Update on Cervical Cancer Screening Barbara S Apgar

Update on Cervical Cancer Screening Barbara S. Apgar, MD, MS Professor of Family Medicine University of Michigan Health System Ann Arbor, Michigan

Disclosures 1. Apgar B, Brotzman G, Spitzer M. Integrated Text and Atlas of Colposcopy. Elsevier Publishers, 2004, 2008. 2. Brotzman G, Spitzer M, Apgar B. Colposcopic Image Library on CD. SABK, Inc. 2004. 3. ASCCP Board of Directors. 2007 -current. Images Apgar B, Brotzman G, Spitzer M. Integrated Text and Atlas of Colposcopy. Elsevier Publishers, 2004, 2008.

The good news is that another cervical cancer guideline conference is NOT planned until 2017.

The bad news is there are women who are not being screened for cervical cancer 1/10 women aged 21 -65 have not been screened in the past 5 years 1/ 4 have no health insurance or PCP Benard VB et al. MMWR Nov 5, 2014; 63: 1 -6

Natural history of HPV infections

Natural hx of HPV infections • Most are transient and do not increase a woman’s risk for cervical cancer. • If oncogenic HPV is persistent, the risk of cervical cancer is increased substantially. – Longer persistence = greater the risk. • The early natural history of HPV infections does not predict the outcomes.

Women >age 30 Women < age 30 Behavior of HPV 16 during 30 months of follow-up (800 infections) Rodriquez AC et al. J Natl Cancer Inst 2008; 100: 513 -517

Progression to Cervical Cancer Apgar, Brotzman, Spitzer oncogenic HPV, persistent for 10 years

Natural History of CIN 3 Mc. Credie MRE et al. Lancet Oncol 2008; 9: 425 -434 • NZ women who were not treated for CIN 3 • Primary outcome = cumulative incidence of invasive cervical or vaginal cancer. • 1063 (86%)/1229 from 1955 -1976) were managed by: – Punch or wedge biopsy, conization

Mc. Credie MRE et al. Lancet Oncol 2008; 9: 425 -431

Key Points on Cervical Carcinogenesis u. Infection of the metaplastic epithelium of the cervical transformation zone with 1 or more oncogenic HPV types starts the process.

Time from HPV infection to CIN 3 is shorter than the decades-long time from first development of a CIN 3 lesion to cancer. HPV infection Apgar, Brotzman, Spitzer HSIL (CIN 3) Cancer

Cumulative relative contributions of 8 most common HPV types by histologic category (38 countries) in women with Cancer. de Sanjose S et al. Lancet Oncol 2010; 11: 1048 -56.

FDA-approved HPV tests • cobas 4800 HPV test. (approved 2011)-Roche • PCR-based. • 14 HR HPV types. • Genotyping for HPV 16 and 18 integrated into the assay. • Concurrently detects remaining 12 HPV types as a group (like hybrid capture- prior HPV test).

Importance of HPV types CYTOPATHOLOGY GYNECOLOGIC REPORT UMHS Labs Diagnosis: Negative for intraepithelial lesion or malignancy. Adequacy: Satisfactory for evaluation. Transformation zone present. Human Papillomavirus Genotype Results: Type 16: Detected Type 18: Not detected Other High Risk Type: Not detected SURGICAL PATHOLOGY REPORT A. Cervix, 12, 3, 5, 7 o'clock, biopsies: HSIL (CIN 3). B. Endocervix, curettage: HSIL (CIN 3)

Finding carcinogenic HPV types does not provide a diagnosis of CIN 3 or cancer It identifies a group of women in whom CIN 3+ is more likely Women testing negative for HPV have extremely low risk of developing cervical cancer over 5 years

Cumulative Incidence of CIN 3+ After Single HPV+ Followed for 10 years Among Women > 30 HPV 16 HPV 18 HR-HPV other than 16/18 HPV neg Khan MJ et al. JNCI 2005; 97: 1072 -9.

Which one of the following tests provides greater reassurance against CIN 3+ over an extended time? 1. Pap test. 2. HPV test. 3. Both are equal. 4. Neither test able to provide reassurance. Apgar, Brotzman, Spitzer

“Reassurance” against cervical cancer is the crucial factor for deciding screening intervals u. Large studies needed to establish actual cancer risks for each possible cytology and HPV result (including negatives) u If effective, each successive screen should lower the subsequent risk. u. Why? u. Because women with disease have already been identified and treated. u. They do not contribute to the overall risk. Katki HA et al. Gynecol Oncol 2011; 12: 663 -72

Kaiser Permanente Northern California data u 2003: shifted to co-testing u. Annual Pap's continued to be available. u 2007: 95% of patients co-tested q 3 yrs. u. Colposcopy after 2 HPV+/negative Paps u 5 year risk of CIN 3+ following a negative co-test was comparable to a 3 -year risk following a negative Pap.

CIN 3+ and Cancer risks 3 and 5 yrs. after negative HPV and cytology Huh WK et al. Interim Clin Guidance. OG 2015; 125(2): 330 -7.

Cumulative Risk Of Invasive Cancer Based on Initial Co-Test Kaiser Study N=331, 818 women > 30 years in prospective cotesting study Katki HA, et al. Lancet Oncol 2011; 12: 663

Important points on 2011 Kaiser data u. Although women Pap and HPV + had the highest risk, those with negative Pap but HPV+ accrued substantial risk for CIN 3+ over 5 years. u. A single negative HPV is sufficient to reassure low risk of cancer and CIN 3+ over 5 years. u. HPV testing predicted future disease much better than Pap. Meijer CLM et al. Gynecol Oncol 2011: 12: Katki HA, et al. Lancet Oncol 2011; 12: 663 -72

Reassurance against future risk of CIN 3 and cancer of a negative screening result u 3 screening strategies. u. Primary Pap testing every 3 years u. Primary HPV testing every 3 years. u. Primary cotesting every 5 years. u 1, 011, 092 women (30 -64 years) in Kaiser Gage JC et al. JNCI 2014; 106(8); dju 153

Important points on 2014 Kaiser data u 3 year risks following an HPV-negative test were LOWER than: u 3 year risk following a Pap-negative test. u 5 year risk following an HPV and Pap-negative cotest. u. A negative Pap provides less reassurance against CIN 3+ than a negative HPV test or negative cotest. Gage JC et al. JNCI 2014; 106(8); dju 153

Which one of the following tests provides greater reassurance against CIN 3+ over extended time? 1. Pap test. 2. HPV test. **** 3. Both are equal. 4. Neither test able to provide reassurance. Apgar, Brotzman, Spitzer

Which test provides greater reassurance against CIN 3+ or cancer? HPV testing improves detection of CIN 3 at baseline resulting in fewer cancers. Castle et al. Obstet Gynecol 2011; 117: 650 -6. Ronco G et al. Lancet Oncol 2010; 11: 249 -57. Katki HA et al. Lancet Oncol 2011; 12: 66 3. Ronco G et al. Lancet 2014; 383: 524 -32. Gage JC et al. JNCI 2014; 106(8); dju 153

2012 Cervical cancer screening guidelines Age 30 -65. Testing with cytology alone every 3 years or co-testing with cytology and testing for high-risk HPV types every 5 years. §Co-testing “preferred” and cytology “acceptable” by all but USPSTF Recommendations NOT intended for women with HIV, immunocompromise, or in utero DES exposure 1. 2. 3. 4. Saslow et al. ACS/ASCCP/ASCP. CA Cancer J Clin 2012; 62: 147 -72 and AJCP 2012; 137: 516 – 542. Moyer VA, et al. USPSTF. Ann Int Med 2012; 156: 880 -91 ACOG Practice Bulletin #131, November 2012 NCCN Cervical Cancer Screening Guideline v. 2 -2012. www. NCCN. org

2012 Consensus Guidelines: Screening Frequency Age 21 -29. Testing with cytology (Pap) alone every 3 years. § Co-testing should NOT be performed for women < age 30. § Reflex HPV testing for ASCUS only. 1. 2. 3. 4. Saslow et al. ACS/ASCCP/ASCP. CA Cancer J Clin 2012; 62: 147 -72 and AJCP 2012; 137: 516 – 542. Moyer VA, et al. USPSTF. Ann Int Med 2012; 156: 880 -91 ACOG Practice Bulletin #131, November 2012 NCCN Cervical Cancer Screening Guideline v. 2 -2012. www. NCCN. org

Co-testing women < 30 years u 50 % of clinicians and clinics do co-testing in women < 30 years (not guideline adherent). u Despite significant rate of transient HPV infections in this age group. u. Bottom line: HPV testing increases cost, anxiety and does not add clinical value in women < age 30 years u. JUST SAY NO! Solomon D et al. CA Cancer J Clin 2007; 105. Kitchener HC et al. Int J Gynecol CA 2008; 18. Berkowitz et al. Obstet Gynecol 2010; 116.

So where are we on decisions about cervical screening since 2012? u. What is the level of protection provided by cotesting every 5 years or Paps every 3 years? u. It has been assumed that any screening strategy would offer at least the same level of protection as annual screening. Balance of benefits and harms

“Best” balance in the 2012 screening guidelines? Harms (numbers of colposcopies) and Benefits (immediate detection of CIN 3 and reduction of CIN 3+ in the interval before the next screening) Paps every 3 years beginning at age 21 = acceptable level of protection (USPSTF). u. Cotesting every 5 years provides protection with decreased colposcopies that is “at least as good” as Paps every 3 years. Moyer VA. USPSTF Rec Statement. Ann Intern Med 2012; 156: 880 -91.

What are the “Harms”? u. Colposcopy referral. u. Treatment of cervical lesions destined to resolve without intervention. u. Burdening women with surveillance of unclear efficacy or endpoint. u 39% of women in surveillance had significant psychological distress. Sharp L et al. JLGTD 2014; 18(2): 142 -50 Sawaya, Kuppermann. Obstet Gynecol 2015: 125: 1 -3 Moyer VA. USPSTF Rec Statement. Ann Intern Med 2012; 156: 880 -91.

How many colposcopies are a “harm”? USPSTF: 3 year vs. 5 year cotesting intervals u. Every cancer prevented by 3 year instead of 5 year intervals, would require additional colposcopies for 92/1000 women and treatment of 3/1000 additional women for CIN 2, 3 u. Every cancer DEATH prevented would require 409 additional colposcopies/1000 women and treatment of 14/1000 additional women for CIN 2, 3 Moyer VA. USPSTF Rec Statement. Ann Intern Med 2012; 156 Sawaya GK, Kupermann M. Obstet Gynecol 2015: 125: 1 -3

Even after knowing the “harms” of more intensive screening…. . u. Significant numbers of patients and providers would not accept the additional lifetime cervical cancer risk conferred by: u. Cotesting every 5 years (0. 74%). u. Cotesting every 3 years (0. 47%). u Some have argued that annual screening should be the benchmark for screening guidelines. u. Annual cytology (0. 25%). Sawaya GK, Kupermann M. Obstet Gynecol 2015: 125: 1 -3. Kinney W et al. Obstet Gynecol 2015: 125: 311 -315.

What if the benchmark is annual screening not cytology every 3 years? Return to annual screening will result in following a large number of lesions that: 1. will resolve on their own 2. have little to do with cancer prevention Kinney W et al. Obstet Gynecol 2015: 125: 311 -15. Sawaya GK, Kupermann M. Obstet Gynecol 2015: 125: 1 -3.

Low vs. high grade cervical lesions Apgar, Brotzman, Spitzer

Changing cotesting from 3 to 5 years u. Publications all agree that going from a 3 to 5 year interval moves screening farther away from protection afforded by annual screening. u. Results in an additional 1/369 women being diagnosed with cancer. u. Results in an additional 1/1639 dying of cancer.

2015 considerations u Recognize that risks associated with annual and 3/5 year intervals are not identical. u Permit a range of acceptable screening options. u. Option of better cancer protection should be an acceptable choice for patients who want it. u Give patients the best available estimates of cancer risk. u Avoid 3 and 5 year intervals unless you have a recall system. u Those involved in guideline construction should be FREE of commercial and intellectual conflicts of interest. *** Kinney W et al. Obstet Gynecol 2015: 125: 311 -315. Sawaya GK, Kupermann M. Obstet Gynecol 2015: 125: 1 -3.

Risk of cervical cancer will decrease over time HPV vaccination u. HPV vaccination. Decreased prevalence of HPV 16 and u. Decreased prevalence of HPV 16 and 18. 18 u. Multiple rounds of cotesting May provide the desired effect of cancer protection using 5 year or even longer intervals in the future.

What do screened women 36 -62 yrs. prefer ? (n-581) u 55% knew that screening guidelines had changed. u 74% believed women should have yearly Pap tests. u. Knowledge of change in guidelines did not equate to increased willingness to follow them. u 68% were willing to extend screening to every 3 years if recommended by health provider. u Only 25% were willing to extend to every 5 years after a negative cotest. Silver MI et al. Obstet Gynecol 2015; 125(2): 317 -329

What do women age 36 -62 years prefer ? u Women with lower incomes more likely to accept longer screening intervals. u. More barriers to scheduling time off work and transportation. Silver MI et al. Obstet Gynecol 2015; 125(2): 317 -329

Concerns of patients 5 5 year cervical cancer screening intervals are a stretch for most women. There is continued reluctance among adult women to accept HPV-based screening algorithms

What concerns do physicians have? u. Losing the well-women annual clinical encounters as a result of less frequent screening. u. Common barrier to use of cotesting and extended screening intervals. u. However……… u 70% of women would continue well-woman visits even if Pap testing was not done. Perkins RB et al. Am J Prev Med 2013; 45: 175 -81. ACOG. Comm Gyn Prac 2012; 120: 421 -424.

So what’s new on the horizon? Cytology every 3 years starting age 21 Cotesting every 5 years starting at age 30 Now consider primary HPV screening starting at age 25

FDA approved HPV testing Roche cobas® 4800 HPV test § FDA approved as HPV test (pool + genotyping) 2012 § FDA approved as primary screening test 4 -24 -2014* * It does not change current medical practice guidelines for cervical cancer screening. These guidelines are developed, reviewed and modified by groups other than the FDA News Release 4/24/14

Primary screening with HPV testing u Screening may begin at age 25. u. Despite 2012 recommendations discouraging HPV testing in women < age 30. u. High prevalence of HPV and transient infections. u Athena trial (GSK sponsored). u 21% had + HPV tests and were referred to colposcopy or placed in surveillance. u. Comparison: 7% to colpo if Pap alone. Huh WK et al. Interim Clin Guidance. OG 2015; 125(2): 330 -7. Moyer VA. USPSTF Rec Statement. Ann Intern Med 2012; 156: 880 -91. Wright TC et al. Athena HPV study. Am J Obstet Gynecol 2012; 206: 46 e 1 -11.

Advantages of HPV-based Primary Screening u High sensitivity for detection of HSIL lesions u High negative predictive value of test u High test reproducibility u Results in higher detection of CIN 2+ in women aged 25 -29 compared to the same strategy starting at age 30 (is this an advantage? ) Wright TC et al. Athena HPV study. Am J Obstet Gynecol 2012; 206: 46 e 1 -11.

Disadvantages of HPV-based Primary Screening u u Prevalence of HPV positivity (14 pooled genotypes) was almost twice as high in women 25– 29 years (21. 1%; 95% CI: 20. 1– 22. 1%) as in women 30– 39 years (11. 6%; 95% CI: 11. 0– 12. 2%). Double number of colposcopies (21%) Unproven long term proof of equivalence of outcome and cost effectiveness compared to consensus guidelines Unknown optimum screening intervals and follow-up protocols

HPV only testing >>>>> More colposcopies if direct referral (especially if starting at age 25) More CIN 2, 3 detected (includes CIN 2 which may regress) More treatments (including for CIN 2) Gage JC et al. JNCI 2014; 106(8); dju 153

Primary HPV Screening Algorithm-2015 b cobas HPV DNA Test A PV n se Routine Follow-up t 3 year intervals H HR 12 mo. Follow-up 31 33 35 39 45 51 52 56 M NIL Other HPV HR DNA Present Cytology 59 58 16 66 18 68 C HP V 1 6/1 Colposcopy 8 P Start at age 25 ≥A S US C- res en t Colposcopy Huh WK, et al, Use of primary high-risk human papillomavirus testing for cervical cancer screening: Interim clinical guidance, Gynecol Oncol (2015), http: //dx. doi. org/10. 1016/j. ygyno. 2014. 12. 022

Cost Estimate by Screening Strategy Simple cost estimate for screening from age 21 to 65 years of age assuming all scheduled screens are completed u Colposcopic exams: linear extrapolation from ATHENA trial rates assuming 50% drop every 3 years x 4 with 50% less frequency in CSG group u Consensus Screening Guideline Primary HPV Screening Cytology tests 11 x $68 = $748 2 x $68 = $136 HPV tests 8 x $131 = $1048 15 x $291 = $4365 Office visits 11 x $125 = $1375 17 x $125 = $2125 15. 4% x $1200 = $185 31% x $1200 = $372 $3356 $6998 Colposcopic Exams Cost Analysis by R K Reynolds, MD. Gynecol Oncol, Univ Michigan 2015

Primary HPV Screening: The Future? What is needed for broad acceptance? u When enough data are available to justify using this test as a replacement for other screening methodologies. e. g. outcome and cost comparisons u When enough data are available to validate this method on more than one platform besides Roche e. g. competition to reduce cost. We are not there yet!

The End…. . Thanks all !