Epidemiology of HPV

and Cervical Neoplasia

Implications for Optimal

Vaccination and Screening

Mark Schiffman, MD, MPH

September 2009

TISSUE: Cervix

• Best understood site of HPV infection

• For all sites where HPV causes cancer:

HPV +

Transformation Zone =

Increased Risk of Carcinogenesis

Transformation

Zone -- Visually

Transformation Zone –

Histologically

10

8

1

9

11

7

5

6

4

2

15

3

13

EXPOSURE: HPVHPV6

HPV11

HPV13

HPV74

HPV44

HPV55

HPV91

HPV7

HPV40

HPV32

HPV42

HPV54

HPV52

HPV67

HPV33

HPV58

HPV16

HPV31

HPV35

HPV34

HPV73

HPV59

HPV18

HPV45

HPV70

HPV39

HPV68

HPV85

HPV26

HPV69

HPV51

HPV82

HPV30

HPV53

HPV56

HPV66

HPV57

HPV2a

HPV27

HPV71

HPV90

HPV61

HPV72

HPV62

HPV81

HPV83

HPV89

HPV84

HPV86

HPV87

HPV28

HPV3

HPV10

HPV29

HPV77

Burk,

Virology 2005

α1, 8, 10, 13

Genital Warts

α2, 3, 4, 15

Commensal

Infections

α5, 6, 7, 9, 11

Carcinoma &

Precursors*Evolutionary

Tree

(millions of

years)

18

16

Peak Ages: 15-25 25-35 45-50

DISEASE: Cervical Precancer and Cancer

Schiffman et al., Lancet, 2007

Epidemiologists Define Critical Steps

From Normal to Cancer

Normal

PrecancerAcute

HPV

Cancer Death

Epidemiologists need

reliable categories

to define risksGenital Warts

The First Step is HPV Infection

• Easily transmitted

• Each infection is

independent

• A woman can have

several, at the same

or different times

• The peak incidence in

a population is usually

at young ages

Normal

Acute

HPV

Rapid Clearance is the Rule

Normal

Acute

HPV

Persistence is Highly Associated

with Risk of CIN3

• Overt Persistence is key risk factor for precancer

• HPV type very important

• Co-factors like smoking or parity less important

PrecancerAcute

HPV

Clearance

90% of New Carcinogenic HPV

Infections Clear. Many but Not All

that Remain Indicate Precancer

Guanacaste Cohort (Rodriguez, JNCI 2007)

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

0-9 mo

10 mo - 4 yr

4 - 8 yr

8 - 12 yr

12 - 18 yr

HPV negative

Other ca-assoctypes

HPV18+

HPV16+

Kaiser Portland HPV Study (23,000 Women)

Cumulative Risk of Cancer, by HPV Type

Just

one

Test!

Prospective Study of Cancer Death

Cancer Death

Sankaranarayanan,

NEJM, 2009

Cancer Death

HPV

neg

8 / 24,000 0

HPV

pos

87 / 2,800 12

8 year follow-up in

randomized trial

(130,000 women)

HPV arm of trial

Clinical Viewpoint: Adapting

to New Knowledge of HPV

Parts of “Old” Prevention Strategy

• Pap smear

• Colposcopic impression if Pap abnormal

• Biopsy if needed

• Treatment of CIN

• Follow-up

• Cytology, colposcopy, and biopsy have

been great successes. We can now do

even better.

HPV Testing: Major Randomized

Trials

• All published in 2007-9– New England Journal, Lancet, J Natl Cancer Institute

• Show screening with HPV tests is more sensitive for early detection of CIN3 than cytology

• HPV testing with or followed by cytology might be useful in some places, or other triage tests might be used

• In low-resource regions, screen-and-treat?

• New IARC Study in India by Sankaranarayanan et al. is a landmark study

Schiffman and Castle, NEJM, 2006

Cervical Cancer Prevention Efforts Should Fit

Age Patterns in Natural History

If We Can Afford to Vaccinate…

Vaccinate before the peak of incidence, because the vaccines are

preventive, and they do not work after infection occurs

If We Screen with Limited Resources…

Screen after the peak of incidence, to improve specificity and positive predictive value

Defining a New Clinical View

of Cervical Carcinogenesis

There are many tests which,

alone or combined, predict

similar levels of risk

Cyto-

ASC-US

HPV-/ASC-US

HPV-/Cyto-All ♀

HPV+/ASC-USLSIL

HSIL

HPV+/Cyto-

~0%

Incre

asin

g 2

-Yr.

Ris

k o

f P

recancer

(CIN

3)

2%

10%

40%

100%Cytology HPV

HPV16+

HPV16-

HPV16/18

HPV16+

or

HPV18+

HPV16- &

HPV18-

“Risks”

Are

Simpler

To Use

Than

Algorithms

Don’t Screen Too Often with HPV ( 5-year cumulative incidence of CIN2+

following first HPV detection)

0

2

4

6

8

10

12

14

16

18

Enrollment First Follow-up 2nd Follow-up Never Positive

Year of First Positive HPV Test

% D

iag

no

sed

wit

h C

IN2+

N 1324 6626

Good risk stratification with enrollment testing

Don’t Screen Too Often with HPV ( 5-year cumulative incidence of CIN2+

following first HPV detection)

0

2

4

6

8

10

12

14

16

18

Enrollment First Follow-up 2nd Follow-up Never Positive

Year of First Positive HPV Test

% D

iag

no

sed

wit

h C

IN2+

N 1324 288 6626

Infections found a year later predict lower risk of CIN2+

Don’t Screen Too Often with HPV ( 5-year cumulative incidence of CIN2+

following first HPV detection)

0

2

4

6

8

10

12

14

16

18

Enrollment First Follow-up 2nd Follow-up Never Positive

Year of First Positive HPV Test

% D

iag

no

sed

wit

h C

IN2+

N 1324 288 203 6626

By year 2, any infections found are new and low-risk

Concluding Predictions

• Even better vaccines

• Decreased role for algorithms based on

cytology, colposcopy and targeted biopsy

• More reliance on HPV-related tests and

risk stratification to define management

• We need good epidemiology to inform new

strategies