

specific survival. Despite competing causes of death and the need to tailor treatment to

patients’ wishes, this suggests the need for greater guideline adherence. However,

research comparing short‐ and longer‐term quality of life after discordant versus standard

treatment is needed to help patients weigh the risks and benefits of treatment options.

117

Chapter 8

Appendix 1: Treatment guidelines per tumour localisation and stage of disease

Hypopharynx

Treatment recommendation Alternative treatment

Tis‐T2 n0 m0 Local radiotherapy, elective radiotherapy of the neck

T1, T2 n1 m0 Locoregional radiotherapy Locoregional surgery

T1, T2 n2 m0 Locoregional radiotherapy with neck dissection for rest tumour

T1,T2 n3 m0 Neck dissection, local radiation therapy, elective radiation of the neck

T3, T4 n0 m0 Chemoradiation* Surgery with post‐operative (chemo)radiation*

T3,T4 n1‐n3 m0 Chemoradiation* with neck dissection for rest tumour

Primary surgical therapy with post‐ operative (chemo)radiation*

Tany Nany m1 Chemoradiation*

Irresectable Chemoradiation*

* Chemotherapy is not mandatory for patients aged 70 and over.

Oropharynx


T1 n0 m0 Local surgery Local radiotherapy

T2 n0 m0 Local surgery with unilateral neck radiation Local radiotherapy with unilateral neck radiation

T1 ,T2 n1 m0 Local radiotherapy with unilateral neck radiation

T3,T4 n0 m0 Local surgery with unilateral neck dissection and locoregional neck radiation

T3,T4 n1 m0 Local surgery with local radiotherapy and bilateral neck radiation

Tany n2, n3 m0 Locoregional surgery with local radiotherapy and bilateral neck radiation

Tany Nany m1 Chemoradiation* Local radiotherapy with bilateral neck radiation

Irresectable Chemoradiation* Local radiotherapy with bilateral neck radiation


Oral cavity


Tany n0 m0 Local surgery with elective treatment of neck(radiation or surgery)

Tany n1 m0 Local surgery with unilateral neck dissection Local surgery with unilateral neck

radiotherapy

Tany n2‐n3 m0 Local surgery with bilateral neck dissection and bilateral neck radiation


Irresectable Chemoradiation*


118


Larynx


T1 n0 m0 Local radiation therapy Local resection

T2 n0 m0 Local radiation therapy and neck radiation

T1,T2 n1 m0 Local radiation therapy and neck radiation

T1,T2 n2,n3 m0 Local radiotherapy with neck dissection and elective neck radiation

T3,T4 n0 m0 (Chemo)radiation* + treatment of neck (radiation or surgery)

T3,T4 n1 m0 (Chemo)radiation* with neck radiation Local surgery, (chemo)radiation* and treatment of neck (radiation/surgery)

T3,T4 n2,n3 m0 Locoregional (chemo)radiation* with neck dissection of rest tumour



Lip


T1‐3 n0 m0 Local surgery or local radiotherapy

T4 n0 m0 Local surgery and local radiotherapy

T1‐3 n1 m0 Local surgery or local radiotherapy with neck dissection

T1‐3 n2,3 m0 Local surgery or local radiotherapy with neck dissection and neck radiation

T4 n1 M0 Local surgery, neck dissection, local radiotherapy

T4 N2,3 m0 Local surgery with neck dissection ad locoregional radiation therapy

Tany Nany M1 (Chemo)radiation*


119

Chapter 8

References 1. Parkin DM. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74‐108. 2. www.ikcnet.nl. 2010. 3. Bernardi D, Barzan L, Franchin G et al. Treatment of head and neck cancer in elderly patients: state of the

art and guidelines. Crit Rev Oncol Hematol 2005;53:71‐80. 4. Reid BC, Alberg AJ, Klassen AC et al. Comorbidity and survival of elderly head and neck carcinoma patients.

Cancer 2001;92:2109‐2116. 5. Argiris A, Karamouzis MV, Raben D, Ferris RL. Head and neck cancer. Lancet 2008;371:1695‐1709. 6. Kamangar F, Dores G, Anderson W. Patterns of cancer incidence, mortality and prevalence across five

continents. J Clin Oncol 2006;24:2137‐2150. 7. van der Schroeff MP, Derks W, Hordijk GJ, de Leeuw RJ. The effect of age on survival and quality of life in

elderly head and neck cancer patients: a long‐term prospective study. Eur Arch Otorhinolaryngol 2007;264:415‐422.

8. Derks W, de L, Jr., Hordijk GJ, Winnubst JA. Elderly patients with head and neck cancer: short‐term effects of surgical treatment on quality of life. Clin Otolaryngol Allied Sci 2003;28:399‐405.

9. Derks W, de Leeuw RJ, Hordijk GJ, Winnubst JA. Quality of life in elderly patients with head and neck cancer one year after diagnosis. Head Neck 2004;26:1045‐1052.

10. Khafif A, Posen J, Yagil Y et al. Quality of life in patients older than 75 years following major head and neck surgery. Head Neck 2007;29:932‐939.

11. Sanabria A, Carvalho AL, Vartanian JG et al. Comorbidity is a prognostic factor in elderly patients with head and neck cancer. Ann Surg Oncol 2007;14:1449‐1457.

12. Derks W, de Leeuw RJ, Hordijk GJ. Elderly patients with head and neck cancer: the influence of comorbidity on choice of therapy, complication rate, and survival. Curr Opin Otolaryngol Head Neck Surg 2005;13:92‐96.

13. Pignon T, Horiot JC, et al. No age limit for radical radiotherapy in head and neck tumours. Eur J Cancer 1996;32A:2075‐2081.

14. Horiot JC. Radiation therapy and the geriatric oncology patient. J Clin Oncol 2007;25:1930‐1935. 15. Boruk M, Chernobilsky B, Rosenfeld RM, Har‐El G. Age as a prognostic factor for complications of major

head and neck surgery. Arch Otolaryngol Head Neck Surg 2005;131:605‐609. 16. Leon X, Quer M, Agudelo D, et al. Influence of age on laryngeal carcinoma. Ann Otol Rhinol Laryngol

1998;107:164‐169. 17. Derks W, de Leeuw L, Jr., Hordijk GJ, Winnubst JA. Reasons for non‐standard treatment in elderly patients

with advanced head and neck cancer. Eur Arch Otorhinolaryngol 2005;262:21‐26. 18. Sarini J, Fournier C, Lefebvre JL et al. Head and neck squamous cell carcinoma in elderly patients: a long‐

term retrospective review of 273 cases. Arch Otolaryngol Head Neck Surg 2001;127:1089‐1092. 19. Sanabria A, Carvalho AL, Vartanian JG et al. Factors that influence treatment decision in older patients

with resectable head and neck cancer. Laryngoscope 2007;117:835‐840. 20. Lewis CM, Hessel AC, Roberts DB et al. Prereferral head and neck cancer treatment: compliance with

national comprehensive cancer network treatment guidelines. Arch Otololaryngol Head Neck Surg 2010;136:1205‐11

21. Grenman R, Chevalier D, Gregoire V et al. Treatment of head and neck cancer: European concensus at the EUFOS congress in Vienna 2007. Eur Arch Otorhinolaryngol 2010;267:1619‐1621


23. Clayman GL, Eicher SA, Sicard MW et al. Surgical outcomes in head and neck cancer patients 80 years of age and older. Head Neck 1998;20:216‐223.

24. Sanabria A, Carvalho AL, Melo RL et al. Predictive factors for complications in elderly patients who underwent head and neck oncologic surgery. Head Neck 2008;30:170‐177.

26. de Cassia Braga RK, Kowalski LP. APACHE II, POSSUM, and ASA scores and the risk of perioperative complications in patients with oral or oropharyngeal cancer. Arch Otolaryngol Head Neck Surg 2003;129:739‐745.

25. Cohen EE, Lingen MW, Vokes EE. The expanding role of systemic therapy in head and neck cancer. J Clin Oncol 2004;22:1743‐1752.

26. Piccirillo JF. Importance of comorbidity in head and neck cancer. Laryngoscope 2000;110:593‐602.

120


121

27. Kusaba R, Sakamoto K, Mori K et al. Laboratory data and treatment outcomes of head and neck tumor patients in the elderly. Auris Nasus Larynx 2001;28:161‐168.

28. Puts MT, Monette J, Girre V et al. Characteristics of older newly diagnosed cancer patients refusing cancer treatments. Support Care Cancer 2010;18:969‐974.

29. Meropol NJ, Egleston BL, Buzaglo JS et al. Cancer patient preferences for quality and length of life. Cancer 2008;113:3459‐3466.

30. Fried TR, Bradley EH, Towle VR, Allore H. Understanding the treatment preferences of seriously ill patients. N Engl J Med 2002;346:1061‐1066.

31. Derks W, De LR, Winnubst J, Hordijk GJ. Elderly patients with head and neck cancer: physical, social and psychological aspects after 1 year. Acta Otolaryngol 2004;124:509‐514.

Part III


Chapter 9

The value of a comprehensive geriatric assessment for patient care in acutely hospitalized older patients with cancer

M.E. Hamaker, B.M. Buurman, B.C. Van Munster, I.M.J.A. Kuper,

C.H. Smorenburg, S.E. de Rooij

The Oncologist 2011;16:1403‐12

Chapter 9

Abstract Introduction: A comprehensive geriatric assessment systematically collects information on

geriatric conditions and is propagated in oncology as a useful tool when assessing older

cancer patients.

Objective: 1) to study the prevalence of geriatric conditions in cancer patients aged ≥65

years, acutely admitted to a general medicine ward, 2) to determine functional decline

and mortality within twelve months after admission, and 3) to assess which geriatric

conditions and cancer‐related variables are associated with twelve‐month‐mortality.

Method: An observational cohort study of 292 cancer patients ≥65 years, acutely admitted

to the general medicine and oncology ward of two university and one secondary teaching

hospital. Baseline assessment included patient characteristics, reason for admission,

comorbidity and geriatric conditions. Follow‐up at three and twelve months was aimed at

functional decline (loss of 1 in ADL‐activities) and mortality.

Results: Median age was 74.9 years, and 95% lived independently; 126 patients (43%) had

metastatic disease. A high prevalence of geriatric conditions was found for IADL

impairment (78%), depressive symptoms (65%), pain (65%), impaired mobility (48%),

malnutrition (46%) and ADL impairment (38%).

Functional decline was observed in 8% and 33% of patients at three and twelve months,

respectively. Mortality rates were 38% at three months and 64% at twelve months.

Mortality was associated with cancer‐related factors only.

Conclusion: In these acutely hospitalized older cancer patients, mortality was only

associated with cancer‐related factors. The prevalence of geriatric conditions in this

population was high. Future research is needed to elucidate if addressing these conditions

can improve quality of life.

126

CGA in hospitalized elderly cancer patients

Introduction Although malignant tumours occur at all ages, cancer disproportionately strikes individuals

aged 65 years and older.1 Data from the National Cancer Institute Surveillance,

Epidemiology, and End Results Program reveal that over half of all newly diagnosed cancer

patients and more than two‐thirds of cancer deaths are in this age group.2 In western

societies, the number of older cancer patients will increase substantially in the coming

decades as a result of increasing life expectancy and ageing of the population. Oncologists

are faced with the challenge of how to determine what treatment is suitable for their

older cancer patients, with their heterogeneity in comorbidity, physical reserve, disability

and geriatric conditions. Age and performance status are too limited to do justice to this

diversity,3,4 and as guidelines for cancer treatment are often based on trials from which

older patients and patients with comorbidity have been excluded,5 these guidelines

cannot automatically be extrapolated to all ages.

Therefore, the use of a comprehensive geriatric assessment (CGA) – a systematic

procedure to appraise the objective health status of older people, focusing on somatic,

functional and psychosocial domains ‐ is frequently propagated in oncology as the tool to

fill in these gaps.4,6‐11 It is thought that identifying those factors associated with poor

outcome will aid in prognostication and decision making regarding treatment for the

individual patient.10,12 Furthermore, modifying the conditions identified with a CGA could

improve outcome and health‐related quality of life,7,13,14 particularly since geriatric

conditions are often missed if they are not specifically looked for.7‐9,15

Although many editorials and review articles endorse the use of CGA in geriatric oncology,

publication of evidence supporting this assumption of an added value of systematic CGA

above usual care is far less frequent.4,5,7‐10,12,14,16‐19 In addition, due to heterogeneity in

study population (inpatients vs. outpatients) and study setting (oncology vs. geriatric

medicine vs. general medicine) as well as the large variation in the extensiveness of the

CGA administered, data remain fragmentary and inconclusive on the association between

CGA and outcome.

One setting in which CGA may be of added value is for older cancer patients requiring

acute hospitalization, for example when an acute illness reveals the presence of

malignancy or due to cancer‐ or treatment‐related complications in the course of the

disease. Independent of the reason for admission, choices need to be made about future

course of treatment during hospitalization. Furthermore, acutely ill patients could be at

increased risk for geriatric conditions and functional decline.

Therefore, we studied the value of a CGA for older cancer patients acutely admitted to the

general medicine and oncology departments of three hospitals. The aim of the present

study was threefold: 1) to study the prevalence of geriatric conditions in cancer patients

aged 65 years and older acutely admitted to a general medicine ward, 2) to determine

127

Chapter 9

functional decline and mortality within twelve months after admission and 3) to assess

which geriatric conditions and cancer‐related variables are associated with twelve‐month

ortality. m

Method

Patients This is an observational substudy of cancer patients who were included in the DEFENCE‐I

and ‐II studies (Development of strategies Enabling Frail Elderly New Complications to

Evade).20,21 The DEFENCE‐I study (n=647) was conducted at the Academic Medical Centre

Amsterdam (AMC), the Netherlands; inclusion ran from November 2002 until March

2006.20 The DEFENCE‐II study (n=639) ran from April 2006 until March 2008 in AMC,

University Medical Centre Utrecht, and Spaarne Hospital Hoofddorp.21

In these prospective cohort studies, all patients aged 65 years and older, acutely admitted

to the general medicine or oncology ward were included. Patients were excluded if 1) they

or their relatives did not give informed consent, 2) they were too ill to participate

according to their attending physician, 3) they came from another ward inside or outside

the hospital, 4) they were transferred to the intensive care unit, coronary care unit or

another ward inside or outside the hospital within 48 hours after admission, or 5) they

were unable to speak or understand Dutch. Inclusion had to take place within 48 hours

after admission. The Medical Ethics Committee of the AMC approved both studies.

In this substudy, only patients with a known malignancy at the time of admission or a

alignancy first diagnosed during admission were included. m

Data collection The methods of the two studies were similar: within 48 hours of admission, a

multidisciplinary evaluation was performed by a geriatric consultation team. This team

consisted of two medical specialists, a geriatric resident, a clinical nurse specialist and two

research nurses trained in geriatric medicine.

Data on social and demographic status were collected. Patients were assessed for the

following geriatric conditions: polypharmacy, malnutrition, incontinence, falls, the ability

to perform (instrumental) activities of daily living (ADL and IADL), and cognitive

impairment two weeks prior to admission, neurosensory deficits, mobility disorders, and

delirium. Furthermore, patients in the DEFENCE‐II study were assessed for the presence of

pain, constipation, pressure ulcers, health status, depressive symptoms and caregiver

burden. Appendix 1 lists the tools used in the assessment. All variables were

dichotomized, using the cut‐offs described in this appendix.

128


Medical history and oncologic treatment prior to, during and after hospital stay were

collected from patients’ medical records by a geriatrician. Based on the treatment during

and after hospitalization, patients were subdivided in two groups: those still receiving

active anti‐cancer treatment (both curative and palliative, i.e. chemotherapy, radiotherapy

and/or surgical therapy) and those receiving supportive or symptomatic care only. The

reason for admission was collected from the discharge report and classified as directly

tumour‐related, treatment‐related or due to another cause.

These reports were also used to derive the Charlson comorbidity index,32 excluding the

current malignancy. The Charlson score is a continuous variable, with scores range from 0‐

1 with higher scores indicating more or more severe comorbidities. 3

Follow up and definition of outcomes Follow‐up consisted of a telephone interview by a research nurse at three and twelve

months after discharge, in which the modified Katz ADL index was re‐administered.

Follow‐up was completed by the same person (patient or primary caregiver) interviewed

at baseline.

Functional decline was defined as a loss of one or more ADL activities at three or twelve

months as compared to premorbid function, two weeks prior to hospital admission. Data

n mortality were collected from the Municipal Data Registry. o

Statistical Analysis Patients receiving active treatment and those receiving supportive care only after

admission were compared with one another, for differences in age, comorbidity, the

presence of geriatric syndromes, as well as for mortality and functional decline. The chi‐

square test and risk analysis were used for nominal and ordinal variables, as well as for

continuous variables with a non‐normal distribution; for continuous variables with a

normal distribution, the Student t‐test was used.

To determine which baseline factors and geriatric conditions were associated with

mortality in the twelve months following admission, a Cox regression analysis was

performed. For each variable, the Cox proportional hazards assumption was tested using

the log minus log plot. Next, a univariate Cox‐regression analysis was performed to

determine which variables were associated with mortality in the twelve months following

admission. Factors with a p‐value <0.20 in the univariate analysis and with less than 20%

missing data, were included in the multivariate analysis. A backward selection procedure

was applied, accepting a p‐value of p<0.05. Kaplan Meier survival plots, with a log‐rank

analysis, were used to determine survival in the twelve months after admission.

The SPSS (Statistical Package for the Social Sciences) version 16.0 was used for the

analyses.

129

Chapter 9

Results

Characteristics of acutely hospitalized older cancer patients In total, 1286 patients were included in the two studies, of which 208 had a known, active

malignancy (16%), and 84 were diagnosed with cancer during admission (7%). Baseline

characteristics are listed in Table 1. Median age was 74.9 years (range 65.0 – 96.2 years)

and 95% of patients lived independently. A total of 27 different types of malignancies

were present (Appendix 2). A total of 126 patients (43%) had metastatic disease.

Table 1: Baseline characteristics of patients with receiving active oncological treatment or supportive care only* (data in n(%) unless otherwise specified)

All patients

n=292 Active treatment

n=150Supportive care

n=137 p‐

value**

Hospital .65 AMC 250 (87.1) 129 (86.0) 121 (88.3) UMCU 16 (5.6) 8 (5.3) 8 (5.8) Spaarne 21 (7.3) 13 (8.7) 8 (5.8)

Study .37 DEFENCE‐I 138 (47.3) 74 (49.3) 47 (46.7) DEFENCE‐II 154 (52.7) 76 (50.7) 49 (53.3)

Median age in years (range) 74.9 (65.0–96.2) 75.5 (65.0–94.7) 73.4 (65.2–96.2) .44 65‐74 years 210 (51.2) 104 (47.3) 101 (55.5) 75‐84 years 71 (34.5) 40 (36.7) 31 (32.1) 85+ years 11 (14.3) 6 (16.0) 5 (12.4)

Female sex 141 (48.8) 80 (53.3) 60 (43.8) .11

Years of education (SD) 10.2 (3.8) 9.5 (3.5) 10.9 (4.0) .003

Living independently 275 (95.0) 141 (94.0) 129 (96.3) .54

Median length of stay in days (range) 8 (1–80) 8 (2–80) 7 (1–42) .84

Charlson score (SD) 1.1 (1.1) 1.1 (0.9) 1.0 (1.2) .55

Number of comorbid conditions (SD) 1.1 (1.1) 1.0 (1.2) 1.1 (1.0) .44

Metastases at inclusion (42.7) 58 (38.7) 64 (47.1) .14

Newly diagnosed at admission (28.2) 68(45.3) 13 (9.5) <0.001

Known cancer diagnosis (71.8) (54.7) 124 (90.5) <0.001

Reason for admission Tumor‐related*** 193 (66.1) 101 (67.3) 91 (66.4) .99 Treatment‐related **** 49 (16.8) 25 (16.7) 24 (17.5) .01 Other 50 (17.1) 23 (15.3) 22 (16.1) .82 SD = standard deviation * due to missing follow‐up, we were unable to classify treatment in five patients. ** adjusted for age and sex; p‐value is considered significant at < 0.05 (in bold). *** tumour‐related causes for admission were asymptomatic jaundice or cholangitis (n=32), haemorrhage or anaemia (n=31), aspecific malaise (n=38), obstruction (n=26), dyspnoea (n=10) or pain (n=12) **** treatment‐related causes for acute admission were chemotherapy related complications (n=33), radiotherapy related complications(n=6), due to a complication of a diagnostic procedure (n=2) or due to a complication of a therapeutic surgical procedure (n=8)

130


Prevalence of geriatric conditions A high prevalence of geriatric conditions was seen, particularly of IADL impairment (77%),

pain (65%), depressive symptoms (65%), polypharmacy (48%), mobility problems (48%),

malnutrition (46%), high caregiver burden (44%) and ADL impairment (38%) (Table 2). On

average, patients had four geriatric conditions while only 9% had no geriatric conditions

The prevalence increased with age, as patients aged 65‐69 years had 2.9 conditions on

average, patients aged 70‐79 years had 4.1 conditions, and those aged 80 and older had

5.5 conditions (p<0.001). Mean Charlson comorbidity index, excluding the current

malignancy, was 1.1 (standard deviation 1.1). Table 2: Prevalence of geriatric conditions in patients receiving active oncological treatment or supportive care only*

All patients

n=292 Active treatment

n=150Supportive care

n=137 p‐value**

Somatic geriatric conditions n (%) Polypharmacy 129/270 (48.0) 67/139 (48.2) 62/131 (47.3) .80 Malnutrition 101/220 (46.0) 47/111 (42.3) 54/111 (48.6) .30 Pain*** 81/125 (64.8) 40/61 (65.6) 41/64 (64.1) .70 Constipation*** 33/149 (22.1) 17/76 (22.4) 16/73 (21.9) .94 Incontinence 66/261 (25.2) 40/138 (29.0) 26/123 (21.1) .29 Pressure ulcers*** 2/135 (1.4) 1/69 (1.4) 1/66 (1.5) .72 Falls 32/254 (12.7) 16/135 (11.9) 16/119 (13.4) .63

Functional geriatric conditions ADL impairment at baseline 104/273 (38.1) 57/142 (40.1) 47/131 (35.9) .94 IADL impairment at baseline 191/250 (76.9) 103/129 (79.8) 88/121 (72.7) .37 Neurosensory deficits 70/268 (26.0) 32/143 (22.4) 38/125 (30.4) .07 Mobility 131/275 (47.9) 78/145 (53.8) 53/130 (40.8) .13

Psychological geriatric conditions Global cognitive impairment 31/201 (15.1) 15/98 (15.3) 16/103 (15.5) .56 Depressive symptoms*** 79/121 (65.3) 40/59 (67.8) 39/162 (62.9) .66 Delirium 61/283 (21.5) 27/148 (18.2) 34/135 (25.2) .10 Low health status score*** 33/120 (28.5) 19/59 (32.2) 14/61 (23.0) .24

Social geriatric conditions High caregiver burden*** 47/108 (43.8) 26/51 (51.0) 21/57 (36.8) .12

Total number of geriatric conditions (SD)

4.0 (2.9) 4.2 (2.9) 4.1 (2.9) .51

Presence of 1 or more geriatric conditions

91.1 91.3 90.5 .80

* due to missing follow‐up data, we were unable to classify five patients ** adjusted for age and sex; p‐value is considered significant if < 0.05. *** only included in DEFENCE‐II study

131

Chapter 9

Differences between patients receiving active vs. supportive care We were unable to classify five patients as either receiving active or supportive care due

to missing follow‐up data. Of the remaining patients, 137 (48%) received only

symptomatic or supportive care; for newly diagnosed patients, this percentage was much

lower than for patients with a known cancer diagnosis (16% vs. 60%, p<0.001). Patients

receiving supportive care only had a higher level of education, and higher Charlson

comorbidity index. Interestingly, they did not differ in age (Table 1) or the presence of

geriatric conditions (Table 2) from those receiving active care.

Table 3: Cox‐regression analysis for mortality at twelve months

Univariate analysis Multivariate analysis *

Hazard ratio


p= * Hazard ratio


p= *

Age 1.01 .99 – 1.03 .48 ‐ Sex .91 .69 – 1.21 .50 ‐ Newly diagnosed cancer 1.18 .87 – 1.61 .29 Comorbidity 1.03 .90 – 1.17 .72 ‐ Metastatic disease 1.89 1.41 – 2.52 <0.001 1.67 1.23 – 2.29 <0.001 Tumor‐related admission 1.79 1.20 – 2.47 <0.001 1.57 1.12 – 2.21 .01 Supportive care only 1.14 .85 – 1.52 .38 Polypharmacy 1.10 .81 – 1.48 .55 Incontinence 1.13 .80 – 1.60 .48 Falls .96 .60 – 1.53 .87 ADL impairment at baseline

1.45 1.08 – 1.98 .02‐

IADL impairment at baseline

1.08 .75 – 1.56 .69

Neurosensory deficits .91 .65 – 1.29 .61 Mobility 1.12 .83 – 1.50 .46 Delirium 1.23 .87 – 1.74 .24

Variables not included in analysis due to >20% missing data

Malnutrition 1.27 .91 – 1.77 .16 Pain 1.11 .95 – 1.28 .18 Constipation 1.41 .88 – 2.28 .15 Pressure ulcers 5.29 1.26 – 22.19 .02 Global cognitive impairment

1.33 .83 – 2.13 .23

Depressive symptoms 1.23 .75 – 2.02 .40 Low health status score 1.24 .75 – 2.04 .41 Caregiver burden 1.32 .82 – 2.12 .25 * Variables with a p‐value <0.20 in the univariate analysis, and with less than 20% missing data, were included in the multivariate analysis; these variables are marked in bold in the p‐value column. Age, sex and comorbidity were also included as potential confounders. A backward selection procedure was applied; a forward analysis selected the same variables.

132


Figure 1: Kaplan Meier survival plotted in relation to active treatment vs. supportive care, after correcting for stage of disease and reason for admission

Outcomes Mortality rates were 11% in hospital, 38 % at three months and 64% at twelve months. Of

patients still alive at follow‐up, 8% experienced functional decline at three months, as

determined by a decline of one or more in ADL abilities in comparison to premorbid

function, two weeks prior to hospitalization. At twelve months, 33% of patients showed

functional decline

Table 3 shows the Cox‐regression analysis for factors associated with twelve‐month‐

mortality. Only metastatic disease (HR 1.67, 95% CI 1.23‐2.29) and a tumour‐related

reason for admission (HR 1.57, 95%CI 1.12‐2.21) were independent predictors of outcome,

while age, sex, and comorbidity were not. Active oncological treatment did not give any

significant survival benefit in comparison to supportive care only (Figure 1); at three

months, mortality rate was 36.7% for patients receiving active treatment vs. 38.7% in

those receiving supportive care only. At twelve months, mortality was 62.7% vs. 65.7%

respectively.

133

Chapter 9

Because several geriatric conditions were not included in the DEFENCE‐I study, we also

performed a separate multivariate analysis using only the data of the DEFENCE‐II study.

This analysis yielded similar results for factors associated with mortality within twelve

onths. m

Discussion To our knowledge, this is the first prospective, in‐patient study addressing the prevalence

of geriatric conditions and their association with outcome in acutely hospitalized older

cancer patients. Recent observations demonstrated that several geriatric conditions were

predictive of outcome in older cancer patients in out‐patient settings.6,11,12 In the present

study, we demonstrated that in our population of 292 acutely hospitalized older cancer

patients, most of whom were living independently prior to admission, geriatric conditions

were highly prevalent, but none of the elements of the CGA were associated with

mortality.

The high prevalence of geriatric conditions we found, was also seen in a hospital‐based,

retrospective study, which reported comparable rates of functional limitations, impaired

mobility, malnutrition and depression.17 Interestingly, we found no differences in the

presence of geriatric conditions between patients receiving active care, and those

receiving supportive care only. Also, receiving active oncologic treatment did not influence

mortality rates compared to receiving supportive care only. Of course, oncologic

treatment is not only aimed at improved survival but also improved quality of life, a factor

not incorporated in this study.

In our study, 8% of cancer patients experienced an increased ADL‐dependency three

months after discharge, and over 33% at twelve months. We found no prior studies that

addressed functional decline after acute hospitalization in older cancer patients. A study

by Covinsky et al. looked at loss of ADL abilities after hospitalization for any acute medical

illness in patients aged 70 and over – without specifically investigating cancer patients.33 In

their cohort, as many as 35‐50% of older patients experienced increased disability

(defined by a loss of one or more ADL‐abilities) three months after acute hospitalization.

The difference with our findings could be explained by differences in the study population,

as the cancer patients included in our analysis were younger, had less comorbidity, less

functional impairment and less cognitive impairment at baseline, all of which are factors

potentially associated with functional decline.34

For acutely hospitalized patients, the value of CGA in predicting mortality seems to be low.

We found that none of the geriatric conditions were associated with mortality within

twelve months of admission. This outcome differs from several other studies in an out‐

patient setting addressing the association between geriatric conditions and outcome;

age,12 comorbidity,12 ADL‐ and IADL‐dependency,11 and depression,6 were found to predict

134


mortality in older cancer patients in this setting. However, in these patients, geriatric

assessment was often administered prior to the onset of oncologic treatment with

curative intent. Therefore, patients were in a different phase of disease than the patients

in our study. Most likely, for these acutely admitted older cancer patients, the presence or

absence of geriatric conditions had little further impact on outcome, probably because of

the severity of the cancer and cancer‐related symptoms: 66% was admitted for a directly

tumour‐related reason, over 43% had metastases, 48% received only supportive care and

64% died within twelve months. This could also explain why mortality was not associated

with age or comorbidity.

In patients with a poor prognosis, the goal of care generally shifts from cure to palliation.

Potentially, identifying geriatric conditions at the time of hospital admission can provide

the treating physician with leads for improving quality of life. For example, in a study

comparing geriatric care for older patients with cancer to care as usual, geriatric care

improved quality of life, although there was no difference in survival.14 Because of the

high prevalence of geriatric conditions in our study, and the fact that geriatric conditions

are easily missed if not specifically looked for,7‐9,15 assessing patients for modifiable

geriatric conditions seems appropriate when aiming to optimize quality of hospital and

palliative care for older cancer patients.

There are several limitations to this study. First of all, the study population forms a

heterogeneous group, with different types of malignancies in different stages of disease.

Furthermore, because two of the three hospitals are tertiary referral centres, there is a

potential referral bias, resulting in a less frail population as well as an overrepresentation

of upper gastro‐intestinal tract tumours – for which the Academic Medical Centre is a

national centre – compared to regional hospitals. This potentially influences the

generalizability of our findings. For example, the low cancer‐specific survival rates of the

upper GI tract malignancies could have decreased the value of the CGA for predicting

mortality. A second limitation is that this is a post‐hoc analysis of two studies, whose

designs were highly but not entirely similar. Some geriatric conditions addressed in the

DEFENCE‐II study were not included in the DEFENCE‐I study, therefore resulting in a

missing data for these items. The effect on the outcomes of this study seems low,

however, as analyses with only the data from the DEFENCE‐II study did not lead to

different results. Due to the high mortality rates at three and twelve months, we were

nable to test which factors at baseline were associated with functional decline. u

In conclusion, our study demonstrates geriatric conditions are highly prevalent in older

cancer patients admitted for an acute illness. Based on prior literature, we assume that

using a CGA for older cancer patients acutely admitted to hospital may have added value

for improving quality of life. None of the elements of the CGA were of value in predicting

135

Chapter 9

mortality, as outcome in this population was only associated with cancer‐related factors.

To elucidate the exact role of CGA, future research comparing quality of life and outcome

in patients receiving either care as usual or specific interventions aimed at modifying the

geriatric conditions identified by a CGA, is needed.

136


Appendix 1: Tools used in the assessment for geriatric conditions

* Tools marked with an asterisk were only administered in the DEFENCE‐II study

Geriatric conditions

Measurement tool Source Range of scores

Cut‐off used

Somatic geriatric conditions Polypharmacy Number of different medications medical chart Continuous ≥ 5 Malnutrition Short Nutritional Assessment

Questionnaire (SNAQ)22* and / or Body Mass Index (BMI)

Patient medical chart

0 – 7 Continuous

≥ 2 < 20

Pain Visual Analogue Scale (VAS) 23* patient** 0 – 10 ≥ 4 Constipation Constipation on physical

examination at admission* medical chart yes or no Yes

Incontinence Presence of incontinence at admission

medical chart yes or no Yes

Falls Two or more falls in the past three months

patient yes or no Yes

Pressure ulcers Observation by the research nurse* nurse yes or no Yes

Functional geriatric conditions ADL functioning Katz ADL index score24 patient/

caregiver*** 0‐6 ≥ 1 is

impairment IADL functioning Modified Katz index25 patient/

caregiver*** 0‐9 ≥ 1 is

impairment Neurosensory deficits

Impairment of hearing and/or vision*, regardless of use of glasses or hearing aid

patient yes or no yes

Mobility Requiring help or the use of a walking aid for mobility

patient yes or no Yes

Health status Euroqol (EQ‐5D)26* patient** utility list of 5 items

present if scored “severe” ≥ 1

Psychological geriatric conditions Cognitive impairment

Informant Questionnaire COgnitive DEcline – Short Form (IQCODE‐SF)27

caregiver 1 – 5 ≥ 3.9

Depressive symptoms

Geriatric Depression Scale‐2 (GDS‐2)28*

patient** 0 – 2 2

Delirium Confusion Assessment Method29 nurse 0 – 4 item 1 and 2 present plus 3 and/or 4

Social geriatric conditions Caregiver burden Experienced Burden of Informal

Care (EDIZ)30* caregiver 0 – 9 ≥ 4

** Not administered if MMSE31 was <16

*** In the DEFENCE‐I study, the Katz was filled out by the patient if MMSE31 was >23. If it was lower, the

primary caregiver filled out the Katz. In the DEFENCE‐II study, the Katz was filled out by the patient if the MMSE was >20; filled out by the patient and double‐checked with the caregiver if MMSE was 16 to 20; and filled out by the caregiver only if MMSE was <16.

137

Chapter 9

Appendix 2: Overview of cancer diagnoses

Number of patients Pancreatic cancer 36 Colon cancer 32 Oesophageal cancer 26 Cholangiocarcinoma 23 Lymphoma 21 Breast cancer 18 Lung cancer 18 Prostate cancer 16 Stomach cancer 15 Bladder cancer 14 Leukaemia 12 Multiple myeloma 10 Rectum cancer 9 Other haematological 9 Ovarian cancer 7 Renal cell carcinoma 7 Unknown origin 5 Hepatocellular carcinoma 3 Melanoma 3 Laryngeal cancer 3 Sarcoma 2 Thyroid cancer 1 Carcinoid 1 Cervical cancer 1 Total: 292

138


References 1. Muss HB. Cancer in the elderly: a societal perspective from the United States. Clin Oncol (R Coll Radiol)

2009;21:92‐8

2. Yancik R, Ries LA. Cancer in older persons. Semin Oncol 2004;31:128‐36

3. Smorenburg CH, Sijp JR. [Breast cancer in the elderly]. Ned Tijdschr Oncol 2006;3:247‐52

4. Extermann M, Overcash J, Lyman GH et al. Comorbidity and functional status are independent in older cancer patients. J Clin Oncol 1998;16:1582‐7

5. Repetto L, Fratino L, Audisio RA et al. Comprehensive geriatric assessment adds information to the Eastern Cooperative Oncology group performance status in elderly cancer patients. J Clin Oncol 2002;20:494‐502

6. Maas HA, Janssen‐Heijnen ML, Olde Rikkert MG, Wymenga AN. Comprehensive geriatric assessment and its impact in oncology. Eur J Cancer 2007;43:2161‐9

7. Extermann M, Meyer J, McGinnis M et al. A comprehensive geriatric intervention detects multiple problems in older breast cancer patients. Crit Rev Hemaol Oncol 2004;49:69‐75

8. Flood KL, Carroll MB, Le CV et al. Geriatric syndromes in elderly patients admitted to an oncology‐acute care for elders unit. J Clin Oncol 2006;24:2298‐303

9. Terrett C, Albrand G, Droz JP, on behalf of the Geriatric Oncology Program. Multidimensional geriatric assessment reveals unkown medical problems in elderly cancer patients. J Clin Oncol 2004;22(14S)abstract 8167.

10. Zagonel V, Fratino L, Piselli et al. The comprehensive geriatric assessment predicts mortality among elderly cancer patients. Proc Am Soc Clin Oncol 2002;21:abstract 1458

11. Terret C, Zulian CG, Naiem A, Albrand G. Multidisciplinary approach to the geriatric oncology patient. J Clin Oncol 2007;25:1876‐81

12. Wedding U, Rohrig B, Klippstein A et al. Age, severe comorbidity and functional impairment independently contribute to poor survival in cancer patients. J cancer Res Clin Oncol 2007;133:945‐50

13. McCorkle R, Strumph NE, Nuamah IF et al. A specialized home care intervention improves survival among older post‐surgical cancer patients. J Am Geriatr Soc 2000;48:1707‐13

14. Rao AV, Hsieh F, Feussner JR, Cohen HJ. Geriatric evaluation and management units in the care of the frail elderly cancer patient. J Gerontol A Med Sci 2005;60A:798‐803

15. Inouye SK, Foreman MD, Mion LC et al. Nurses' recognition of delirium and its symptoms: comparison of nurse and researcher ratings. Arch Intern Med 2001 Nov 12;161(20):2467‐73

16. Serraino D, Fratino L, Zagonel V, for the GIOGer study group. Prevalence of functional dysability among elderly patients with cancer. Crit Rev Oncol Hematol 2001;39:269‐73

17. Retornaz F, Seux V, Pauly V, Soubeyrand J. Geriatric assessment and care for older cancer inpatients admitted in acute care for elders unit. Crit Rev Oncol Hematol 2008;68:165‐71

18. Retornaz F, Seux V, Sourial N et al. Comparison of the health and functional status between older inpatients with and without cancer admitted to a geriatric/internal medicine unit. J Gerontol A Med Sci 2007;62A:917‐22

19. Hurria A, Lichtman SM, Gardes J et al. Identifying vulnerable older adults with cancer. J Am Geriatr Soc 2007;55:1604‐8

20. Buurman BM, Munster BC van, Korevaar JC et al. Prognostication in acutely admitted older patients by nurses and physicians. J Gen Intern Med 2008;21:20‐29

21. Buurman BM, Hoogerduijn JG, van Gemert EA et al. Systematic Comprehensive Geriatric Assessment in older hospitalised patients at high and low risk for functional decline. [submitted]

22. Kruizenga HM, Seidell JC, de Vet HC et al. Development and validation of a hospital screening tool for malnutrition: the short nutritional assessment questionnaire (SNAQ). Clin Nutr 2005;24:75‐82

23. Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity scale: what is moderate pain in millimetres? Pain 1997;72:95‐7

24. Katz S, Ford AB, Moskowits RW et al. Studies of illness in the aged. The index of ADL: a standardized measure of biological and psychological function. JAMA 1963;185:914‐9

139

Chapter 9

140

25. Weinberger M, Samsa GP, Schmader K et al. Comparing proxy and patients' perceptions of patients' functional status: results from an outpatient geriatric clinic. J Am Geriatr Soc 1992;40:585‐8

26. The EuroQol Group. EuroQol‐a new facility for the measurement of health‐related quality of life. Health Policy 1990:16;199‐208.

27. Jorm AF, Jacomb PA. The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): socio‐demographic correlates, reliability, validity and some norms. Psychol Med 1989;19:1015‐22

28. Arroll B, Khin N, Kerse N. Screening for depression in primary care with two verbally asked questions: cross sectional study. BMJ 2003;327:1144‐6

29. Inouye SK, van Dyck CH, Alessi CA et al. Clarifying confusion: the confusion assessment method. A new method for detection of delirium. Ann Intern Med 1990;113:941‐8.

30. Pot AM, van DR, Deeg DJ. [Perceived stress caused by informal caregiving. Construction of a scale]. Tijdschr Gerontol Geriatr 1995;26:214‐9

31. Folstein MF, Folstein SE, McHugh PR. "Mini‐mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189‐98.

32. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373‐83.

33. Covinsky KE, Palmer RM, Fortinsky RH et al. Loss of independence in activities of daily living in older adults hopitalized with medical illnesses. J Am Geriatr Soc 2003;51:451‐8

34. Boyd CM, Ricks M, Fried LP et al. Functional decline and recovery of activities of daily living in hospitalized, disabled older women: the Women's Health and Aging Study I. J Am Geriatr Soc 2009;57:1757‐66

Chapter 10

Baseline comprehensive geriatric assessment is associated with toxicity and survival in elderly metastatic breast cancer patients receiving single‐agent chemotherapy

Results from the OMEGA study of the Dutch Breast Cancer Trialists’ Group

M.E. Hamaker, C. Seynaeve, A.N.M. Wymenga, H. van Tinteren, J.W.R. Nortier, E. Maartense, H. de Graaf, F.E. de Jongh, J.J. Braun, M. Los, J.G. Schrama,

A.E. van Leeuwen‐Stok, S.M. de Groot, C.H. Smorenburg

[submitted]

Chapter 10

Abstract Aim: To evaluate the association between baseline comprehensive geriatric assessment

(CGA) and toxicity in elderly metastatic breast cancer patients treated with first‐line

palliative chemotherapy.

Patients and Methods: Metastatic breast cancer patients (≥ 65 years) were randomized

between pegylated liposomal doxorubicin or capecitabine. CGA included instrumental

activities of daily living (IADL), cognition using the mini mental state examination (MMSE),

mood using the geriatric depression scale (GDS), comorbidity using the Charlson index,

polypharmacy and nutritional status using the body mass index. Frailty on CGA was

defined as 1+ of the following: IADL≤13, MMSE ≤23, GDS ≥5, BMI ≤20, ≥5 medications or

Charlson ≥2. The Groningen Frailty Index (GFI) was used as a frailty screening tool (cut‐off

for frailty ≥ 4).

Results: Of the 78 patients that were randomized (median age 75.5 years, range 65.8‐86.8

years), 73 were evaluable for CGA; 52 (71%) had one or more geriatric conditions. Only

19% of patients without geriatric conditions experienced grade 3‐4 chemotherapy‐related

toxicity compared to 56% of patients with two geriatric conditions and 80% of those with

three or more (p=0.03). Polypharmacy was the only individual factor associated with

toxicity (p=0.001). GFI had a sensitivity of 69% for frailty on CGA and a specificity of 76%,

and was not predictive of outcome.

Conclusion: In this study of elderly patients with metastatic breast cancer, the number of

geriatric conditions correlated with grade 3‐4 chemotherapy‐related toxicity, suggesting

that a CGA should be incorporated as a standard of care if chemotherapy is considered for

elderly cancer patients.

142

CGA predicts toxicity in elderly metastatic breast cancer patients

Introduction Although breast cancer occurs at all ages, it disproportionally strikes those aged 65 years

and older. In 2010, 17% of the Dutch population was aged 65 years and older,1 but 39% of

newly diagnosed breast cancers and 59% of breast cancer‐related deaths occurred in this

age group.2 With the imminent ageing of society, the number of older breast cancer

patients is expected to rise greatly in the coming decades. These patients have been

underrepresented in the clinical trials on which treatment recommendations have been

based.3,4 As a result, many questions remain regarding optimal treatment for older breast

cancer patients, including whether older patients derive the same benefit from

chemotherapy as younger patients, as well as how to determine whether an older patient

will be able to tolerate such treatment. As metastatic breast cancer is still an incurable

disease, major goals of chemotherapy especially in elderly patients are relief of symptoms

and prolongation of the time period with a reasonable quality of life.

Ageing is very much an individual process, in which differences in genetic predisposition

and life‐style will become apparent, intercurrent and chronic diseases as well as geriatric

conditions leave their mark, and the speed of decline of physiologic reserves varies from

person to person.5,6 As a result, the elderly form a highly heterogeneous population. To

allow for description of health status despite this heterogeneity in contributing factors,

geriatric medicine uses the concept of frailty,7 which can be seen as the final common

pathway of ageing resulting from the cumulative decline across multiple organ systems,

causing a diminished resistance to stressors. The presence of frailty can be determined

using the comprehensive geriatric assessment (CGA), which is a systematic procedure to

assess a patient’s health status, focusing on the somatic, psychosocial and functional

domains.8 An alternative approach is to use a short frailty screening tool specifically

designed for this purpose, such as the Groningen Frailty Index.9

Although it has been demonstrated that a CGA is helpful in identifying previously

unrecognized health issues in older cancer patients,5,10‐12 the exact role of the CGA or a

frailty screening tool in decision making regarding treatment for older cancer patients still

needs to be clarified. In this study, we evaluated the value of frailty according to CGA or

the GFI, as well as individual geriatric conditions, for predicting chemotherapy‐related

toxicity and overall survival in metastatic breast cancer patients treated with first‐line

alliative chemotherapy. p

Methods The Dutch Breast Cancer Trialists Group (BOOG) OMEGA study (trial number NTR897,

BOOG 2006‐02) is a multi‐centre, randomized clinical trial regarding the tolerance and

efficacy of first line, single‐agent palliative chemotherapy for metastatic breast cancer

143

Chapter 10

patients aged 65 years or older, with incorporation of a comprehensive geriatric

assessment.

Patients were recruited between April 2007 and September 2011, by 25 Dutch hospitals.

Eligibility criteria included: a) they were female, aged 65 years or older with proven

metastatic breast cancer and eligible for first‐line palliative chemotherapy for metastatic

breast cancer; b) with European Cooperative Oncology Group (ECOG) performance status

0 to 2, or 3 (in case of pain or another pre‐existent disabling disease was determinative of

performance); and c) had acceptable bone marrow, cardiac, liver and renal functions.

Patients were excluded in case of anthracycline or capecitabine resistance (i.e.

development of recurrent or metastatic disease while on adjuvant anthracycline or

capecitabine therapy, or within 12 months of the completion of this therapy in the

adjuvant setting) evidence of uncontrolled cerebral metastases, another malignancy

within the previous 5 years, or inability of providing informed consent or complying with

regular follow‐up. All patients gave informed consent prior to inclusion.

Eligible patients were subsequently randomized between pegylated liposomal doxorubicin

(Caelyx®) 45 mg/m2 administered intravenously once every four weeks for a maximum of

six cycles, or capecitabine 2000 mg/m2 (Xeloda®), given orally on days 1 to 14, every three

weeks for a maximum of eight cycles. Toxicity was graded according to the Common

Toxicity Criteria (CTC) of the National Cancer Institute (NCI), version 3.13 In addition to

baseline data on socio‐economic factors, performance status and quality of life (using the

QLQ‐C30)14 , this study included a baseline comprehensive geriatric assessment of the

following six geriatric conditions: comorbidity as assessed by the Charlson comorbidity

index (cut‐off score ≥2),15 functional status as assessed by the Lawton & Brody scale for

instrumental activities of daily living (IADL, cut‐offs for partial dependence 14‐27, full

functional dependence ≤13),16 number of medications being used (cut‐off for

polypharmacy ≥5), nutritional status using the body mass index (BMI, cut‐off for

undernutrition ≤20 kg/m2), cognition as assessed by the mini‐mental state examination

(MMSE, cut‐off for cognitive impairment ≤23),17 and mood as assessed by the geriatric

depression scale (GDS, cut‐off for severe depressive symptoms ≥10, moderate depressive

symptoms 5‐9).18 Frailty on CGA was defined as the presence of one or more of the

following: full IADL dependence, comorbidity score ≥2, polypharmacy, cognitive

impairment, undernutrition and/or moderate to severe depressive symptoms. In addition

to the CGA, the Groningen Frailty Index (GFI)9 was used as a frailty screening tool, with a

cut‐off of 4 or more for frailty. The GFI was designed in the Netherlands as a frailty

screening instrument, and consists of 15 questions covering physical functioning, visual

and hearing impairment, cognition, medication, social support and depression.9

144


Statistical analysis All variables were categorized into two or three groups, as appropriate. To assess

differences between groups, the chi‐square test was used. A p‐value of <0.05 was

considered significant.

A univariate logistic regression analysis was performed to test for association between

baseline characteristics and grade 3‐4 chemotherapy‐related toxicity; a subsequent

multivariable analysis was omitted due to the fact that only one factor was associated in

the univariate analysis. To determine which factors were associated with mortality, a

multivariable Cox regression analysis was performed. For each variable, the Cox

proportional hazards assumption was tested using the log minus log plot, after which

factors with a p‐value <0.10 in the univariate analysis were subsequently entered in the

multivariable analysis. In addition, Kaplan Meier survival plots with a log‐rank analysis

were used.

The SPSS (Statistical Package for the Social Sciences) version 19.0 was used for all analyses.

Table 1: Baseline characteristics of patients included in the OMEGA study

n= (%)

Number of patients 73

Chemotherapy arm Doxorubicin Capecitabine

37 (51%) 36 (49%)

Age 65‐69 years 70‐74 years 75‐80 years 80+ years

17 (23%) 16 (22%) 30 (41%) 10 (14%)

Marital status

Married Widowed Single Missing

42 (57%) 22 (30%) 8 (11%) 1 (1%)

Current living situation Independent Residential care Missing

69 (96%) 3 (3%) 1 (1%)

Performance status 0 1 2+

21 (29%) 36 (49%) 16 (22%)

Quality of life (global score)14 Poor (0‐33) Average (34‐66) Good (67‐100) Missing

15 (21%) 15 (21%) 42 (58%) 1 (1%)

145

Chapter 10

Results

Baseline characteristics In total, 78 patients were randomized. The study was prematurely closed due to slow

accrual and supply problems with liposomal doxorubicin. As five patients had missing

values in one or more CGA variables; these patients were excluded from further analyses.

Of the 73 remaining patients, 37 received liposomal doxorubicin and 36 capecitabine

(Table 1). Median age was 75.9 years (range 65.8‐86.8 years). Over half of patients was

married (42 patients, 57%) and almost all patients lived independently (69 patients, 96%).

Performance status was 0‐1 in 57 patients (78%) and 2 in 15 (21%). According to QLQ‐C30,

42 patients experienced good global quality of life (58%), 15 average (21%) and 15

patients expressed poor quality of life (21%) at baseline.

Results from the geriatric assessments are listed in Table 2. Comorbidity was limited, with

64 patients having a Charlson score of 0 or 1 (88%). Most common geriatric conditions

were partial IADL dependency (65 patients, 89%), polypharmacy (37 patients, 51%), and

depressive symptoms (24 patients, 33%), while undernutrition (5 patients, 7%) and

cognitive impairment (5 patients, 7%) were less prevalent. Overall, only 21 patients had no

geriatric conditions (29%), 31 patients had one geriatric condition (42%), 16 patients had

two (22%) and five patients (7%) had three or more of the six examined geriatric

conditions. Table 2: Baseline geriatric assessment

Geriatric condition n= (%)

Charlson comorbidity index15 0‐1 64 (88%) 2+ 9 (12%)

Polypharmacy 0‐4 types of medication 36 (49%) 5+ types of medication 37 (51%)

Body Mass index (BMI) Undernutrition (BMI <20 kg/m2) Normal (BMI 20‐30 kg/m2) Overweight (BMI >30 kg/m2)

4 (5%) 64 (88%) 5 (7%)

IADL16 Fully independent (28‐30 points) 8 (11%) Partially dependent (14‐27 points) 65 (89%) Fully dependent (≤13 points) 0

Cognition (MMSE17) No cognitive dysfunction (24+ points) 68 (93%) Cognitive dysfunction (≤23 points) 5 (7%)

Mood (GDS18) No depressive symptoms (0‐4 points) 49 (67%) Mild depressive symptoms (5‐9 points) 21 (29%) Severe depressive symptoms (10+ points) 3 (4%) IADL instrumental activities of daily living, MMSE mini mental state examination, GDS geriatric depression scale

146


Based on the GFI, 41 patients (56%) were considered frail. The GFI was not able to

accurately predict outcome of CGA. Of the 52 patients deemed frail according to the CGA

(one or more geriatric conditions present), 36 were also frail based on the GFI. Of the 21

patients without geriatric conditions, 5 were considered frail based on the GFI. Thus, the

GFI had a sensitivity of 69% for frailty on CGA and a specificity of 76%, with a positive

redictive value of 88% and negative predictive value of 50%. p

Toxicity Eight patients experienced no chemotherapy‐related toxicity (11%); maximum

chemotherapy‐related toxicity was grade 1 for 14 patients (19%), grade 2 for 24 (33%),

grade 3 for 26 (36%) and grade 4 in one patient (1%). The most frequently occurring grade

3‐4 toxicities for liposomal doxorubicin were fatigue (n=5), stomatitis (n=4) and hand‐foot

syndrome (n=4). For capecitabine, the most prevalent grade 3‐4 toxicities were fatigue

(n=5) and hand‐foot syndrome (n=6). The incidence of grade 3‐4 toxicity did not differ

between the two types of chemotherapy (odds ratio (OR) for capecitabine versus

liposomal doxorubicin 0.93, 95% confidence interval (CI) 0.32‐2.67, p=1.00).

Figure 1: Association between number of geriatric conditions and percentage of patients with grade 3‐4 toxicity

147

Chapter 10

Table 3: Association between baseline factors and grade 3‐4 chemotherapy‐related toxicity

Univariate analysis* % of patients experiencing grade 3‐4 toxicity OR 95% CI p=

Chemotherapy arm Doxorubicin Capecitabine

14/37 (38%)13/36 (36%)

10.93

0.32‐2.67

1.00

Age >75 years <75 years ≥ 75 years

12/33 (36%)15/40 (38%)

11.05

0.40‐2.73

0.92

Performance status 0 1 2+

5/21 (24%)15/36 (42%)7/16 (44%)

12.252.43

0.61‐9.65

0.49‐12.98

0.25 0.29

Comorbidity 0‐1 2+

22/64 (34%)5/9 (56%)

12.36

0.46‐13.16

0.28

Polypharmacy no yes

6/36 (17%)21/37 (57%)

16.38

1.99‐23.47

0.001

Nutritional status (BMI) ≤ 20 20‐30 ≥ 30

1/4 (25%)24/64 (38%)2/5 (40%)

11.791.85

0.13‐98.32 0.06‐156.5

1.00 1.00

IADL independent dependent

2/8 (25%)25/65 (38%)

11.86

0.30‐20.24

0.70

Cognition normal impairment

25/68 (37%) 2/5 (40%)

11.14

0.09‐10.71

1.00

Depressive symptoms no mild severe

15/49 (31%)11/21 (52%)1/3 (33%)

12.431.13

0.77‐8.10

0.02‐23.31

0.11 1.0

* As only one factor demonstrated a p‐value <0.10, no multivariate analysis was performed. OR odds ratio, CI confidence interval, IADL instrumental activities of daily living, BMI body mass index

The proportion of patients that experienced grade 3‐4 toxicity increased rapidly with

increasing number of geriatric conditions (Figure 1): only 19% of patients without geriatric

conditions experienced grade 3‐4 chemotherapy‐related toxicity compared to 56% of

patients with two geriatric conditions and 80% of those with three or more (p=0.03). The

number of geriatric conditions remained predictive after correcting for potential

confounders such as age, performance status and chemotherapy arm (p=0.04).

The association between individual CGA components and grade 3‐4 chemotherapy‐related

toxicity can be found in Table 3. Polypharmacy was the only individual factor predictive of

grade 3‐4 toxicity, which occurred in 57% of patients using five or more types

of.medication compared to 17% of patients using 4 or less, resulting in an OR of 6.38

(95%CI.1.99‐23.47, p=0.002); this association was not altered by correction for potential

confounders. In addition, GFI was not associated with toxicity (OR 1.98, 95%CI 0.67‐6.13,

p=0.22). Grade 3‐4 chemotherapy related toxicity was experienced by 44% of patients GFI‐

frail and 28% of GFI‐fit patients.

148


Table 4: Multivariable Cox regression analyses for mortality, both for cumulative frailty (based on total number of geriatric conditions) and for individual geriatric conditions

Univariate analysis Multivariable analysis

HR 95% CI p= * HR 95% CI p= *

Chemotherapy arm PEGdoxorubicin capecitabine

10.85 0.50‐1.47

0.57

Age <75 years ≥ 75 years

12.04 1.16‐3.57

0.01 12.27 1.25‐4.13

0.01

Performance status 0 1 2

11.212.82

0.62‐2.341.31‐6.07

0.01 11.111.81

0.55‐2.23 0.74‐4.42

0.78 0.19

Comorbidity 0‐1 2+

11.72 0.84‐3.53

0.14

Polypharmacy no yes

11.41 0.82‐2.44

0.21

Nutritional status (BMI)

≤ 20 20‐30 ≥ 30

10630.67

0.19‐2.040.15‐3.01

0.74

IADL independent dependent

12.1 0.85‐5.43

0.10

Cognition normal impairment

13.74 1.43‐9.73

0.004 11.88 0.65‐5.45

0.24

Depressive symptoms

none mild severe

10.933.89

0.49‐1.751.41‐13.28

0.05 10.743.67

0.34‐1.64 0.88‐15.3

0.46 0.07

* Variables with a p‐value <0.10 in the univariate analysis were included in the multivariate analysis; these variables are marked in bold in the univariate p‐value column. HR hazard ratio, CI confidence interval, IADL instrumental activities of daily living, BMI body mass index

Association between frailty and survival After a median follow‐up of 32 months, 56 patients died. Median survival for fit patients

was 19.9 months versus 10.3 months for frail patients (p=0.04, Figure 2a). However, after

correction for age, performance status and chemotherapy type, this association was no

longer significant (hazard ratio (HR) 1.70, 95%CI 0.75‐3.84, p=0.2).

Of the individual geriatric conditions, both cognitive impairment (HR 3.76, 95%CI 1.43‐

9.73, p=0.004) and severe depressive symptoms (HR 3.89, 95%CI 1.14‐13.28, p=0.05) were

associated with overall mortality (Table 4), as well as age and performance status. In the

multivariable model, age over 75 years remained associated with higher mortality (HR

2.27, 95%CI 1.25‐4.13,p=0.01, Figure 2b) while severe depressive symptoms showed a

borderline significant association (HR 3.67, 95%CI 0.88‐15.3, p=0.07).

GFI was not associated with survival (HR 1.06, 95%CI 0.61‐1.82, p=0.85).

149

Chapter 10

Figure 2: Kaplan Meier survival plots for overall survival A. Fit (no geriatric conditions) versus frail patients (one or more geriatric conditions)

B. Patients younger and older than 75 years

150


Discussion In this study of elderly metastatic breast cancer patients receiving first‐line single‐agent

palliative chemotherapy, the number of geriatric conditions was a strong independent

predictor of severe chemotherapy‐related toxicity. Of individual geriatric conditions, only

polypharmacy was associated with toxicity while all other conditions were not. In this

study, the frailty screening tool GFI was not predictive of toxicity. For survival, older age

was the most important predictive factor.

Frailty has been defined as the final common pathway of ageing, caused by the cumulative

decline across multiple organ systems, and resulting in a decreased resistance to

stressors,7 such as chemotherapy. This suggests that it is the accumulation of deficits that

is determinative of a patient’s vulnerability to the development of adverse outcomes

when exposed to stressors, while the particulars of each deficit are of secondary

importance. In line with this perspective, we found that the number of geriatric conditions

was associated with toxicity but this association was seen for only one of the six individual

geriatric conditions. This confirms that it is the accumulation of deficits in geriatric

domains – irrespective of which geriatric domains – that is contributing to the vulnerability

of patients to develop toxicity.

To our knowledge only two prior studies have investigated the value of geriatric

assessments in breast cancer patients receiving palliative chemotherapy for predicting

toxicity of chemotherapy.19‐21 Similar to our study, patients (n=152) in the PELICAN

study19,20 were randomized between capecitabine and liposomal doxorubicin. In

accordance with our findings, an association between overall frailty on CGA (defined as

the presence of one or more geriatric conditions, irrespective of which conditions) and

toxicity was also found and additionally, polypharmacy was the only individual geriatric

condition associated with toxicity. The DOGMES study21 assessed the association between

individual geriatric conditions and toxicity in 60 metastatic breast cancer patients

receiving liposomal doxorubicin; this study found comorbidity – in particular the presence

of diabetes – to be associated with hematologic toxicity. Assessment of polypharmacy was

not included in their study.

No prior studies have assessed the association between overall frailty on CGA and survival

in patients with metastatic breast cancer. However, this association has been addressed in

six studies which included patients with other cancer types or heterogeneous cohorts of

patients with multiple types of malignancies, different stages of disease and various

treatment regimens.22‐28 Despite these disparities, all six studies found that frailty

diagnosed by CGA was predictive of mortality, irrespective of the underlying geriatric

condition(s) that caused the frailty. While we also found an association between frailty on

CGA and survival, this association was no longer significant after correction for age and

performance status, possibly due to the small number of patients. Two prior studies have

151

Chapter 10

assessed the predictive value of individual geriatric conditions for survival in metastatic

breast cancer patients. One study found that IADL function and comorbidity (assessed

using the cumulative illness rating scale – geriatric version (CIRS‐G)29 which takes severity

of comorbidity into account) were predictive19,20 while the other found that no particular

geriatric condition was associated with mortality.30 Neither found an association between

depressive symptoms and survival.

The predictive value of a CGA for chemotherapy‐related toxicity suggests that for optimal

decision making regarding treatment risks and potential benefits, a CGA should be

incorporated into the standard assessment of all elderly cancer patients if chemotherapy

is considered. One important barrier to such routine implementation of CGA is the fact

that performing a CGA can be time‐consuming. This has led cancer specialists to search for

possible short‐cuts, such as the use of frailty screening tools.31 If such a tool had sufficient

predictive value for relevant outcome measures, it could replace the CGA altogether. The

Groningen Frailty Index, which was used in our study, is one of the tools currently being

examined in this context. We did not find an association between GFI and outcome;

however, two other studies using GFI did find an association with survival32 and toxicity33.

Therefore, the value of GFI for predicting outcome deserves further examination. In

addition to frailty screening tools, two tools aimed specifically at predicting toxicity of

chemotherapy have been developed in the past year, using clinical, geriatric and

biochemical parameters;34,35 further studies are now undertaken to validate these tools in

other patient cohorts and will eventually elucidate their predictive value compared to a

more extensive CGA.

Frailty screening tools have also been used to select which patients should proceed with a

CGA and for which patients more elaborate assessment is unnecessary. However, the

sensitivity of GFI for the presence of geriatric conditions on CGA was only 69%, with a

specificity of 77%. With a resultant negative predictive value of around 50%, this means

that over half of patients deemed fit according to the GFI will be found to have geriatric

conditions after CGA. Therefore, GFI does not appear to be an adequate frailty screening

tool in this setting. Thus far, none of the currently available frailty screening tools has

demonstrated sufficient discriminative power to separate fit older cancer patients that are

able to receive standard cancer treatment based upon the complete treatment schedule,

from vulnerable patients that should subsequently receive a CGA to guide tailoring of their

treatment regimen.36

This study using geriatric assessments in elderly metastatic breast cancer patients treated

with palliative chemotherapy has several limitations. When the study was designed, we

followed the recommendations formulated by the International Society of Geriatric

Oncology in 2005,37 which state that a CGA for older cancer patients should consist of

assessment of cognitive function, mood and some measure of functional status, such as

152


IADL or ADL. More recently, studies using an elaborate CGA have shown that ADL and IADL

function may both be of value, as other measures such as a formal assessment of

nutritional status and mobility32,38‐41 which we did not include. Furthermore, the

discrepancy between a high prevalence of polypharmacy but low Charlson index scores

suggests that this may not be the optimal tool for quantifying comorbidity burden; in

several recent studies, the CIRS‐G – which includes a broader scope of diseases and a

measure of their severity – was shown to be a more suitable tool.20,29,42‐45 In addition, due

to difficulties in the accrual of elderly patients, the patient cohort is relatively small. On

the other hand, an important strength of our study is the homogeneity of the study

population, as only patients with a single type of malignancy and similar stage of disease

were included using a limited number of treatment regimens.

Further research is needed to confirm the value of cumulative frailty in predicting toxicity

of chemotherapy. To avoid previous issues caused by heterogeneity in study populations,

such studies will be most useful in populations with a specific type of malignancy, a similar

hase of the treatment process and a limited number of treatment regimens. p

In conclusion, this randomized study on first‐line single‐agent chemotherapy in elderly

metastatic breast cancer patients, baseline CGA demonstrated a strong predictive value

for grade 3‐4 toxicity of chemotherapy, suggesting that if chemotherapy is considered for

elderly patients, a geriatric assessment should be routinely incorporated in the

ssessment and decision‐making process. a

Financial support: the OMEGA study received unrestricted financial support from the

Dutch Cancer Society and Amgen BV the Netherlands, Janssen‐Cilag BV the Netherlands,

MSD the Netherlands.

153

Chapter 10

References 1. www.cbs.nl. 3‐2‐2012. 2. www.cijfersoverkanker.nl. 2‐3‐2012. 3. Hutchins LF, Unger JM, Crowley JJ et al. Underrepresentation of patients 65 years of age or older in cancer‐

treatment trials. N Engl J Med 1999;341:2061‐2067. 4. Peto R, Davies C, Godwin J et al. Comparisons between different polychemotherapy regimens for early

breast cancer: meta‐analyses of long‐term outcome among 100,000 women in 123 randomised trials. Lancet 2012;379:432‐444.


6. Louwman WJ, Vulto JC, Verhoeven RH et al. Clinical epidemiology of breast cancer in the elderly. Eur J Cancer 2007;43:2242‐2252.

7. Ferrucci L, Guralnik JM, Studenski S et al. Designing randomized, controlled trials aimed at preventing or delaying functional decline and disability in frail, older persons: a consensus report. J Am Geriatr Soc 2004;52:625‐634.


9. Slaets JP. Vulnerability in the elderly: frailty. Med Clin North Am 2006;90:593‐601. 10. Extermann M, Meyer J, McGinnis M et al. A comprehensive geriatric intervention detects multiple

problems in older breast cancer patients. Crit Rev Oncol Hematol 2004;49:69‐75. 11. Terret C, Albrand G, Droz JP. Multidisciplinary geriatric assessment reveals unknown medical problems in

elderly cancer patietns. J Clin Oncol 2004;22:a8167 12. Repetto L, Fratino L, Audisio RA et al l. Comprehensive geriatric assessment adds information to Eastern

Cooperative Oncology Group performance status in elderly cancer patients: an Italian Group for Geriatric Oncology Study. J Clin Oncol 2002;20:494‐502.

13. Trotti A, Colevas AD, Setser A et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13:176‐181.

14. Aaronson NK, Ahmedzai S, Bergman B et al. The European Organization for Research and Treatment of Cancer QLQ‐C30: a quality‐of‐life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365‐376.


16. Lawton MP, Brody EM. Assessment of older people: self‐maintaining and instrumental activities of daily living. Gerontologist 1969;9:179‐186.

17. Folstein MF, Folstein SE, McHugh PR. "Mini‐mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189‐198.

18. Sheikh JI, Yesavage JA, Brooks JO et al. Proposed factor structure of the Geriatric Depression Scale. Int Psychogeriatr 1991;3:23‐28.

19. De Wit M, Honecker F, Wedding U et al. Incorporation of a comprehensive geriatric assessment (CGA) into a randomized phase III trial for metastatic breast cancer (MBC): The PELICAN study. Journal of Clinical Oncology Conference: 2010;Suppl 1:28.

20. De Wit M, Harbeck N, Scholz M et al. Incorporation of a Comprehensive Geriatric Assessment (CGA) into a randomized phase III trial for metastatic breast cancer: The PELICAN Study. Journal of Clinical Oncology Conference: 2009;Suppl1:20.

21. Sostelly A, Falandry C, Pejuade‐Lauraine E et al. Predicting chemotherapy induced hematotoxicity in elderly paients with metastatic breast cancer treated by pegylated liposomal doxorubicin. J.Clin.Oncol. e19740. 2011.

22. Arnoldi E, Dieli M, Mangia M et al. Comprehensive geriatric assessment in elderly cancer patients: an experience in an outpatient population. Tumori 2007; 93(1):23‐25.

23. Bamias A, Lainakis G, Kastritis E et al. Biweekly carboplatin/gemcitabine in patients with advanced urothelial cancer who are unfit for cisplatin‐based chemotherapy: report of efficacy, quality of life and geriatric assessment. Oncology 2007; 73(5‐6):290‐297.

154


155

24. Basso U, Tonti S, Bassi C et al. Management of Frail and Not‐Frail elderly cancer patients in a hospital‐based geriatric oncology program. Crit Rev Oncol Hematol 2008; 66(2):163‐170.

25. Basso U, Falci C, Brunello A et al. Prognostic value of multidimensional geriatric assessment on survival of a prospective cohort of 880 elderly cancer patients. J.Clin.Oncol. S9065. 2011.

26. Poon D, Lee HH, Chan LL et al An exploratory analysis of comprehensive geriatric assessment results and overall survival in 233 consecutive elderly cancer patients. Journal of Clinical Oncology 2009;27 (15 SUPPL 1):9504.


28. Brunello A, Monfardini S, Falci C et al. Prognostic role of comprehensive geriatric assessment (CGA): a prospective study of a cohort of 1038 elderly cancer patients. J Geriatr Oncol 2011;Suppl 1:36.

29. Miller MD, Paradis CF, Houck PR et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res 1992;41:237‐248.

30. Freyer G, Lortholary A, Delcambre C et al. Unexpected toxicities in elderly patients treated with oral idarubicin in metastatic breast cancer: the GINECO experience. Clin Oncol (R Coll Radiol) 2004; 16:17‐23.

31. Maas HA, Janssen‐Heijnen ML, Olde Rikkert MG et al. Comprehensive geriatric assessment and its clinical impact in oncology. Eur J Cancer 2007;43:2161‐2169.

32. Aaldriks AA, Maartense E, le CS et al. Predictive value of geriatric assessment for patients older than 70 years, treated with chemotherapy. Crit Rev Oncol Hematol 2011;79:205‐212.

33. Van Fraeyenhove F, Baitar A, De VM et al. Prediction of chemotherapy toxicity by the groningen frailty index (GFI) and the comprehensive geriatric assessment (CGA) in elderly cancer patients (PTS): An interim analysis. Annals of Oncology 2010;Oct suppl:185.

34. Extermann M, Boler I, Reich RR et al. Predicting the risk of chemotherapy toxicity in older patients: The Chemotherapy Risk Assessment Scale for High‐Age Patients (CRASH) score. Cancer 2012;118:3377‐86.

35. Hurria A, Togawa K, Mohile SG et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicentre study. J Clin Oncol 2011;29:3457‐3465.

36. Hamaker ME, Jonker JM, de Rooij SE et al C. Frailty screening tools for predicting outcome of a comprehensive geriatric assessment in older cancer patients. Lancet oncology in print. 2012.


38. Biesma B, Wymenga AN, Vincent A et al. Quality of life, geriatric assessment and survival in elderly patients with non‐small‐cell lung cancer treated with carboplatin‐gemcitabine or carboplatin‐paclitaxel: NVALT‐3 a phase III study. Ann Oncol 2011;22:1520‐1527.

39. Honecker FU, Wedding U, Rettig K et al. Use of the Comprehensive Geriatric Assessment (CGA) in elderly patients (pts) with solid tumors to predict mortality. Journal of Clinical Oncology 2009;27 suppl 1: 9549.

40. Kanesvaran R, Li H, Koo KN, Poon D. Analysis of prognostic factors of comprehensive geriatric assessment and development of a clinical scoring system in elderly asian patients with cancer. J Clin Oncol 2011;29:3620‐3627.

41. Ramsdale EE, Polite N, Bylow KA et al. Relationship between components of the comprehensive geriatric assessment, chemotherapy dose intensity, and overall survival in a colorectal cancer cohort age 65 and older. J Clin Oncol 2011;S9000.

42. Kristjansson SR, Jordhoy MS, Nesbakken A et al. Which elements of a comprehensive geriatric assessment (CGA) predict post‐operative complications and early mortality after colorectal cancer surgery? Journal of Geriatric Oncology 2010;1:57‐65

43. Marinello R, Marenco D, Roglia D et al. Predictors of treatment failures during chemotherapy: A prospective study on 110 older cancer patients. Arch Gerontol Geriatr 2009;48:222‐226.

44. Soubeyran P, Rainfray M, Mathoulin‐Pelissier S et al. Screening of elderly patients with cancer for early death risk. Results of a prospective multicentric study of 364 patients under chemotherapy. Crit Rev Oncol Hematol 2006;S23.

45. Wedding U, Rohrig B, Klippstein A et al. Age, severe comorbidity and functional impairment independently contribute to poor survival in cancer patients. J Cancer Res Clin Oncol 2007; 133:945‐950.

Chapter 11

The value of geriatric assessments in predicting treatment tolerance and all‐cause

mortality in older cancer patients

M.E. Hamaker, A.G. Vos, C.H. Smorenburg, S.E. de Rooij, B.C. van Munster

The Oncologist 2012 Aug 31. [Epub ahead of print]

Chapter 11

Abstract Background: Awareness on the use geriatric assessments for older cancer patients is

increasing. The aim of this review was to summarize all available evidence on the

association between geriatric assessments and relevant oncologic outcomes.

Method: A systematic search in Medline and Embase of studies on geriatric assessment in

oncology, focusing on the association between baseline assessment and outcome.

Results: Literature search identified 2008 reports; 51 publications from 37 studies were

selected for inclusion in the review. The quality of studies was heterogeneous and

generally poor. A median of five geriatric conditions was assessed per study (interquartile

range 4‐8). Little consistency was found in the results of the studies. Furthermore,

different tools appear to be predictive depending on the outcome measure: frailty,

nutritional status and comorbidity assessed by the CIRS‐G were predictive for all‐cause

mortality; frailty was predictive for toxicity of chemotherapy; cognitive impairment and

ADL‐impairment for chemotherapy completion; and IADL impairment for perioperative

complications.

Conclusion: Although various geriatric conditions appear to be of some value in predicting

outcome in elderly cancer patients, results are too inconsistent to guide treatment

decisions. Further research is needed to elucidate the role of geriatric assessments in the

oncologic decision‐making process for these patients.

158

Review CGA and outcome in older cancer patients


aged 65 years and older,1 and the number of elderly cancer patients will increase

substantially in the coming decades as a result of increasing life expectancy and ageing of

the population. Oncologists are faced with the challenge of determining the optimal

treatment for these patients, with their heterogeneity in comorbidity, physical reserve,

disability and geriatric conditions. In this context, a myriad of editorials and review articles

have been published, endorsing the use of a comprehensive geriatric assessment (CGA) in

geriatric oncology.2‐9 A CGA is a systematic procedure used to objectively appraise the

health status of older people, focusing on somatic, functional and psychosocial domains,2

aimed at constructing a multidisciplinary treatment plan. Its value in geriatric medicine

has been proven extensively,10 but outside this field, the evidence is more scarce.

Oncology studies comparing treatment choices in patients that are considered fit or frail

on the basis of a CGA have shown that frail patients receive less intensive treatment or

receive no treatment at all.11,12 Although this shows that standard medical assessment

overlaps in part with geriatric assessment, an additional value of the latter is its ability to

identify previously unrecognised but potentially modifiable health issues, such as

depressive symptoms, cognitive or functional impairment and malnutrition.4,5,7 In

addition, some studies are now using CGA to assess patients for trial eligibility or to

allocate them to alternative treatments regimens.13,14 However, the legitimacy of such

decision‐making protocols has been insufficiently proven thus far. It remains unclear how

to translate data from the CGA to clinical practice; should geriatric assessment only be

used to classify patients as fit, vulnerable or frail, or do individual geriatric conditions have

predictive value for relevant patient outcomes?

Therefore, the aim of this systematic review is to summarize all available evidence on the

association between CGA – its individual domains as well as the summarized assessment

of vulnerability – and clinically relevant outcomes, such as all‐cause mortality,

chemotherapy toxicity, chemotherapy completion, perioperative complications and

diotherapy tolerance. ra

Methods

Search strategy and article selection Our aim was to identify cohort studies which investigated the association between

baseline geriatric assessment and outcome in cancer patients, independent of age, cancer

type or stage of disease. For this purpose, a geriatric assessment was defined as an

assessment using validated assessment tools, composed of two or more of the following

distinct domains: cognitive function, mood/depression, nutritional status, activities of

159

Chapter 11

daily living (ADL), instrumental activities of daily living (IADL), comorbidity, polypharmacy,

mobility/falls, and frailty. Studies only using non‐validated assessment tools or non‐

validated subscales of validated assessment tools were excluded. We also excluded

studies that included other patient groups in addition to cancer patients, or studies using a

treatment protocol in which the outcome of the geriatric assessment determined

treatment choice.

For outcome, the following items were defined: all‐cause mortality, toxicity of

chemotherapy, chemotherapy completion, perioperative complications, and radiotherapy

completion and toxicity.

We performed the following search in both Medline and Embase on February 15th 2012:

(("Geriatric Assessment"[Mesh]) OR (geriatric assessment*[tiab])) AND

(("Neoplasms"[Mesh]) OR (neoplasm*[tiab] OR cancer*[tiab] OR tumour[tiab] OR

tumours[tiab] OR tumour[tiab] OR tumors[tiab] OR oncolog*[tiab] OR malignan*[tiab])).

No limits in age, language or publication date were applied to the search.

In addition, conference abstracts for the 2007 to 2011 scientific meetings of the American

Society of Clinical Oncology (ASCO), European Society of Medical Oncology (ESMO),

International Society of Geriatric Oncology (SIOG), American Geriatric Society (AGS) and

European Geriatric Medicine Society (EUGMS) were hand‐searched for studies on geriatric

assessments in cancer patients to identify additional eligible studies.

The titles and abstracts of all studies retrieved by the search were assessed by one

investigator (MH) to determine which were eligible for further investigation. All potentially

relevant articles were subsequently screened as full text by two authors (MH and AV). In

case only an abstract was available, we attempted to find a final report of the study by

searching Embase and Medline using the names of first, second and/or final authors as

well as key words from the title. Also, in case of insufficient data in the original

manuscript, the authors were contacted for additional information, for example on the

tools used in the geriatric assessment.

Finally, references of included publications were cross‐referenced to retrieve any

dditional relevant citations. a

Data extraction Data regarding study design and results were independently extracted by two

investigators (MH and AV) for each eligible study. Items that were extracted were the type

of study, study setting, study population (cancer type, cancer stage, cancer treatment),

content of geriatric assessment and assessment tools used, outcome measures examined,

methods of statistical analysis, and the reported results on the association between

geriatric assessments and the outcome measures.

160


Quality assessment The methodological quality of each of the studies was independently assessed by two

reviewers (MH and AV). Disagreements among the reviewers was discussed during a

consensus meeting and in case of persisting disagreement, the assistance of a third

reviewer (BvM) was enlisted. We used a standardised list of 16 criteria to assess the

methodological quality of the included studies. This list was a modified version of the

checklist used by Kuijpers et al,15 based on the theoretical considerations and

ethodological aspects described by Altman16 (Appendix 1a). m

Data synthesis and analysis As a result of heterogeneity in study designs, diversity of patient populations and the wide

variety in content of the geriatric assessment, a formal meta‐analysis was not possible.

Therefore, we summarized the study results to describe our main outcomes of interest. If

necessary, reciprocal odds ratios or hazard ratios were calculated for optimizing

comparability of data. When applicable, subgroup summaries were made based on the

tools used in the assessment of the geriatric conditions.

Figure 1: Search results and study selection

All studies n=2008 Medline n= 893 Embase n= 962 Conference abstracts n= 153

Duplicates n=614

Exclusion n=1343 Not original research n=566 Not oncology n=340 No geriatric assessment n=148 No relevant outcome n=143 No association GA with outcome n=122 GA guided treatment n= 13 Study announcement only n= 3 Non‐cancer patients included n= 2 Assessment unclear n= 2 Non‐validated assessment tools n= 4 for individual domains

Cross referencing yielded no additional studies Inclusion: 51 publications from 37 studies

161

Chapter 11

Table 1: Characteristics of included studies

Author Publication year

Full text

Setting/department Study population

Aaldriks 17,18

2010, 2011 yes medical oncology dept various

Aparicio19,20

2011 no medical oncology/ gastroenterology dept

metastatic colon cancer

Arnoldi 21 2007 yes various various

Audisio22 2003 yes surgical oncology dept various

Bamias 23 2007 yes clinical therapeutics dept irresectable bladder cancer

Basso 24 2008 yes medical oncology ward various

Biesma 25‐27

2007, 2009, 2011 yes medical oncology dept NSCLC

Brunello 28,29

2010,2011 no medical oncology dept various

Brunello 30 2008 no medical oncology dept metastatic renal cell cancer

Castagneto 31 2004 yes oncology dept bladder cancer

De Wit 32,33

2009, 2010 no medical oncology dept metastatic breast cancer

Extermann36 2011 yes medical oncology dept various

Freyer 34 2005 yes medical oncology dept advanced ovarian cancer

Freyer 35 2004 yes medical oncology dept metastatic breast cancer

Hamaker 37 2011 yes general medicine ward various

Honecker 38 2009 no internet based registry various solid tumours

Hurria 39,40

2010, 2011 yes medical oncology dept various solid tumours

Hurria 41 2006 yes medical oncology dept breast/lung/prostate cancer

Kanesvaran 42‐44

2010, 2011, 2011 yes geriatric oncology clinic various

Karampeazis 45 2011 no medical oncology dept advanced NSCLC

Klepin46 2011 no medical oncology dept acute myologenous leukaemia

Kothari 47,48

2010, 2011 yes thoracic surgery dept various

Kristjansson 49‐51

2008, 2010, 2010 yes surgery dept colorectal cancer

Maione 52 2005 yes medical oncology dept advanced NSCLC

Marinello 53 2009 yes geriatric/oncology unit lung/colon/breast cancer

PACE‐participants 54,55

2006, 2008 yes surgical oncology dept various

Pilnik 56 2010 no medical oncology dept lung cancer

Poon 57 2009 no national cancer centre various

Ramsdale 58 2011 no oncology dept colorectal cancer

Sostelly 59 2011 no medical oncology dept metastatic breast cancer

Soubeyran 60 2006 no medical oncology dept various

Tahir 61 2010 no breast cancer clinic early breast cancer

Tredan 62 2006 yes medical oncology dept advanced ovarian cancer

Tucci 63 2009 yes medical oncology dept diffuse large cell lymphoma

v Fraeyenhove 64 2010 no medical oncology dept various

Wedding 65,66

2007, 2010 yes medical oncology ward various

Zagonel 67 2002 no medical oncology dept various

* CT chemotherapy CTRT chemoradiation ** M all‐cause mortality, tox chemotherapy toxicity,

162


Table 1: Characteristics of included studies

Types of treatment*

Number of patients

Median age in years (range)

Number of assessed conditions

Summary score used

Outcome measures examined**

CT 202 77.2 (71‐92) 3 M CC

CT 123 80 (75‐91)

4 tox CC

various 153 76 (70‐91) 5 + M

surgery 72 77 (70‐92) 5 S

CT 32 75.5 (57‐84) 4 + M

CT 117 75 (70‐92) 7 M CC

CT 182 74 (70‐87) 8 M tox CC

CT 1038 77 (70‐92) 6 + M

CT 28 73.6 (70‐81) 7 + tox

CT 25 76 (71‐87) 3 M

CT 152 61 (22‐85) 4 + M tox CC

CT 518 75.5 (70‐92) 6 tox

CT 83 76 (70‐90) 3 M tox

CT 26 70+ 6 M

various 292 74.9 (65‐96) 8 M

various 1130 76.3 (69‐95) 7 M

CT 500 73 (65‐91) 5 tox

CT 20 75 (66‐84) 4 tox

unclear 249 77 (70‐94) 8 M

CT 131 74 (65‐92) 5 tox

CT 74 70 (±6.2) 4 M

surgery 60 76 (?) 5 S

surgery 182 80 (70‐94) 7 + M

CT 566 74 (70‐84) 3 M

CT 110 75 (70‐87) 4 M tox CC

surgery 460 76.9 (70‐95) 5 S

CT/CTRT 130 ? 4 tox

various 233 77 (70‐93) 7 + M

CT 38 72 (65‐89) 5 M CC

CT 60 ? 4 tox

CT 364 77.5 (70‐99) 7 M

unclear 124 82 (70‐94) 5 M

CT 155 75.5 (70‐90) 5 M

CT 84 73 (66‐89) 3 + M

CT 21 71.2 (66‐86) 8 + tox

CT 427 ?(18‐80+) 3 M

various 252 72 (65‐94) 2 M

CC chemotherapy completion, S surgery NSCLC = non‐small cell lung cancer

163

Chapter 11

Results

Study characteristics The literature search identified 1855 citations (893 from Medline and 962 from Embase),

of which 61 were duplicates. Hand‐searching of conference abstracts yielded another 153

potentially relevant publications. Details on the search and reasons for exclusion can be

found in Figure 1. After exclusion of 1343 publications, 51 publications from 37 studies

were included in this review.17‐67 Cross‐referencing yielded no additional results.

The characteristics of these 37 studies are summarized in Table 1. The first publication is

from 2002, but more than half of the studies were published in the last two years.17‐67 All

but one study consisted of prospective cohorts.24 The median sample size was 152

patients (range 20‐1130 patients).17‐67 Study populations were heterogeneous, with only

half focusing on a specific type of cancer,19,20,23,25‐27,30‐35,45,46,49‐52,56,58,59,61‐63 of which eleven

also focused on a specific cancer stage (30% of all studies).19,20,23,30,32‐35,45,52,59,61‐63

Furthermore, while 26 studies focused on chemotherapy (70%)17‐20,23‐36,39‐41,45,46,52,53,56,58‐

60,62‐66 and four studies focused on surgery (11%),22,47‐51,54,55 seven studies included

patients receiving multiple treatment modalities (19%).21,37,38,42‐44,57,61,67

The median number of geriatric conditions that were assessed was five (interquartile

range 4 – 8, Table 1).17‐67 Table 2 gives an overview of the geriatric conditions included in

the studies and the method of assessment,68‐93 a more detailed overview per study can be

found in Appendix 2. Ten studies summarized results of geriatric assessment in a summary

score (27%).21,23,24,28‐30,32,33,49‐51,57,63,64 Two used the cumulative number of geriatric

conditions57,64 as a summary and eight defined patients as frail if they were ADL‐

dependent, had 3 or more comorbidities (or one severe comorbid conditions) or one or

more geriatric conditions.21,23,24,28‐30,32,33,49‐51,63

The association between geriatric assessment and all‐cause mortality was assessed in 25

out of 37 studies (68%, Table 1),17,18,21,23‐29,31‐35,37,38,42‐44,46,49‐53,57,58,60‐67 chemotherapy

toxicity in 13 (35%)19,20,25‐27,30‐36,39‐41,45,53,56,59,64 and chemotherapy completion in seven

(19%).17‐20,24‐27,32,33,53,58 Four studies focused on the association between geriatric

assessment and perioperative complications (11%).22,47‐51,54,55 No studies were found on

eriatric assessment in relation to radiotherapy. g

Study quality The quality of the studies was heterogeneous, with a median score of 9 out of the sixteen

items on the quality checklist (interquartile range 7‐11). Reviewer agreement was >95%

for all aspects. In‐ and exclusion criteria and patient population were clearly described in

22 and 26 studies, respectively. The participation rate ‐ i.e. the percentage of potential

164


Table 2: Content of geriatric assessments in included studies

Condition No. of studies assessing condition (%)

Assessment tool used for assessing condition

No. of studies using tool

Instrumental activities of daily living (IADL)

32 (86%) Lawton & Brody68 NEADL69 PAT‐D93

30 1 1

Comorbidity 32 (86%) Charlson70 CIRS‐G71 Satariano72 No. of conditions

12*

12 2 7

Activities of daily living (ADL) 31 (84%) Barthel73 Katz 74 OARS76 PAT‐D

93 Unclear

6 22 1 1 1

Cognition 26 (70%) MMSE75 IQcode77 Blessed

78 SPMSQ79

23* 2 1 1

Mood/depression 24 (65%) GDS80 HADS

81 PANAS82 SCID84 CES‐D

91

18* 4 1 1 1

Polypharmacy 13 (35%) No. of pills 12

Nutritional status 9 (24%) MNA83 Determine85 SNAQ86 NHI NHC

**

6 1 1 1

Mobility 9 (24%) TUG90 SPPB92

7 2

Frailty screening 6 (16%) GFI87 VES‐1388 Fried89

3 2 1

* some studies used more than one tool to assess the domain ** no reference available NEADL Nottingham extended activities of daily living, PAT‐D Pepper Assessment Tool for Disability, CIRS‐G cumulative illness rating scale‐geriatrics, OARS Older American Resources and Services, MMSE mini mental state examination, IQCODE informant questionnaire on cognitive decline in the elderly, SPMSQ short portable mental status questionnaire, GDS geriatric depression scale, HADS hospital anxiety and depression scale, SCID structured clinical interview for DSM IV CES‐D Centre for Epidemiologic Studies‐Depression scale MNA mini nutritional assessment, SNAQ short nutritional assessment questionnaire, NHI NHC national health initiative nutritional health checklist, TUG timed get‐up‐and‐go SPPB short physical performance battery, GFI Groningen frailty index, VES‐13 vulnerable elders scale‐13

165

Chapter 11

participants that received a geriatric assessment ‐ was only described in nine studies and

although 21 studies listed the duration of follow‐up, only eight described the number of

patients lost to follow‐up or compared completers with non‐completers. Only thirteen

studies described the completeness of data. Outcome reporting was of better quality, with

28 studies providing data from univariable analyses for the association of geriatric

assessments with outcome measures and 24 presenting some form of prognostic model.

However, reporting of associations was differed notably between studies, with some

presenting odds ratios, others hazard ratios and others only p‐values to indicate a

statistical significance without reporting on the actual odds/hazard ratio or confidence

interval. This complicated any comparison of data and hindered combining data for a

formal meta‐analysis. Furthermore, three studies did not appear to have sufficient

numbers for their multivariable analyses. Full results for the quality assessment can be

und in Appendix 1b. fo

All‐cause mortality The predictive value of geriatric assessments for all‐cause mortality was reported in 25

studies (Table 3). Six studies addressed the association of a summary score with mortality:

all six found that frail patients showed poorer overall survival (100%).21,23,24,28,29,57,63 In

these studies, median survival was between 1.6 and 3.7 times longer for fit patients

compared to frail subjects. Likewise, frailty assessed with a formal frailty screening tool

was found to be associated with mortality, in three out of four studies (75%).17,18,23,25,58

Nutritional status was found to be associated with mortality in all four studies in which it

was assessed (100%).17,18,42,50,58,60 For comorbidity, initial analysis revealed that only six out

of sixteen studies found an association with mortality (38%). However, when subdividing

according to the assessment method used, only one out of five studies using the Charlson

comorbidity index38 and none of the four studies using the number of comorbid conditions

found an association, while four out of five studies using the CIRS‐G found comorbidity

was associated with mortality (80%).33,50,53,60,65 For one study, the results for comorbidity

were not clearly reported. Of the 14 studies addressing cognitive function, only two found

an association between cognition and mortality.38,61 Only four out of fourteen studies

found an association between ADL‐impairment and mortality (29%),25,38,46,61,67 and six out

of sixteen reported finding an association for IADL‐impairment (38%).25,33,38,46,52,67 Results

for mood/depression, mobility and polypharmacy were inconclusive with approximately

equal numbers of studies that did and did not find an association. All of these results were

not altered when correcting for the assessment tool that was used.

166


Table 3: Association of geriatric assessment with all‐cause mortality

Author

Number of

patients

Can

cer type

Summary score

Cognition

Mood/depression

Mobility

ADL

IADL

Nutritional status

Frailty screening

Comorbidity

Polypharmacy

De Wit 32,33

152 Breast + + + + – –

Freyer 35 26 Breast – – – – – –

Tahir 61 124 Breast + + – – + + – – – –


182 Colorectal – – – – – – – – + + + + – –

Ramsdale 58 38 Colorectal – – – – – – + + – –

Biesma 25‐27

182 Lung – – + + + + + + + + + ?

Maione 52 566 Lung – – + + – –

Tucci 63 84 Lymphoma +

Klepin46 74 AML – – ++ – – ++

Bamias 23 32 Bladder + –

Castagneto 31 25 Bladder – – –

Freyer 34 83 Ovarian – – – – + +

Tredan 62 151 Ovarian – – + + – – ± – –

Aaldriks 17,18

202 Various – – + + + +

Arnoldi 21 153 Various +

Basso 24 117 Various +

Brunello28 29 1038 Various + +

Hamaker 37 292 Various – – – – – – – – – – –

Honecker 38 1130 Various + + + + + + +

Kanesvaran 42‐44

249 Various – – + + – – – – + + – – – –

Marinello 53 110 Various – – – – – – + +

Poon 57 233 Various + + ±

Soubeyran 60 364 Various – – – – – – – – – – + + – –

Wedding 65,66

437 Various – – ± + +

Zagonel 67 252 Various + + + +

++ / ‐ ‐ : significant / no association in multivariable analysis, respectively. Of note, there was little uniformity across studies in the confounders and variables included in these analysis + / ‐: significant / no association in univariable analysis, respectively, no multivariable analysis performed or factor not included in multivariable analysis ?: association not described in the publication ±: association only present in subgroup of patients but not all patients (I)ADL: (instrumental) activities of daily living AML: acute myologenous leukaemia

167

Chapter 11

Toxicity of chemotherapy Results for toxicity of chemotherapy and chemotherapy completion are listed in Table 4.

For toxicity, the score summarizing geriatric assessment was found to be associated with

toxicity of chemotherapy in two out of three studies (66%), but these only reported

univariable results.33,64 Similarly, two out of three studies found an association between

toxicity and a frailty screening tool.56,64 For all other geriatric conditions, results were quite

variable across studies. Polypharmacy was associated with toxicity in two out of four.33,56

Comorbidity was associated with toxicity of chemotherapy in only three out of ten

studies.53,56,59 Method of assessing for comorbidity did not influence results. Toxicity of

chemotherapy was associated with impaired cognition in 17% of studies, depressed mood

in 13%, impaired mobility in 33%, ADL impairment in 0%, and IADL‐impairment in 18%,

spectively (Table 4). re

Chemotherapy completion For completion of chemotherapy, impaired cognitive function was found to be associated

with less completion or the need for dose reduction in two out of three studies (66%,

Table 4).17‐20 ADL‐impairment showed similar results (association in two out of three

studies).25,58 Furthermore, three out of five studies.33,53,58 found that comorbidity was

predictive of lower completion rates (60%). Two of these used the CIRS‐G and one the

Charlson comorbidity index to assess comorbidity; the two studies that did not find an

association both used the Charlson comorbidity index. One study addressed nutritional

status, and found an association in their multivariable analysis.17,18 Data was inconclusive

for the summary score (association in one of two studies, 50%), and negative for

epressed mood, impaired mobility, IADL impairment and the presence of frailty (Table 4). d

Perioperative complications Four studies addressed the association between geriatric assessment and perioperative

complications (Table 4). Only one study assessed the association of a summary score and

found it to be associated with perioperative complications (100%).49 This association was

found for IADL‐impairment in three out of the four studies.47,49,54 For depressed mood,

results were inconclusive with only two out of four studies finding an association.47,49 For

ADL‐impairment, polypharmacy, nutritional status, cognitive function and comorbidity, no

r little association was found. None of the studies used a frailty screening tool. o

Radiotherapy toxicity/completion No studies were identified that addressed the association between geriatric assessments

and toxicity or completion of radiotherapy. One study assessed patients receiving

chemotherapy or chemoradiation, but did not report separately on the latter group.56

168


Table 4: Association of geriatric assessment with treatment complications/completion

Outcome

Author

Can

cer type

Number of

patients

Summary score

Cognition

Mood/depression

Mobility

ADL

IADL

Nutritional status

Frailty screening

Comorbidity

Polypharmacy

De Wit 32,33

Breast 152 + – – – +

Sostelly 59 Breast 60 – – – – – – + + + +

Aparicio 19,20

Colorectal 123 + + – – + + – –

Biesma 25‐27

Lung 182 + + – – – – – – – – – –

Karampeazis 45 Lung 131 – – – – –

Pilnik 56 Lung 130 – – + +

Brunello 30 Renal cell 28 –

Freyer 34 Ovarian 83 – – – – – –

Castagneto 31 Bladder 25 – – –

Extermann36 Various 518 ± – – ± ± – – – –

Hurria 39,40

Various 500 – – – – + + + + ?

Hurria 41 Various 20 – – – –

Marinello 53 Various 110 – – – – – – + +

CHEM

OTHER

APY

TOXICITY

Van Fraeyenhove 64 Various 21 + +

De Wit 32,33

Breast 152 – – – + –

Aparicio 19,20

Colorectal 123 + + – – – – – –

Ramsdale 58 Colorectal 38 – – + + + + ? – – + +

Biesma 25‐27

Lung 182 – – – – + + + + – – – –

Aaldriks 17,18

Various 202 + + + + – –

Basso 24 Various 117 + C

HEM

OTHER

APY

COMPLETION

Marinello 53 Various 110 – – – – – – + +


Colorectal 182 + + – – + + – – + + – – + + – –

Audisio 22 Various 72 – – + – –

Kothari 47,48

Various 60 + – + –

PER

IOPER

ATIVE

COMPLICATIONS

PACE‐participants 54

Various

460

– –

– –

– –

+ +

– –

++ / ‐ ‐ : significant / no association in multivariable analysis, respectively. Of note, there was little uniformity across studies in the confounders and variables included in these analysis + / ‐: significant / no association in univariable analysis, respectively; no multivariable analysis performed or factor not included in multivariable analysis ?: association not described in the publication ±: association only present in for a particular type of toxicity but not all toxicity (I)ADL: (instrumental) activities of daily living

169

Chapter 11

Discussion In this review on the value of geriatric assessments in predicting treatment tolerance and

all‐cause mortality in older cancer patients, little consistency was found between the

results of the various studies. Interestingly, different geriatric conditions appear to be

predictive for the primary outcome measures: frailty, nutritional status and comorbidity

(when measured with CIRS‐G) for all‐cause mortality; frailty for toxicity of chemotherapy;

cognitive function and ADL‐impairment for chemotherapy completion; and IADL‐

impairment for perioperative complications. However, the only truly consistent finding

was the association between a summary score of the geriatric assessment and mortality.

The studies included in this systematic review were heterogeneous in design, content and

reported outcomes. In addition, reporting was frequently too inadequate to assess

potential sources of bias. It was often unclear whether outcome of geriatric assessment

was known to the treating physician, allowing differences in overall survival to be caused

by the reception of suboptimal oncologic treatment based on the outcome of geriatric

assessment (and subsequently the assumption that patients would not be able to tolerate

standard treatment). Another potential bias is that the patients participating in studies

focusing on chemotherapy and surgery were already preselected as suitable for this

treatment by their physician. Thus, while many geriatric conditions were not predictive of

toxicity, one cannot conclude that patients should receive chemotherapy irrespective of –

for example – their cognitive status or IADL‐score.

These factors limited our possibilities of performing a formal meta‐analysis and drawing

definitive conclusions. One method to solve some of these issues would be to perform an

individual patient data analysis using the original data of included studies. A second

limitation of this review is that it focuses on studies assessing multiple geriatric conditions.

Studies focusing on single conditions or including multiple conditions but not identified as

geriatric assessment would not have been selected from Medline or Embase with our

search strategy. Despite these limitations, this review does provide a thorough overview

of the currently available evidence on the value of geriatric assessment for predicting

relevant outcomes in older cancer patients.

The results of this review have several clinical implications. First of all, although various

geriatric conditions appear to have some predictive value for each of the four outcome

measures, the lack of consistency in the findings does not support excluding patients from

certain treatment options based solely on their score on a geriatric assessment tool.

A second clinical implication of this review is that while current geriatric assessments used

in oncology primarily focus on cognitive function, mood/depression, and functional

limitations, less frequently examined geriatric conditions, such as malnutrition and

polypharmacy appear to be of similar or even greater predictive value and are potentially

modifiable; therefore, their assessment should not be omitted. Also, it appears that

170


assessment of comorbidity without including a measure for the severity of these

conditions is not useful, and therefore, we recommend using the CIRS‐G rather that the

Charlson comorbidity index, despite that fact that the former is more time‐consuming.71,94

Interestingly, assessment of mobility – which is one of the cornerstones of geriatric

medicine – is rarely included. Given its predictive value in the general geriatric population,

this element of CGA deserves further exploration.95

Various factors may have contributed to the variation in the results of the included

studies. First of all, it appears that the choice of assessment tool influences outcome, as

illustrated by the fact that comorbidity assessed by CIRS‐G was associated with mortality

but when assessed with the Charlson, it was not. Heterogeneity in patient populations will

also have contributed to the variation in study outcomes: not only do different elements

of the CGA appear to be predictive depending on the outcome measure that is examined,

but it is possible that the specific characteristics and prognosis of a malignancy will also

affect the predictive value of various geriatric syndromes. All these factors mean that

finding that one optimal assessment tool that will be predictive of all outcome measures

in all patient populations and all treatment settings may not be feasible.

On the other hand, the results of our systematic review suggest that in predicting

outcome, it may be more important to determine whether or not a patient is frail than to

determine what makes him or her frail. This fits with the definition of frailty as the final

common pathway of ageing,96 where the presence of deficits in geriatric domains is the

determinative factor while the particulars of each deficit are of secondary importance. If

this be the case, a short frailty screening tool could potentially suffice in allocating a

patient to standard treatment of tailored care, and the time‐consuming process of a

formal geriatric assessment could be avoided.40,47,97 This does require that this tool has a

high sensitivity for frailty, allowing the assessor to trust that those patients deemed fit,

actually are fit.88 Those who are not, should then receive further assessment to ascertain

their ability to tolerate treatment. However, there is still much debate on the precise

definition of frailty and how it should be measured, and as yet, there is insufficient

evidence on the quality of the various screening tools in predicting fitness in this particular

setting to endorse one tool over the others.98

Ultimately, in limiting the use of a systematic geriatric assessment in oncology practice to

a decision‐making tool, the potential benefit of using the CGA to optimize care for elderly

patients with cancer is overlooked. For example, although a cognitive disorder does not

necessarily predict chemotherapy toxicity, it potentially means that a patient may not

respond adequately in case of complications, or will not take oncologic or supportive

medication as prescribed; these patients may require extra monitoring or home health

care. Similarly, addressing previously undiagnosed depressive symptoms or malnutrition

can improve a patient's resilience when undergoing treatment. A geriatric assessment

171

Chapter 11

could thus be seen as a starting point for further treatment and care, for improving not

only the outcomes addressed in this review, but also quality of life or functional capacity.

However, as a formal comprehensive geriatric assessment is time‐consuming, and a

geriatric consultation is often a scarce commodity, it may be useful to develop screening

tools that are particularly suitable for finding those patients at high risk for having geriatric

onditions that are modifiable or require intervention.99 c

In conclusion, this systematic review shows that although different geriatric conditions

appear to be predictive for each of the major outcome measures, currently available

evidence is too inconsistent to guide clinical decision making. Many questions remain

unanswered and will require further exploration. To elucidate the impact of the various

geriatric conditions on treatment tolerance and outcome for older cancer patients, future

clinical research should use broad geriatric assessments that address all geriatric

conditions, and include ‘geriatric’ outcome measures, such as functional capacity, in

addition to standard oncologic outcomes. Furthermore, research should focus on

validating screening tools that predict fitness rather than frailty, and on applying geriatric

assessment as an intervention aimed at optimizing a patient's resilience during treatment,

rather than as a decision‐making tool only.

172


Appendix 1a: Criteria for quality assessment, based on Kuijpers et al15 and Altman

16

Criterion Score* Positive (+) if

Study population

A Inception cohort + – ? moment of inclusion is related to same point in the course of disease or treatment

B Descriptions of inclusion and exclusion criteria

+ ? clearly described

C Description of study population + ? setting, age and sex, cancer type and cancer treatment patients is described

Participation

D Participation rate + – ? less than 10% of potential patients did not receive geriatric assessment

E Information about participants vs. non‐participants

+ – description of non‐participants/reason for not participating is available

Follow‐up

F Prospective data collection + – ? prospective data collection or a historical cohort in which the determinants have been measured before the outcome was determined

G Duration + – length of follow‐up has been described

H Loss‐to‐follow‐up + – ? loss to follow‐up is less than 10%

I Information of completers vs. loss to follow up

+ – ? description of those patients lost to follow‐up is available

Geriatric assessment

J Standardized assessment of geriatric conditions

+ ? clear description of tools used for geriatric assessment and, if applicable, cut‐off values described

K Completeness of geriatric assessment data

+ – ? less than 10% missing data

Outcome

L Treatment decision blinded to outcome of geriatric assessment

+ – ? blinding process is clearly described

M Standardized assessment of relevant outcome criteria

+ – ? definitions of the used outcome measures are stated or standard definitions are used

N Outcome assessment blinded for baseline geriatric assessment

+ – ? blinding of outcome assessor for baseline geriatric assessment is mentioned

Data presentation

O Appropriate analysis + – ? univariable crude estimates are provided for the association of CGA (components) with outcome measure

P Prognostic model presented + – an attempt is made to determine a set of prognostic markers with the highest prognostic value

Q Sufficient numbers + – number of cases in the multivariable analysis is at least ten times the number of independent variables in the analysis

* evaluation is positive ( + ) if sufficient information is available and criteria in Appendix 1a are met; negative ( – ) if sufficient information is available but there is potential for bias due to inadequate design or conduct; ? signifies insufficient information for evaluating criterion. na = not applicable

173

Chapter 11

Appendix 1b: Quality assessment of included studies

Publication Study population Participation Follow‐up Geriatric assessment

Outcome Data presentation

Author

Inception cohort

Description of in–

and exclusion

Description of

patient population

Participation rate

Data on

participants vs.

Prospective data

collection

Description of

follow–up duration

Loss to follow–up

Data on completers

vs. lost to follow‐up

Standardized

assessmen

t of

geriatric conditions

Completeness of

data collection

Treatm

ent blinded

for GA outcome

Standardization of

outcome measures

Outcome

assessmen

t blinded

Appropriate

li

Prognostic m

odel

presented

Sufficient numbers

Aaldriks 17,18 + + + ? – + + ? – + + ? + ? + + +

Aparicio 19,20 + ? + – – + + ? – + ? ? + ? – + +

Arnoldi 21 – ? ? ? – + + ? – + ? ? + na + + +

Audisio 22 + + + + – + + + – + + ? + + + – na

PACE‐participants 54 + + + ? – + + ? – + ? + + ? + + +

Bamias 23 + + + + + + + ? – + ? + + ? + – na

Basso24 + + + + + – + ? – + ? ? + ? + – na

Biesma 25‐27 + + + + + + + ? – + + + + ? + + +

Brunello 28,29 – ? + ? – + + ? – + + ? + ? + + +

Brunello 30 + ? ? ? – + – ? – + + ? – ? ? – na

Castagento 31 + + + + – + + + + + + ? + ? – – na

De Wit 32,33 + ? + – – + – ? – ? ? + ? ? + + +

Extermann36 + + + + + + + + + – + + + ? + + +

Freyer 34 + + + ? – + – ? – + + ? + ? + + +

Freyer 35 + + + ? – + – ? – ? – ? + ? + – na

Hamaker 37 – + + ? – + + + – + – + + na + + +

Honecker 38 – ? + ? – + + ? – ? ? ? + ? + – na

Hurria 39,40 + ? ? ? – + – ? – + ? ? ? ? – + +

Hurria 41 + + + + + + – ? – + + ? + ? + – na

Kanesvaran 42‐44 + + ? ? – + – ? – + + ? + na + + +

Karampeazis 45 + ? ? ? – + – ? – ? ? ? + ? + – na

Klepin46 + + + ? – + + ? – + ? + + na ? + –

Kothari 47,48 + + + ? – + – ? – + ? + – + – – –

Kristjansson 49‐51 + + + – + + + + + + – + + ? + + +

Maione 52 + + + ? – + + ? – + + + + na + + +

Marinello 53 + + + ? – + + + + + + ? + ? + + +

Pilnik 56 ? + ? ? – + – ? – ? ? ? + ? ? – na

Poon 57 – ? ? ? – + – ? – ? ? ? + ? + + +

Ramsdale 58 + ? + ? – + – ? – ? ? ? + ? + + –

Sostelly 59 + ? ? ? – + – ? – ? ? ? + ? ? + ?

Soubeyran 60 + ? + ? – + – ? – ? ? ? + na + + +

Tahir 61 – ? ? ? – + + ? – + ? ? + na – + +

Tredan 62 + + + + + + – + + + – + + ? + + +

Tucci63 + + + + – + – + + + + ? + ? + – na

Van Fraeyenhove 64 + ? ? ? – + + ? – + ? ? + ? + – na

Wedding 65,66 + + + ? – + + ? – + – + + na + + +

Zagonel 67 + ? ? ? – + + ? – + ? ? + na + + +

na: not applicable

174


Appendix 2: Content of geriatric assessments

Author

Cognition

Mood/

depression

Mobility

ADL

IADL

Nutritional

status

Frailty

screening

Comorbidity

Polypharmacy

Composition of

summary score

(if ap

plicab

le)*

Aaldriks 17,18

MMSE IQcode

MNA GFI

Aparicio 19,20

MMSE GDS Lawton Charlson

Arnoldi 21 MMSE GDS Barthel Lawton Charlson A

Audisio 22 MMSE GDS Katz Lawton Satariano

Bamias 23 Katz Lawton VES‐13 Charlson A

Basso 24 MMSE GDS Katz Lawton CIRS‐G + A

Biesma 25‐27

MMSE GDS15PANAS

TUG Katz Lawton GFI Charlson CIRS‐G

Brunello 28,29

MMSE GDS Katz Lawton CIRS‐G + A

Brunello 30 MMSE GDS Katz Lawton CIRS‐G + A

Castagneto 31 GDS Katz Lawton

De Wit 32,33

Barthel Lawton CIRS‐G + A

Extermann36 MMSE GDS Lawton MNA CIRS‐G +

Freyer 34 MMSE # +

Freyer 35 MMSE HADS TUG Lawton # +

Hamaker 37 IQcode Katz Lawton SNAQ Charlson +

Honecker 38 MMSE SCID TUG Barthel Lawton Charlson +

Hurria 39,40

Blessed HADS TUG OARS Lawton #

Hurria 41 GDS Katz Lawton Charlson

Kanesvaran 42‐44

MMSE GDS15 Katz Lawton Determine Charlson +

Karampeazis 45 MMSE GDS15 Katz Lawton CIRS‐G

Klepin46 CES‐D SPPB PAT‐D PAT‐D

Kothari 47,48

MMSE GDS15 Katz Lawton NHI NHC


MMSE GDS30 Barthel NEADL MNA CIRS‐G + A

Maione 52 Katz Lawton Charlson

Marinello 53 SPMSQ Katz Lawton CIRS‐G

PACE‐participants 54 MMSE GDS Katz Lawton Satariano

Pilnik 56 X Lawton # +

Poon 57 MMSE GDS TUG Katz Lawton MNA Charlson B

Ramsdale 58 MMSE SPPB Katz VES‐13 Charlson

Sostelly 59 HADS Katz Lawton Fried #

Soubeyran 60 MMSE GDS15 TUG Katz Lawton MNA CIRS‐G

Tahir 61 MMSE GDS Barthel Lawton #

Tredan 62 MMSE HADS Lawton # +

Tucci 63 Katz CIRS‐G A

Van Fraeyenhove 64 MMSE GDS TUG Katz Lawton MNA GFI Charlson B

Wedding 65,66

Barthel Lawton CIRS‐G

Zagonel 67 Katz Lawton

* composition of summary score: A – frail if ADL‐dependent, 3+ comorbidities or 1+ severe comorbidities, or 1+ geriatric conditions; B – number of geriatric conditions; x assessed but method not stated; # number of conditions See next page for further information

175

Chapter 11

Continuation of Appendix 2: NEADL Nottingham extended activities of daily living, PAT‐D Pepper Assessment Tool for Disability, CIRS‐G cumulative illness rating scale‐geriatrics, OARS Older American Resources and Services, MMSE mini mental state examination, IQCODE informant questionnaire on cognitive decline in the elderly, SPMSQ short portable mental status questionnaire, GDS geriatric depression scale, HADS hospital anxiety and depression scale, SCID structured clinical interview for DSM IV CES‐D Centre for Epidemiologic Studies‐Depression scale MNA mini nutritional assessment, SNAQ short nutritional assessment questionnaire, NHI NHC national health initiative nutritional health checklist, TUG timed get‐up‐and‐go SPPB short physical performance battery, GFI Groningen frailty index, VES‐13 vulnerable elders scale‐13

176


References 1. Muss HB. Cancer in the elderly: a societal perspective from the United States. Clin Oncol (R Coll Radiol )

2009; 21:92‐98. 2. Extermann M, Aapro M, Bernabei R et al. Use of comprehensive geriatric assessment in older cancer

patients: recommendations from the task force on CGA of the International Society of Geriatric Oncology (SIOG). Crit Rev Oncol Hematol 2005;55:241‐252.

3. Rodin MB, Mohile SG. A practical approach to geriatric assessment in oncology. J Clin Oncol 2007;25:1936‐1944.


5. Extermann M, Meyer J, McGinnis M et al. A comprehensive geriatric intervention detects multiple problems in older breast cancer patients. Crit Rev Oncol Hematol 2004;49:69‐75.

6. Terret C, Zulian GB, Naiem A, Albrand G. Multidisciplinary approach to the geriatric oncology patient. J Clin Oncol 2007;25:1876‐1881.

7. Terret C, Albrand G, Droz JP. Multidisciplinary geriatric assessment reveals unknown medical problems in elderly cancer patients. J Clin Oncol 2004;22[14S]:8167.

8. Gosney M. Contribution of the geriatrician to the management of cancer in older patients. Eur J Cancer 2007; 43:2153‐2160.

9. Brunello A, Sandri R, Extermann M. Multidimensional geriatric evaluation for older cancer patients as a clinical and research tool. Cancer Treat Rev 2009;35:487‐492.

10. Ellis G, Whitehead MA, O'Neill D et al. Comprehensive geriatric assessment for older adults admitted to hospital. Cochrane Database Syst Rev 2011;7:CD006211.

11. Barthelemy P, Heitz D, Mathelin C et al. Adjuvant chemotherapy in elderly patients with early breast cancer. Impact of age and comprehensive geriatric assessment on tumor board proposals. Crit Rev Oncol Hematol 2011;79:196‐204.

12. Klepin HD, Gajra A, Hardt N et al. Predictors of primary dose reduction among patients aged 65 years and older receiving adjuvant chemotherapy. J Clin Oncol 2011:S9003.

13. Tirelli U, Fratino L, Balzarotti M et al. Comprehensive geriatric assessment‐adapted chemotherapy in elderly patients (>70 years) with diffuse large B‐Cell non‐hodgkin's lymphoma (DLBCL): Final results and long term follow‐up. Blood 2009;114.

14. Corre R, Chouaid C, Barlesi F et al. Study ESOGIA‐GFPC 08‐02: Phase III, randomized multicentre trial involving subjects over age 70 with stage IV non small cell lung cancer, comparing a "classical" strategy of treatment allocation, based on performance status and age, with an "optimized" strategy allocating the same treatments according to a simplified geriatric assessmnet screening scale plus a more thorough geriatric assessment if necessary. ASCO meeting abstract 2011;TPS219.

15. Kuijpers T, van der Windt DA, van der Heijden GJ, Bouter LM. Systematic review of prognostic cohort studies on shoulder disorders. Pain 2004;109:420‐431.

16. Altman DG. Systematic reviews of evaluations of prognostic variables. BMJ 2001;323:224‐228. 17. Aaldriks AA, Maartense E, le Cessie S et al. Predictive value of geriatric assessment for patients older than

70 years, treated with chemotherapy. Crit Rev Oncol Hematol 2011;79:205‐212. 18. Aaldriks AA, Maartense E, Giltay EJ et al. The impact of geriatric assessment on the outcome of older breast

cancer patients treated with chemotherapy. J Geriatr Oncol 2011;2suppl1:S29. 19. Aparicio T, Jouve JL, Teillet L et al. Geriatric factors to predict toxicity and dose‐intensity reduction in FFCD

2001‐02 phase III study comparing a first‐line chemotherapy of LV5FU2 or FOLFIRI in treatment of metastatic colorectal cancer in elderly patients. J Clin Oncol 2011;S9111.

20. Aparicio T, Jouve JL, Teillet F et al. Geriatric factors to predict toxicity and dose‐intensity reduction in FFCD 2001‐02 phase III study comparing a first line chemotherapy with LV5FU2 or FOLFIRI in treatment of metastatic colon cancer in elderly patients. J Geriatr Oncol 2011;2suppl1:S61.

21. Arnoldi E, Dieli M, Mangia M et al R. Comprehensive geriatric assessment in elderly cancer patients: an experience in an outpatient population. Tumori 2007;93:23‐25.

22. Audisio RA, Gennari R, Sunouchi K et al l. Preoperative assessment of cancer in elderly patients: a pilot study. Support Cancer Ther 2003;1:55‐60.

177

Chapter 11

23. Bamias A, Lainakis G, Kastritis E et al. Biweekly carboplatin/gemcitabine in patients with advanced urothelial cancer who are unfit for cisplatin‐based chemotherapy: report of efficacy, quality of life and geriatric assessment. Oncology 2007;73:290‐297.

24. Basso U, Tonti S, Bassi C et al. Management of Frail and Not‐Frail elderly cancer patients in a hospital‐based geriatric oncology program. Crit Rev Oncol Hematol 2008;66:163‐170.

25. Biesma B, Wymenga AN, Vincent A et al l. Quality of life, geriatric assessment and survival in elderly patients with non‐small‐cell lung cancer treated with carboplatin‐gemcitabine or carboplatin‐paclitaxel: NVALT‐3 a phase III study. Ann Oncol 2011;22:1520‐1527.

26. Wymenga M, Biesma B, Vincent A et al. Platinum‐based combination chemotherapy in the treatment of older non‐small cell lung cancer (NSCLC) patients (pts): Is there a role for Complete Geriatric Assessment (CGA)? Final results from the prospective multicentre NVALT‐3 study. Journal of Clinical Oncology 2009; e20547.

27. Wymenga M, Biesma B, Vincent A et al. Can baseline complete geriaric assessment predict toxicity in elderly non‐small cell lung cancer patients? J Clin Oncol 2007;S7537.

28. Brunello A, Monfardini S, Falci C et al. Prognostic role of comprehensive geriatric assessment (CGA): a prospective study of a cohort of 1038 elderly cancer patients. J Geriatr Oncol 2011;2Suppl1: S36.

29. Basso U, Falci C, Brunello A et al. Prognostic value of multidimensional geriatric assessment on survival of a prospective cohort of 880 elderly cancer patients. J Clin Oncol 2011;S9065.

30. Brunello A, Sacco C, Barile C et al. Sunitinib in elderly patients (>=70 years) with metastatic renal cell carcinoma (MRCC): Multicentre analysis of tolerability and efficacy. Annals of Oncology 2008;viii:209.

31. Castagneto B, Zai S, Marenco D et al l. Single‐agent gemcitabine in previously untreated elderly patients with advanced bladder carcinoma: response to treatment and correlation with the comprehensive geriatric assessment. Oncology 2004;67:27‐32.

32. De Wit M, Honecker F, Wedding U et al. Incorporation of a comprehensive geriatric assessment (CGA) into a randomized phase III trial for metastatic breast cancer (MBC): The PELICAN study. Journal of Clinical Oncology 2010;suppl:1.

33. De Wit M, Harbeck N, Scholz M et al. Incorporation of a Comprehensive Geriatric Assessment (CGA) into a randomized phase III trial for metastatic breast cancer: The PELICAN Study. Journal of Clinical Oncology 2009; 27:9551.

34. Freyer G, Geay JF, Touzet S et al l. Comprehensive geriatric assessment predicts tolerance to chemotherapy and survival in elderly patients with advanced ovarian carcinoma: a GINECO study. Ann Oncol 2005;16:1795‐1800.

35. Freyer G, Lortholary A, Delcambre C et al. Unexpected toxicities in elderly patients treated with oral idarubicin in metastatic breast cancer: the GINECO experience. Clin Oncol (R Coll Radiol ) 2004; 16:17‐23.


37. Hamaker ME, Buurman BM, van Munster BC, et al. The Value of a Comprehensive Geriatric Assessment for Patient Care in Acutely Hospitalized Older Patients with Cancer. Oncologist 2011;16:1403‐12

38. Honecker FU, Wedding U, Rettig K et al. Use of the Comprehensive Geriatric Assessment (CGA) in elderly patients (pts) with solid tumors to predict mortality. Journal of Clinical Oncology 2009;27:a9549.

39. Hurria A, Togawa K, Mohile SG et al. Predicting chemotherapy toxicity in older adults with cancer: A prospective 500 patient multicentre study. Journal of Clinical Oncology 2010;28:1.

40. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicentre study. J Clin Oncol 2011;29:3457‐3465.

41. Hurria A, Fleming MT, Baker SD et al. Pharmacokinetics and toxicity of weekly docetaxel in older patients. Clin Cancer Res 2006;12:6100‐6105.

42. Kanesvaran R, Li H, Koo KN, Poon D. Analysis of prognostic factors of comprehensive geriatric assessment and development of a clinical scoring system in elderly asian patients with cancer. J Clin Oncol 2011;29:3620‐3627.

43. Kanesvaran R, Li H, Koo K, Poon D. An analysis of prognostic factors of comprehensive geriatric assessment (CGA) and development of a clinical scoring system in elderly Asian cancer patients. Journal of Clinical Oncology 2010;28:1.

44. Kanesvaran R, Koo KN, Chen W, Poon D. An analysis of the prognostic value of handgrip strength and its incorporation into the comprehensive geriatric assessment in elderly Asian patients with cancer. J Clin Oncol 2011: S9093.

178


45. Karampeazis A, Vamvakas L, Agelaki S et al l. Baseline comprehensive geriatric assessment (CGA) and prediction of toxicity in eldelry non‐small cell lung cancer (NSCLC) patients receiving chemotherapy. J Clin Oncol 2011:e19656.

46. Klepin HD, Tooze JA, Geiger AM et al. Geriatric assessment predicts overall survival among older adults receiving induction chemotherapy for acute myologenous leukemia. J Geriatr Oncol 2011; 2suppl1:S32.

47. Kothari A, Phillips S, Bretl T et al Components of geriatric assessments predict thoracic surgery outcomes. J Surg Res 2011;166:5‐13.

48. Weigel T, Kothari A, Bretl T et al. An abbreviated thoracic onco geriatric assessment (TOGA) and its components predict outcomes of esophagectomies. Annals of Surgical Oncology 2010:17:S128.

49. Kristjansson SR, Nesbakken A, Jordhoy MS et al. Comprehensive geriatric assessment can predict complications in elderly patients after elective surgery for colorectal cancer: a prospective observational cohort study. Crit Rev Oncol Hematol 2010;76:208‐217.

50. Kristjansson SR, Jordhoy MS, Nesbakken A et al. Which elements of a comprehensive geriatric assessment (CGA) predict post‐operative complications and early mortality after colorectal cancer surgery? Journal of Geriatric Oncology 2010;1:57‐65.

51. Kristjansson SR, Nesbakken A, Jordhoy MS, Wyller TB. Comprehensive geriatric assessment predicts complications in elderly patients operated for colon cancer. Crit Rev Oncol Hematol 2008;68:S24.

52. Maione P, Perrone F, Gallo C et al. Pretreatment quality of life and functional status assessment significantly predict survival of elderly patients with advanced non‐small‐cell lung cancer receiving chemotherapy: a prognostic analysis of the multicentre Italian lung cancer in the elderly study. J Clin Oncol 2005;23:6865‐6872.

53. Marinello R, Marenco D, Roglia D et al. Predictors of treatment failures during chemotherapy: A prospective study on 110 older cancer patients. Arch Gerontol Geriatr 2009;48:222‐226.

54. Audisio RA, Pope D, Ramesh HS et al. Shall we operate? Preoperative assessment in elderly cancer patients (PACE) can help. A SIOG surgical task force prospective study. Crit Rev Oncol Hematol 2008;65:156‐163.

55. Ramesh HS, Pope D, Stotter A et al. Testing CGA components to predict 30 day surgery outcome in elderly cancer patients. Crit Rev Oncol Hematol 2006;S21.

56. Pilnik NG, Ibero MM, Carri D, Mareca O. Influence of some clinical parameters in the tolerance to treatment in elderly patients with advanced lung cancer. Journal of Clinical Oncology 2010;15suppl:1.

57. Poon D, Lee HH, Chan LL et al. An exploratory analysis of comprehensive geriatric assessment results and overall survival in 233 consecutive elderly cancer patients. Journal of Clinical Oncology 2009;27suppl1:9504.

58. Ramsdale EE, Polite N, Bylow KA et al. Relationship between components of the comprehensive geriatric assessment, chemotherapy dose intensity, and overall survival in a colorectal cancer cohort age 65 and older. J Clin Oncol 2011;S9000.

59. Sostelly A, Falandry C, Pejuade‐Lauraine E et al. Predicting chemotherapy induced hematotoxicity in elderly paients with metastatic breast cancer treated by pegylated liposomal doxorubicin. J Clin Oncol 2011:e19740.

60. Soubeyran P, Rainfray M, Mathoulin‐Pelissier S et al. Screening of elderly patients with cancer for early death risk. Results of a prospective multicentric study of 364 patienets under chemotherapy. Crit Rev Oncol Hematol 2006;S23.

61. Tahir M, Pretorius R, Robinson T et al. Optimising the management of breast cancer in older patients. European Journal of Surgical Oncology 2010;36:1117‐1118.

62. Tredan O, Geay JF, Touzet S et al. Carboplatin/cyclophosphamide or carboplatin/paclitaxel in elderly patients with advanced ovarian cancer? Analysis of two consecutive trials from the Groupe d'Investigateurs Nationaux pour l'Etude des Cancers Ovariens. Ann Oncol 2007;18:256‐262.


64. Van Fraeyenhove F, Baitar A, De VM et al. Prediction of chemotherapy toxicity by the groningen frailty index (GFI) and the comprehensive geriatric assessment (CGA) in elderly cancer patients (PTS): An interim analysis. Annals of Oncology 2010;21

65. Wedding U, Rohrig B, Klippstein A et al. Age, severe comorbidity and functional impairment independently contribute to poor survival in cancer patients. J Cancer Res Clin Oncol 2007;133:945‐950.

66. Winkelmann N, Petersen I, Kiehntopf M et al. Results of comprehensive geriatric assessment effect survival in patients with malignant lymphoma. J Cancer Res Clin Oncol 2011;137:733‐738.

179

Chapter 11

67. Zagonel V, Fratino L, Piselli P et al. The comprehensive geriatric assessment predicts mortality among elderly cancer patients. Proc Am Thorac Soc 2002;21:abstract 1458.

68. Lawton MP, Brody EM. Assessment of older people: self‐maintaining and instrumental activities of daily living. Gerontologist 1969;9:179‐186.

69. Nouri FM, Lincoln NB. An extended activities of daily living index for stroke patients. Clin Rehabil 1987;1:301‐305.


71. Miller MD, Paradis CF, Houck PR et al. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res 1992; 41:237‐248.

72. Satariano WA, Ragland DR. The effect of comorbidity on 3‐year survival of women with primary breast cancer. Ann Intern Med 1994;120:104‐110.

73. Barthel D. Functional evaluation: the Barthel index. Md State Med J 1965;14:61‐65. 74. Katz S, Ford AB, Moskowitz RW et al. Studies of illness in the aged. The index of ADL: a standardized

measure of biological and psychological function. JAMA 1962;185:914‐919. 75. Folstein MF, Folstein SE, McHugh PR. "Mini‐mental state". A practical method for grading the cognitive

state of patients for the clinician. J Psychiatr Res 1975;12:189‐198. 76. Fillenbaum GG, Smyer MA. The development, validity, and reliability of the OARS multidimensional

functional assessment questionnaire. J Gerontol 1981;36:428‐434. 77. Jorm AF, Jacomb PA. The Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE): socio‐

demographic correlates, reliability, validity and some norms. Psychol Med 1989;19:1015‐1022. 78. Blessed G, Tomlinson BE, Roth M. The association between quantitative measures of dementia and of

senile change in the cerebral grey matter of elderly subjects. Br J Psychiatry 1968; 114(512):797‐811. 79. Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly

patients. J Am Geriatr Soc 1975;23:433‐441. 80. Sheikh JI, Yesavage JA, Brooks JO et al. Proposed factor structure of the Geriatric Depression Scale. Int

Psychogeriatr 1991;3:23‐28. 81. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361‐370. 82. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative

affect: the PANAS scales. J Pers Soc Psychol 1988;54:1063‐1070. 83. Vellas B, Guigoz Y, Garry PJ et al. The Mini Nutritional Assessment (MNA) and its use in grading the

nutritional state of elderly patients. Nutrition 1999;15:116‐122. 84. First MG, Spitzer RL, Gibbon M, Williams JB. Structured Clinical Interview for DSM‐IV Axis I Disorders.

Clinical version, administration booklet. New York: Biometrics Research Department, New York State Psychiatric Institute; 1997.

85. Posner BM, Jette AM, Smith KW, Miller DR. Nutrition and health risks in the elderly: the nutrition screening initiative. Am J Public Health 1993;83:972‐978.

86. Kruizenga HM, Seidell JC, de Vet HC et al. Development and validation of a hospital screening tool for malnutrition: the short nutritional assessment questionnaire (SNAQ). Clin Nutr 2005;24:75‐82.

87. Slaets JP. Vulnerability in the elderly: frailty. Med Clin North Am 2006;90:593‐601. 88. Saliba D, Orlando M, Wenger NS et al Identifying a short functional disability screen for older persons. J

Gerontol A Biol Sci Med Sci 2000;55:M750‐M756. 89. Fried LP, Tangen CM, Walston J et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci

Med Sci 2001;56:M146‐M156. 90. Mathias S, Nayak US, Isaacs B. Balance in elderly patients: the "get‐up and go" test. Arch Phys Med Rehabil

1986; 67:387‐389. 91. Radloff LS. The CES‐D scale. Applied Phychological Management 1977;1:385‐401. 92. Guralnik JM, Simonsick EM, Ferrucci L et al. A short physical performance battery assessing lower extremity

function: association with self‐reported disability and prediction of mortality and nursing home admission. J Gerontol 1994;49:M85‐M94.

93. Rejeski WJ, Ip EH, Marsh AP et al. Measuring disability in older adults: the International Classification System of Functioning, Disability and Health (ICF) framework. Geriatr Gerontol Int 2008;8:48‐54.

94. Lee L, Cheung WY, Atkinson E, Krzyzanowska MK. Impact of comorbidity on chemotherapy use and outcomes in solid tumors: a systematic review. J Clin Oncol 2011;29:106‐117.

95. Studenski S, Perera S, Patel K et al. Gait speed and survival in older adults. JAMA 2011;305:50‐58.

180


181

96. Ferrucci L, Guralnik JM, Studenski S et al. Designing randomized, controlled trials aimed at preventing or delaying functional decline and disability in frail, older persons: a consensus report. J Am Geriatr Soc 2004;52:625‐634.

97. Overcash JA, Beckstead J, Moody L et al. The abbreviated comprehensive geriatric assessment (aCGA) for use in the older cancer patient as a prescreen: scoring and interpretation. Crit Rev Oncol Hematol 2006;59:205‐210.

98. Hamaker ME, Jonker JM, de Rooij SE et al. Frailty screening tools for predicting outcome of a comprehensive geriatric assessment in older cancer patients. Lancet oncology in print. 2012.


Chapter 12

Frailty screening tools for predicting outcome of a comprehensive geriatric assessment in older cancer patients:

a systematic review

M.E. Hamaker, J.M. Jonker, S.E. de Rooij, A.G. Vos, C.H. Smorenburg, B.C. van Munster

Lancet Oncology 2012;13:e437‐e444

Chapter 12

Abstract Aim: To assess which frailty screening tools demonstrate the best sensitivity and

specificity for predicting the presence of impairments on comprehensive geriatric

assessment (CGA) in older cancer patients.

Method: A systematic search in Medline and Embase and hand‐search of conference

abstracts, for studies on the association between frailty screening tools and CGA in older

cancer patients.

Results: Literature search identified 4440 reports, of which 22 publications from 14

studies, assessing seven different frailty screening tools, were included in the review.

Median sensitivity and specificity of the screening tools for frailty on CGA were

respectively: Vulnerable Elders Survey‐13 (VES‐13) 68%/78%, Geriatric 8 (G8) 87%/61%,

Triage Risk Screening tool (TRST, cut‐off 1+) 91%/±45%, Groningen Frailty Index (GFI) ±50%

/±75%, Fried frailty criteria, ±30%/±90%, Barber 59%/79%, and abbreviated CGA (aCGA)

51%/97%. However, even in case of the highest sensitivity, the negative predictive value

was only 60%.

Conclusion: G8 and TRST demonstrated the highest sensitivity for frailty, but had a poor

specificity and negative predictive value. These findings suggest that for now, it may be

beneficial for all older cancer patients to receive a complete geriatric assessment as the

currently available frailty screening tools have insufficient discriminative power in

selecting patients for further assessment.

184

Frailty screening tools in geriatric oncology


aged 65 years and older,1 and the number of elderly cancer patients will increase

substantially in the coming decades as a result of increasing life expectancy and ageing of

the population. Cancer specialists are faced with the challenge of determining the optimal

treatment for these patients, with their heterogeneity in comorbidity, physical reserve,

disability and geriatric conditions.

For this purpose, two concepts of geriatric medicine are being incorporated in geriatric

oncology: the concept of frailty and the comprehensive geriatric assessment. Frailty is

considered as a state of decreased physiological reserves, arising from cumulative deficits

in multiple physiological systems, resulting in a diminished resistance to stressors.2,3 As

cancer and its treatment both form significant stressors, requiring patients to encroach on

their reserves, the concept of frailty appears particularly relevant for older cancer

patients. As yet, there is no consensus on its operationalization. The original definition of

frailty as formulated by Fried et al. focuses primarily on physical weakness and wasting,

but many other definitions and criteria have been postulated, incorporating different

aspects of ageing that contribute to diminishing reserves.4‐6

In geriatric oncology, the comprehensive geriatric assessment (CGA) is used to detect

disabilities and geriatric conditions that potentially contribute to frailty. A CGA is a

systematic procedure used to objectively appraise the health status of older people,

focusing on somatic, functional and psychosocial domains,7 and its value in geriatric

medicine has been proven extensively.8 However, as a CGA is time‐consuming, research is

now focusing on screening tools to separate fit older cancer patients that are able to

receive standard cancer treatment based upon the complete treatment schedule, from

vulnerable patients that should subsequently receive a CGA to guide tailoring of their

treatment regimen.9

In this context, the sensitivity of a frailty screening tool is considered to be of prime

importance; this will allow the treating physician to trust that frail patients will correctly

be identified by the screening tool.10,11 However, to optimize the time‐saving potential of a

two‐stepped approach, a good specificity of the screening tool is also required to ensure

that the number of fit patients incorrectly identified as frail on the screening tool and thus

unnecessarily receive a CGA, will be limited. To determine which screening tool best

meets both criteria, we performed a systematic review to assess the sensitivity and

specificity of frailty screening tools in predicting the presence of impairments on a CGA.

185

Chapter 12

Methods

Search strategy and selection criteria Our aim was to identify cohort studies which investigated the association between an

established frailty screening tool and a more complete CGA in cancer patients,

independent of age, cancer type or stage of disease.

The following search was performed on December 28th 2011, in both Medline and

Embase: ((("Geriatric Assessment"[Mesh]) OR (gfi[tiab] OR groningen frailty index[tiab])

OR (tfi[tiab] OR tilburg frailty index[tiab]) OR (isar[tiab] OR identification seniors at

risk[tiab]) OR (G8[tiab]) OR (fried[tiab]) OR (barber[tiab]) OR (edmonton[tiab]) OR

(saop[tiab] OR senior adult oncology program[tiab]) OR (triage risk screening tool[tiab])

OR (runciman rowland questionnaires[tiab]) OR (ves 13 OR vulnerable elderly survey[tiab])

OR (abbreviated comprehensive geriatric assessment[tiab] OR acga[tiab]) OR (geriatric

assessment*[tiab])) OR ((screening tool*[tiab]) AND (elderl*[tiab] OR geriatri*[tiab] OR old

age[tiab]))) AND (("Neoplasms"[Mesh]) OR (neoplasm*[tiab] OR cancer*[tiab] OR

tumour[tiab] OR tumours[tiab] OR tumor[tiab] OR tumors[tiab] OR oncolog*[tiab] OR

malignan*[tiab])).

No language limits or date ranges were applied.

In addition, conference abstracts of the scientific meetings from 2007‐2011 of the

American Society of Clinical Oncology (ASCO), European Society of Medical Oncology

(ESMO), International Society of Geriatric Oncology (SIOG), American Geriatric Society

(AGS) and European Geriatric Medicine Society (EUGMS) were hand‐searched for studies

on CGA in cancer patients to identify additional eligible studies.

We defined a frailty screening tool as a tool designed to assess the concept of frailty,

irrespective of the population or purpose for which this tool is intended. A CGA was

defined as an assessment using validated assessment tools investigating at least three of

the following domains: cognitive function, mood/depression, nutritional status, activities

of daily living (ADL), instrumental activities of daily living (IADL), comorbidity,

polypharmacy, mobility/falls, and/or social support. Studies were not eligible for inclusion

in this review if the study cohort included non‐cancer patients. We also excluded studies if

the CGA was only performed in a subgroup of patients selected by the outcome of the

frailty screening tool(s). For studies which included more than one screening tool,

eligibility was assessed separately for each tool.

The titles and abstracts of all studies retrieved by the searches were assessed by one

reviewer (MH) to determine which warrant further examination. All potentially relevant

articles were subsequently screened as full text by two authors (MH and JJ). If only an

abstract was available, an effort was made to find the final report of the study by

searching Embase and Medline using the names of first, second and/or final authors as

186


well as key words from the title. Also, in case of insufficient data in the original

manuscript, the authors were contacted for additional information, for example on the

ols used in the geriatric assessment or the cut‐off value of the screening tool. to

Data extraction Data regarding study design and results were independently extracted by two

investigators (MH and JJ) for each eligible study. Items that were extracted are the type of

study, study setting, study population (cancer type, cancer stage, cancer treatment, the

timing of screening), the frailty screening tools used including cut‐off values, the content

of the CGA and the assessment tools used, as well as the outcomes in terms of association

etween frailty screening tool and CGA. b

Quality assessment The methodological quality of each of the studies was also assessed independently by two

reviewers (MH and JJ), using the QUADAS‐2 tool (Quality Assessment of Studies of

Diagnostic Accuracy included in Systematic reviews, Appendix 1) as developed and revised

by Whiting et al.12,13 Disagreements among the reviewers were discussed during a

consensus meeting and in case of persisting disagreement, the assistance of a third

reviewer (BvM) was enlisted.

To address the suitability of the CGA to detect frailty, we used the definition formulated

by the International Society of Geriatric Oncology, which states that at minimum, a CGA

for older cancer patients should include assessment of functional status, cognition and

ood.7 m

Data synthesis and analysis We summarized the study results to describe our main outcomes of interest. If necessary,

sensitivity, specificity, positive and negative predictive values were calculated based on

the results listed in the publication, or – in case of insufficient published information – on

additional data obtained from the authors.

187

Chapter 12

Table 1: Studies on the association between frailty assessment and comprehensive geriatric assessment

Publication Patients

Author

Publication year

Abstract(A) or full

text (F)

Setting/departm

ent

Study population

Number of patients

% m

ale

Median age

in years

(range)

Baitar14‐16

2011 A Unclear Various cancer types 135 ? 77 (66‐97)

Kellen17 2010 F Medical

oncology/ general practice

Various cancer types, irrespective of stage of disease or treatment

113 60% 77 (SD 4)

Kenis18 2009 A Medical

oncology Newly diagnosed cancer or progressive disease, during admission

140 ? 76.5 (SD 5.1)

Kenis19 2011 A Medical

oncology Various cancer types, patients considered for start of chemotherapy

843 37% 76.9 (SD 5.1)

Kristjansson20 2008 A Surgery Pre‐operative patients with

colorectal cancer considered for surgery

74 46% 80 (71‐94)

Luciani21 2010 F Medical

oncology Various cancer types, prior to start first‐line chemotherapy

419 55% 76 (70‐97)

Mohile22,23

2007 F Medical oncology

Prostate cancer patients receiving hormonal treatment

50 100% 78 (70‐92)

Molina‐Garrido

24‐26

2010 F Medical oncology

Early stage breast cancer prior to start chemotherapy

41 0% 74.5 (SD 5.1)

Molina‐Garrido

27

2011 A Medical oncology

Unclear 58 ? Not stated

Monfardini28 2010 A Medical

oncology Patients with breast cancer newly referred to medical oncologist

150 ? 76 (70‐94)

Owusu29 2010 F Medical

oncology Various cancer types, first visit to outpatient clinic with newly diagnosed cancer

117 18% 72 (IQR 69‐80)

Pottel30‐32

2011 F Radiotherapy Head and neck cancer patients prior to start radiotherapy

51 84% 72 (65‐86)

Soubeyran33 2008 A Medical

oncology Various cancer types, prior to start of first‐line chemotherapy

364 59% 77 (70‐99)

Soubeyran34,3

5

2011 A Medical oncology

Unclear 1425 30% 78.2 (70‐98)

* A more detailed overview of the content of the geriatric assessment can be found in Appendix 3

188


Table 1: Studies on the association between frailty assessment and comprehensive geriatric assessment

Screening tool Geriatric assessment (CGA)*

Frailty screening

tool used (cut‐off)

Cognition

Mood

ADL

IADL

Nutritional status

Social support

Comorbidity

Polypharmacy

Mobility/falls

Cut‐off value for

frailty on CGA (nr of

impairm

ents unless

otherw

ise stated)

GFI (4+) G8 (≤14)

+ + + + + + + + 2+

aCGA (1+)** VES13 (3+) GFI (4+)

+ + + + Cognitive impairment or 2+ more other impairments

GFI (4+) G8 (≤14) TRST (1+/2+)

+ + + + + + 2+

TRST (1+) G8 (≤14)

+ + + + + + + + 2+

Fried (3+)

+ + + + + + 1+

VES‐13 (3+)

+ + + + + + 1+

VES‐13 (3+)

+ + + + + + 2+

VES‐13 (3+) Barber (1+)

+ + + + + + + 2+

Fried (3+) VES‐13 (3+)

+ + + + + + + 2+

VES‐13 (3+)

+ + + + + + ADL‐dependent, 3+ comorbidities or 1+ severe comorbidities

VES‐13 (3+)

+ + + + + + + + 2+

G8 (≤14) VES‐13 (3+)

+ + + + + + 2+

G8 (≤14)

+ + + + + + + 1+

G8 (≤14) VES‐13 (3+)

+ + + + + + + 1+

** In the original aCGA publications, screening yielded subscores per domain but not an overall score. This cut‐off value in this study was defined by its researchers, based on the aggregated results of the subscores. GFI Groningen Frailty Index, G8 Geriatric 8, aCGA abbreviated comprehensive geriatric assessment, VES13 Vulnerable Elders’ Survey‐13, TRST Triage risk screening tool, (I)ADL instrumental activities of daily living

189

Chapter 12

Results

Characteristics of included studies The literature search yielded 3943 citations (1769 from Medline and 2174 from Embase)

and an additional 497 studies on geriatric assessment in cancer patients were identified in

conference abstracts. For one of the identified abstracts, the full text publication came out

after the search date; however, as this contained additional, useful information, we

included this manuscript as well. After exclusion of 1279 duplicates and 3139 studies for

other reasons (Appendix 2), a total of 22 publications from 14 studies were included in this

review.14‐35

The characteristics of these 14 studies are summarized in Table 1. The first publication is in

2007 but most were published in the past two years.14‐35 For seven studies, full text

reports were available,17,21‐26,29‐32 but for the seven remaining studies, conference

abstracts were the only publications.14‐16,18‐20,27,28,33‐35 Median sample size was 117 patients

(range 41‐1425). Median age of included patients ranged from 72 to 80 years.14‐35 All but

two studies20,30‐32 were performed in the medical oncology department. Half of the studies

included patients with various cancer types17‐19,21,27,29,33 while two did not elaborate on the

type of patients that were included,27,34,35 the other five studies focused on one cancer

type only (two on breast cancer,24‐26,28 and one each on prostate cancer,22,23 colorectal

cancer20 and head and neck cancer30‐32).

Seven different frailty screening tools were assessed: Vulnerable elders survey‐135 was the

most frequently examined (VES‐13, 9 studies) followed by the Geriatric 836 (G8, six

studies), Groningen Frailty Index4 (GFI, 3 studies), Triage Risk Screening Tool37 (TRST) and

Fried frailty criteria6 (2 studies each), and the abbreviated CGA38 (aCGA) and Barber24 (one

study each). Table 2 gives an overview of the domains these tools focus on, and their

relative weight. All tools address functional status, and most (five out of seven tools) also

focus on some aspect of psychosocial functioning, but for other geriatric conditions there

is much more variation (Table 2). The aCGA and G8 are the only tools designed specifically

for assessment of frailty in older cancer patients (Table 2).

The content of the CGA varied from four to eight geriatric conditions (Table 1); the median

number of examined conditions was seven.14‐35 All studies included assessment of

cognitive function and activities of daily living (ADL), although the method of assessment

varied.14‐35 Instrumental ADL functioning was examined in twelve studies,14‐17,19,21‐35 mood

in ten studies,14‐20,28‐35 nutritional status in nine,14‐16,19‐21,24‐27,30‐35 mobility/falls history in six

studies14‐16,19,29,33‐35 and social support in seven.14‐16,18,22‐29 Comorbidity was examined in

thirteen studies14‐16,18‐35 and polypharmacy in eight.18‐27,29 Frailty on CGA was defined as

the presence of one or more geriatric conditions in four studies20‐22,33‐35 and two or more

190


Table 2: Relative weight of geriatric conditions in the frailty screening tools (in % of total points per tool)

Geriatric domains GFI4 G836 VES‐135 aCGA38 Fried6 Barber24 TRST37

Functional status 27% 11% 60% 60% 60% ADL impairment 13% 20% 33% IADL impairment 7% 40% 11% mobility/falls 7% 20%

Psychosocial domain

40% 11% 40%

Cognitive disorder

7% 27% 20%

Mood/anxiety 13% 13% Social support 20% 11%

Neuro‐sensory deficits

13% 22%

Nutritional status/weight loss

7% 46% 20%

Polypharmacy 7% 6% 20%

Comorbidity

Recent hospitalization

11% 20%

Geriatric syndromes 20%

Self‐reported health 7% 11% 10% 20% 11%

Age 11% 30%

Optimal score 0 17 0 * 0 0 0

Poorest score 15 0 10 * 5 9 5

Standard cut‐off value

4+ ≤14 3+ * 3+ 1+ 2+

Population for which tool was designed

Various Cancer patients

Community‐dwelling elderly

Cancer patients

No specific population

Primary care patients

Patients in emergency room

* no overall scoring of aCGA is available; subscores for each geriatric domain are calculated. GFI Groningen Frailty Index, G8 Geriatric 8, VES‐13 Vulnerable elders’ survey‐13, aCGA abbreviated comprehensive geriatric assessment, TRST triage risk screening tool; (I)ADL (instrumental) activities of daily living

in eight studies.14‐16,18,19,22‐27,29‐32 The remaining two studies applied other definitions,

assigning different weights to the various assessed conditions (Table 1).17,28

Study populations showed a wide variation in the prevalence of frailty as diagnosed by the

CGA; a median of 68% of patients was considered frail (range 28‐94%, Table 3).14‐35 This

wide range cannot be explained solely by differences in the cut‐off used for the definition

of frailty on CGA, as in studies using the cut‐off of 1+ prevalence of frailty ranged from 28

to 80%20‐22,33‐35 while in studies using a cut‐off of 2+, the range was 43% to 88%.14‐16,18,19,22‐

27,29‐32 According to frailty screening tools, the median prevalence of frailty was 49% (range

12‐83%).14‐35

191

Chapter 12

Quality assessment Results of the quality assessment can be found in Figure 1. Reviewer agreement was over

95% for all aspects. For patient selection, the risk of bias was generally considered low,

and little concerns existed about the applicability of selection criteria. Due to the lack of

information on the independent scoring of the frailty screening tool and the CGA, the risk

of bias was frequently unclear for both the index test as well as the reference test.

Variation in the content of the CGA resulted in some concerns about the applicability of

study results and may also have introduced risk of bias. Flow and timing of the studies

yielded few concerns. Full results of the quality assessment can be found in Appendix 1b.

Figure 1: Quality assessment of included studies, using the QUADAS‐2 assessment tool (Appendix 1a).

13 The

complete assessments per study can be found in Appendix 1b.

192


Frailty screening tools vs. comprehensive geriatric assessment Table 3 gives an overview of the sensitivity and specificity of the frailty screening tools

compared to frailty as detected by CGA, while Figure 2 demonstrates the relationship

between sensitivity and false‐positives per screening tool. For the VES‐13, sensitivity

ranged from 39% to 88% with a median of 68%.17,21‐32,34,35 Thus, a median of 32% of frail

patients were not recognized as frail by the VES‐13. The specificity was better, ranging

from 62% to 100%, with a median of 78%. Thus, the VES‐13 yielded 22% false‐positives.

For the G8, the sensitivity ranged from 77% to 92%, with a median of 87% but the

specificity ranged from 39% to 75%, with a median of 61%.14‐16,18,19,30‐35 The TRST (using a

cut‐off of 1+) showed a sensitivity of 91% with a specificity of between 43 and 50%.18,19

For the GFI, sensitivity ranged from 39% to 62%, and the specificity was between 69% and

87%.14‐18 For the Fried frailty criteria, the sensitivity was between 25% and 37%, with a

specificity of 86% to 96%.20,27

Table 3: Sensitivity and specificity of frailty screening tools for outcome of (summarized) comprehensive geriatric assessment (CGA)

Screening tool

(cut‐off) Study

% frail on screening

tool

% frail on CGA

Sensitivity (%)

Specificity (%)

Positive predictive value (%)

Negative predictive value (%)

aCGA (1+) Kellen (2010)17 36 68 51 97 97 48

Barber (1+) Molina‐Garrido (2010)24‐26

42 56 59 79 77 63

Kristjansson (2008)20 12 38 25 96 78 67

Fried (3+) Molina Garrido (2011)

27 35 88 37 86 95 16

Baitar (2011)15 75 44 92 39 55 85

Kenis (2009)18 76 79 80 40 83 35

Kenis (2011)19 74 73 87 61 86 63

Pottel (2011)30‐32

67 69 86 75 88 71

Soubeyran (2008)33 82 94 85 65 97 22

G8 (≤14)

Soubeyran (2011)34,35

68 80 77 64 90 41

Baitar (2011)14,16

44 44 62 69 62 69

Kellen (2010)17 31 68 39 86 86 40 GFI (4+)

Kenis (2009)18 48 79 57 87 94 36

Kenis (2009)18 83 79 92 50 87 63

TRST (1+) Kenis (2011)

19 82 73 91 43 81 64

TRST (2+) Kenis (2009)18 50 79 64 100 100 43

Kellen (2010)17 49 68 61 78 85 48

Luciani (2010)21 54 28 87 62 45 93

Mohile (2007)22,23

50 60 73 86 89 67

Molina‐Garrido (2010)24‐26

29 56 55 100 100 66

Molina‐Garrido (2011)27 35 88 39 100 100 18

Monfardini (2010)28 46 44 68 71 65 74

Owusu (2010)29 56 43 88 69 68 88

Pottel (2011)30‐32

39 69 57 100 100 52

VES‐13 (3+)

Soubeyran (2011)34,35

60 80 69 74 92 37

GFI Groningen Frailty Index, G8 Geriatric 8, VES‐13 Vulnerable elders’ survey‐13, aCGA abbreviated comprehensive geriatric assessment, TRST triage risk screening tool

193

Chapter 12

Figure 2: Sensitivity and 1‐specificity of the screening tools for predicting outcome of comprehensive geriatric assessment

GFI Groningen Frailty Index, G8 Geriatric 8, VES‐13 Vulnerable elders’ survey‐13, aCGA abbreviated comprehensive geriatric assessment, TRST triage risk screening tool

Interestingly, as the prevalence of frailty in most study populations was high, even the

screening tools with the highest sensitivity to frailty still yielded negative predictive values

of around 60% (Table 3). This means that four out of every ten patients considered fit

fter frailty screening, will be diagnosed as frail after CGA. a

Frailty screening tools vs. individual geriatric conditions Although frailty screening tools are generally used to predict overall frailty, a few studies

have addressed the association between these tools and individual geriatric conditions

(Table 4); four studies addressed the latter association for VES‐13,17,22,23,29‐32 while one

study each looked at aCGA,17 GFI17 and G8.30‐32 Comparison of results is compromised,

however, by the different methods of presentation of the data.

Overall, VES‐13 appears to be most strongly associated with ADL and IADL functioning

(area under the curve (AUC) between 0.81 and 0.91, sensitivity between 67% and 83% and

specificity between 61% and 89%.)17,22,23,29‐32 Association between VES‐13 and cognitive

disorders, impaired mobility and malnutrition was fair (AUC 0.79‐0.81, 0.78‐0.87 and 0.78

194


Table 4: Association between screening tools and individual geriatric domains

Screening tool

Study

Cognition

Mood

ADL

IADL

Nutritional status

Comorbidity

Social support

Mobility

Polypharmacy

aCGA Kellen (2010)

17

Sens 23% Spec 100%

Sens 69% Spec 92%

Sens 97% Spec 47%

Sens 92% Spec 69%

G8 Pottel (2011)

30‐32

AUC 0.73 AUC 0.78 AUC 0.71 AUC 0.67 AUC 0.95

AUC 0.74 AUC 0.74

GFI Kellen (2010)

17

Sens 47%Spec 76%

Sens 39% Spec 86%

Kellen (2010)

17

Sens 76% Spec 63%

Sens 67% Spec 89%

Mohile (2007)

22,23

Sens 75% Spec 58%

Sens 83% Spec 61%

Sens 76% Spec 68%

Sens 76% Spec 64%

Sens 33% Spec 46%

Sens 70% Spec 67%

Owusu (2010

29

AUC 0.81 AUC 0.66 AUC 0.81 AUC 0.91 AUC 0.73 AUC 0.57 AUC 0.78

AUC 0.72

VES‐13

Pottel (2011)

30‐32 AUC 0.79 AUC 0.74 AUC 0.87 AUC 0.86 AUC

0.78AUC 0.70 AUC

0.87

aCGA abbreviated comprehensive geriatric assessment, G8 Geriatric 8, GFI Groningen frailty index, VES‐13 Vulnerable Elders Survey‐13; AUC area under the curve

respectively) 22,23,29‐32 G8 showed a strong association with malnutrition (AUC 0.95), but

was of lesser value for predicting the presence of other geriatric conditions. 30‐32 aCGA

showed a strong association with ADL and IADL impairment (sensitivity 97% and 92%,

specificity 47% and 69% respectively), but the sensitivity for cognitive dysfunction or

depressive symptoms was low (23% and 69% respectively). 17 GFI demonstrated poor

sensitivity for functional impairment (ADL impairment 47%, IADL impairment 39%) but

asonable specificity (76% and 86% respectively). 17 re

Discussion A useful frailty screening tool in geriatric oncology should have a high sensitivity, thus

allowing the assessor to trust that those patients deemed fit actually are fit, and a

sufficient specificity so that the time‐consuming process of a full CGA is optimally

utilized.10,11,39 In this systematic review on the discriminative power of frailty screening

tools for outcome of full CGA, we found that those tools with the highest sensitivity lacked

specificity and vice versa. In addition, even for the screening tools with the highest

sensitivity, the negative predictive value (i.e. the proportion of patients considered fit on

the screening tool that were also considered fit after full CGA) was around 63%, meaning

over a third of patients were unjustly considered fit after frailty screening.

195

Chapter 12

Given the content of the frailty screening tools and the population and purpose for which

they were developed (Table 2), the lack of sensitivity of some of these screening tools

when used to predict outcome of CGA is hardly surprising. For example, the VES‐135 and

Fried frailty criteria6 focus strongly on functional status (Table 2). Therefore, it is not

surprising that these tools are not very successful in identifying impairments in other

geriatric domains. For other screening tools, such as the GFI, the lack of discriminative

power could be explained by the fact that they were designed to diagnose frailty in a

different patient population than that in which it is being used in geriatric oncology (Table

2). In fact, most of the screening tools that are being used in geriatric oncology were not

designed or validated for this particular setting; only the G8 and aCGA developed

specifically for use in older cancer patients. As the aCGA was designed to identify which

individual geriatric domains require further assessment, the lack of sensitivity for overall

frailty (median 51%) as well as for two of the four individual domains for which it was

designed (Table 4) is disappointing. Ultimately, the G8 and the TRST ‐ developed to assess

functional impairment in older patients admitted to the emergency department37 – had

the highest sensitivity for frailty on CGA (87 and 91% respectively) but both lacked

specificity (61% and 45% respectively).

Thus, no screening tool currently used in geriatric oncology combined adequate sensitivity

and specificity. In addition, for tools with the highest sensitivity, the percentage of

patients diagnosed as frail after screening was around 70%, and – based on the negative

predictive value of these tools – one‐third of the remaining 30% was incorrectly diagnosed

as fit. Considering these findings, the question rises whether there is any benefit in a two‐

stepped approach in which a CGA is preceded by a screening tool. The time‐saving

potential of this approach will be limited if the prevalence of frailty is high, and potentially

does not outweigh the risk of incorrectly identifying patients as fit and delivering care‐as‐

usual where a more cautious approach would have been better.

Another aspect for which the use of a screening tool prior to CGA could be beneficial is to

identify which geriatric domains require further assessment; this could save time by

allowing the CGA to be limited to those domains only. Several studies in this review

addressed this issue (Table 4); however, the available data are limited. In addition, these

studies examined the association between the entire screening tool and a specific geriatric

domain while it would be more useful to assess the sensitivity of each individual screening

question for the presence or absence of the specific geriatric conditions that the question

inquires after. This aspect warrants further exploration in future research.

This systematic review provides a valuable overview of all currently available evidence on

the use of frailty screening tools in geriatric oncology but it also has several limitations.

First of all, we limited our search to studies performed in older cancer patients, thus

excluding available evidence from other patient populations. However, the prevalence of

196


geriatric conditions and frailty will differ greatly in various contexts and the discriminative

power of frailty screening tools should therefore be assessed in the context that it will be

used in. Another limitation of this review is that no full text reports have been published

for half of the included studies, and we had to rely on conference abstracts as the only

source of information on the execution and results of the study. In particular, this limited

our ability to analyse the predictive value of the frailty screening tools for individual

domains. Moreover, the content of the CGA differed considerably between studies, as did

the cut‐off value that was used to define frailty. The definition of frailty that is used will

influence the prevalence of frailty in a study population and similarly the sensitivity and

specificity of screening tools in predicting that frailty. This lack of golden standard for

assess frailty assessment decreased the comparability of study results and subsequently

hampered the execution of a formal meta‐analysis, thus leaving uncertainty about the

relative quality of the different screening tools for detecting frailty in older cancer

patients. In 2005, the International Society of Geriatric Oncology recommended that, at

minimum, frailty assessment for older cancer patients should include an evaluation of

functional status, cognition and mood.39 In wake of growing evidence that geriatric

conditions, such as nutritional status, polypharmacy, comorbidity and social support may

also be significant for older cancer patients, perhaps an update of these recommendations

should be undertaken. Moreover a consensus in the scales to use, including their cut‐off

values, would result in greater uniformity in clinical practice and research, and allow for

better comparison between studies and patient populations.

Ultimately, establishing whether a patient is fit or frail is not a goal in itself, but a method

for optimizing and tailoring oncologic and non‐oncologic treatment in elderly cancer

patients. Therefore, future research should focus on a more elaborate exploration of the

value of frailty and individual geriatric conditions for predicting and improving clinical

utcomes such as quality of life, survival, treatment tolerance and functional decline. o

In conclusion, although the Geriatric 8 and Triage Risk Screening Tool demonstrated the

best sensitivity for frailty on full CGA in older cancer patients, they had a poor specificity

and negative predictive value. Perhaps it will be possible to develop targeted screening

tools with better sensitivity and specificity once the relative importance of individual

geriatric domains and the benefit of appropriate interventions and follow‐up are fully

elucidated in this patient population. Until such a time, it may be beneficial for all older

patients to receive a complete geriatric assessment, since the two‐stepped approach –

using frailty screening tools to select patients for CGA – has insufficient discriminative

power.

197

Chapter 12

Appendix 1a: Quality Assessment of studies of Diagnostic Accuracy included in Systematic reviews‐2 (QUADAS‐

2) tool12,13

Domain 1: Patient Selection Risk of Bias: Could the selection of patients have introduced bias? 1: Was a consecutive or random sample of patients enrolled? 2: Was a case–control design avoided? 3: Did the study avoid inappropriate exclusions? Applicability: Are there concerns that the included patients and setting do not match the review question? Domain 2: Index Test Risk of Bias: Could the conduct or interpretation of the index test have introduced bias?

1: Were the index test results interpreted without knowledge of the results of the reference standard?

2: If a threshold was used, was it prespecified? Applicability: Are there concerns that the index test, its conduct, or its interpretation differ from the review question? Domain 3: Reference Standard Risk of Bias: Could the reference standard, its conduct, or its interpretation have introduced bias? 1: Is the reference standard likely to correctly classify the target condition?

For this purpose, we used the definition formulated by the International Society of Geriatric Oncology, which states that at minimum, a CGA for older cancer patients should include assessment of functional status, cognition and mood.

40 2: Were the reference standard results interpreted without knowledge of the results of the index test?

Applicability: Are there concerns that the target condition as defined by the reference standard does not match the research question? Domain 4: Flow and Timing Risk of Bias: Could the patient flow have introduced bias? 1: Was there an appropriate interval between the index test and reference standard? 2: Did all patients receive the same reference standard? 3: Were all patients included in the analysis?

198


Appendix 1b: Overview of quality assessment according to the QUADAS‐2 tool per study

Risk of bias Concerns about applicability Author

Patient selection

Index test

Reference test

Flow and timing

Patient selection

Index test

Reference test

Baitar (2011)14‐16 Unclear Unclear Unclear Unclear Low Low Low

Kellen (2010)17 Low Unclear Unclear Low Low Low Low

Kenis (2009)18 Low Low Low Low Low Low Low

Kenis (2011)19 Low Unclear Unclear Low Low Low Low

Kristjansson (2008)20 Low Unclear Unclear Low Low Low Low

Luciani (2010)21 Low Unclear Unclear Low Low Low Low

Mohile (2007)22,23 Low High High Low Low Low High

Molina‐Garrido (2010)24‐26 Low Unclear High Low Low Low High

Molina‐Garrido (2011)27 Unclear Unclear High Low Unclear Low High

Monfardini (2010)28 Low Low Low Low Low Low Low

Owusu (2010)29 Low Unclear Unclear Low Low Low Low

Pottel (2011)30‐32 Low Unclear Unclear Low Low Low Low

Soubeyran (2008)33 Low High Unclear Unclear Low Low Low

Soubeyran (2011)34,35 Low Unclear Unclear High Low Low Low

Appendix 2: Search and study selection

All studies n= 4440 Medline n= 1769 Embase n= 2174 Conference abstracts n= 497

Duplicates n= 1279

Exclusion n= 3139 Not original research n= 747 Not oncology n= 444 No comprehensive geriatric assessment (CGA) n= 1340 No screening tool n= 566 No association screening tool with CGA n= 22 Non‐cancer patients included n= 8 CGA not performed in all patients n= 3 Substantial overlap with another publication n= 2 Insufficient data for analysis n= 5 Retrospective tool development n= 2

Inclusion: 22 publications from 14 studies

199

Chapter 12

Appendix 3: Complete overview of comprehensive geriatric assessment per study

Author

Frailty screening tool

used (cut‐off)

Cognition

Mood

ADL

IADL

Nutrition

Mobility/falls

Social

Comorbidity

Polypharmacy

Others

Baitar (2011)

14‐16

GFI (4+) G8 (≤14)

MMSE GDS Katz Lawton TUG MOS‐ SSS

Charlson

Kellen (2010)

17

aCGA (1+) VES13 (3+) GFI (4+)

MMSE GDS Barthel Lawton

Kenis (2009)

18

GFI (4+) G8 (≤14) TRST (1+/2+)

MMSE GDS Katz + + CIRS‐G + Age

Kenis (2011)

19

TRST (1+) G8 (≤14)

MMSE GDS Katz Lawton MNA Fall history

Charlson + Fatigue (Mob‐T)

Kristjansson (2008)

20

Fried (3+) MMSE GDS Barthel MNA CIRS‐G +

Luciani (2010)

21

VES‐3 (3+) MMSE Katz Lawton MNA CIRS‐G + Social status

Mohile (2007)

22,23

VES‐3 (3+) SPMSQ, Pfeiffer

Katz Lawton MOS‐SSS

Charlson + SPPB

Molina‐ Garrido (2010)

24‐26

VES‐3 (3+) Barber (1+)

Pfeiffer Barthel Lawton NSI Gijon Charlson +

Molina‐ Garrido (2011)

27

Fried (3+) VES‐13 (3+)

Pfeiffer Barthel Lawton NSI Gijon Charlson +

Monfardini (2010)

28

VES‐3 (3+)

MMSE GDS Katz Lawton CIRS‐G +

Owusu (2010)

29

VES‐3 (3+)

MMSE GDS Katz Lawton TUG, fall history

MOS‐SSS

Charlson + Visual/hearing impairment

Pottel (2011)

30‐32

G8 (≤14) VES‐3 (3+)

MMSE GDS Katz Lawton MNA Tinetti CIRS‐G

Soubeyran (2008)

33

G8 (≤14) MMSE GDS Katz Lawton MNA TUG CIRS‐G Quality of life (QLQ‐C30)

Soubeyran (2011)

34,35

G8 (≤14) VES‐3 (3+)

MMSE GDS Katz Lawton MNA TUG CIRS‐G

GFI Groningen Frailty Index, G8 Geriatric 8, VES‐13 Vulnerable elders survey‐13, aCGA abbreviated comprehensive geriatric assessment, TRST triage risk screening tool + assessed without the use of a specific assessment tool (I)ADL instrumental activities of daily living, MMSE mini mental state examination, SPMSQ short portable mental status questionnaire, GDS geriatric depression scale, MNA mini nutritional assessment, NSI nutrition screening initiative TUG Timed up and go, MOS‐SSS medical outcomes study – social support survey CIRS‐G cumulative illness rating scale‐geriatric, SPPB short physical performance battery QLQ‐C30 quality of life questionnaire

200


References 1. Muss HB. Cancer in the elderly: a societal perspective from the United States. Clin Oncol (R Coll Radiol )

2009;21:92‐98. 2. Ferrucci L, Guralnik JM, Studenski S et al. Designing randomized, controlled trials aimed at preventing or

delaying functional decline and disability in frail, older persons: a consensus report. J Am Geriatr Soc 2004;52:625‐634.

3. Rockwood K, Mitnitski A. Frailty defined by deficit accumulation and geriatric medicine defined by frailty. Clin Geriatr Med 2011;27:17‐26.

4. Slaets JP. Vulnerability in the elderly: frailty. Med Clin North Am 2006;90:593‐601. 5. Saliba D, Orlando M, Wenger NS et al. Identifying a short functional disability screen for older persons. J

Gerontol A Biol Sci Med Sci 2000;55:M750‐M756. 6. Fried LP, Tangen CM, Walston J et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci

Med Sci 2001;56:M146‐M156. 7. Extermann M, Aapro M, Bernabei R et al. Use of comprehensive geriatric assessment in older cancer

patients: recommendations from the task force on CGA of the International Society of Geriatric Oncology (SIOG). Crit Rev Oncol Hematol 2005;55:241‐252.

8. Ellis G, Whitehead MA, O'Neill D et al. Comprehensive geriatric assessment for older adults admitted to hospital. Cochrane Database Syst Rev 2011;7:CD006211.


10. Molina‐Garrido MJ, Guillen‐Ponce C. Overvaluation of the vulnerable elders survey‐13 as a screening tool for vulnerability. J Clin Oncol 2011;29:3201‐3202.

11. Falci C, Brunello A, Monfardini S. Detecting functional impairment in older patients with cancer: is vulnerable elders survey‐13 the right prescreening tool? An open question. J Clin Oncol 2010;28:e665‐e666.

12. Whiting P, Rutjes AW, Reitsma JB et al. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003;3:25.

13. Whiting PF, Rutjes AW, Westwood ME et al l. QUADAS‐2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529‐536.

14. Baitar A, Van Fraeyenhove F, Galdermans D et al. Evaluation of the groningen frailty indicator as a screening tool for frailty in older cancer patients (PTS). European Journal of Cancer 2011:46:S282.

15. Baitar A, Vandebroek A, De DE et al. Evaluation of the g8 questionnaire as a screening tool for frailty in older cancer patients (PTS). European Journal of Cancer 2011;47:S275.

16. Schrijvers D, Baitar A, de Vos M et al P. Evaluation of the Groningen Frailty Index as a screening tool in elderly patients (PTS): an interim analysis. Annals of Oncology 2010;21(Suppl 8):556PD.

17. Kellen E, Bulens P, Deckx L et al. Identifying an accurate pre‐screening tool in geriatric oncology. Crit Rev Oncol Hematol 2010;75:243‐248.

18. Kenis C, Schuermans H, van Cutsem E et al. Screening for a geriatric risk profile in older cancer patients; a comparative study of the predictive validity of three screening tools. Crit Rev Oncol Hematol 2009;72(suppl 1):S22.

19. Kenis C, Decoster L, Vanpuyvelde K et al l. Comparison of two screening tols in older cancer patients. J Geriatr Oncol 2011;2(suppl 1):S39.

20. Kristjansson SR, Jordhoy MS, Nesbakken A, Wyller TB. A comparison of two methods to measuring frailty in elderly patients with colorectal cancer. Crit Rev Oncol Hematol 2008;68(suppl1):S30.

21. Luciani A, Ascione G, Bertuzzi C et al l. Detecting disabilities in older patients with cancer: comparison between comprehensive geriatric assessment and vulnerable elders survey‐13. J Clin Oncol 2010;28:2046‐2050.

22. Mohile SG, Bylow K, Stadler WW, Rodin M. A pilot study of the Vulnerable Elders Survey (VES‐13) as compared to Comprehensive Geriatric Assessment (CGA) to examine geriaric domains in elder prostate cancer patients at risk for decline. J Clin Oncol 2006;24(suppl):8539.

23. Mohile SG, Bylow K, Dale W et al. A pilot study of the vulnerable elders survey‐13 compared with the comprehensive geriatric assessment for identifying disability in older patients with prostate cancer who receive androgen ablation. Cancer 2007;109:802‐810.

24. Molina‐Garrido MJ, Guillen‐Ponce C. Comparison of two frailty screening tools in older women with early breast cancer. Crit Rev Oncol Hematol 2011;79:51‐64.

201

Chapter 12

202

25. Molina‐Garrido M, Guillen‐Ponce C, Santiago J et al. Barber test as a screening tool in multidimensional geriatric evaluation in elderly women with early breast cancer. J Clin Oncol 2009;155:e20645.

26. Molina‐Garrido MJ, Guillen‐Ponce C, Guirado‐Risueno M et al Usefulness of the Barber test is compared with the comprehensive geriatric assessment for identifying disability in outpatients older than 65 years with breast cancer who receive chemotherapy. Crit Rev Oncol Hematol 2008;68(suppl2):S33.

27. Molina‐Garrido M, Guillen‐Ponce C, Munoz M et al. A comparison of the fried criteria and the VES‐13 questionnaire as screening instruments for frail elderly patients diagnosed with cancer. European Geriatric Medicine 2011;2(suppl 1):S34.

28. Monfardini S, Basso U, Fiduccia P et al. Can the short screening test Vulnerable Elders Survey 13 (VES‐13) substitute for the time‐consuming comprehensive geriatric assessment (CGA) to identify vulnerable/frail elderly breast cancer patients? Journal of Clinical Oncology 2010;28.

29. Owusu C, Koroukian SM, Schluchter M et al. Screening older cancer patients for a Comprehensive Geriatric Assessment: A comparison of three instruments. J Geriatr Oncol 2011;28:121‐129.

30. Pottel L, Boterberg T, Pottel H et al. Determination of an adequate screening tool for identification of vulnerable elderly head and neck cancer patients treated with radio(chemo)therapy. J Clin Oncol 2011;di:10.1016/j/jgo.2011.11.006.

31. Debruyne PR, Boterberg T, Pottel H et al. Can the Vulnerable Elders Survey‐13 and/or the G8 adequatley identify elderly patients with head and neck cancer in need of a comprehensive geriatric assessment? J Clin Oncol 2011:e19670.

32. Pottel L, Boterberg T, Pottel H et al l. Could the combined test "VES‐13 + (max‐G8) represent an interesting alternative screening tool for identification of potential vulnerable elderly head and neck cancer patients? J Geriatr Oncol 2011;2(suppl1):S21.

33. Soubeyran P, Bellera CA, Gregoire F et al. Validation of a screening test for elderly patients in oncology. J Clin Oncol 2008;26(15S):20568.

34. Soubeyran P, Bellera C, Goyard J et al. Validation of the G8 geriatric screening tool for elderly cancer patients and comparions with the VES‐13: final results of the ONCODAGE project. J Geriatr Oncol 2011;2(S1):S17.

35. Soubeyran P, Bellera C, Goyard J et al. Validation of the G8 screening tool in geriatric oncology; the ONCODAGE project. J Clin Oncol 2011;29(15S):9001.

36. Bellera C, Rainfray M, Mathoulin‐Pelissier S, Soubeyran I. Validation of a screening tool in geriatric oncology: the Oncodage project. Crit Rev Oncol Hematol 2008;68(suppl):O2.

37. Meldon SW, Mion LC, Palmer RM et al. A brief risk‐stratification tool to predict repeat emergency department visits and hospitalizations in older patients discharged from the emergency department. Acad Emerg Med 2003;10:224‐232.




Part IV

General discussion, summary

and acknowledgements

Chapter 13 General discussion

Chapter 13

Introduction In this thesis, we addressed current treatment practices and decision making in elderly

cancer patients as well as the consequences of these decisions for clinical outcome. In

addition, we investigated the value of a comprehensive geriatric assessment in the

decision‐making process for these patients. In this final chapter, the results of these

studies will be placed in a broader perspective and some issues and implications of these

ndings will be discussed, including some directions for future research. fi

Research in elderly cancer patients: clinical trials and observational cohort studies If we want to know the optimal treatment for older cancer patients, we must design

studies that specifically target these patients, as it is incorrect to assume that optimal

treatment for a younger patient population will be just as suitable for the older group.

However, designing clinical trials for elderly patients is complicated by multiple factors,

including the heterogeneity of the population resulting in a great variability in the ability

to tolerate treatment, as well as hesitance of older patients to participate in a trial,

especially in case of randomization (Chapter 5). These factors may affect the accrual of

elderly patients, or mean the patients that are included in clinical trials do not form a

representative sample of the target population. Therefore, studies in real‐life populations

are needed to supplement trial data.

One option is to use observational cohorts as was done in this thesis, where we examined

cohorts of patients with breast cancer, colon cancer and head and neck cancer, to address

the issue of guideline adherence and differences in treatment choices between younger

and older patients (Chapter 1‐3 and 6‐8). These have provided us with data on real‐life

clinical practice and in particular, on the reasons why certain treatment choices are made;

these data could not have been obtained in a clinical trial format. However, as all these

studies were done retrospectively, data on potentially relevant confounders such as the

presence of geriatric syndromes or decreased functional capacity were not available. Since

these data could also have influenced treatment choices or prognosis, it will be useful to

repeat these studies prospectively, with incorporation of baseline data on frailty and

geriatric syndromes. In fact, it could be argued that for any research in elderly cancer

patients – irrespective of the chosen study design – reporting of geriatric baseline data is

as important as the tumour‐related data or other patient characteristics. Without these

data, it is not possible to compare study results or extrapolate their findings to the

dividual patient.1 in

206

General Discussion

Non‐standard or sub‐standard treatment: more than mere semantics For decades, the issue of guideline adherence for the elderly was barely addressed in

oncologic literature, as older age was a generally accepted contra‐indication for cancer

treatment. In the last decades of the twentieth century, cancer specialists came to realize

that this automatic exclusion of older patients did not do them justice.2 In wake of studies

demonstrating that standard cancer treatment could be given successfully to selected

elderly,3 older patients not receiving standard care were stated to receive “substandard”

or even “inappropriate” treatment.2,4 However, despite the obviously positive

development that age itself is no longer considered to be a criterion on which to base

decisions, it should be emphasized that for the elderly patient discordance with standard

practice or guideline‐recommended treatment does not automatically imply that

treatment is suboptimal. As discussed in Chapters 1‐4 and 6‐8, valid reasons for deviation

from guidelines exist. Furthermore, although there is general consensus that age should

not be the primary reason for withholding treatment,5,6 it would be unfair to say that age

does not matter at all. For instance, the process of ageing means that a person gradually

loses some of his or her physiological reserves, influencing their treatment tolerance. In

addition, much cancer treatment is aimed at preventing future cancer‐related

complications; as ageing limits the remaining life‐expectancy of a patient, it may also limit

the benefit of such treatment.

Using the terms “substandard treatment” or “undertreatment” to describe non‐standard

treatment suggests that any tailoring of treatment to the patient’s situation should be

avoided. However, overtreatment of frail elderly patients may be as harmful as

undertreatment. Therefore, tailor‐made care should be the standard of care for older

patients. The important questions that remain are what factors should guide such tailoring

nd how to reliably determine and balance potential benefit with potential harm. a

Using geriatric concepts in oncology With this in mind, cancer specialists are looking towards the concepts and methods of

geriatric medicine. One such concept is frailty, which is a state of decreased physiological

reserves, arising from cumulative decline across multiple physiological systems, resulting

in a diminished resistance to stressors;7 it can be seen as the final common pathway of

ageing. As both cancer and its treatment form significant stressors, which require patients

to encroach on their reserves, the concept of frailty appears particularly relevant for older

cancer patients. Another geriatric concept that has been adopted in oncology is the

comprehensive geriatric assessment (CGA), which is a systematic appraisal of the health

status of older individuals, focusing on somatic as well as psychosocial and functional

domains;8 it is used to detect disabilities and geriatric conditions that potentially

contribute to frailty.

207

Chapter 13

However, whereas geriatricians consider the CGA to be the entirety of their consultation

with a patient, focusing on as many as twenty geriatric conditions,9,10 in geriatric oncology

the term CGA is used to describe any procedure in which one or more screening tests for

individual geriatric conditions is used. For example, the International Society of Geriatric

Oncology, in their 2005 guidelines for the use of geriatric tools in the assessment of older

cancer patients,11 states that a CGA should consist of a mini‐mental state examination

(MMSE), geriatric depression scale (GDS) and Katz activities of daily living (ADL) scale. In

other studies that describe using a CGA for older cancer patients, this consisted of only

one ADL or instrumental ADL scale.12 However, malnutrition, polypharmacy, addressed

and unaddressed comorbidity and mobility disorders have also been demonstrated to be

quite common in older cancer patients (Chapter 9 and 10) and deserve evaluation.

Limiting a CGA to only a few geriatric conditions could mean that relevant frailty remains

unrecognized, as it has been demonstrated that many geriatric conditions are easily

missed if they are not specifically looked for.13 For example, one study comparing the

physician’s assessment of a patient’s fitness for chemotherapy with the assessment as

done by a CGA, revealed that 20% of patients deemed fit by their oncologist had more

than three geriatric conditions.14

These newly identified health issues could form the starting point for interventions aimed

at optimizing quality of life, survival and treatment tolerance. Studies that incorporated a

geriatric consultation into standard oncologic care for older cancer patients found that

over 70% of patients subsequently receive at least one non‐oncologic intervention.15‐17

The effect of such interventions remains to be evaluated, as thus far only one study has

addressed this issue. In this study by Rao et al. adding geriatric care to standard in‐patient

cancer care for 99 hospitalized elderly cancer patients resulted in a significant decrease in

the amount of emotional limitations, social dysfunction and bodily pain that these

patients experienced at three months; the effect on pain was still significant one year after

hospital discharge.18 This aspect of CGA in cancer care certainly warrants further

investigation.

Another purpose for which CGA could be used in geriatric oncology is to guide treatment

decisions. Cancer treatment should be tailored to the estimated remaining life‐

expectancy, and as demonstrated in Chapter 10 and 11, frailty – defined by the presence

of one or more geriatric conditions – is an independent predictor of survival, even after

correcting for age, cancer type, stage and treatment choice. Interestingly, the presence of

frailty seems to be more important than the individual geriatric conditions that cause the

frailty, supporting the concept of frailty as a final common pathway of the ageing process.

In addition to predicting survival, frailty also appears to be of value in predicting treatment

tolerance, in particular toxicity of chemotherapy (Chapter 10 and 11), but these findings

still require further confirmation. Despite this uncertainty, studies that have incorporated

208

General Discussion

CGA into standard cancer care demonstrate that knowledge of the presence of geriatric

conditions resulted in an altered treatment plan in 21‐49% of patients;15‐17,19 these

alterations consisted of temporary delays to optimize non‐oncologic issues as well as both

decreased and increased intensity of treatment, meaning that some patients actually were

less frail than initially thought by the cancer specialist.

In summary, the value of CGA has been proven for identifying previously unrecognized

health issues, which can guide treatment decisions and can also be used to implement

multidisciplinary interventions to optimize a patient’s health status and quality of life. We

therefore strongly urge all cancer specialists to incorporate a CGA into the standard

ssessment of all older cancer patients. a

Frailty screening tools An important obstacle to incorporating the CGA into oncologic practice is the fact it is time

consuming. This has understandably led cancer specialists to search for a short screening

tool or checklist that could stratify patients as fit or unfit for treatment, rendering a full

assessment unnecessary.20 An alternative approach is to use the tool to select patients for

further assessment.21 However, currently available screening tools lack discriminative

power and the prevalence of one or more geriatric conditions is so high that almost all

older cancer patients will potentially benefit from further assessment (Chapter 12). As

capturing the heterogeneity and nuances of the elderly cancer population in a few simple

questions is unlikely to become a reality, future research should focus on other ways in

which the full CGA with subsequent interventions could be incorporated into the cancer

care process. Although this will be time consuming, it will be balanced out by the gains in

terms of avoiding under‐ and overtreatment, preventing complications and adverse

eatment outcomes as well as improving quality of life. tr

Patient preferences and decision making Another factor that could influence treatment decisions for older cancer patients is a

patient’s personal preferences. In several studies included in this thesis (Chapter 1‐4 and

8), patient preference was an important reason for non‐adherence to treatment

guidelines. In addition, patient’s refusal of trial participation was an important factor in

the poor accrual for our clinical trial on chemotherapy for metastatic breast cancer

(Chapter 5). It is possible that elderly patients truly do not want cancer treatment, and are

unwilling to participate in trials. However, older patients tend to take a more passive role

in the decision‐making process,22 and because they are less likely to access alternative

sources of information,23 their decisions will be greatly influenced by the information that

he or she has been offered by the cancer specialist.24 In addition, studies have shown that

the endorsement of a clinical trial by the treating physician greatly increases the likelihood

209

Chapter 13

of trial participation.25,26 Therefore, an interesting but as yet unanswered question is how

much of what is stated to be the patient’s preference is actually a reflection of the

physician’s preference.

Various studies have demonstrated that to the older patient, quality of life and

maintaining functional independence are much more important than overall survival.27

However, these types of outcome measures are generally not included in clinical trials.28 In

order for future studies to supply truly relevant information for the older patient, we

highly recommend incorporating geriatric outcome measures, such as cognitive function

and functional capacity, as well as quality‐adjusted survival analyses like the TWiST, the

ime Without Symptoms or Toxicity.29 T

Directions for future research and patient care Given the currently available data on the value of CGA in geriatric oncology, we believe

that some form of geriatric assessment should be incorporated into standard cancer care

for older patients. However, the precise role of the comprehensive geriatric assessment

for both patient care and decision making needs to be determined. A further exploration

of the ability of CGA to predict treatment tolerance is warranted; however, to avoid

previous issues with heterogeneity in patient population and limited content of CGA,

these studies should be conducted in single‐tumour cohorts and include a broad range of

geriatric conditions. Another important research question still to be addressed is whether

implementing interventions aimed at the conditions identified by CGA can improve a

patient’s ability to tolerate treatment, and in line with this, whether this will improve

quality of life and/or survival.

As stated before, research that addresses older cancer patients should incorporate

baseline frailty information. Given the heterogeneity of the elderly population, these data

allow for comparison between studies and enable translation of outcomes to the

individual patient. In addition, future studies in older cancer patients should focus not only

on survival and somatic treatment complications, but also include relevant patient‐

centred outcome measures to allow weighing the benefits with the risks of a particular

eatment for the individual patient. tr

In conclusion, for older cancer patients, tailor‐made care should be the standard of care,

striking the golden mean between undertreatment and overtreatment and fully taking

into account the heterogeneity of this patient population. The comprehensive geriatric

assessment will provide invaluable information about a patient’s overall health status, but

its exact place within the decision‐making process still remains to be defined.

210

General Discussion

References 1. Audisio RA, Van LB. When reporting on older patients with cancer, frailty information is needed. Annals of

Surgical Oncology 2011;18:4‐5 2. Byrne A, Carney DN. Cancer in the elderly. Curr Probl Cancer 1993;17:145‐218. 3. Valentini V, Morganti AG, Luzi S et al. Is chemoradiation feasible in elderly patients? A study of 17 patients

with anorectal carcinoma. Cancer 1997;80:1387‐1392. 4. Bergman L, Dekker G, van Kerkhoff EH et al. Influence of age and comorbidity on treatment choice and

survival in elderly patients with breast cancer. Breast Cancer Res Treat 1991;18:189‐198. 5. Audisio RA, Bozzetti F, Gennari R, et al. The surgical management of elderly cancer patients;

recommendations of the SIOG surgical task force. Eur J Cancer 2004;40:926‐938. 6. Lichtman SM, Wildiers H, Chatelut E et al. International Society of Geriatric Oncology Chemotherapy

Taskforce: evaluation of chemotherapy in older patients‐‐an analysis of the medical literature. J Clin Oncol 2007;25:1832‐1843.

7. Ferrucci L, Guralnik JM, Cavazzini C et al l. The frailty syndrome: a critical issue in geriatric oncology. Crit Rev Oncol Hematol 2003;46:127‐137.


9. Buurman BM, Hoogerduijn JG, de Haan RJ et al l. Geriatric conditions in acutely hospitalized older patients: prevalence and one‐year survival and functional decline. PLoS One 2011;6:e26951.

10. Ellis G, Langhorne P. Comprehensive geriatric assessment for older hospital patients. Br Med Bull 2004; 71:45‐59.



13. Extermann M, Meyer J, McGinnis M et al. A comprehensive geriatric intervention detects multiple problems in older breast cancer patients. Crit Rev Oncol Hematol 2004; 49:69‐75.

14. Wedding U, Kodding D, Pientka L et al. Physicians' judgement and comprehensive geriatric assessment (CGA) select different patients as fit for chemotherapy. Crit Rev Oncol Hematol 2007; 64:1‐9.

15. Chaibi P, Magne N, Breton S et al l. Influence of geriatric consultation with comprehensive geriatric assessment on final therapeutic decision in elderly cancer patients. Crit Rev Oncol Hematol 2011;79:302‐307.

16. Caillet P, Canoui‐Poitrine F, Vouriot J et al. Comprehensive Geriatric Assessment in the Decision‐Making Process in Elderly Patients With Cancer: ELCAPA Study. J Clin Oncol 2011;29:3636‐3642.

17. Horgan AM, Leighl NB, Coate L et al. Impact and Feasibility of a Comprehensive Geriatric Assessment in the Oncology Setting: A Pilot Study. Am J Clin Oncol 2011.

18. Rao AV, Hsieh F, Feussner JR, Cohen HJ. Geriatric evaluation and management units in the care of the frail elderly cancer patient. J Gerontol A Biol Sci Med Sci 2005;60:798‐803.

19. Girre V, Falcou MC, Gisselbrecht M et al. Does a geriatric oncology consultation modify the cancer treatment plan for elderly patients? J Gerontol A Biol Sci Med Sci 2008; 63:724‐730.



22. Fallowfield L. Participation of patients in decisions about treatment for cancer. BMJ 2001;323:1144. 23. Pinquart M, Duberstein PR. Information needs and decision‐making processes in older cancer patients. Crit

Rev Oncol Hematol 2004;51:69‐80. 24. Pierce PF. Deciding on breast cancer treatment: a description of decision behavior. Nurs Res 1993;42:22‐28. 25. Townsley CA, Chan KK, Pond GR et al. Understanding the attitudes of the elderly towards enrolment into

cancer clinical trials. BMC Cancer 2006;6:34.

211

Chapter 13

212

26. Wright JR, Whelan TJ, Schiff S et al. Why cancer patients enter randomized clinical trials: exploring the factors that influence their decision. J Clin Oncol 2004;22:4312‐4318.

27. Yellen SB, Cella DF, Leslie WT. Age and clinical decision making in oncology patients. J Natl Cancer Inst 1994;86:1766‐1770.

28. Reuben DB, Tinetti ME. Goal‐oriented patient care‐‐an alternative health outcomes paradigm. N Engl J Med 2012;366:777‐779.

29. Goldhirsch A, Gelber RD, Simes RJ et al. Costs and benefits of adjuvant therapy in breast cancer: a quality‐adjusted survival analysis. J Clin Oncol 1989;7:36‐44.

Chapter 14

Summary / Samenvatting

Chapter 14

Summary: Decision making in geriatric oncology

In Western societies, the population is ageing. Between 2012 and 2030, the proportion of

Dutch inhabitants aged 75 years and over will double. This will have significant

implications for health care, as many diseases that predominantly affect the elderly will

become more prevalent. This is also true for cancer.

Optimal cancer care for the elderly ideally strikes the golden mean between

overtreatment and undertreatment and must be tailored to the individual patient. Cancer

specialists are now collaborating with geriatricians to optimize this decision‐making

process. The aim of this thesis was to study current treatment practice in older cancer

patients with breast cancer, colon cancer and head and neck cancer (Part I and II) and to

investigate the potential value of geriatric concepts and methods in this decision‐making

process (Part III).

In Part I, current treatment practice and decision making is studied in older patients with

breast cancer. In Chapter 1, we examined a cohort of 166 patients aged 70 years and

older, newly diagnosed with breast cancer between 2002 and 2004 at the Medical Centre

Alkmaar (MCA). We found that 74% of these patients were diagnosed and treated in

accordance with national guidelines. However, guideline adherence decreased with age:

only 40% of patients aged 90 years and older received guideline recommended treatment

compared to 88% of patients aged 70‐79 years. The deviations from guidelines were

primarily deliberate adjustments to comorbid diseases and the patient’s preferences.

To assess whether implementation of a weekly, multidisciplinary breast cancer meeting to

discuss all new patients in 2006 improved guideline adherence, we extended the previous

study until 2007 and focused on the 232 patients aged 70 years and older who were

diagnosed with an early stage of breast cancer within that period (Chapter 2). For early

stage breast cancer, surgery is the primary treatment. However, contrary to what we

expected to find, the percentage of patients that received surgery decreased from 94% in

2002 to 76% in 2007 and this decrease was even more explicit in patients aged 80 years

and older, in which this proportion decreased from 80% to 33%. Although guideline

adherence was not altered by the initiation of the multidisciplinary meetings, we did find

that awareness of guidelines improved over time and discordance was more clearly

motivated in the patient’s charts.

In Chapter 3, we took a more detailed look at a cohort of 187 patients aged 75 years and

older with early breast cancer for which surgery was omitted despite guideline

recommendations, diagnosed between 1990 and 2008 at five Dutch hospitals. Most of

these patients received hormone treatment instead. Earlier studies have demonstrated

that this kind of treatment can prevent disease progression in many patients for some

years. However, for patients with a limited life‐expectancy, a temporary solution may

214


suffice. In our study, patients lived for an average of 2.1 years after the breast cancer was

diagnosed. We found that 34% of patients died of breast cancer; this is quite similar to

what is seen in studies of older breast cancer patients receiving guideline recommended

treatment. For 21% of the patients in our study, breast cancer was stated as an underlying

disease in patients who died of other causes. However, for the remaining patients, breast

cancer was not thought to be relevant at the time of death, suggesting that hormone

treatment had been successful in suppressing the cancer for the remainder of the

patient’s lifetime. Interestingly, in one‐third of patients, omission of surgery was not for

medical reasons but at the patient’s own request.

Chapters 1 to 3 focus on treatment decisions that were made in the hospital or out‐

patient clinic. However, this does require referral to a hospital by the primary care

physician. In Chapter 4, we sent out a questionnaire to elderly‐care physicians (ECPs), who

provide the primary care for patients living in nursing homes, to inquire after their

decisions regarding patients with possible breast cancer. Almost 60% of respondents

stated that they had encountered at least one such patient in the past year; of these, one‐

third was not referred to a hospital. The most important reasons for not referring the

patient were end‐stage dementia, the patient’s or family’s preference and limited life‐

expectancy. Referral was often felt to be too burdensome for these very vulnerable

patients. In this chapter, potential alternatives to hospital‐based care are discussed.

Optimal decision making is based on evidence that is obtained from clinical trials focusing

specifically on the target population, in which treatment is randomly assigned so that

potential confounding and bias is as limited as possible. However, even when a clinical

trial is designed for older patients, it may be difficult to accrue sufficient numbers of

patients. We encountered this issue in the OMEGA study of the Dutch Breast Cancer

Trialists’ Group (BOOG), which was designed for frail older patients with metastatic breast

cancer (for more detail, see Chapter 10), and therefore examined possible obstacles to

accrual (Chapter 5). The primary barrier we identified was that a patient’s overall health

status was considered either too fit or too frail for inclusion, as well as the patient’s refusal

to participate or preference for a particular type of treatment. These findings demonstrate

how difficult it is to design a study for older cancer patients and emphasize the need for

educating patients on the potential benefits of trial participation.

In Part II, current treatment practices and decision making is studied in two other types of

cancer: colon cancer and head and neck cancer. In Chapter 6, we compared treatment for

colon cancer in 183 patients aged 70 years and older with 103 patients younger than 70

years. Of the older patients, 91% received surgery in accordance with guidelines compared

to 100% of the younger age group. For chemotherapy, these proportions were 32% and

85% respectively. The primary reason for deviation from guidelines was age, as well as

comorbidity and poor general health. Despite the fact that only fit patients received

215

Chapter 14

standard treatment, older patients experienced more post‐operative complications and

mortality than younger patients; for chemotherapy, toxicity was equal despite careful

selection and less aggressive regimens for older patients. These results suggest that older

age does affect the ability of a patient to tolerate treatment and can therefore not be

discounted in the decision‐making process.

There are many different causes for iron deficiency but in 2‐15% of patients, it is caused by

gastro‐intestinal cancer. To diagnose such cancers, endoscopy is needed, which is a

burdensome examination; if cancer is found, surgery is often the next step. As not all

patients are fit enough for surgery, or do not want this treatment, it is sometimes decided

to forego further examination when iron‐deficiency is found. In Chapter 7, we looked at a

group of 471 patients aged 80 years and older with iron‐deficiency anaemia, diagnosed at

the laboratory of the Gelre hospital in Apeldoorn. For 59% of these patients, no anaemia‐

related diagnostic procedures were performed. This was more likely in case of female sex,

older age, less severe anaemia and in the absence of gastrointestinal symptoms. Initial

work‐up revealed 46 malignancies (9 upper and 37 lower gastrointestinal tract). In

addition to these, another 16 patients were diagnosed with colon cancer during follow‐up,

primarily in patients who received no initial diagnostic procedures. Of these patients, 9

were subsequently treated surgically; for these patients, the delay in diagnosis could have

resulted in disease progression and thus decreased the likelihood of curative treatment or

increased the risk of complications. By contrast, a delay in diagnosis is less relevant if the

underlying disease is incurable or requires treatment that is considered too burdensome

for the individual patient. Median survival varied greatly, depending on the initial

diagnosis. For colon cancer patients receiving surgery, perioperative mortality was high

(15%) but after 1.3 years, benefit of surgical treatment became apparent.

Chapter 8 focuses on 606 patients with head and neck cancer of all ages, diagnosed and

treated at the Medical Centre Alkmaar. Head and neck cancer can have tremendous

impact on a patient’s quality of life, as the disease affects some of our most basic body

functions, such as breathing, swallowing and speech. However, the burden of treatment

can be just as great. An additional complicating factor is that head and neck cancer is

often related to alcohol use and smoking and as a result, patients frequently suffer from

multiple lifestyle‐related diseases simultaneously. These diseases can affect the ability of

patient to tolerate treatment. In our study, we found that 91% of patients were treated in

accordance with guidelines. Primary reasons for discordant treatment were comorbidity

and lack of cooperation in patients aged younger than 70 years and patient’s refusal of

treatment for older patients. Five‐year overall survival was 64% for patients treated in

accordance with guidelines, compared to only 15% in case of discordant treatment,

emphasizing the importance of guideline adherence in these patients.

216


In Part III, the value of two concepts of geriatric medicine is examined for oncologic

decision making. One such concept is frailty, which is a state of diminished functional

reserves, caused by a cumulative decline across multiple organ systems, resulting in a

decreased resistance to stressors. Frailty can be seen as the final common pathway of

ageing. As both cancer and its treatment represent significant stressors, which require a

patient to use their reserves to get through it, the concept of frailty seems particularly

relevant to oncology. A second concept is the comprehensive geriatric assessment or CGA,

which is a systematic assessment of a patient’s health status, focusing on their physical,

functional and psychosocial well‐being. It can be used to detect the presence of geriatric

syndromes, such as dementia, depression, malnutrition, disability, impaired mobility etc.,

as well as frailty.

In Chapter 9, the CGA is used in a group of 292 patients aged 65 years and older, with any

type of cancer, acutely admitted to the internal medicine wards of three hospitals. In

these patients, we saw that geriatric syndromes were frequently present in older cancer

patients. For example, almost half of patients suffered from polypharmacy (the use of

more than five types of medications), impaired mobility and malnutrition, while more than

two‐thirds suffered from depressive symptoms, functional limitations and pain. Detecting

these conditions can form a starting point for interventions aimed at optimizing general

health and quality of life. However, the prognosis of these patients was not associated

with these geriatric conditions; for survival, having metastatic disease or cancer‐related

complications that required hospital admission were much more important factors.

Chapter 10 describes the results of the OMEGA study, in which 78 patients aged 65 years

and older with metastatic breast cancer were treated with two types of palliative

chemotherapy. Metastatic breast cancer cannot be cured, and therefore the aim of this

treatment is to prolong survival and maintain an optimal quality of life. By contrast,

serious toxicity of chemotherapy can decrease quality of life and is potentially fatal.

Therefore, predicting toxicity can be quite useful in deciding which patients should or

should not receive palliative chemotherapy. In our study, all patients received a CGA at

baseline. We found that the chance of experiencing serious toxicity due to the

chemotherapy increased steadily with the number of geriatric conditions that a patient

had, rising from 18% in patients without geriatric problems, to over 50% in case two

conditions were present and over 80% in case of three or more.

Chapters 11 is a systematic review of other studies that have assessed the predictive value

of CGA for survival and treatment tolerance, both for chemotherapy as well as oncologic

surgery. We found 37 other studies that addressed this issue; however, little consistency

was found in their results. It appears that different geriatric syndromes were useful for

predicting each of the different outcome measures: frailty (defined as having one or more

geriatric syndromes), nutritional status and comorbidity were predictive of survival; frailty

217

Chapter 14

was also predictive of toxicity of chemotherapy; cognitive impairments and difficulties in

performing the basic daily activities (such as eating, washing or using the toilet) were

associated with not completing all planned cycles of chemotherapy; and difficulties in

performing instrumental daily activities (such as using the telephone or taking medication)

were associated with perioperative complications. The inconsistency in study results could

be caused by the great variety in the study populations (in terms of tumour types, stage of

disease and treatment modalities) that were included in the different studies. Thus,

although various geriatric conditions appear to be of some value for predicting outcome in

elderly cancer patients, the results as yet are too inconsistent to guide treatment

decisions.

Performing a CGA is time‐consuming, which has led cancer specialists to search for short

frailty screening tools that can be used to select which patients should receive a full

assessment. We performed a systematic review to determine the sensitivity and

specificity of the currently available screening tools for the outcome of the complete CGA

(Chapter 12). We identified 14 studies, assessing seven different frailty screening tools.

We found that the Geriatric 8 (G8) and Triage Risk Screening Tool had the best sensitivity

for frailty on full CGA, but these tools lacked specificity; similarly, tools with a good

specificity for frailty, such as the Fried Criteria and abbreviated CGA (aCGA) severely

lacked sensitivity. These findings suggests that all older patients should receive a complete

geriatric assessment as the two‐stepped approach – using frailty screening tools to select

patients for CGA – has insufficient discriminative power.

The general discussion in Chapter 13 elaborates on the observed results and discusses the

plications of these findings for patient care and for future research. im

In conclusion, the studies in this thesis have shown that for older cancer patients, tailor‐

made care should be the standard of care, striking the golden mean between

undertreatment and overtreatment and fully taking into account the heterogeneity of this

patient population. The comprehensive geriatric assessment will provide valuable

information about a patient’s overall health status, but its exact place within the decision‐

making process still remains to be defined.

218


Samenvatting: Besluitvorming in de geriatrische oncologie

In een vergrijzend Nederland zal het aantal inwoners ouder dan 75 jaar in de komende 20

jaar verdubbelen, met grote gevolgen voor de gezondheidszorg. Zo zullen allerlei ziekten

die vooral ouderen treffen, waaronder kanker, steeds vaker voorkomen. Een optimale

kankerbehandeling van de oudere patiënt moet aangepast zijn aan het individu; alleen zo

kan de balans gevonden worden tussen onderbehandeling en overbehandeling.

Dit proefschrift gaat over het zoeken van die balans. In de eerste twee delen is gekeken

naar de huidige besluitvorming en invulling van de zorg voor oudere patiënten met

borstkanker, darmkanker en hoofd‐halskanker. Het laatste deel van dit proefschrift

beschrijft de toegevoegde waarde van een geriatrische werkwijze in het complexe

besluitvormingsproces rond de kankerbehandeling bij ouderen. Geriaters zijn

ziekenhuisspecialisten die zich richten op het zorgvuldig in kaart brengen en optimaliseren

van de gezondheid van oudere patiënten met meervoudige gezondheidsproblematiek.

Deel I richt zich op borstkanker bij ouderen. Hoofdstuk 1 beschrijft een groep van 166

patiënten van 70 jaar en ouder, bij wie tussen 2002 en 2004 in het Medisch Centrum

Alkmaar borstkanker werd vastgesteld. Daarbij bleek dat 74% van deze patiënten

overeenkomstig de Nederlandse behandelrichtlijnen werd onderzocht en behandeld.

Echter, hoe ouder de patiënt, hoe vaker van de geadviseerde behandeling werd

afgeweken: slechts 40% van de patiënten boven de 90 jaar werd behandeld volgens de

richtlijnen. Het afwijken van de richtlijnen gebeurde veelal bewust, vanwege bijkomende

ziekten of de voorkeur van de patiënt.

Om te beoordelen of het invoeren van een wekelijkse, multidisciplinaire

borstkankerbespreking in 2006 een verbetering gaf in het volgen van de richtlijn, werd het

voorgaande onderzoek verlengd tot 2007, waarbij de studie zich beperkte tot de 232

patiënten ouder dan 70 jaar bij wie een vroeg stadium van borstkanker werd vastgesteld

(Hoofdstuk 2). Bij een vroeg stadium van borstkanker staat chirurgie centraal in de

behandeling. In tegenstelling tot wat verwacht was, daalde het percentage patiënten dat

werd geopereerd van 94% in 2002 tot 76% in 2007. Deze daling was het meest

uitgesproken bij patiënten boven de 80 jaar: voor hen daalde het percentage van

respectievelijk 80% naar 33%. Hoewel het volgen van de richtlijn niet verbeterde door de

invoering van de borstkankerbespreking, bleken behandelaars de richtlijn wel bewuster te

hanteren; afwijkingen hiervan werden ook beter werd gemotiveerd in de patiëntenstatus.

In Hoofdstuk 3 is meer gedetailleerd gekeken naar een groep van 187 borstkanker‐

patiënten, van 75 jaar en ouder, bij wie van de geadviseerde chirurgische behandeling

werd afgezien. Het merendeel van deze patiënten kreeg als vervanging hormonale

behandeling. Eerder onderzoek heeft uitgewezen dat deze behandeling de borstkanker

219

Chapter 14

tijdelijk kan onderdrukken maar het niet kan genezen; echter, voor patiënten met een

beperkte levensverwachting kan een tijdelijke oplossing soms volstaan. De patiënten in

ons onderzoek leefden gemiddeld nog 2,1 jaar nadat de borstkanker was vastgesteld en

34% overleed ten gevolge ervan. Voor bijna de helft van de patiënten was de borstkanker

klinisch niet relevant op het moment van overlijden. Dit suggereert dat de

hormoontherapie de kanker voldoende heeft onderdrukt voor de resterende levensjaren

van de patiënt. Hormonale behandeling lijkt dus een effectief en veilig alternatief voor

opereren bij geselecteerde oudere patiënten, maar er moet wel rekening gehouden

worden met de mogelijke negatieve gevolgen van het weglaten van chirurgie op de

langere termijn, bijv. doordat de tumor ongevoelig wordt voor de hormoonbehandeling.

Een interessant bijkomend gegeven is dat bij één‐derde van de patiënten het weglaten

van de operatie niet om medische redenen gebeurde maar op verzoek van de patiënt.

Hoofdstuk 1 tot en met 3 hebben betrekking op de behandeling van borstkanker in het

ziekenhuis. Daaraan vooraf gaat een verwijzing naar het ziekenhuis. Hoofdstuk 4 beschrijft

de uitkomsten van een enquête gericht op het verwijsbeleid van specialisten

ouderengeneeskunde bij verdenking op borstkanker. Specialisten ouderengeneeskunde

zijn de artsen die de basiszorg leveren aan patiënten die verblijven in een verpleeghuis.

Bijna 60% van de respondenten meldde in het afgelopen jaar tenminste één patiënt met

een mogelijke borstkanker te hebben behandeld; hiervan werd een‐derde niet

doorverwezen voor nadere diagnostiek of behandeling. De belangrijkste redenen daarvoor

waren een eindstadium van dementie, de wens van de patiënt of de familie en een

beperkte levensverwachting. Het bezoek aan het ziekenhuis werd veelal te belastend

geacht voor deze kwetsbare patiëntengroep. In dit hoofdstuk worden ook mogelijke

alternatieven voor ziekenhuiszorg besproken.

Medici baseren hun behandelkeuzes bij voorkeur op bewijzen verkregen uit klinische

studies. Het sterkste bewijs komt uit zogeheten gerandomiseerde studies, waarbij

patiënten niet op basis van patiëntkenmerken maar juist door toeval aan een bepaalde

behandeling worden toegewezen. Dit zorgt ervoor dat mogelijke bijkomende factoren en

patiënteigenschappen zo gelijkmatig mogelijk over de verschillende behandelgroepen

verdeeld worden. Om diverse redenen, waaronder ethische, is deze vorm van onderzoek

niet altijd mogelijk. In dat geval zijn beschrijvende studies, zoals weergegeven in de

voorgaande hoofdstukken, een alternatief; deze kunnen waardevolle aanvullende

informatie verschaffen. Ook als het wel mogelijk is om een gerandomiseerde studie te

ontwerpen, kan het moeilijk zijn om hiervoor voldoende patiënten te werven. Dat was ook

het geval bij de OMEGA studie, die werd opgezet voor kwetsbare ouderen met een

uitgezaaide vorm van borstkanker (zie ook Hoofdstuk 10). Vanwege teleurstellende

patiëntenaantallen werd onderzoek ingezet naar mogelijk obstakels in het werven van

patiënten (Hoofdstuk 5). Het belangrijkste obstakel bleek te zijn dat de conditie van de

220


patiënt als te goed of te slecht werd beschouwd voor deelname. Daarnaast wilden veel

patiënten zelf niet mee doen, of hadden zij een sterke voorkeur voor één van de twee

behandelopties waardoor randomisatie van de behandeling niet gewenst was. Deze

uitkomsten laten zien hoe moeilijk het is om een goede studie te ontwerpen voor ouderen

en benadrukken de noodzaak van voorlichting aan patiënten over het belang van

deelname aan wetenschappelijk onderzoek.

In Deel II is de huidige behandeling en besluitvorming bij twee andere soorten kanker

onderzocht: darmkanker en hoofd‐halskanker. In Hoofdstuk 6 werd de behandeling van

darmkanker bij 183 patiënten ouder dan 70 jaar vergeleken met die van 103 patiënten

jonger dan 70 jaar. Van de oudere patiënten kreeg 91% een chirurgische behandeling

zoals geadviseerd door de richtlijnen, vergeleken met 100% van de jongere patiënten.

Voor chemotherapie waren deze percentages respectievelijk 32% en 85%. De belangrijkste

redenen om af te wijken van de richtlijnen waren leeftijd, bijkomende ziektes en de

algehele conditie van de patiënt. Ondanks het feit dat alleen vitale patiënten de standaard

behandeling kregen, ontwikkelden ouderen patiënten meer complicaties na de operatie

en was de sterfte in de eerste maand na de ingreep aanzienlijk hoger. Bij chemotherapie

traden evenveel bijwerkingen op ondanks zorgvuldige selectie van geschikte patiënten en

minder agressieve behandeling voor de oudere patiënten. Deze resultaten tonen aan dat

leeftijd wel degelijk van belang is voor het vermogen van de patiënt om een behandeling

te doorstaan; dit aspect mag dus niet buiten beschouwing kunnen worden gelaten in de

besluitvorming.

Eén van de symptomen van darmkanker is ijzergebrek. Omdat ijzer een belangrijke

bouwstof is van rode bloedcellen, zal ijzergebrek uiteindelijk leiden tot bloedarmoede, ook

wel anemie genoemd. Bij 2‐15% van de patiënten met ijzergebreksanemie is kanker in het

maagdarmkanaal de oorzaak. Kanker veroorzaakt schade aan de binnenkant van de maag

of darm, waar al dan niet ongemerkt bloed uit kan siepelen; hierdoor raakt de

ijzervoorraad in het lichaam uitgeput. Om deze vormen van kanker vast te stellen wordt

de binnenkant van het maagdarmkanaal bekeken tijdens een zogeheten coloscopie (darm)

of gastroscopie (maag). Als daarbij kanker wordt vastgesteld, is chirurgisch ingrijpen de

volgende stap. Omdat niet alle patiënten fit genoeg zijn voor een operatie, of deze

behandeling niet willen, wordt soms afgezien van verdere diagnostiek als een

ijzergebreksanemie wordt vastgesteld. In Hoofdstuk 7 is gekeken naar een groep van 471

patiënten van 80 jaar en ouder waarbij in het laboratorium van het Gelre ziekenhuis in

Apeldoorn ijzergebreksanemie werd vastgesteld. Bij 59% van deze patiënten werd geen

verder onderzoek verricht; dit gebeurde vaker bij vrouwen, bij een hogere leeftijd, bij

minder ernstige bloedarmoede en bij het ontbreken van maagdarmklachten. Bij 61

patiënten werd maagdarmkanker vastgesteld, waarvan 16 pas in tweede instantie omdat

aanvankelijk was gekozen werd voor een terughoudend diagnostisch beleid. Bij patiënten

221

Chapter 14

met darmkanker was de sterfte ten gevolge van de operatie 15% en pas na 1.3 jaar werd

de toegevoegde waarde van opereren ten opzichte van patiënten bij wie van deze

operatie werd afgezien zichtbaar in de overlevingsgrafieken.

Hoofdstuk 8 richt zich op 606 patiënten met hoofd‐halskanker, die behandeld werden in

het Medisch Centrum Alkmaar. Hoofd‐halskanker heeft vaak een ingrijpend effect op de

kwaliteit van leven van een patiënt, omdat het een verstoring kan geven van heel basale

lichaamsfuncties zoals slikken, ademen en spreken. De belasting van de behandeling kan

echter ook zeer groot zijn. Een bijkomende complicerende factor is dat hoofd‐halskanker

vaak een gevolg is van alcoholmisbruik en roken. Hierdoor treden in deze patiëntengroep

vaak gelijktijdig meerdere levensstijl‐gerelateerde ziekten op; deze ziekten kunnen de

belastbaarheid van de patiënten sterk beïnvloeden. In ons onderzoek zagen wij dat 91%

van alle patiënten behandeld werd volgens de richtlijnen, met een leeftijdsgerelateerde

afname. De belangrijkste redenen om van de richtlijnen af te wijken bij jongere patiënten

waren bijkomende ziektes en gebrek aan medewerking van de patiënt terwijl voor oudere

patiënten het afwijken van de richtlijn vooral op verzoek van de patiënt gebeurde. Na vijf

jaar was 64% van de patiënten die volgens de richtlijn werden behandeld nog in leven,

maar slechts 15% van de patiënten met een afwijkende behandeling; dit toont aan hoe

belangrijk het volgen van de richtlijnen voor deze patiëntengroep is.

In Deel III is de mogelijke toegevoegde waarde onderzocht van twee concepten uit de

geriatrie voor de oncologische besluitvorming. Het eerste concept is frailty, ofwel

kwetsbaarheid; in de geriatrie wordt dit gedefinieerd als een toestand van verminderde

reserves, die ontstaat door de opeenstapeling van veroudering van diverse

orgaansystemen, met als gevolg een verminderd vermogen om ziektes op te vangen.

Omdat zowel kanker als kankerbehandeling een forse belasting betekenen voor het

lichaam en een patiënt dwingen om zijn reserves bij te zetten, lijkt het concept van frailty

bijzonder relevant in de oncologie. Een tweede concept is het uitgebreid geriatrisch

onderzoek, ofwel comprehensive geriatric assessment (CGA), waarbij op een

systematische wijze de gezondheidstoestand van een patiënt in kaart wordt gebracht, met

aandacht voor lichamelijke, psychologische, sociale en functionele aspecten. Het wordt

gebruikt om geriatrische problemen op te sporen, zoals dementie, stemmingsstoornissen,

ondervoeding, beperkingen in de zelfredzaamheid of mobiliteit en ook frailty.

In Hoofdstuk 9 werd dit CGA gebruikt in een groep van 292 patiënten van 65 jaar en ouder

met een actieve vorm van kanker die via de spoedeisende hulp werden opgenomen op

één van de interne geneeskunde afdelingen van drie ziekenhuizen. Hierbij bleek dat

geriatrische problematiek veelvuldig voorkomt. Bij bijna de helft van deze patiënten was

sprake van ondervoeding, mobiliteitsproblemen, een overbelaste mantelzorger en

polyfarmacie (het gebruik van meer dan vijf verschillende soorten medicatie) terwijl bij

meer dan twee‐derde sprake was van stemmingsstoornissen, pijn en beperkte

222


zelfredzaamheid. Het opsporen van deze aandoeningen kan een aangrijpingspunt zijn voor

het optimaliseren van de gezondheidstoestand en de kwaliteit van leven van een patiënt.

Ons onderzoek toonde echter ook aan dat voor deze specifieke patiëntenpopulatie

geriatrische factoren niet bepalend zijn voor de prognose; hiervoor zijn

kankergerelateerde factoren van veel groter belang.

Hoofdstuk 10 beschrijft de resultaten van de OMEGA studie, waarin 78 patiënten van 65

jaar en ouder met uitgezaaide borstkanker werden behandeld met twee soorten

chemotherapie. Genezing van uitgezaaide kanker is niet mogelijk en het doel van deze

chemotherapie is dan ook vooral om de kwaliteit van leven in de resterende levenstijd te

verbeteren. Omdat ernstige bijwerkingen van de chemotherapie die kwaliteit juist nadelig

kunnen beïnvloeden en tot vroegtijdig overlijden kunnen leiden, zou het waardevol zijn als

vooraf voorspeld kon worden bij welke patiënten ernstige bijwerkingen te verwachten

zijn. In ons onderzoek kregen alle patiënten voor aanvang van de chemotherapie een CGA.

Daarbij werd gezien dat de kans op ernstige bijwerkingen bijna lijnrecht steeg met het

aantal geriatrische aandoeningen bij de patiënt; van patiënten zonder zulke aandoeningen

had slechts 18% ernstige bijwerkingen, maar dit percentage steeg naar 50% bij twee

geriatrische aandoeningen en 80% bij drie of meer. Bovendien bleek ook de overleving van

de patiënten gerelateerd aan geriatrische problematiek.

Hoofdstuk 11 geeft een overzicht van 37 eerder gepubliceerde studies naar de

voorspellende waarde van het CGA voor overleving en belastbaarheid – zowel wat betreft

chemotherapie als chirurgie ‐ die tot nu toe verricht zijn bij oudere kankerpatiënten.

Helaas zijn de uitkomsten van deze 37 studies tegenstrijdig. Daarnaast lijkt het erop dat

verschillende factoren voorspellend zijn per bestudeerde uitkomstmaat: frailty

(gedefinieerd als de aanwezigheid van één of meer geriatrische problemen), ondervoeding

en comorbiditeit waren voorspellend voor overleving; frailty was ook voorspellend voor

de toxiciteit van chemotherapie; cognitieve stoornissen en beperkingen in de

zelfredzaamheid waren voorspellend voor het voltooien van alle geplande chemokuren;

en beperkingen in de zelfredzaamheid waren voorspellend voor complicaties na

oncologische operaties. Het gebrek aan eenduidigheid in de resultaten van deze studies

zou het gevolg kunnen zijn van het brede scala aan patiëntenpopulaties waarin deze

onderzoeken verricht werden. Al met al lijken geriatrische problemen wel voorspellende

waarde te hebben voor de oudere patiënt met kanker maar is het nog te vroeg om de

besluitvorming rond de kankerbehandeling uitsluitend hierop te baseren.

Omdat het uitvoeren van een CGA tijdrovend is, zijn kankerbehandelaars op zoek gegaan

naar een korte vragenlijst of screening tool die gebruikt kan worden om patiënten te

selecteren voor uitgebreider onderzoek. De ideale screeningstool heeft een zeer hoge

sensitiviteit, wat betekent dat alle kwetsbare patiënten geïdentificeerd worden, en ook

een hoge specificiteit, wat betekent dat zo min mogelijk fitte patiënten ten onrechte als

223

Chapter 14

224

kwetsbaar worden aangewezen. Hoofdstuk 12 geeft een overzicht van de 14 studies die

tot op heden over dit onderwerp verricht zijn. De Geriatric 8 (G8) en Triage Risk Screening

Tool hadden de beste sensitiviteit voor kwetsbaarheid, maar misten specificiteit;

omgekeerd hadden die vragenlijsten met een goede specificiteit een te geringe

sensitiviteit. Gezien deze uitkomsten lijkt een een twee‐traps‐CGA, waarin een

screeningstool wordt gebruikt om patiënten te selecteren voor uitgebreider onderzoek,

niet haalbaar met de nu beschikbare vragenlijsten.

In Hoofdstuk 13 wordt uitgebreider ingegaan op de resultaten van de voorgaande

onderzoeken en besproken welke consequenties deze hebben voor de patiëntenzorg en

toekomstig onderzoek. Samenvattend blijkt uit de studies in dit proefschrift dat zorg op

maat de norm zou moeten zijn in de behandeling van ouderen met kanker, waarbij de

balans gezocht wordt tussen onderbehandeling en overbehandeling en waarbij rekening

wordt gehouden met de verscheidenheid aan patiënten die in deze populatie wordt

gezien. Het uitgebreid geriatrisch onderzoek kan belangrijke aanvullende informatie

verschaffen over de gezondheidstoestand van de patiënt maar de exacte rol die dit

onderzoek kan vervullen in de besluitvorming rond de behandeling van ouderen met

kanker moet nog worden bepaald.

Author Affiliations

Author Affiliations

T. Acampo, MD Department of Geriatric Medicine, Gelre Hospitals, Apeldoorn

W.P. Achterberg, MD PhD Department of public Health and Primary Care, Leiden University Medical Centre, Leiden

E. Bastiaannet, MSc PhD Department of Surgery and Department of Gerontology and Geriatrics, Leiden University Medical Centre, Leiden

E. Boven, MD PhD Department of Medical Oncology, Vrije Universiteit Medical Centre, Amsterdam

J.J. Braun, MD PhD Department of Internal Medicine Vlietland Hospital, Schiedam

N.G. Breeuwsma, MD Department of Pathology, Medical Centre Alkmaar, Alkmaar

R.J. Bun, MD Department of Oral and Maxillofacial Surgery, Medical Centre Alkmaar, Alkmaar

B.M. Buurman, MSc PhD Department of Internal Medicine, section of Geriatric Medicine, Academic Medical Centre, Amsterdam

D. Evers, MD Department of Internal Medicine, Haga Hospital, the Hague

L. van der Geest Comprehensive Cancer Centre, the Hague

H. de Graaf, MD PhD Department of Medical Oncology, Medical Centre Leeuwarden, Leeuwarden

S.M. de Groot Dutch Breast Cancer Trialists' Group BOOG/Comprehensive Cancer Centre, Amsterdam

V.C. Hamelinck, MSc Department of Surgery and Department of Gerontology and Geriatrics, Leiden University Medical Centre, Leiden

F.E. de Jongh, MD PhD Department of Internal Medicine, Ikazia Hospital, Rotterdam

J.M. Jonker, MD Department of Geriatric Medicine, Slotervaart Hospital, Amsterdam

I.M.J.A. Kuper, MD Department of Geriatric Medicine, Slotervaart Hospital, Amsterdam

G.T. de Kuyper, MD Department of Otolaryngology, Medical Centre Alkmaar, Alkmaar

A.E. van Leeuwen‐Stok, PhD Dutch Breast Cancer Trialists' Group BOOG, Amsterdam

G.J. Liefers, MD PhD Department of Surgery, Leiden University Medical Centre, Leiden

228

Author Affiliations

M. Los, MD PhD Department of Internal Medicine, St. Antonius Hospital, Nieuwegein

E. Maartense, MD PhD Department of Internal Medicine, Reinier de Graaf Hospital, Delft

B.C. van Munster, MD PhD Department of Internal Medicine, section of Geriatric Medicine, Academic Medical Centre, Amsterdam

Department of Geriatric Medicine, Gelre Hospitals, Apeldoorn

J.W.R. Nortier, MD PhD Department of Medical Oncology, Leiden University Medical Centre, Leiden

H.A. Paling, MD Department of Geriatric Medicine, Gelre Hospitals, Apeldoorn

J.E.A. Portielje, MD PhD Department of Internal Medicine, Haga Hospital, The Hague

A. Pronk, MD PhD Department of Surgery, Diakonessenhuis, Utrecht

J.A. Remijn, PhD Department of Clinical Chemistry and Hematology, Gelre Hospitals, Apeldoorn

S.E. de Rooij, MD PhD Department of Internal Medicine, section of Geriatric Medicine, Academic Medical Centre, Amsterdam

J.G. Schrama, MD PhD Department of Internal Medicine, Spaarne Hospital, Hoofddorp

W.H. Schreurs, MD PhD Department of Surgery, Medical Centre Alkmaar, Alkmaar

C. Seynaeve, MD PhD Department of Medical Oncology, Erasmus University Medical Centre – Daniel den Hoed Cancer Centre, Rotterdam

M. Soesan, MD Department of Internal Medicine, Slotervaart Hospital, Amsterdam

H.J. van Slooten, MD PhD Department of Pathology, Medical Centre Alkmaar, Alkmaar

C.H. Smorenburg, MD PhD Department of Internal Medicine, Medical Centre Alkmaar, Alkmaar

H. van Tinteren Antoni van Leeuwenhoek Hospital/NKI, Amsterdam

C.R. Tulner, MD PhD Department of Geriatric Medicine, Slotervaart Hospital, Amsterdam

S.A.C van Tuyl, MD PhD Department of Gastro‐enterology, Diakonessenhuis, Utrecht

229

Author Affiliations

230

J.M. Uppelschoten, MD Department of Radiotherapy, Medical Centre Alkmaar, Alkmaar

A.G. Vos, MD Department of Internal Medicine, Diakonessenhuis, Utrecht

W. van de Water, MD Department of Surgery, Leiden University Medical Centre, Leiden

B.M. Wiarda, MD Department of Radiology, Medical Centre Alkmaar, Alkmaar

M. Wymenga, MD PhD

Department of Medical Oncology, Medisch Spectrum Twente, Enschede

E.S. van der Zaag, MD PhD Department of Surgery, Gelre Hospitals, Apeldoorn

A.M. Zeilemaker, MD Department of Surgery, Rijnland Hospital, Leiderdorp

Curriculum Vitae

Curriculum Vitae

Marije Emilie Hamaker was born on January 26th 1976 in Assen, the Netherlands. For most

of her youth, she lived in a little village in Drenthe called Lhee, but spent her early teens in

Hilo, Hawaii. After graduating from high school in 1994, she spent a year travelling to

South America and teaching at a primary school in Zimbabwe. From 1995 to 1996 she

studied psychology and cultural anthropology at the University of Amsterdam before

finally getting into medical school in 1996 at the same university. She completed her

medical training in 2003. Having considered almost every specialty she came into contact

with as a medical student, Marije finally settled on geriatric medicine, first as an AGNIO

and then as of 2005, as a resident. During a rotation at the oncology department of the

Medical Centre Alkmaar, she met Carolien Smorenburg, whose enthusiasm for geriatric

oncology was the inspiration for this thesis.

Since completing her training as a geriatrician in the fall of 2009, Marije has been working

at the Diakonessenhuis in Utrecht and Zeist. She is the secretary of the special interest

group Geriatric Oncology of the Dutch Society for Geriatric Medicine (NVKG) and a

member of the Dutch Geriatric Oncology Foundation (GeriOnNe).

Acknowledgements

(Dankwoord)

Acknowledgements (Dankwoord)

Als achtjarig meisje leerde ik voor het eerst over statistiek van mijn grootvader, Hugo

Christiaan Hamaker (1905‐1993). Hij had een grote bak met gele en rode kraaltjes en een

schep met 10 bij 10 gaatjes. Spelenderwijs bracht hij mij daarmee de beginselen bij van

steekproeven nemen en kansberekening. Zo leerde ik dat statistiek iets leuks is (en niet

lleen maar moeilijk en ongrijpbaar). Wellicht is het toen allemaal begonnen. a

Het vervolg liet wel een tijdje op zich wachten. Hoewel het toeval in de wetenschap zo

veel mogelijk moet worden buitengesloten, was dit proefschrift zonder toeval niet tot

stand gekomen. Terugkijkend op de zes jaar waarin dit boekje zich langzaam vulde – eerst

als losse stukjes onderzoek en sinds 2010 als officieel promotietraject – realiseer ik mij dat

het vooral toevallige ontmoetingen zijn geweest die mij tot dit hier hebben gebracht. Je

kunt nog zo veel willen, of nog zo hard werken, maar zonder de juiste mensen kom je er

iet. Bij een aantal van hen wil ik graag stilstaan. n

Carolien Smorenburg – wie had kunnen denken dat er een proefschrift voort zou komen

uit mijn terloopse vraag in het najaar van 2006 of je nog ideeën had voor “iets” met

kanker en ouderen. Jouw enthousiasme voor de geriatrische oncologie blijft inspirerend.

Maar daarnaast ben jij voor mij een voorbeeld geweest van het soort dokter dat ik wilde

worden. Jouw vermogen om zonder veel woorden je betrokkenheid en empathie op een

patiënt over te brengen vind ik nog altijd heel bijzonder. Dank je wel voor je gulheid en

lles wat jij voor mij mogelijk hebt gemaakt. a

Professor Sophia de Rooij – dankzij jou (of is het nu u?) had ik in de zomer van 2009

onverwachts een promotieplan. Eigenlijk kwam ik praten over een ander onderzoek maar

jij was duidelijk: met al twee gepubliceerde studies binnen de geriatrische oncologie was

het zonde om met iets nieuws te beginnen. Ik was die dag voor het eerst getuige van jouw

creativiteit en indrukwekkende vermogen om vage ideeën om te zetten in concrete

onderzoeksvoorstellen: binnen een paar minuten had ik een blaadje vol met plannen. Het

duurde even voordat onze wederzijdse verwachtingen over de invulling van onze

samenwerking op één lijn zaten maar we zijn een goed team geworden. Heel leuk dat ik je

p de valreep als promotor mag noteren! o

Barbara van Munster – wat ben ik blij dat jij je aan mij hebt opgedrongen! Dank je wel

voor al je steun, vertrouwen en begeleiding, voor je uitleg over de politiek van het

publiceren en vooral voor je kritische blik. Jouw vermogen om de vinger te leggen op de

zwakke plekken in mijn plannen en artikelen is de kwaliteit van dit proefschrift en van mij

als onderzoeker zeer ten goede gekomen. Uiteindelijk kregen we er een “methodologically

236


impeccable” van een reviewer voor terug, een compliment dat vooral jou toekomt. Er

ggen nog diverse plannen op de plank en ik verheug mij op onze verdere samenwerking. li

Professor Marcel Levi – dank u wel dat u deze promotie mogelijk heeft gemaakt. Hoewel

uw naamverbintenis met dit proefschrift aanvankelijk vooral praktische redenen had,

realiseerde ik mij pas later hoe betrokken u bent bij de discussie over (te lang)

doorbehandelen vs onderbehandelen. Ik verheug mij erop die discussie op 30 november

oort te zetten. v

Ik wil graag alle leden van de leescommissie hartelijk bedanken voor hun bereidheid dit

anuscript inhoudelijk te beoordelen en te opponeren. m

Bij het opstellen van een lijstje met mede‐auteurs besefte ik mij pas met hoeveel

verschillende mensen ik heb mogen werken: het bleken er uiteindelijk 44 te zijn! Allemaal

heel veel dank voor de prettige samenwerking. In het bijzonder wil ik noemen: Judith

Jonker, oud‐collega, sparring partner, mede‐auteur en vriendin: het is leuk hoe we al deze

aspecten hebben weten te combineren. Dat er nog vele gezellige etentjes, onderzoeks‐

projecten en nieuwe inzichten mogen volgen! Esther Bastiaannet, ik kon aanvankelijk

moeilijk geloven dat iemand zo zonder dubbele agenda wetenschap kon bedrijven. Het

was me een waar genoegen om met je samen te werken en ideeën uit te wisselen. To be

continued...? Alinda Vos: wat heb jij hard gewerkt voor die review, daar kan ik je niet

genoeg voor bedanken. Je bent een groot talent en ik ben heel benieuwd waar dat

allemaal toe gaat leiden. Johanneke Portielje, ik hoop op nog vele stevige, vruchtbare

discussies en samenwerkingsprojecten binnen en buiten GeriOnNe. Ook wil ik Bianca

Buurman en alle medewerkers van de DEFENCE studie bedanken voor het mogen

gebruiken van hun data, evenals Caroline Seynaeve, Elise van Leeuwen, Harm van

interen, Steffen de Groot en alle medewerkers van de OMEGA studie. T

Ja

n Willem Broek – heel veel dank voor de prachtige lay‐out! Ben er superblij mee.

Voor alle leden van de special interest group Geriatrische Oncologie: het ontwikkelen van

een geriatrische visie op oncologische zorg en deze naar buiten brengen blijkt niet altijd

eenvoudig maar samen staan we sterk! Het is elke keer weer leuker om daarover van

gedachten te wisselen. Dank daarvoor, en dank ook voor jullie input bij de review over de

voorspellende waarde van het CGA. Dr. Huub Maas – als partner‐in‐crime in de

geriatrische oncologie ben je steeds aan de zijlijn aanwezig geweest. Dank je wel voor het

geven van het goede voorbeeld! Ons artikel is er niet echt van gekomen dus bij deze nog

237


maar eens een porretje. Truus Schuurman – in 2003 was jij als mijn zaalsupervisor in het

Slotervaartziekenhuis het bewijs dat een pittig karakter en hoge hakken prima samengaan

met de zorg voor kwetsbare ouderen. Ik vind het heel leuk dat wij elkaar nu via de SIG

eer tegen zijn gekomen. w

Dr. Victor Umans – zonder de stimulans om (toch enigszins tegen mijn zin) in 2006 met dat

case‐report over het Eisenmenger syndroom aan de slag te gaan, waren mijn

wetenschappelijke ambities waarschijnlijk in hun winterslaap verzonken gebleven.

Grappend zei ik toen dat als er ooit een proefschrift zou komen u een eervolle vermelding

verdiende maar ik denk niet dat één van ons in die mogelijkheid geloofde. Bij dezen wil ik

nogmaals heel hartelijk danken voor dat zetje richting onderzoek. u

Mijn (oud)collega’s uit het Medisch Centrum Alkmaar en het AMC: het was fijn om altijd

gezelligheid te vinden als ik langskwam voor een afspraak of om statussen in te zien. Dank

wel daarvoor! je

Dan al mijn collega’s in het Diakonessenhuis: wat heb ik het getroffen! Prettige collega’s,

goede sfeer, een geweldig team om in te werken… Wat wil een mens nog meer? Richard

Faaij, ik heb veel respect voor de ruimte die jij mij het gegeven om me te bemoeien met

“jouw kindje”. Dank je wel voor al je flexibiliteit die het voortzetten van mijn

promotietraject mogelijk heeft gemaakt. Meike Prins, ooit mijn coassistent, toen

vriendinnen geworden, nu mijn collega en wie weet volgend jaar ook mijn maat: ik

verheug mij bijzonder op onze samenwerking in het uitbouwen van ons geriatrisch

centrum midden‐Nederland. Peter Thunnissen, René van der Griend en Daan ten Bokkel

Huinink: doen jullie mee? Frans Boerenboom, Bas van Tuyl en Apollo Pronk: op het

snijvlak van jullie specialismen en het mijne ligt een goudmijn aan onderzoeksvragen. Ik

verheug mij op de kans om dit samen verder te verkennen. Femke Meulendijks, Patricia

Hamers, Lizelot(t)e van Beek: heel leuk om het stokje nu aan jullie door te geven! Rob

Gallas, jij verdient een speciale vermelding: jouw overtuiging dat het voor mij niet

haalbaar zou zijn om starten als vrijgevestigde te combineren met promotieonderzoek, is

uitermate prikkelend geweest en mijn wens jouw ongelijk te bewijzen heeft mij door

eerdere moedeloze momenten heen geholpen. m

Bij de verdediging van mijn proefschrift word ik bijgestaan door mijn paranimfen: Nadine

Vieleers en Karen Farrington. Lieve Nadine – het was geen vriendschap op het eerste

gezicht toen wij elkaar ontmoeten tijdens de introductieweek van geneeskunde in 1996.

Gelukkig bleken onze cultuurverschillen prima overbrugbaar door een

gemeenschappelijke interesse in reality series en ATWT en je bent een van mijn beste

238


239

vriendinnen geworden. Het blijft voor mij bijzonder dat jij lang voordat ik het wilde horen

al besloten had dat geriatrie het beste bij mij paste. Ik geloof nog steeds niet dat ik daar

zelf ooit aan gedacht zou hebben. Zonder jou was er voor mij geen geriatrie geweest en

dus ook geen geriatrische oncologie: dat alleen zou reden genoeg zijn geweest om je te

vragen mijn paranimf te zijn… Lieve Karen – in gedachten zie ik ons nog zitten (eind 1995)

bij jou op de bank, papieren op schoot, rode pen in de hand, geconcentreerde

(gepijnigde?) blik, krassend in elkaars werkstukken. Elke drie weken moesten wij vijf

A4tjes volschrijven over een antropologisch boek en dus hebben we heel wat keren zo

gezeten. Samen met jou leerde ik de lol van het schrijven en aangezien onze docent vooral

keek naar het uiterlijk van het werkstuk, verdien jij veel van de eer van alles wat ik toen

geleerd heb. Om die reden, en alles wat we in de 17 jaar vriendschap erna hebben

gedeeld, wilde ik graag dat jij mijn paranimf zou zijn. Dank jullie beiden voor jullie

riendschap en steun, en natuurlijk ook de hulp bij het maken van de stellingen. v

Twee andere vriendinnen mogen hier niet ontbreken: Karien Hoogenboezem – in alle

drukte van opleiding, stages, onderzoek en werk ben jij met je gezin al vele jaren het

rustpunt. Na zo’n weekje kletsen, spelletjes spelen en knuffelen met mijn neefjes kan ik

weer met nieuwe energie aan de slag. Het is heerlijk een plek te hebben waar ik mij zo

welkom weet. Anneke Geel – it’s so great to have my friend of over 25 years here for this

occasion. Thank you so much for coming! I just realized that the day of my defense will be

the first time that all four of my oldest friends are together in one place. Makes the day

xtra special… e

Voor mijn ouders: jullie zijn een prachtig bewijs dat leeftijd niets zegt over hoe oud je

bent. Dank je wel voor alles wat jullie me hebben meegegeven. Het voelt goed om te

weten dat ik altijd op jullie kan rekenen. En natuurlijk ook bedankt voor de hulp met de

Nederlandse samenvatting. Jasper, het was zo fijn op te groeien met iemand zoals jij als

grote broer. Dat zou ik ieder meisje gunnen. Lief dat je vanuit Lagos komt om erbij te zijn.

_ _ _ _ _ _

Een promotie lijkt een afsluiting. Dat is het ook wel, natuurlijk. Maar er liggen nog zoveel

onbeantwoorde vragen … Hoe meer we kunnen, des te prangender wordt de vraag of we

dat ook moeten inzetten, en zo ja, wanneer en bij wie. Alle reden voor mij om door te

gaan met onderzoek. Wie doet er mee?

decision making in geriatric oncology · between overtreatment and undertreatment and must be...

Documents