The Influence of Human Papillomavirus Genotypes on Visual Screening and Diagnosis of Cervical Precancer and Cancer Jose Jeronimo, MD,1 Pooja Bansil, MPH,1 Melissa Valdez, MPH,1 Le-Ni Kang, BS,2 Fang-Hui Zhao, MD, PhD,2 You-Lin Qiao, MD, PhD,2 Wen Chen, PhD,2 Xun Zhang, MD,2 Proma Paul, MHS,1 Ping Bai, MD,2 Roger Peck, BS,1 Jing Li, PhD,2 Feng Chen, BS,2 Mark H. Stoler, MD,3 and Philip E. Castle, PhD, MPH4 Objective: To examine the influence of human papillomavirus (HPV) genotypes on the sensitivity of visual inspection with acetic acid (VIA) for screening, and colposcopy for diagnosis of cervical intraepithelial neoplasia grade 2 (CIN2) or more severe (CIN2+). Materials and Methods: Women aged 25 to 65 years from China (n = 7,541) were screened with 6 tests (careHPV and Hybrid Capture 2 on self- and clinician-collected specimens; HPV-16, HPV-18, HPV-45 E6 detection; and VIA). Biopsies from women with a diagnosis of CIN2+ underwent testing for 25 HPV genotypes using SPF10/LiPA. Human papillomavirus genotyping results were classified according to broad categories of cancer risk. Results: Among the 143 women with a diagnosis of CIN2+, the percentage who were HPV16 positive increased with increasing severity of diagnosis: 33.3% for CIN2 (n = 39), 69.1% for CIN3 (n = 94), and 90% for cancer (n = 10). There was a higher percentage of HPV-16 in women with abnormal colposcopic impression (p = .007) and positive VIA (p = .02) than normal colposcopy and negative VIA, respectively. Colposcopy and VIA were more sensitive to detect CIN2+ among HPV-16– and/or HPV18–positive women than HPV-16–/HPV-18–negative women (67.4% vs 43.1%, p = .008, for colposcopy; and 53.3% vs 37.3%, p = .08, for VIA). Conclusions: Human papillomavirus type 16 is related to more clear visual acetowhite changes in the epithelium. Therefore, we should expect a reduction of the performance of VIA for cervical cancer screening to identify women with CIN2+, and reduction of the performance of colposcopy to diagnose CIN2+, in vaccinated populations.
prophylactic HPV vaccination and molecular testing for detection of HPV. Although prophylactic HPV vaccination may be the ultimate cervical cancer prevention strategy, it does not treat preexisting HPV infections and related precancer conditions. Additionally, the current generation of HPV vaccines target 2 of the 14 genotypes associated with cervical cancer (HPV-16 and HPV-18) and therefore will not prevent approximately 25% to 30% of cervical cancers. Thus, cervical cancer screening will be needed in the foreseeable future to prevent cervical cancer. We took advantage of a population-based study of lower-cost cervical cancer screening conducted in rural China to examine the clinical importance of HPV genotypes in the visualization of cervical precancer lesions, during screening by visual inspection with acetic acid (VIA) and colposcopy in the management of HPVscreen–positive women.1 Because women were exhaustively screened with 6 screening tests (including VIA) and all screenpositive women underwent a rigorous colposcopic evaluation, we assume that most women with cervical precancer and cancer in the study population had a diagnosis. The primary objective of this study was to examine the relationship of causal HPV genotypes on the sensitivity of VIA for screening, and colposcopy for diagnosis of cervical intraepithelial neoplasia more severe than grade 2 (CIN2+). As studies have shown that HPV-16–positive CIN2/3 was more common in younger women than in older women,2–7 we evaluated the effect of HPV status and median age on the sensitivity of CIN2+ diagnosis.
Key Words: cervical cancer, colposcopy, cervical screening, HPV, China (J Lower Gen Tract Dis 2015;19: 220–223)
MATERIALS AND METHODS
T
he identification of human papillomavirus (HPV) as the primary cause of cervical cancer has led to the development of novel primary and secondary cervical cancer prevention strategies: 1
PATH, Seattle, WA; 2Department of Epidemiology, Cancer Institute and Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, People's Republic of China; 3Department of Pathology, University of Virginia, Charlottesville, VA; and 4Global Cancer Initiative, Chestertown, MD Reprint requests to: Jose Jeronimo, MD, PATH, PO Box 900922, Seattle, WA 98109. E-mail: [email protected] The START-UP project was funded by a grant from the Bill & Melinda Gates Foundation. Conflict of interest: No conflicts of interest for Pooja Bansil, Melissa Valdez, Le-Ni Kang, Dr. Zhao, Dr. Qiao, Dr. Chen, Dr. Zhang, Proma Paul, Dr. Bai, Roger Peck, Dr. Li, and Feng Chen. Dr. Castle has received commercial HPV tests for research at a reduced or no cost from Roche, QIAGEN, Norchip, and MTM. He is a paid consultant for BD, GE Healthcare, Roche, Gen-Probe/Hologic, and Cepheid. He is compensated as a member of a Merck Data and Safety Monitoring Board for HPV vaccines. Dr. Jeronimo was the director of the study and received all the tests used in the study as a donation from the manufacturing companies (QIAGEN and Arbor Vita Corporation). Dr. Stoler is a consultant in a clinical trial design and implementation and/or has served as an expert pathologist for clinical trials for Merck, Roche, Becton-Dickinson, QIAGEN, Gen-Probe, Hologic, Ventana Medical Systems, and mtm Laboratories. © 2015, American Society for Colposcopy and Cervical Pathology
220
This study was nested within the Screening Technologies to Advance Rapid Testing—Utility and Program Planning project. In 2010 and 2011, 7,541 women aged 25 to 65 years living in Yangcheng, Xinmi, and Tonggu counties in China were enrolled. Women who were sexually active, not pregnant, had an intact uterus, and had no history of CIN, cervical cancer, or pelvic radiation were eligible for the present study. Details on participant recruitment have been published elsewhere.1 The institutional review boards of the Cancer Institute/Hospital, Chinese Academy of Medical Sciences (CICAMS), and PATH approved the screening study. All participants provided written informed consent before entry into the study. Participants in the screening study were given an education session about cervical cancer before the start of the study procedures. First, women were asked to complete a short risk-factor questionnaire administered by study personnel. Then, women were given instructions on how to self-collect a vaginal specimen; the procedure was completed in a private room. Next, women underwent a routine pelvic examination by female clinicians, at which time 2 cervical specimens were collected, the first into a dry tube for OncoE6 testing and the second into careHPV collection medium (QIAGEN) for HPV deoxyribonucleic acid (DNA) testing. Finally, visual inspection after 5% acetic acid (VIA) was done and the results recorded.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Copyright © 2015 American Society for Colposcopy and Cervical Pathology. Unauthorized reproduction of this article is prohibited.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Women who tested positive for any of the 6 screening tests performed (VIA, HPV-E6, Hybrid Capture 2, and careHPV on clinicians-collected and self-collected specimens) were referred to colposcopy, and approximately a 10% random sample of the women who tested negative for all screening tests (screennegative women) underwent a rigorous colposcopic evaluation that included using a 4-quadrant biopsy protocol as previously described.1 As dictated by the CICAMS and PATH institutional review boards, women who had no visible lesions had their screening result disclosed, and if screen negative on all 6 screening tests, no biopsies were taken. Two pathologists at CICAMS (X.Z. and J.L.) read the slides independently, and for any disagreements, a joint review was conducted to reach a consensus diagnosis. A third pathologist (M.H.S.) reviewed all high-grade (CIN2+) diagnoses, and disagreements were further resolved by a joint review.
Human Papillomavirus Genotyping of Biopsies The sandwich technique was used to cut paraffin sections from one cervical specimen for hematoxylin and eosin staining and HPV DNA analysis. Hematoxylin and eosin staining was reviewed by a pathologist (M.H.S.) and compared to the original impressions. Three 4-μmol/L paraffin sections were incubated in dry heat blocks for 24 hours with 250-μL proteinase K solution (1 mg/mL proteinase K, 45-mmol/L Tris–HCl, 0.9 mmol/L EDTA, and 0.45% Tween 20; pH 8.0) at 70°C + 0.5°C. Tubes were then incubated for 10 minutes at 95°C to inactivate proteinase K. A 10-μL aliquot of the DNA from each sample was used for HPV genotyping by SPF10 polymerase chain reactionDEIA-LiPA25 version 1 (Labo Biomedical Products, The Netherlands) based on SPF10-licensed Innogenetics technology), which uses SPF10 primers for DNA amplification by polymerase chain reaction3 and detected HPV genotypes by reverse hybridization. This system can detect 25 HPV types (HPV-6, HPV-11, HPV-16, HPV-18, HPV-31, HPV-33 to HPV-35, HPV-39, HPV40, HPV-42 to HPV-45, HPV-51 to HPV-54, HPV-56, HPV-58, HPV-59, HPV-66, HPV-68, HPV-70, and HPV-74) by using the reverse hybridization assay. For this analysis, HPV genotyping results were categorized according to cancer risk: HPV-16 positive, HPV-16 negative and HPV-18 positive, HPV-16 negative and HPV-18 negative and positive for other carcinogenic HPV types, negative for all carcinogenic HPV types and positive for noncarcinogenic HPV genotypes, HPV positive but no type detected, or HPV negative. The HPV test results were also categorized as HPV-16 and/or HPV-18 positive, or HPV-16 and HPV-18 negative (HPV-16/HPV-18 status).
Statistical Analyses We calculated the distribution of HPV categories by diagnosis (CIN2, CIN3, or cancer) and for CIN2+. We then stratified the categorized HPV results on whether the colposcopic impression was normal or abnormal and the VIA was negative or positive. Finally, the sensitivity of colposcopy and VIA for CIN2+ by HPV status, median age of CIN2+, and joint categories of each was calculated and differences in proportions between groups were tested for statistical differences using the Fisher exact test. We tested for differences in the median age of CIN2/3 by HPV-16/HPV-18 status using the Kruskal-Wallis nonparametric test. Analyses were done using Stata version 12.1 (Stata Corp, College Station, TX). p < .05 was considered statistically significant.
RESULTS The sociodemographic characteristics, risk factors, screen test positivity, and the clinical performance of each screening test © 2015, American Society for Colposcopy and Cervical Pathology
HPV and Cervical Visualization
have been described elsewhere.4 Table 1 shows the distribution of HPV categories and diagnoses of CIN2, CIN3, and cancer. We observed an increasing trend in the percent HPV-16 positive with severity of diagnosis: 33.3% for CIN2, 69.1% for CIN3, and 90.0% for cancer. The results of genotyping were stratified on whether or not there was visual evidence of abnormalities by colposcopy or VIA (Table 2). Overall, the distribution of HPV categories for CIN2+ was significantly different between normal and abnormal colposcopic impression (p = .007) and negative and positive VIA (p = 0.02), with a higher percentage HPV16 positive in visually positive CIN2+. Notably, the distribution of HPV categories for CIN3 was different for colposcopically positive (HPV-16: 78.6% and other carcinogenic: 14.3%) versus colposcopically negative (HPV-16: 55.3% and other carcinogenic: 42.1%) (p = .009) and VIA positive (HPV-16: 77.3% and other carcinogenic: 13.6%) versus VIA negative (HPV-16: 62.0% and other carcinogenic: 36.0%) (p = .01). We found that HPV-16–positive CIN2/3 was more common in younger women than in older women. In our study, the median age of HPV-16–positive CIN2/3 was younger than for CIN2/3 caused by other types (44 vs 48 years, respectively; p = .002) (data not shown). The sensitivity for CIN2+ was higher in HPV-16– and/or HPV-18–positive women than HPV-16– and HPV–18-negative women for colposcopy (67.4% vs 43.1%; p = .008, respectively) and VIA (53.3% vs 37.3%: p = 0.08, respectively) (Table 3). The sensitivity of colposcopy for CIN2+ decreased from 72.0% for HPV-16- or HPV18-positive CIN2+ in women younger than 46 years to 38.7% for HPV16- and HPV18-negative CIN2+ in women aged 46 years and older (p = .005) (odds ratio (OR), 4.0, 95% confidence interval [CI] = 1.4-12). Likewise, the sensitivity of VIA for CIN2+ decreased from 62.0% for HPV-16- or HPV18–positive CIN2+ in women younger than 46 years to 22.6% for HPV-16– and HPV-18–negative CIN2+ in women aged 46 years and older (p = .0006) (odds ratio, 5.6; 95% confidence interval [CI], 1.8–18.0).
DISCUSSION In this study, we found that the percent of HPV-16–positive women with CIN3 and cancer was greater than expected. The percent HPV-16 positive for CIN3 was 69.1% compared to 58.2% (p = .001) reported by Guan et al.,8 and the percent HPV-16 positive for cancer was also higher than expected from de Sanjose et al.9 (90% vs 60%, respectively; p = .04). Although we cannot explain the cause of the higher-than-expected HPV-16 positivity in CIN3 and cancer here, these findings raise the possibility that some unknown or unmeasured bias was introduced into our study. We also had a limited number of CIN2+ cases. To confirm these findings and better quantify the type-specific differences, data TABLE 1. Distribution of HPV Risk Categories by Diagnosis for Women With CIN2 or CIN2+ CIN2
CIN3
Cancer
Total
n (%)
n (%)
n (%)
n (%)
HPV-16 13 (33.3) 65 (69.1) 9 (90.0) 87 (60.8) HPV-18 5 (12.8) 0 (0.0) 0 (0.0) 5 (3.5) Other carcinogenic 18 (46.2) 24 (25.5) 0 (0.0) 42 (29.4) Noncarcinogenic 2 (5.1) 3 (3.2) 0 (0.0) 5 (3.5) Negative 1 (2.6) 2 (2.1) 1 (10.0) 4 (2.8) Total 39 (100.0) 94 (100.0) 10 (100.0) 143 (100.0)
221
Copyright © 2015 American Society for Colposcopy and Cervical Pathology. Unauthorized reproduction of this article is prohibited.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Jeronimo et al.
TABLE 2. Distribution of HPV Risk Categories of Women With CIN2 or CIN2+ by Colposcopic Impression (A) and VIA (B) Colposcopy positive CIN2 A HPV-16 HPV-18 Other carcinogenic Noncarcinogenic Negative Total
CIN3
Colposcopy negative
Cancer
Total
CIN2
HPV-16 HPV-18 Other Carcinogenic Noncarcinogenic Negative Total
Cancer
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
9 (42.9) 2 (9.5) 8 (38.1) 1 (4.8) 1 (4.8) 21 (100.0)
44 (78.6) 0 (0.0) 8 (14.3) 2 (3.6) 2 (3.6) 56 (100.0)
7 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 7 (100.0)
60 (71.4) 2 (2.4) 16 (19.0) 3 (3.6) 3 (3.6) 84 (100.0)
4 (22.2) 3 (16.7) 10 (55.6) 1 (5.6) 0 (0.0) 18 (100.0)
21 (55.3) 0 (0.0) 16 (42.1) 1 (2.6) 0 (0.0) 38 (100.0)
2 (66.7) 0 (0.0) 0 (0.0) 0 (0.0) 1 (33.3) 3 (100.0)
27 (45.8) 3 (5.1) 26 (44.1) 2 (3.4) 1 (1.7) 59 (100.0)
CIN2
CIN3
Cancer
Total
CIN2
CIN3
Cancer
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
7 (41.2) 2 (11.8) 6 (35.3) 1 (5.9) 1 (5.9) 17 (100.0)
34 (77.3) 0 (0.0) 6 (13.6) 3 (6.8) 1 (2.3) 44 (100.0)
6 (85.7) 0 (0.0) 0 (0.0) 0 (0.0) 1 (14.3) 7 (100.0)
47 (56.0) 2 (2.4) 12 (14.3) 4 (4.8) 3 (3.6) 68 (81.0)
6 (27.3) 3 (13.6) 12 (54.5) 1 (4.5) 0 (0.0) 22 (100.0)
31 (62.0) 0 (0.0) 18 (36.0) 0 (0.0) 1 (2.0) 50 (100.0)
3 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 3 (100.0)
40 (67.8) 3 (5.1) 30 (50.8) 1 (1.7) 1 (1.7) 75 (100.0)
VIA positive
B
CIN3
VIA negative
from equally rigorous studies should be pooled or studies of colposcopy should be done in populations vaccinated against HPV16 and HPV-18. Of note, there is significant misclassification by relying on HPV genotyping of cytologic specimens to determine attribution10 or colposcopically directed biopsies (vs 4-quadrant microbiopsies) to ascertain disease.11 We found that HPV-16/ HPV-18–positive and specifically HPV-16–positive CIN2+ was associated with having a more visually obvious CIN2+ lesion. We noted that the sensitivity of colposcopy and VIA was affected by HPV type and age, such that the sensitivity could vary by 2-fold between subgroups. Unlike previous studies, two of which found similar to HPV-16–related effects on colposcopy12,13 and one that did not,2 our study was not biased by referral to colposcopy of only women with cytologic abnormalities. Our findings, and those of the 2 previous positive studies,12,13 also suggest that HPV-16–related precancer is found earlier than precancer caused by other HPV types.14 The implication from our results is that in populations vaccinated against HPV-16 and HPV-18 and therefore lesions will be due almost entirely to types other than HPV-16 and HPV-18,
screening using VIA or diagnosis using colposcopy will be less sensitive for CIN2+. To compensate for the loss in colposcopic sensitivity, more biopsies could be taken randomly from areas of the cervix that might not qualify strictly as abnormal. We took advantage of our study design in which most women with CIN2+ in a screening population went to colposcopy and underwent a 4-quadrant microbiopsy protocol to investigate the effects of HPV type on visualization of CIN2+ lesions whether or not lesions could be seen. However, we noted some limitations in our analysis. Cytologic testing was not included as an additional cervical cancer screening test, thus limiting the generalizability of our results to women screened with cytology alone. However, this limitation would not apply to women tested with HPV DNA tests. In conclusion, secondary prevention through screening, diagnosis, and timely treatment of precancerous lesions is predicted to become less efficient in the absence of HPV-16 or HPV-18. First, the lower prevalence of disease in general will reduce the predictive value of all tests and diagnostic procedures.15 Human papillomavirus tests in the absence of HPV-16 are less predictive of CIN2+ and CIN3+.16 Second, subjective screening tests are
TABLE 3. Impact of Age and HPV-16/HPV-18 Status of Women With CIN2+ on the Sensitivity of Colposcopy (A) and VIA (B) for CIN2+ HPV-16/HPV-18+ A. Colposcopy
prophylactic HPV vaccination and molecular testing for detection of HPV. Although prophylactic HPV vaccination may be the ultimate cervical cancer prevention strategy, it does not treat preexisting HPV infections and related precancer conditions. Additionally, the current generation of HPV vaccines target 2 of the 14 genotypes associated with cervical cancer (HPV-16 and HPV-18) and therefore will not prevent approximately 25% to 30% of cervical cancers. Thus, cervical cancer screening will be needed in the foreseeable future to prevent cervical cancer. We took advantage of a population-based study of lower-cost cervical cancer screening conducted in rural China to examine the clinical importance of HPV genotypes in the visualization of cervical precancer lesions, during screening by visual inspection with acetic acid (VIA) and colposcopy in the management of HPVscreen–positive women.1 Because women were exhaustively screened with 6 screening tests (including VIA) and all screenpositive women underwent a rigorous colposcopic evaluation, we assume that most women with cervical precancer and cancer in the study population had a diagnosis. The primary objective of this study was to examine the relationship of causal HPV genotypes on the sensitivity of VIA for screening, and colposcopy for diagnosis of cervical intraepithelial neoplasia more severe than grade 2 (CIN2+). As studies have shown that HPV-16–positive CIN2/3 was more common in younger women than in older women,2–7 we evaluated the effect of HPV status and median age on the sensitivity of CIN2+ diagnosis.
Key Words: cervical cancer, colposcopy, cervical screening, HPV, China (J Lower Gen Tract Dis 2015;19: 220–223)
MATERIALS AND METHODS
T
he identification of human papillomavirus (HPV) as the primary cause of cervical cancer has led to the development of novel primary and secondary cervical cancer prevention strategies: 1
PATH, Seattle, WA; 2Department of Epidemiology, Cancer Institute and Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, People's Republic of China; 3Department of Pathology, University of Virginia, Charlottesville, VA; and 4Global Cancer Initiative, Chestertown, MD Reprint requests to: Jose Jeronimo, MD, PATH, PO Box 900922, Seattle, WA 98109. E-mail: [email protected] The START-UP project was funded by a grant from the Bill & Melinda Gates Foundation. Conflict of interest: No conflicts of interest for Pooja Bansil, Melissa Valdez, Le-Ni Kang, Dr. Zhao, Dr. Qiao, Dr. Chen, Dr. Zhang, Proma Paul, Dr. Bai, Roger Peck, Dr. Li, and Feng Chen. Dr. Castle has received commercial HPV tests for research at a reduced or no cost from Roche, QIAGEN, Norchip, and MTM. He is a paid consultant for BD, GE Healthcare, Roche, Gen-Probe/Hologic, and Cepheid. He is compensated as a member of a Merck Data and Safety Monitoring Board for HPV vaccines. Dr. Jeronimo was the director of the study and received all the tests used in the study as a donation from the manufacturing companies (QIAGEN and Arbor Vita Corporation). Dr. Stoler is a consultant in a clinical trial design and implementation and/or has served as an expert pathologist for clinical trials for Merck, Roche, Becton-Dickinson, QIAGEN, Gen-Probe, Hologic, Ventana Medical Systems, and mtm Laboratories. © 2015, American Society for Colposcopy and Cervical Pathology
220
This study was nested within the Screening Technologies to Advance Rapid Testing—Utility and Program Planning project. In 2010 and 2011, 7,541 women aged 25 to 65 years living in Yangcheng, Xinmi, and Tonggu counties in China were enrolled. Women who were sexually active, not pregnant, had an intact uterus, and had no history of CIN, cervical cancer, or pelvic radiation were eligible for the present study. Details on participant recruitment have been published elsewhere.1 The institutional review boards of the Cancer Institute/Hospital, Chinese Academy of Medical Sciences (CICAMS), and PATH approved the screening study. All participants provided written informed consent before entry into the study. Participants in the screening study were given an education session about cervical cancer before the start of the study procedures. First, women were asked to complete a short risk-factor questionnaire administered by study personnel. Then, women were given instructions on how to self-collect a vaginal specimen; the procedure was completed in a private room. Next, women underwent a routine pelvic examination by female clinicians, at which time 2 cervical specimens were collected, the first into a dry tube for OncoE6 testing and the second into careHPV collection medium (QIAGEN) for HPV deoxyribonucleic acid (DNA) testing. Finally, visual inspection after 5% acetic acid (VIA) was done and the results recorded.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Copyright © 2015 American Society for Colposcopy and Cervical Pathology. Unauthorized reproduction of this article is prohibited.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Women who tested positive for any of the 6 screening tests performed (VIA, HPV-E6, Hybrid Capture 2, and careHPV on clinicians-collected and self-collected specimens) were referred to colposcopy, and approximately a 10% random sample of the women who tested negative for all screening tests (screennegative women) underwent a rigorous colposcopic evaluation that included using a 4-quadrant biopsy protocol as previously described.1 As dictated by the CICAMS and PATH institutional review boards, women who had no visible lesions had their screening result disclosed, and if screen negative on all 6 screening tests, no biopsies were taken. Two pathologists at CICAMS (X.Z. and J.L.) read the slides independently, and for any disagreements, a joint review was conducted to reach a consensus diagnosis. A third pathologist (M.H.S.) reviewed all high-grade (CIN2+) diagnoses, and disagreements were further resolved by a joint review.
Human Papillomavirus Genotyping of Biopsies The sandwich technique was used to cut paraffin sections from one cervical specimen for hematoxylin and eosin staining and HPV DNA analysis. Hematoxylin and eosin staining was reviewed by a pathologist (M.H.S.) and compared to the original impressions. Three 4-μmol/L paraffin sections were incubated in dry heat blocks for 24 hours with 250-μL proteinase K solution (1 mg/mL proteinase K, 45-mmol/L Tris–HCl, 0.9 mmol/L EDTA, and 0.45% Tween 20; pH 8.0) at 70°C + 0.5°C. Tubes were then incubated for 10 minutes at 95°C to inactivate proteinase K. A 10-μL aliquot of the DNA from each sample was used for HPV genotyping by SPF10 polymerase chain reactionDEIA-LiPA25 version 1 (Labo Biomedical Products, The Netherlands) based on SPF10-licensed Innogenetics technology), which uses SPF10 primers for DNA amplification by polymerase chain reaction3 and detected HPV genotypes by reverse hybridization. This system can detect 25 HPV types (HPV-6, HPV-11, HPV-16, HPV-18, HPV-31, HPV-33 to HPV-35, HPV-39, HPV40, HPV-42 to HPV-45, HPV-51 to HPV-54, HPV-56, HPV-58, HPV-59, HPV-66, HPV-68, HPV-70, and HPV-74) by using the reverse hybridization assay. For this analysis, HPV genotyping results were categorized according to cancer risk: HPV-16 positive, HPV-16 negative and HPV-18 positive, HPV-16 negative and HPV-18 negative and positive for other carcinogenic HPV types, negative for all carcinogenic HPV types and positive for noncarcinogenic HPV genotypes, HPV positive but no type detected, or HPV negative. The HPV test results were also categorized as HPV-16 and/or HPV-18 positive, or HPV-16 and HPV-18 negative (HPV-16/HPV-18 status).
Statistical Analyses We calculated the distribution of HPV categories by diagnosis (CIN2, CIN3, or cancer) and for CIN2+. We then stratified the categorized HPV results on whether the colposcopic impression was normal or abnormal and the VIA was negative or positive. Finally, the sensitivity of colposcopy and VIA for CIN2+ by HPV status, median age of CIN2+, and joint categories of each was calculated and differences in proportions between groups were tested for statistical differences using the Fisher exact test. We tested for differences in the median age of CIN2/3 by HPV-16/HPV-18 status using the Kruskal-Wallis nonparametric test. Analyses were done using Stata version 12.1 (Stata Corp, College Station, TX). p < .05 was considered statistically significant.
RESULTS The sociodemographic characteristics, risk factors, screen test positivity, and the clinical performance of each screening test © 2015, American Society for Colposcopy and Cervical Pathology
HPV and Cervical Visualization
have been described elsewhere.4 Table 1 shows the distribution of HPV categories and diagnoses of CIN2, CIN3, and cancer. We observed an increasing trend in the percent HPV-16 positive with severity of diagnosis: 33.3% for CIN2, 69.1% for CIN3, and 90.0% for cancer. The results of genotyping were stratified on whether or not there was visual evidence of abnormalities by colposcopy or VIA (Table 2). Overall, the distribution of HPV categories for CIN2+ was significantly different between normal and abnormal colposcopic impression (p = .007) and negative and positive VIA (p = 0.02), with a higher percentage HPV16 positive in visually positive CIN2+. Notably, the distribution of HPV categories for CIN3 was different for colposcopically positive (HPV-16: 78.6% and other carcinogenic: 14.3%) versus colposcopically negative (HPV-16: 55.3% and other carcinogenic: 42.1%) (p = .009) and VIA positive (HPV-16: 77.3% and other carcinogenic: 13.6%) versus VIA negative (HPV-16: 62.0% and other carcinogenic: 36.0%) (p = .01). We found that HPV-16–positive CIN2/3 was more common in younger women than in older women. In our study, the median age of HPV-16–positive CIN2/3 was younger than for CIN2/3 caused by other types (44 vs 48 years, respectively; p = .002) (data not shown). The sensitivity for CIN2+ was higher in HPV-16– and/or HPV-18–positive women than HPV-16– and HPV–18-negative women for colposcopy (67.4% vs 43.1%; p = .008, respectively) and VIA (53.3% vs 37.3%: p = 0.08, respectively) (Table 3). The sensitivity of colposcopy for CIN2+ decreased from 72.0% for HPV-16- or HPV18-positive CIN2+ in women younger than 46 years to 38.7% for HPV16- and HPV18-negative CIN2+ in women aged 46 years and older (p = .005) (odds ratio (OR), 4.0, 95% confidence interval [CI] = 1.4-12). Likewise, the sensitivity of VIA for CIN2+ decreased from 62.0% for HPV-16- or HPV18–positive CIN2+ in women younger than 46 years to 22.6% for HPV-16– and HPV-18–negative CIN2+ in women aged 46 years and older (p = .0006) (odds ratio, 5.6; 95% confidence interval [CI], 1.8–18.0).
DISCUSSION In this study, we found that the percent of HPV-16–positive women with CIN3 and cancer was greater than expected. The percent HPV-16 positive for CIN3 was 69.1% compared to 58.2% (p = .001) reported by Guan et al.,8 and the percent HPV-16 positive for cancer was also higher than expected from de Sanjose et al.9 (90% vs 60%, respectively; p = .04). Although we cannot explain the cause of the higher-than-expected HPV-16 positivity in CIN3 and cancer here, these findings raise the possibility that some unknown or unmeasured bias was introduced into our study. We also had a limited number of CIN2+ cases. To confirm these findings and better quantify the type-specific differences, data TABLE 1. Distribution of HPV Risk Categories by Diagnosis for Women With CIN2 or CIN2+ CIN2
CIN3
Cancer
Total
n (%)
n (%)
n (%)
n (%)
HPV-16 13 (33.3) 65 (69.1) 9 (90.0) 87 (60.8) HPV-18 5 (12.8) 0 (0.0) 0 (0.0) 5 (3.5) Other carcinogenic 18 (46.2) 24 (25.5) 0 (0.0) 42 (29.4) Noncarcinogenic 2 (5.1) 3 (3.2) 0 (0.0) 5 (3.5) Negative 1 (2.6) 2 (2.1) 1 (10.0) 4 (2.8) Total 39 (100.0) 94 (100.0) 10 (100.0) 143 (100.0)
221
Copyright © 2015 American Society for Colposcopy and Cervical Pathology. Unauthorized reproduction of this article is prohibited.
Journal of Lower Genital Tract Disease • Volume 19, Number 3, July 2015
Jeronimo et al.
TABLE 2. Distribution of HPV Risk Categories of Women With CIN2 or CIN2+ by Colposcopic Impression (A) and VIA (B) Colposcopy positive CIN2 A HPV-16 HPV-18 Other carcinogenic Noncarcinogenic Negative Total
CIN3
Colposcopy negative
Cancer
Total
CIN2
HPV-16 HPV-18 Other Carcinogenic Noncarcinogenic Negative Total
Cancer
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
9 (42.9) 2 (9.5) 8 (38.1) 1 (4.8) 1 (4.8) 21 (100.0)
44 (78.6) 0 (0.0) 8 (14.3) 2 (3.6) 2 (3.6) 56 (100.0)
7 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 7 (100.0)
60 (71.4) 2 (2.4) 16 (19.0) 3 (3.6) 3 (3.6) 84 (100.0)
4 (22.2) 3 (16.7) 10 (55.6) 1 (5.6) 0 (0.0) 18 (100.0)
21 (55.3) 0 (0.0) 16 (42.1) 1 (2.6) 0 (0.0) 38 (100.0)
2 (66.7) 0 (0.0) 0 (0.0) 0 (0.0) 1 (33.3) 3 (100.0)
27 (45.8) 3 (5.1) 26 (44.1) 2 (3.4) 1 (1.7) 59 (100.0)
CIN2
CIN3
Cancer
Total
CIN2
CIN3
Cancer
Total
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
n (%)
7 (41.2) 2 (11.8) 6 (35.3) 1 (5.9) 1 (5.9) 17 (100.0)
34 (77.3) 0 (0.0) 6 (13.6) 3 (6.8) 1 (2.3) 44 (100.0)
6 (85.7) 0 (0.0) 0 (0.0) 0 (0.0) 1 (14.3) 7 (100.0)
47 (56.0) 2 (2.4) 12 (14.3) 4 (4.8) 3 (3.6) 68 (81.0)
6 (27.3) 3 (13.6) 12 (54.5) 1 (4.5) 0 (0.0) 22 (100.0)
31 (62.0) 0 (0.0) 18 (36.0) 0 (0.0) 1 (2.0) 50 (100.0)
3 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 3 (100.0)
40 (67.8) 3 (5.1) 30 (50.8) 1 (1.7) 1 (1.7) 75 (100.0)
VIA positive
B
CIN3
VIA negative
from equally rigorous studies should be pooled or studies of colposcopy should be done in populations vaccinated against HPV16 and HPV-18. Of note, there is significant misclassification by relying on HPV genotyping of cytologic specimens to determine attribution10 or colposcopically directed biopsies (vs 4-quadrant microbiopsies) to ascertain disease.11 We found that HPV-16/ HPV-18–positive and specifically HPV-16–positive CIN2+ was associated with having a more visually obvious CIN2+ lesion. We noted that the sensitivity of colposcopy and VIA was affected by HPV type and age, such that the sensitivity could vary by 2-fold between subgroups. Unlike previous studies, two of which found similar to HPV-16–related effects on colposcopy12,13 and one that did not,2 our study was not biased by referral to colposcopy of only women with cytologic abnormalities. Our findings, and those of the 2 previous positive studies,12,13 also suggest that HPV-16–related precancer is found earlier than precancer caused by other HPV types.14 The implication from our results is that in populations vaccinated against HPV-16 and HPV-18 and therefore lesions will be due almost entirely to types other than HPV-16 and HPV-18,
screening using VIA or diagnosis using colposcopy will be less sensitive for CIN2+. To compensate for the loss in colposcopic sensitivity, more biopsies could be taken randomly from areas of the cervix that might not qualify strictly as abnormal. We took advantage of our study design in which most women with CIN2+ in a screening population went to colposcopy and underwent a 4-quadrant microbiopsy protocol to investigate the effects of HPV type on visualization of CIN2+ lesions whether or not lesions could be seen. However, we noted some limitations in our analysis. Cytologic testing was not included as an additional cervical cancer screening test, thus limiting the generalizability of our results to women screened with cytology alone. However, this limitation would not apply to women tested with HPV DNA tests. In conclusion, secondary prevention through screening, diagnosis, and timely treatment of precancerous lesions is predicted to become less efficient in the absence of HPV-16 or HPV-18. First, the lower prevalence of disease in general will reduce the predictive value of all tests and diagnostic procedures.15 Human papillomavirus tests in the absence of HPV-16 are less predictive of CIN2+ and CIN3+.16 Second, subjective screening tests are
TABLE 3. Impact of Age and HPV-16/HPV-18 Status of Women With CIN2+ on the Sensitivity of Colposcopy (A) and VIA (B) for CIN2+ HPV-16/HPV-18+ A. Colposcopy
