HPV Testing at Princess Alexandra Hospital

1
Thames Valley Cytology Society17 June 2009

• HPV Testing at Princess Alexandra Hospital
• Janet Leake
• Pathologist

2
Itinerary

• Background
• Sentinel sites
• Forming the business case
• Early results from PAH

3
HPV and Cancer
4
HPV and Cancer

• Persistent infection with high risk HPV subtype
  is necessary but not sufficient for
  carcinogenesis
• Sensitivity of HR-HPV is high (?too high)
• Specificity is low (especially in the young)

5
Natural History

• Elimination (median 8 -14 months)
• Most prevalent after sexual debut
• Co-existence of more than one subtype (especially
  lt 30s)

6
HPV Subtypes

• 66 of CIN 2 associated with HPV16 or HPV18
• Frequency of most prevalent 16, 52, 18 31 ,
  51, 39
• Infection with more than one type in lt30 year
  olds
• Bivalent vaccination ineffective in approx one
  third

7
Distribution of HR HPV subtypes in cervical cancer
8
HPV Testing

• Qiagen (formerly digene) hc 2
• Cervista
• Greiner

9
Qiagen (Digene) hc2
10
Qiagen hc2

• HR-HPV types 16/18/31/33/35/39/45/51/52/56/58/59/
  68
• Detection at 1pg/ml, approx 5,000 DNA copies/ml,
  standard for clinical relevance

11
hc 2

• Longest track record, FDA approval and data in
  literature
• Relatively simple
• Highly consistent and reproducible

12
hc 2

• No information on individual subtypes
• Risk of false negative on acellular sample
• Needs bulk runs to be cost effective 1 kit
  provides 88 tests (with 8 controls)

13
Cervista (Hologic)

• Cervista HR 14 type specific probes. PCR
• Cervista 16/18
• FDA approval March 2009 for ASC-US (B/L) in women
  30
• Smaller batch size (28)

14
PapilloCheck(Greiner)

• PCR based, type specific primers
• Simple kit
• Identifies 18 HR subtypes 6 LR subtypes
• Consistent reproducible results published

15
NHSCSP LBC/HPV pilot

• 5,654 women 20-64
• Triage of low grade abnormalities
• 45.6 borderline HR-HPV positive
• 82.6 mild dysk HR-HPV positive

16
HR-HPV status by age for low grade
17
Impact on Colposcopy and Lab

• Referrals more than doubled
• Protocol modified under 35s had repeat cytology
  and HPV at 6 months
• Repeat cytology fell by 74

18
Impact on patients

• Opportunity for early referral, treatment of
  smaller lesion
• Fewer HPV women default from colposcopy than
  from cytological follow up of LG abnormality
• Increased detection of CIN2

19
ARTISTIC

• A Randomized Trial in Screening to Improve
  Cytology
• NHSCSP 24 510 women enrolled
• Brit J Cancer 2006 95 56-61
• Brit J Cancer 2008 98 1704-1709

20
hc2 by age (screening population)
21
hc2 by cytology
22
HPV status clinical implications

• HR HPV is highly sensitive (gt99) for CIN 2
  or worse false negative rate negligibly low
• BUT specificity is relatively low, especially in
  the young false positive rate for significant
  pathology relatively high
• 10 year Cumulative incidence of CIN 3 or worse is
  15-20 for persistent HR-HPV

23
Protective effect of negative HC2versus negative
cytology for CIN 2
24
NHSCSP early implementersSentinel Site
Laboratories April 2008N

• 6 labs 2 clusters of 3
• Northern Sheffield, Manchester, Liverpool
• Southern Bristol, Northwick Park, Norfolk and
  Norwich
• Testing carried out in Manchester and Bristol
• Triage of low grade abnormalities
• Test of cure post treatment

25
Low Grade Triage Sentinel Sites
26
Follow up arm/Test of Cure

• 6/12 follow up

Cytology normal
Cytology abnormal
HPV test
HPV negative
HPV positive
Refer to colposcopy
Return to normal recall
27
Challenges at PAH

• Not a sentinel site! Strict adherence to NHSCSP
  guidelines on referral and surveillance
• Could it be cost effective?
• Informed consent of women involved

28
Low Grade Referrals to PAH

• Approx. 300 per annum
• Of these, about 2/3 are over age of 30
• Many clinically inflammatory
• Of those biopsied approx 25 will have
  significant pathology (CIN 2) Which ones?

29
Costing

• Staff components
• BMS band 4
• BMS band 5
• AP/Pathologist (in putting and validation)
• Consumables
• Qiagen kits
  
• pipettes and sundries

30
Unit cost v run batch size
Cost
Number of tests/run
31
Potential Savings

• Reduced re-referrals to colposcopy for persistent
  low grade abnormalities. (Colposcopy tariff is
  approx. 250)
• Freeing of slots for new referrals.
• ?Reduced negative biopsy rate
• (Quality improvement due to increased detection
  of CIN 2)

32
PAH Criteria 1

• Low grade referrals aged 30 and over
• 6 month follow up after LLETZ high grade
  histology, any age test of cure
• Borderline cannot exclude high grade any age

33
PAH Criteria 2

• 12 month follow up of clinically normal/untreated
  HR-HPV positive low grade referrals
• Clinically indicated egg. Cytology/histology
  non-correlation, ?HIV, miscellaneous management
  problems

34
Low Grade Algorithm 1
HPV negative
Negative colposcopy No biopsy, or biopsy with no
CIN
CIN 1
NHSCSP Cytology follow up
Conservative management
35
Low Grade Algorithm 2
HPV positive
Negative colposcopy No biopsy, or biopsy with no
CIN
Biopsy or treat
Colposcopy or biopsy positive
NHSCSP Cytology follow up with repeat HPV at 12
months
Manage clinically as appropriate
36
Implementing

• Patient information and opt out phone line
• Communication with all smear takers
• Communication with QA, contractor services
• Training and education colposcopists,
  gynaecologists, GUM
• Request generation at time of reporting pap
• Failsafe of requesting

37
Early Results

• 52 analysed 2 inappropriate - excluded
• 31 Low grade referral, of which 24 were hc 2
• 18 Post treatment, of which 2 were hc 2
• 1 Other (had CIN 2 in 2004, neg cytol polyp)

38
Unexpected volumes and costs

• Cost 35 based on 10/run, 2 weekly run i.e.
  approx 250/annum
• Number performed 1st Jan 31st May 300
• Spoilt runs (not in costing)
• Larger runs consumable efficiency not paralleled
  by manpower
• ? Net effect on biopsy rate

39
hc 2 status for Low Grade and Follow up
40
Hc 2 status for Low Grade by age
41
Follow up group correlations

• All had negative concurrent cytology
• Both of hc 2 positive had a history of CIN 3,
  completely excised.
• The only uncertainly excised CIN had negative hc
  2 and negative cytology.

42
Low Grade hc 2 status by age
43
Low Grade triage

• High hc 2 positive rate in lt40s is 92.8
• In 40 65s, 66.6 62.5
• Borderline and mild not separated.

44
Low Grade Corellations

• 12 of 24 LG referrals were biopsied
• 7 CIN 1
  
• 2 HPV changes no CIN
  
• 2 -negative
• 1 CIN 3, age 22, 8XS
• None of 7 hc 2 negative patients were biopsied

45
Initial Impressions

• Very high hc 2 rates among low grades lt 40
• Less than 10 reduction of LG referrals lt 40
• In over 40s could triage out a least one third of
  low grade referrals
• Big saving in 10 year follow up of treated HG

46
The Future

• Definite role for follow up
• Possible role for LG triage, but ?age 40
• Possible increased specificity in selection of
  candidates for biopsy, increased sensitivity
  using triple approach i.e. Cytology, HPV test and
  colposcopy

47
The End