systematic review of lhrh agonists for the adjuvant treatment of early breast cancer

ARTICLE IN PRESS

The Breast (2005) 14, 181–191

THE

BREAST

KEYWORDLHRH agonBreast CarAdjuvant t

0960-9776/$ - sdoi:10.1016/j.b

�CorrespondiE-mail addr

www.elsevier.com/locate/breast

REVIEW

Systematic review of LHRH agonists for theadjuvant treatment of early breast cancer

Rohini Sharma, Jane Beith, Anne Hamilton�

Department of Medical Oncology, Sydney Cancer Center, Missenden Road, Camperdown,2050, NSW, Australia

Received 31 August 2004; received in revised form 10 January 2005; accepted 3 February 2005

Sists;cinoma;herapy

ee front matter & 2005reast.2005.02.001

ng author. Tel.: +61 2 9ess: Anne.Hamilton@cs

Summary There is increasing use of luteinising hormone-releasing hormone(LHRH) agonists in the adjuvant treatment of breast cancer (J. Clin. Oncol. 19(2)(2001) 343). However, few mature studies are available and there is uncertaintyregarding the optimal use of these agents. We performed a systematic review toaddress the role of LHRH agonists in the adjuvant treatment of pre-menopausalwomen with early breast cancer. As ovarian suppression is unlikely to benefit womenwith ER-negative tumours, the review is limited to women with ER-positive disease.The objectives of this review were to address the following issues; the role of LHRHagonists compared to tamoxifen (TAM), LHRH agonists in place of chemotherapy andLHRH agonists integrated into chemo-hormonal regimens. We identified 11randomised trials. In three trials, adjuvant suppression of ovarian function usingLHRH agonists, with or without TAM, had similar benefits at 5–6 years follow-up interms of disease-free survival (DFS) and overall survival (OS) to adjuvant CMFchemotherapy (J. Clin. Oncol 20(24) (2002) 4628; J. Natl. Cancer Inst. 95(24) (2003)1833; Anticancer Res. 22 (2002) 2325; In: San Antonio Breast Cancer Symposium, SanAntonio, TX, 2003, Abstr 40). These findings suggest that ovarian suppression usingLHRH agonists (7TAM) is a reasonable alternative to CMF chemotherapy in womenwith oestrogen receptor (ER) positive tumours. The role of chemotherapy in additionto LHRH agonists is not clearly defined and mature results of four trials are awaited(J. Clin. Oncol. 20(24) (2002) 4621; J. Clin. Oncol. 18(14) (2000) 2718; Proc. Am. Soc.Clin. Oncol. 2000, Abstr 279; Proc. Am. Soc. Clin. Oncol. 20 (2001) Abstr 104; Proc.Am. Soc. Clin. Oncol. 2001, Abstr. 1777). Data is also inadequate at the time ofpublication to inform decisions about the efficacy of LHRH agonists in comparisonwith TAM for the treatment of ER-positive early breast cancer (Proc. Am. Soc. Clin.Oncol. 21 (2001) Abstr. 103; Eur. J. Surg. Oncol. 28(5) (2002) 505; Proc. Am. Soc.Clin. Oncol. 22 (2003), Abstr. 15).& 2005 Elsevier Ltd. All rights reserved.

Elsevier Ltd. All rights reserved.

515 5894; fax: +61 2 9519 1546..nsw.gov.au (A. Hamilton).

www.elsevier.com/locate/breast

ARTICLE IN PRESS

R. Sharma et al.182

Background

Breast cancer is a major cause of morbidity andmortality. Approximately 25% of women diagnosedand treated for breast cancer will eventually die ofthe disease, those younger than 35 years of age andthose who have hormone sensitive disease havingan even worse prognosis.1 Approximately 60% ofbreast tumours in pre-menopausal women arehormone sensitive and these patients are candi-dates for hormonal treatment.2 This can beachieved by blocking oestrogen receptors withdrugs such as tamoxifen (TAM), suppressing oestro-gen synthesis with LHRH3 analogues, or by ovarianablation either, surgically or using radiotherapy.

Chemotherapy induces amenorrhoea in approxi-mately 60% of pre-menopausal women who receiveadjuvant treatment.4 Disease-free survival (DFS) isimproved in women with hormone-sensitive diseasewho become amenorrhoeic following chemother-apy than in those who do not.5,6 This suggests thatthe efficacy of chemotherapy in pre-menopausalwomen may in part be endocrine mediated.Because of the lack of efficacy of hormonal therapydemonstrated in ER-negative women in these trials,ER-negative women were excluded from the cur-rent systematic review.7–9

There is an increasing role for LHRH agonists inthe adjuvant treatment of early breast cancer. As

Table 1 Studies addressing integration of LHRH agonists

Study name Research group Design

LHRH vs. TAMNorwegian study Norway (1) (7

(2) (7

ZIPP International (1) (7(2) (7(3) (7(4) (7

LHRH vs. LHRH+TAMZIPP International (1) (7

(2) (7(3) (7(4) (7

E5188/INT 0101 USA (1) CA(2) CA3) CAF

TAM vs. LHRH+TAMZIPP International (1) (7

2) (73) (74) (7

the amenorrhoea induced by LHRH agonists is notpermanent, they are an important pharmacologicalgroup for the treatment of younger women whowish to maintain their fertility. However, there isuncertainty regarding their optimal role. Thissystematic review locates, organises and sum-marises the available evidence to identify what isknown now, what will become known soon, andwhat needs to be studied in the future. Inparticular this review will address the efficacy ofLHRH agonists compared to TAM, the use of LHRHagonists in place of chemotherapy and the integra-tion of LHRH agonists into chemo-hormonal regi-mens (Table 1).

Methodology

A search was performed by the Secretariat of theCochrane Collaboration Breast Cancer ReviewGroup (see Search strategy and selection criteria)to identify randomised controlled trials eligible forinclusion. The eligibility criteria were applied toeach trial by two independent reviewers. Anydiscrepancies were resolved by a third reviewer.Trial methods were rated independently by 2reviewers using standardised criteria.12

A single reviewer extracted data describing thetrial and patients’ baseline characteristics. Two

into adjuvant hormonal therapy.

Publication

chemo)-G Soreide et al.5

chemo)-TAM Full publicationMedian follow-up 87 months

chemo)-nil Baum et al.10

chemo)-G Abstractchemo)-TAM Follow-up not reportedchemo)-G+TAM



F Davidson et al.11

F-G Abstract-G+TAM Median follow-up 9.6 yr



ARTICLE IN PRESS

Systematic review of LHRH agonists for the adjuvant treatment of early breast cancer 183

independent reviewers extracted data on out-comes, including overall survival, relapse-freesurvival (RFS) or DFS, quality of life (validated ortrial-specific instruments) and toxicity.

Median survival times, hazard ratios (HR) foroverall and DFS were extracted from the reports orestimated from the survival curves if not reported.Updated data presented at ASCO 2003 and SanAntonio Breast Cancer Symposium 2003 is included.Although unpublished, this information is widelyknown and likely to influence decision-making. Noformal meta-analysis was attempted due to thelack of mature survival data.

Results

Eleven randomised trials were identified that haveaddressed the role of LHRH agonists in the adjuvanttreatment of pre-menopausal women with ER-positive early breast cancer: five were reportedonly as abstracts.10,11,13–15 and six with fullpublications.5,16–20 Studies had high ratings for trialquality, and their conclusions were judged unlikelyto be affected by bias. Readers are referred to thetables for information regarding stratification in-formation and global quality rating (Table 4).

Substitution of LHRH agonists forchemotherapy

Three studies were identified that compared CMFchemotherapy with a LHRH agonist as adjuvanttreatment for early breast cancer; the ZEBRA study,IBSCG 8 and the Table study, including a total of3204 women.16–18 All three studies planned toadminister six cycles of a 28-day schedule of CMF.IBSCG administered classical CMF while the Tablestudy cyclophosphamide administered intrave-nously on day one and day eight whilst the Zebraallowed investigators to choose between an in-travenous or oral route of administration (Table 2).All three trials have been fully published.

The largest of the three trials is the ZEBRA study,which reported on 1614 of 1640 women with stage II,node positive breast cancer, 1189 of whom had ER-positive disease.16 Patients were randomised togoserelin (3.6mg depot sc q28d� two years)(n ¼ 797) or CMF (cyclophosphamide 500mg/m2 ivd1 and 8, or 100mg/m2 orally d1–14, methotrexate40mg/m2 iv d1 and 8, 5FU 600mg/m2 iv d1 and 8)q28d� 6 cycles (n ¼ 817). At 6-years, CMF wassuperior to goserelin in terms of DFS, but not OS.However, in a protocoled sub-analysis, a highly signi-ficant interaction was observed between treatment

and ER status (P ¼ 0:0016). In patients with ER-positive tumours, goserelin was equivalent to CMFfor DFS (HR 1.01; 95% CI 0.84–1.2; P ¼ 0:94) and OS(HR 0.99; 95% CI 0.76–1.28; P ¼ 0:92). However, inthe ER-negative population goserelin was inferior toCMF for DFS and OS. The data for OS was not matureat the time of publication.

Rates of amenorrhoea were higher in patientstreated with goserelin compared to CMF (95% vs.58.6%) after 6 months of treatment. However, at 3years, only 22.6% of patients who had receivedgoserelin remained amenorrhoeic compared with76.9% of patients treated with CMF. The incidenceof menopausal side-effects, hot flushes, vaginaldischarge and vaginal soreness, were similar in bothgroups. However, these side-effects resolved afterceasing goserelin but persisted in the CMF arm. Nograde 3 or 4 symptoms were reported.21 In aseparate study, bone mineral density (BMD) of thelumbar spine (L2–L4) and neck of femur wasassessed in 96 patients from eight centres.22 At 2years, the mean percentage BMD loss at the lumbarspine was worse in women receiving goserelin was�10.5% compared with �6.5% for women receivingCMF (P ¼ 0:0005). BMD loss at the neck of femurwas �6.4% and �4.5% in women receiving goserelinand CMF, respectively (P ¼ 0:04). However, at 3years, mean percentage BMD loss for goserelin andCMF at the lumbar spine was �6.2% and �7.2%(P ¼ 0:26) and at the neck of femur was �3.1% and�4.6%, (P ¼ 0:48). Quality of life was assessed in 86centres.23 An initial improvement in quality of lifefrom baseline was seen in women receivinggoserelin but no difference was seen after 6 monthsof treatment.

IBSCG 8 randomised 1063 patients with nodenegative breast cancer to receive goserelin alone(3.6mg sc monthly for 24 months) (n ¼ 346),classical CMF for six courses (n ¼ 360), or six cyclesof CMF followed by goserelin for 18 months(n ¼ 357).17 A fourth observation arm was discon-tinued in 1992 with 42 patients. Overall, the studyfound no statistically significant differences be-tween the three treatment arms in terms of DFSand OS. However, differences were noted insubgroups defined by ER status. Patients with ERnegative tumours receiving CMF alone had animproved 5 year disease free survival compared tothose randomised to goserelin alone (5-year DFS73% compared to 84%; HR 1.52, 95% CI 0.89–2.58;P ¼ 1:2). No statistically significant benefit wasseen with the addition of goserelin to chemother-apy (5-year DFS 88%). However, patients with ER-positive tumours, 5-year DFS was equivalent forboth those receiving CMF alone and goserelin (5-year DFS 81%; HR 0.97, 95% CI 0.66–1.42; P ¼ 0:86).

ARTICLE IN PRESS

Table 2 Studies addressing integration of LHRH agonists into chemo-hormonal therapy.

Study name Researchgroup

Design Publication

LHRH vs. ChemoZEBRA International (1) G Jonat et al.16

(2) CMF Full publicationMedian follow-up 6 yr

IBCSG 8 International (1) G International Breast Cancer Study Group.200317

(2) CMF Full Publication(3) CMF-G Median follow-up 5.7 yr(4) Nil

LHRH/TAM vs. ChemoItalian breast Italy (1) Ov Supp+T Boccardo F et al. 200020

Cancer adjuvant (2) CMF Full publicationStudy Median follow-up 76 monthsAustrian breast and

colorectal cancer studyAustria (1) G+T Jakesz R et al. 200219

(2) CMF Full publicationMedian follow-up 60 mo

FASG06 (French study) France (1) Tr+T Roche H et al. 200013

(2) FEC50 AbstractMedian follow-up 54 mo

LHRH vs. Chemo-LHRHIBCSG 8 International (1) G International Breast Cancer Study

(2) CMF Group.17

(3) CMF-G Full publication(4) nil Median follow-up 5.7 yr

Chemo vs. Chemo-LHRHIBCSG 8 International (1) G International Breast Cancer Study Group.17

(2) CMF Abstract(3) CMF-G(4) nil Median follow-up 5.7 yr

E5188/INT 0101 USA (1) CAF Davidson et al.11

(2) CAF-G Abstract(3) CAF-G/T Median follow-up 9.6 yr

ZIPP International (1) (+/�chemo)-nil Baum et al.10

(2) (+/�chemo)-G Abstract(3) (+/�chemo)-T Follow-up not reported(4) (+/�chemo)-G/T

Chemo vs. Chemo-LHRH+TAME5188/INT 0101 USA (1) CAF Davidson et al.11

(2) CAF-G Abstract(3) CAF-G+T Median follow-up 9.6 yr

MAM-1 GOCSI Italy (1) CMF or A-CMF Bianco et al.14

(2) CMF-G+T or A-CMF-G+T

AbstractMedian follow-up 5 yr

U Pretoria South Africa (1) CMF Falkson et al.15

(2) CMF-B AbstractMedian follow up not reported

ZIPP International (1) (7 chemo)-nil Baum et al.10

(2) (7 chemo)-G Abstract(3) (7 chemo)-T Follow-up not reported(4) (7 chemo)-G+T

R. Sharma et al.184

ARTICLE IN PRESS


Sequential therapy provided a statistically non-significant improvement compared to either mod-ality alone (5-year DFS 86%). The benefits ofovarian suppression appeared most beneficial inyounger women with ER-positive tumours; howeverthis was based on an unplanned, retrospectivesubgroup analysis.

In IBSCG 8, amenorrhoea was induced in themajority of patients within three months ofcommencement of goserelin and resumed at theend of treatment. In contrast, patients whoreceived CMF amenorrhoea was induced in 50% ofpatients and regular menses resumed in approxi-mately 15% of patients at the end of treatment.Grade 3 or worse toxicities were experienced by18.8% of patients receiving chemotherapy, mainlyleucopoenia, neutropaenia and nausea/vomiting.By contrast 3.9% of those receiving goserelinexperienced grade 3 or worse toxicities, primarilyweight gain.

The final study identified is the TABLE study inwhich 527 of 599 women were assigned to receiveCMF (cyclophosphamide 500mg/m2 orally d1–14,methotrexate 40mg/m2 iv d1 and 8, and 5FU600mg/m2 iv d1 and 8) q28d� six cycles(n ¼ 117) or leuprorelin acetate 11.25mg every 3months for 2 years (n ¼ 110). An interim analysis of227 patients has been fully published and theseresults were updated in an abstract presented atSan Antonio, 2003.18,24 Eligibility criteria wereamended in March 1998 such that only women withreceptor positive disease were eligible. The fullypublished interim analysis showed no statisticallysignificant difference in progression free survival orOS between the two treatment arms after 2 years.The 2-year disease free survival was 59.1% forwomen receiving leuprorelin and 45.3% for womenreceiving CMF. These results remain unchanged inthe updated results.

All women treated with leuprorelin becameamenorrhoeic during the treatment period com-pared to 90.4% treated with CMF. As expectedtreatment related side-effects were more com-monly reported by those receiving CMF (99.7%)than leuprorelin acetate (95.2%). The most fre-quently reported side-effects in the leuprorelinarm were hot flushes (83.3%), weight gain (79.6%)and sweating (77.6%). In patients treated withCMF the commonest reported side-effects werethose attributable to chemotherapy. The over-all self-assessment of tolerability by patientswas markedly better after 3 and 6 months oftreatment in the leuprorelin group comparedwith the CMF group. No significant differencein self-assessment of tolerability was noted after2 years.

These studies illustrate that goserelin is aseffective as CMF chemotherapy in the adjuvanttreatment of ER positive early breast cancertumours and is therefore a reasonable alternative.As discussed goserelin has the further advantage ofhaving fewer distressing side-effects than conven-tional chemotherapy. Furthermore, adjuvant treat-ment with LHRH agonists results in reliable ovarianoestrogen suppression as indicated by the highincidence of amenorrhoea. As indicated in theZEBRA study, goserelin was substantially moreeffective in inducing amenorrhoea than CMF che-motherapy. In addition, the onset of amenorrhoeawas rapid and reversible.

Integration of LHRH agonists into adjuvanthormonal therapy

Three eligible trials were identified that addressedthe integration of LHRH agonists into adjuvanthormonal therapy, the ZIPP trial, the Norwegianstudy and the Intergroup study (INT 010).5,10,11

These trials included 4534 women and addressedthe efficacy of LHRH agonists alone or concurrentlywith TAM. The combinations and schedules studieswere similar but the duration of LHRH agonisttherapy varied between trials (Table 3). All threetrials used TAM given as 20mg daily for 5 years andgoserelin given as a subcutaneous injection every28 days for 2 years except INT 0101, whichcontinued goserelin for 5 years.

The largest of these studies was the ZIPP trial.10

This trial used a 2� 2 factorial design in which 2710eligible women with early stage invasive breastcancer were randomised to receive goserelin for 2years (n ¼ 1354) or no adjuvant treatment(n ¼ 1356). In addition, 1800 patients were rando-mised to receive TAM for 2 years (control n ¼ 899;TAM n ¼ 901). Patients may have received TAMelectively or by a second randomisation. After amedian follow-up of 66 months, patients rando-mised to receive goserelin had a significantlyimproved RFS (relative risk [RR] ¼ 0.8; 95% CI0.69–0.92; Po0:001) and OS (RR 0.82, 95% CI0.67–0.99; P ¼ 0:04) compared to those not whodid not receive goserelin. Subgroup analysis sug-gests goserelin has a greater benefit in patientswith ER positive tumours who did not receive priorchemotherapy, however, tests for heterogeneitywere not significant. In the 860 women randomisedto receive TAM, and in the 895 women who receivedTAM electively, the addition of goserelin signifi-cantly improved event free survival (EFS) comparedwith TAM alone (RR 0.76 and 0.70, respectively). Itshould be noted that this subgroup analysis does

ARTICLE IN PRESS

Table 3 Type of LHRH agonist used and duration of use in each study.

Study name LHRH agonist Dose Duration of treatment

ZIPP10 Goserelin Unknown UnknownNorwegian study21 Goserelin 3.6mg 2 yrItalian Breast Cancer Adjuvant Study20 Goserelin 3.6mg 2 yrAustrian Breast and Colorectal Cancer Study19 Goserelin 3.6mg 3 yrFASG06 (French study)13 Triptorelin 3.75mg 3 yrZEBRA16 Goserelin 3.6mg 2 yrIBCSG 817 Goserelin 3.6mg 2 yr or 18moU Pretoria15 Buserelin Unknown UnknownE5188/INT 010111 Goserelin 3.6mg 5 yrMAM-1 GOCSI14 Goserelin Unknown 2 yr

Table 4 Demographic information of clinical trials.

STUDY Nevaluable

MEANAGE

Mastectomy(%)

Radiotherapy(%)

ERpositive(%)

LNpositive(%)

Stratification Globalqualityrating

ZIPP� 2710 44 56INT 0101� 1504 59ZEBRA 1614 53 65 74 99 Centre ATABLE 227 44 60 5 93 99 AIBCSG 8 1063 45 44 51 68 0 ER status,

radiotherapy,institution

A

Italian study 244 45 55 100 100 86 B1Austrian 1034 36 68 49 Tsize, node number,

type of surgery,tumour grade, ERstatus

B1

Roche� 333Norwegian 320 45 80 44 75 28 B1MAM-1 GOCSI� 466 466 18U Pretoria� 147 145

�Unable to give global quality rating as reported in abstract form

R. Sharma et al.186

not control for ER status (33% ER-negative, and 23%ER status unknown), or for the use of chemotherapy(43% of patients received chemotherapy). OS hasnot been reported. No comparison of goserelin withTAM has yet been reported. Results from compar-isons of individual arms may not be presented inany detail in future reports (Table 4).

The Norwegian study compared the use of LHRHagonists or TAM as adjuvant therapy.5 320 eligiblepatients with stage II breast cancer were rando-mised to either to receive TAM (20mg daily)(n ¼ 161) or goserelin (3.6mg depot sc q28d)(n ¼ 159) for 2 years. No statistically significantdifference in time to disease recurrence (P ¼ 0:56)or OS (P ¼ 0:42) between goserelin and TAM wasdemonstrated. However, it must be noted that the

study has limited statistical power because of itssmall sample size.

INT 0101 is the third study that allows thecomparison of single agent LHRH agonist and LHRHagonist given in combination with TAM.11 This largestudy has only been reported in abstract form.Thousand five hundred and four premenopausalwomen with node positive, receptor positive breastcancer were randomised to receive six 28-daycycles of CAF chemotherapy (cyclophosphamide100mg/m2 orally q14, doxorubicin 30mg/m2

iv d1 and 5-fluorouracil [5FU] 500mg/m2 iv d1and 8) followed by goserelin for 5 years (3.6mg scq28) (CAF-Z) or CAF followed by goserelin andTAM (20mg daily� 5 years) (CAF-ZT). The three-way randomisation and statistical design allows

ARTICLE IN PRESS


comparison of CAF to CAF-Z or CAF-ZT. The additionof TAM to CAF-Z resulted in a statistically significantimprovement in DFS at a median follow-up of 9.6years (HR 0.73, 1-sided P-value o0.01). Theaddition of Z after CAF did not improve DFS(HR0.93, 1-sided P-value ¼ 0.25). No statisticallysignificant difference was seen in OS between thetwo arms, HRs 0.91 (P ¼ 0:21) and 0.88 (P ¼ 0:14).A retrospective subset analysis suggests that theaddition of goserelin to CAF may be beneficial inthose women o40 years, those who are notamenorrhoeic after CAF, and in those who hadpre-menopausal oestradiol levels prior to therapy.Final analyses of the impact of amenorrhoea,patient age and serum hormone levels on clinicaloutcome are awaited.

Taken together these studies demonstrate theefficacy of adjuvant goserelin, with or without TAM,in the adjuvant treatment of women with earlybreast cancer. What remains unclear is whetherpremenopausal patients who retain menses afterchemotherapy should be offered TAM alone, or TAMplus LHRH agonists. As discussed, these papers havea number of limitations and the full publications oftwo papers are not yet available. However, basedon the current available evidence combined TAMand LHRH analogue may be regarded as a treatmentoption for premenopausal women with endocrine-responsive disease. These studies only used goser-elin and whether these results can be extrapolatedto other LHRH agonists is unclear. Furthermore, theduration of LHRH treatment has not been welldefined as each trial using a different time period,2, 3 and 5 years (Table 3). This needs to bedelineated.

Integration of LHRH agonists intoadjuvant chemo-hormonal therapy

We identified three studies that compared the useof LHRH agonists given concurrently with TAMcompared to chemotherapy alone, enrolling 1611women.13,19,20 The LHRH agonist used variedbetween the trials as did the duration of treatment(Table 2).

The largest reported study is the Austrian Study,which randomised 1034 of 1099 patients to receiveeither goserelin (3.6mg depot sc q28d� threeyears) (n ¼ 511) with concurrent TAM (20mg dai-ly� 5 years) or CMF (cyclophosphamide 600mg/m2

iv d1 and 8, methotrexate 40mg/m2 iv d1 and 8,and 5FU 600mg/m2 iv d1 and 8) q28d� six cycles)(n ¼ 523).19 At a median follow-up of 5 years, therewas a significant advantage for the concurrent use

of goserelin and TAM over CMF in terms of relapsefree survival (81% vs. 76%, P ¼ 0:037). There was anon-significant advantage for the hormonal arm inOS; however, this data is not yet mature.

The Italian study was similar in design and againcompared combination endocrine therapy withCMF.20 This study was published in full andrandomised 244 patients to receive either CMF(cyclophosphamide 100mg/m2 orally d1–14, meth-otrexate 40mg/m2 iv d1 and 8, and 5FU 600mg/m2

iv d1 and 8) q28d� six cycles (n ¼ 120) or TAM(30mg/d� 5 years) combined with ovarian abla-tion/suppression (n ¼ 124). Ovarian ablation/sup-pression was achieved by either surgicaloophorectomy, ovarian irradiation through a15 cm� 15 cm pelvic port for a total dose of 15Gyor medical castration with goserelin (3.6mg depotsc q28d� 2 years). After a median follow-up of 76months, no statistically significant difference inDFS or OS was demonstrated (HR 0.94, 95% CI0.6–1.47; P ¼ 0:8). Unfortunately due to difficultiesin recruitment, this study is underpowered, espe-cially with regards to survival and therefore resultsneed to be interpreted with caution.

In the Italian study all women treated withgoserelin (n ¼ 70) became amenorrhoeic duringthe treatment period. Fourteen of these patients(20%) recommenced menstruation after disconti-nuation of treatment. Of the 106 women treatedwith CMF, 72 (68%) of women became amenorrhoeicduring treatment and remained so after cessationof treatment. Menopausal symptoms were moreprevalent in the TAM+ovarian ablation/suppressionarm than the CMF arm.

The final study is the French study. Unlike theother two papers the full publication was notavailable.13 This is an underpowered study thatrandomised 333 pre-menopausal patients to trip-torelin (3.75mg im every month) and TAM (30mgdaily) for 3 years (n ¼ 164) or FEC50 (epirubicin50mg/m2 iv d1 and 8, cyclophosphamide 500mg/m2, and 5FU 500mg/m2 iv d1 and 8) q21d� sixcycles (n ¼ 169). After a median follow-up of 54months, there was no statistically significantdifference in disease free survival between theendocrine arm (91.7%) and the chemotherapy arm(80.9%) (P ¼ 0:12). Similarly, no statistically sig-nificant difference was seen in overall survival wasseen (97% vs. 92.9%, P ¼ 0:18). The French study isthe only randomised comparison of LHRH agonisttherapy with an anthracycline. However, a criticismof this study has been the dose of epirubicinused. In the French Adjuvant Study Group thataddresses the issue of the optimum anthracyclinedose, 100mg/m2 of epirubicin was shown tohave improved efficacy compared to 50mg/m2,

ARTICLE IN PRESS

R. Sharma et al.188

therefore any conclusions from the French studyhave to drawn with caution.25

As presented the evidence for the addition ofLHRH agonists to conventional cytotoxic agentsremains unclear. However, data from the IBSCGestablished that young women (o35 years old),with ER positive tumours, who do not achieveamenorrhoea with cytotoxic chemotherapy, havesignificantly worse DFS and OS than older preme-nopausal women.26 As CMF chemotherapy inducesamenorrhoea in approximately 40% of youngerwomen the endocrine effects of chemotherapyalone may not be sufficient in this population groupand therefore additional endocrine treatment maybe considered.4

Two further studies addressed the role of LHRHagonists in differing circumstances. The MAM-1GOCSI study is only available in abstract form andthe doses of regimens used were not reportedmaking any meaningful comparison difficult.14 How-ever, this remains an important paper given theinclusion of an anthracycline. Four hundred andsixty-six patients randomised to receive (a) CMF, (b)doxorubicin followed by CMF, (c) CMF followed bygoserelin and (d) TAM for 2 years and doxorubicinfollowed by CMF followed by goserelin plus TAM for 2years. The study compared (a) and (c) versus (b) and(d) and (a) and (b) versus (c) and (d). After medianfollow-up of 5 years there was no significantdifference in the rate of relapse between anthracy-cline-containing arms and arms not containinganthracycline (HR 0.86; P ¼ 0:42). The addition ofcombined endocrine therapy significantly reducedthe risk of relapse (HR 0.71; P ¼ 0:04) comparedwith chemotherapy alone. The addition of combinedhormonal therapy significantly reduced the risk ofrelapse (HR 0.71; P ¼ 0:04), but no difference in OS

Table 5 Ongoing clinical trials of LHRH agonists.

Group Study PI N

German GABG-IVA Kaufmann

German GABG-IVB Kaufmann

UKCCR Johnson

International breastcancer study group

SOFT Francis Pa


TEXT Pagani C


PERCHE Nasi

was demonstrated. Corresponding HRs for OS were0.79 for anthracycline compared with non-anthra-cycline arms (P ¼ 0:31) and 0.86 for combinedendocrine therapy compared with chemotherapyalone (P ¼ 0:52).

Finally a study conducted by the University ofPretoria randomised 145 of 147 premenopausalwomen with node positive breast cancer to receiveeither CMF (n ¼ 75) or CMF followed by depo-buserelin (n ¼ 72) for an unspecified time dura-tion.15 The results of this small study are onlyavailable in abstract form and again no significantdifference between median disease free intervalfor patients receiving CMF alone or combinationtherapy was reported (6.2 years vs. 6.8 years).Survival data is not complete and the full publica-tion is awaited.

Future research

As discussed there is evidence supporting the roleof LHRH agonists in the adjuvant setting inpremenopausal women with early breast cancer,however, there are a number of relevant questionsthat still remain unanswered, many of which arebeing addressed by a number of ongoing studies(Table 5).

The IBSCG Suppression of Ovarian Trial (SOFT)addresses the issue of whether women who retainmenses despite chemotherapy benefit from LHRHagonists, as suggested by INT 0101.11 This trial willrandomise patients to receive TAM alone, TAM plusovarian function suppression (either by triptorelin,bilateral oophorectomy or ovarian irradiation) orovarian function suppression plus exemestane.

otes Randomisation

CMF vs. goserelin

Chemo-goserelin vs. chemo

TAM vs. chemo-TAM vs. OvSupp+TAM

re-menopausalfter chemo

Tam vs. LHRH+TAM vs.LHRH+exemestane

hemo permitted LHRH+TAM vs. LHRH+exemestane

Chemo+LHRH+TAM orexemestane vs. LHRH+TAM orexemestane

ARTICLE IN PRESS


The issue of whether chemotherapy adds tooptimal adjuvant endocrine therapy in premeno-pausal women is being addressed by the PERCHEtrial (Premenopausal Endocrine-Responsive Che-motherapy Trial). Despite the presented data notrial as yet investigates the substitution of TAM byLHRH agonists and this needs further clarification.

The role of aromatase inhibitors in the treatmentof adjuvant breast cancer also needs furtherinvestigation, particularly in those women madepost-menopausal by either chemotherapy or ovar-ian suppression/ablation. Anastrazole has beenshown to be superior to TAM in the adjuvanttreatment of post-menopausal women with breastcancer and there is a rationale for combininganastrazole with goserelin.27 This has been ad-dressed in premenopausal women with advanceddisease where the addition of anastrazole togoserelin in women who had progressed on goser-elin plus TAM was found to be beneficial.27 The roleof aromatase inhibitors in the adjuvant setting isbeing addressed by the complementary Tamoxifenand Exemestane Trial (TEXT). This study willcompares triptorelin plus TAM versus triptorelinplus exemestane.

The majority of the studies discussed usedgoserelin as the LHRH agonist of choice and itremains unclear if the results presented can beextrapolated to other LHRH agonists and this needsto be further defined. The optimal duration oftreatment with LHRH agonists is unknown and noone trial specifically compares different durationsof LHRH agonist treatment. The trials discussed inthis review each used LHRH agonists for varyingtime periods, most commonly 2–3 years (Table 3).In INT 0101 patients were treated with goserelin for5 years and this was well tolerated by patients.There is a need, however, for further studiesaddressing treatment duration.

Conclusions

Whilst the role of LHRH agonists in the adjuvanttreatment of pre-menopausal women with earlybreast cancer is not clearly defined, a number ofconclusions can be drawn from the available data.IBCSG 8 investigated the use of ovarian suppressionboth as a replacement for and as a supplement tocytotoxic chemotherapy. This is also addressed byfour other studies, the ZEBRA study, the TABLEstudy and the Italian and the Austrian studies.These four studies compared the use of LHRHagonists, with or without the addition of TAM, toCMF chemotherapy. All trials show equal or superior

outcomes for the non-chemotherapy containingarms, although like IBSCG8 none of the trials weredesigned with sufficient power to demonstrateformal equivalence between study arms. However,based on the available information it appears thatovarian suppression with LHRH agonists provides atleast comparable results to CMF given in premeno-pausal women with ER-positive breast cancer.Furthermore, LHRH agonists were well toleratedand had a more tolerable side-effect profiledifferent compared to CMF chemotherapy.

As discussed, the role of LHRH agonists inaddition to chemotherapy remains unclear. Data isalso inadequate with regards to the use of LHRHagonists instead of TAM for the treatment of ER-positive early breast cancer.

The major limitation of the trials discussed, isthe use of CMF chemotherapy as standard therapy.CMF has largely been superseded by anthracyclinebased regimens and it is unclear whether theseresults can be extrapolated to the anthracyclinebased regimens. Furthermore, as mentioned pre-viously, the duration LHRH agonist treatment variedbetween trials, so there is uncertainty with regardsto the optimal duration of treatment. Individualtrial numbers are small and variability in bothentry criteria of each trial and the treatment usedcan make meaningful comparison of trial resultsdifficult.

Despite these difficulties, several publishedinternational guidelines have established the roleof LHRH agonists in the treatment of hormonesensitive early breast cancer.28–30 The St GallenConsensus recommends ovarian ablation with orwithout TAM for women with node-negative,medium to high risk disease, and ovarian ablationin combination with TAM for patients with nodepositive disease.29 The European Society of Mastol-ogy Guidelines on Endocrine Therapy of BreastCancer further states that premenopausal womenwith early breast cancer should be offered treat-ment with a LHRH agonist for 2 years with orwithout TAM for 5 years as an alternative toadjuvant cytotoxic therapy.30

Search strategy and selection criteria

A search of the specialised register maintainedby the Secretariat of the Cochrane Collabora-tion Breast Cancer Review Group was per-formed. Details of the search strategy appliedby the Group to create the register, and theprocedure used to code references, are de-scribed in the Group’s module on the Cochrane

ARTICLE IN PRESS

R. Sharma et al.190

Library. The register also includes trials thatare ongoing or unreported. A total of 32references were identified. The eligibilitycriteria for the review were prospectivelydefined. Studies were included if they wererandomised, reported in English and includedonly women with early breast cancer receivingLHRH agonists an adjuvant setting and repre-sented a comparison of LHRH agonist-contain-ing with non-LHRH agonist-containingregimens.

Acknowledgments

Based on a review commissioned and supported bythe National Breast Cancer Centre of Australia.

Appendix A. List of abbreviations

Time to recurrence—time from randomisation orrecurrence

Overall survival—time from randomisation todeath of any cause

� Negative+ PositiveER Oestrogen receptorHR Hazard ratioIV intravenousLN Lymph nodePO oralSC subcutaneousTAM TamoxifenQoL Quality of life

References

1. Ferlay JBF, Pisani P, Parkin DM. GLOBOCAN 2000: CancerIncidence, Mortality and Prevalence Worldwide, Version 1.0.IARC CancerBase No. 5, IARCPress. 2001.

2. Pujol P, Daures JP, Thezenas S, et al. Changing estrogen andprogesterone receptor patterns in breast caner carcinomaduring the menstrual cycle and menopause. Cancer1998;83:698–705.

3. Klijn JG, Blamey RW, Boccardo F, Tominaga T, Duchateau L,Sylvester R. Combined TAM and luteinising hormone-releas-ing hormone (LHRH) agonist versus LHRH agonist alone inpre-menopausal advanced breast cancer: a meta-analysis offour randomised trial. J Clin Oncol 2001;19(2):343–53.

4. Bines J, Oleske DM, Cobleigh MA. Ovarian function in pre-menopausal women treated with adjuvant chemotherapy forbreast cancer. J Clin Oncol 1996;14:1718–29.

5. Soreide JA, Varhaug JE, Fjosne HE, Erikstein B, Jacobsen AB,Skovland E, et al. Adjuvant endocrine treatment (goserelinvs. TAM) in pre-menopausal patients with operable nodepositive stage II breast cancer: a prospective randomisednational multicenter study. Eur J Surg Oncol 2002;28(5):505–10.

6. Pagani O, O’Neil A, Castiglione M, et al. Prognostic impact ofamenorrhoea after adjuvant chemotherapy in pre-meno-pausal breast cancer patients with axillary node involve-ment: results of the International Breast Cancer Study Group(IBCSG) Trial VI. Eur J Cancer 1998;34:632–40.

7. Early Breast Cancer Trialists’ Collaborative Group. Ovarianablation in early breast cancer: Overview of the randomisedtrails. Lancet 1996;348:1189–96.

8. Scottish Cancer Trials Breast Group (SCTBG) and ICRF BreastUnit, Guy’s Hospital, London. Adjuvant ovarian ablationversus CMF in pre-menopausal women with pathologicalstage II breast carcinoma: the Scottish trial. Lancet 1993;341:1293–9.

9. Ejlertsen B, Dombernowsky P, Mouridsen HT, Kamby C, KjaerM, Rose C, et al. Comparable effect of ovarian ablation (OA)and CMF chemotherapy in pre-menopausal hormone recep-tor positive breast cancer patients. Proc Am Soc Clin Oncol1999;18 (Abstr 248).

10. Baum M, Houghton J, Sawyer W, Odling-Smee W, RutqvistLE, Nordenskjold B, et al. Management of pre-menopausalwomen with early breast cancer: is there a role forgoserelin? Proc Am Soc Clin Oncol 2001;21 (Abstr 103).

11. Davidson NE, O’Neill A, Vukov A, Osborne CK, Martino S,White D, et al. Chemohormonal therapy in pre-meno-pausal node-positive, receptor positive breast cancer: AnEastern Cooperative Oncology Group phase III intergrouptrial (E5188, INT-0101). Proc Am Soc Clin Oncol 2003;22(Abstr 15).

12. Liddle JWM, Irwig L. Method for evaluating researchand guideline evidence. Sydney: NSW Health Department;1996.

13. Roche H, Kerbrat P, Bonneterre J, Fargeot P, Fumoleau P,Monnier A, et al. Complete hormonal blockade versuschemotherapy in pre-menopausal early stage breast cancerpatients (pts) with positive hormone-receptor (HR+) and 1–3node-positive (N+) tumor: results of the FASG 06 trial. ProcAm Soc Clin Oncol 2000 (Abstr 279).

14. Bianco AR, Costanzo R, Di Lorenzo G, Adamo V, Altavilla G,D’Aprile M, et al. The MAM-1 GOCSI trial: a randomised trialwith factorial design of chemo-endocrine adjuvant treat-ment in node positive (N+) early breast cancer (ebc). ProcAm Soc Clin Oncol 2001;20 (Abstr 104).

15. Falkson CI, Jooste R, Falkson HC, Falkson G, Rens D, vanAarde R, et al. Cyclophosphamide, methotrexate andfluorouracil (CMF) plus or minus depo-buserelin in pre-menopausal women with lymph node positive breast cancer.Proc Am Soc Clin Oncol 2001 (Abstr 1777).

16. Jonat W, Kaufmann M, Sauerbrei R, Blamey R, Cuzik J,Namer M, et al. Goserelin versus cyclophosphamide,methotrexate, and fluorouracil as adjuvant therapy in pre-menopausal patients with node-positive breast cancer: thezoladex early breast cancer research association study. JClin Oncol 2002;20(24):4628–35.

17. International Breast Cancer Study Group. Adjuvant che-motherapy followed by goserelin versus either modalityalone for premenopausal lymph node-negative breastcancer: a randomised trial. J Nat Cancer Inst 2003;95(24):1833–46.

18. Schmid P, Untch M, Wallwiener D, Kosse V, Bondar G,Vassiljev L, et al. Cyclophosphamide, methotrexate and

ARTICLE IN PRESS


fluorouracil (CMF) versus hormonal ablation with leuprorelinacetate as adjuvant treatment of node-positive, premeno-pausal breast cancer patients: preliminary results of theTABLE-study (Takeda Adjuvant Breast Cancer Study withLeuprorelin acetate). Anticancer Res 2002;22:2325–32.

19. Jakesz R, Hausmaninger H, Kubista E, Gnant M, Menzel C,Bauernhofer T, et al. Randomised adjuvant trial of TAM andgoserelin versus cyclophosphamide, methotrexate, andfluorouracil: evidence for the superiority of treatment withendocrine blockade in pre-menopausal patients with hor-mone responsive breast cancer—Austrian breast and color-ectal cancer study group trial 5. J Clin Oncol 2002;20(24):4621–7.

20. Boccardo F, Rubagotti A Amoroso D, Mesiti M, Romeo D,Sismondi P, et al. Cyclophosphamide, methotrexate andfluorouracil versus TAM plus ovarian suppression as adjuvanttreatment of oestrogen receptor-positive pre-/peri meno-pausal breast cancer patients: results of the Italian breastcancer adjuvant study group 02 randomised trial. J ClinOncol 2000;18(14):2718–27.

21. Nystedt M, Berglund G, Bolund C, Fornander T, Rutqvist LE.Side effects of adjuvant endocrine treatment in premeno-pausal breast cancer patients: a prospective randomisedstudy. J Clin Oncol 2003;21(9):1836–44.

22. Blamey R, Fogelman I. Reversibility of bone mineral densityloss in pre-/peri menopausal patients with node-positiveearly breast cancer after treatment with ZoladexTM: resultsfrom the ZEBRA study. Breast Cancer Res Treatment 2001;69(Abstr 244).

23. De Haes JC, Olschewski M, Jonat W, Kaufmann M, Jonat W.Quality of life (QL) assessment in an international rando-mised trial: temporary ovarian ablation vs. CMF in earlybreast cancer. Eur J Cancer 1994;30(Suppl 2).

24. Untch M, Kahlert S, Kosse V, Bondar G, Kienle E, Maubach L,Schmid P, Wallwiener D, et al. LHRH-analogue therapy withleuprorelin-acetate three-month depot offers similar effi-cacy to conventional adjuvant CMF chemotherapy inreceptor positive, node positive Breast Cancer patients. In:San Antonio breast cancer symposium, San Antonio, TX,2003 (Abstr 40).

25. French Adjuvant Study Group. Benefit of a high-doseepirubicin regimen in adjuvant chemotherapy for node-positive breast cancer patients with poor prognostic factors:5-year follow-up results of French adjuvant study group 05randomized trial. J Clin Oncol 2001;19:602–11.

26. Aebi S, Belber S, Castiglione-Gertsch M, Gelber RD, Collins J,et al. Is chemotherapy alone adequate for young womenwith oestrogen-receptor-positive breast cancer? Lancet2000;355:1869–74.

27. The ATAC (Arimidex, Tamoxifen Alone or in Combination)Trialists’ Group. Anastrazole alone or in combination withtamoxifen versus tamoxifen alone for adjuvant treatment ofpostmenopausal women with early breast cancer: firstresults of the ATAC randomised trial. Lancet2002;359:2131–9.

28. National Institutes of Health Consensus Development Panel.National Institutes of Health Consensus Development Con-ference Statement: Adjuvant therapy for breast Cancer.JNCI 2001;93:979–89.

29. Goldhirsch A, Glick JH, Gelber RD, Coates AS, Senn H- J.Meeting highlights: international consensus panel on thetreatment of primary breast cancer. Seventh InternationalConference on Adjuvant Therapy of Primary Breast Cancer. JClin Oncol 2001;19:3817–27.

30. Blamey RW. Guidelines on endocrine therapy of breastcancer—EUSOMA. Eur J Cancer 2001;38:615–34.

systematic review of lhrh agonists for the adjuvant treatment of early breast cancer

Documents