Obstetrics & Gynecology:
An Analysis of High-Risk Human Papillomavirus DNA-Negative Cervical Precancers in the ASCUS-LSIL Triage Study (ALTS)
Castle, Philip E. PhD, MPH1; Cox, J Thomas MD2; Jeronimo, Jose MD1; Solomon, Diane MD3; Wheeler, Cosette M. PhD4; Gravitt, Patti E. PhD, MS5; Schiffman, Mark MD, MPH1
From the 1Division of Cancer Epidemiology and Genetics and 3Division of Cancer Prevention, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, Maryland; 2Gynecology and Colposcopy Clinic Student Health Service, University of California, Santa Barbara, California; 4Departments of Molecular Genetics and Microbiology and Obstetrics and Gynecology, University of New Mexico Health Sciences Center, School of Medicine, Albuquerque, New Mexico; and 5Department of Epidemiology, Johns Hopkins University, Baltimore, Maryland.
Supported by the National Cancer Institute, National Institutes of Health Department of Health and Human Services contracts CN-55153, CN-55154, CN-55155, CN-55156, CN-55157, CN-55158, CN-55159 and CN-55105. Also supported in part by intramural research program of the National Cancer Institute. Some of the equipment and supplies used in these studies were donated or provided at reduced cost by Digene Corporation, Gaithersburg, MD; Cytyc Corporation, Marlborough, MA; National Testing Laboratories, Fenton, MO; DenVu, Tucson, AZ; TriPath Imaging, Inc., Burlington, NC; and Roche Molecular Systems Inc., Alameda, CA.
The authors thank the ALTS Group Investigators for their help in planning and conducting the trial and Information Management Services, Inc., Rockville, Maryland, for data management and programming support.
Presented in part at the 24th International Papillomavirus Conference, Beijing, China, November 3–9, 2007.
Corresponding author: Philip E. Castle, PhD, MPH, Division of Cancer Epidemiology and Genetics, National Cancer Institute, 6120 Executive Boulevard. Room 5004, EPS MSC 7234, Bethesda, MD 20892-7234; e-mail: email@example.com.
Financial Disclosure Dr. Cox received honoraria from and was a consultant to Digene Corporation (Gaithersburg, MD). Dr. Gravitt is a consultant for and has received research funding from Roche Molecular Diagnostics (Pleasanton, CA), who manufactured the HPV Linear Array test used in this study. The other authors have no potential conflicts of interest to disclose.)
OBJECTIVE: To describe women diagnosed with cervical intraepithelial neoplasia-grade 3 (CIN-3) diagnosed over the 2-year duration of the atypical squamous cells of undetermined significance (ASCUS) and low-grade squamous intraepithelial lesion (LSIL) Triage Study (ALTS) that tested negative for high-risk human papillomavirus (HPV) at enrollment.
METHODS: Clinical center pathologists and quality control pathology group reviewed all histology; any CIN-3 diagnosis on biopsy or loop electrosurgical excision procedure (n=621) by at least one pathology review over the duration of ALTS led to inclusion in this analysis. Enrollment cervical specimens were tested for high-risk HPV DNA by two HPV assays; results were combined to minimize simple testing errors. We compared the characteristics of baseline high-risk HPV-negative (n=33) to baseline high-risk HPV-positive (n=588) cumulative diagnosed CIN-3.
RESULTS: High-risk HPV-negative CIN-3 cases were less likely to have a second, confirming diagnosis of CIN-3 (24% compared with 56%) by the other pathology group, were more likely to be diagnosed later in follow-up, and more likely to be referred into ALTS because of an ASCUS Pap test rather than an LSIL Pap. Upon review of case histories of the 33 baseline high-risk HPV-negative CIN-3 (5.3% of all cases), there was evidence that these cases were due to incident (new) cases (n=12, 1.9%), non–high-risk HPV (n=5, 0.8%), misclassified histology (n=8, 1.3%), and false-negative high-risk HPV (n=8, 1.3%).
CONCLUSION: In any sizeable population, even among women with evidence of cytologic abnormalities, there will be a few cases of cervical precancer that will test high-risk HPV negative for one or more reasons.
LEVEL OF EVIDENCE: II
Click on the links below to access all the Data Supplements for this article.
Please note that Data Supplement files may launch a viewer application outside of your web browser.
Carcinogenic or “high-risk” human papillomavirus (high-risk HPV) testing has become the standard triage in the US for women with atypical squamous cells of undetermined significance (ASC-US) cytology and is designated as a stand-alone follow-up option in a number of postcolposcopy and posttreatment clinical management scenarios.1,2 Primary cervical screening guidelines now also include the option of cotesting women aged 30 years and older with the Papanicolaou (Pap) test and a high-risk HPV test and extending screening intervals to 3 years if both are negative.1–3 Additionally, some are advocating moving to screening women in this age group with the more sensitive high-risk HPV test alone, followed by a more specific test (eg, cytology) for women testing high-risk HPV positive.4 This paradigm shift to molecular testing in primary cervical screening and abnormal Pap test management has been based on the nearly absolute requirement of persistent infections by high-risk HPV genotypes in the carcinogenesis process leading to cervical cancer and its immediate precursor lesion, cervical intraepithelial neoplasia grade 3 (CIN-3).5
The ASCUS LSIL Triage Study (ALTS), a multicenter, randomized clinical trial designed to compare management strategies for women with low-grade squamous intraepithelial lesion (LSIL) and equivocal (ASC-US) cytologic abnormalities,6–9 confirmed findings of previous studies that high-risk HPV testing has a high sensitivity (>90%) to identify women with CIN-3 and a corresponding high level of reassurance of the absence of CIN-3 among high-risk HPV-negative women.10,11
Dependence on a single test result as the only determinant of screening interval and follow-up options requires confidence that the test will accurately reflect risk. Cervical intraepithelial neoplasia 3 is almost invariably caused by high-risk HPV, and women are generally high-risk HPV positive several years before a CIN-3 diagnosis. However, no test has 100% accuracy, and HPV testing is no exception. Increasing use of HPV testing has resulted in occasional reports of “false-negative” HPV test results for women with CIN-3.12,13 If testing high-risk HPV positive supposedly identifies the subset of women who are at risk of CIN-3 and cancer, and testing high-risk HPV negative provides reassurance against CIN-3 and cancer, what is the meaning of a high-risk HPV-negative test occuring within a couple of years before a CIN-3 diagnosis? What are the possible explanations for these HR-HPV negative CIN-3? Most importantly, how clinically relevant are these relatively rare occurrences?
In ALTS, the 2-year risk of CIN-3 among high-risk HPV-negative women was quite low,11 less than 2%. To better understand high-risk HPV-negative CIN-3, we described the occurrence of such cases in ALTS. We combined the test results of two well-validated HPV tests to minimize simple testing errors, thereby focusing our analysis on other sources of error. We compared CIN-3 diagnosed throughout the 2-year duration of ALTS among participants who at baseline tested high-risk HPV negative to those who tested positive to characterize differences by high-risk HPV status and to elucidate what percentage of cases might be 1) incident cases, 2) due to non–high-risk HPV genotypes, 3) false-positive pathology, and 4) false-negative for high-risk HPV (despite dual HPV testing). The goal of the study was to provide insight into the cause of many of the cases of high-risk HPV-negative CIN-3 and their clinical relevance.
PARTICIPANTS AND METHODS
ALTS was a randomized trial comparing three management strategies for 5,060 women with ASC-US (n=3,488) or LSIL (n=1,572): 1) immediate colposcopy arm (referral to colposcopy regardless of enrollment test results); 2) HPV triage (referral to colposcopy if enrollment HPV result by Hybrid Capture 2 [Digene Corporation, Gaithersburg, MD] was positive or missing, or if the enrollment cytology was high-grade squamous intraepithelial lesion [HSIL]); or 3) conservative management (referral to colposcopy if enrollment cytology was HSIL). At enrollment, all women underwent a pelvic examination with collection of two cervical specimens; the first specimen in PreservCyt for ThinPrep cytology (Cytyc Corporation, Marlborough, MA) and the second in specimen transport medium (Digene Corporation). Women in all three arms of the study were reevaluated by cytology every 6 months for 2 years and sent to colposcopy if cytology was HSIL. An exit examination with colposcopy was scheduled for all women, regardless of study arm or prior procedures, at the completion of the follow-up. We refer readers to other references for details on randomization, examination procedures, patient management, and laboratory and pathology methods.6–9 The National Cancer Institute and local institutional review boards approved the study and all participants provided written informed consent.
Testing was blinded to all clinical data, and testing data were not used in making diagnoses. Two HPV DNA tests were performed on residual clinical specimens collected at enrollment and follow-up.
Hybrid Capture 2 using probe set B, a pooled probe DNA test for one or more high-risk HPV genotypes (HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68), was performed on residual Preserv-Cyt specimens after cytology slides were prepared. A positive test does not identify which high-risk HPV genotype(s) are present. Hybrid Capture 2 is also well-known to cross-react with untargeted HPV genotypes14,15 including HPV 66, recently classified as a high-risk HPV genotype.16
Human papillomavirus genotyping was performed using Line Blot Assay (Roche Molecular Systems, Alameda, CA), which is an L1-based polymerase chain reaction assay that uses a primer set designated PGMY09/11. Line Blot Assay was conducted on the specimen transport medium specimen for detection of 27 (HPV 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51–59, 66, 68, 73, 82, 83, and 84) or 38 HPV (the 27 HPV genotypes listed plus non-high-risk HPV genotypes 61, 62, 64, 67, 69–72, 81, 82 variant, and 89) genotypes at enrollment and during follow-up.15
Enrollment specimens collected from women referred because of an ASC-US Pap test were retrospectively tested using Linear Array (Roche Molecular Systems), a commercialized, research-use-only version of Line Blot Assay that tests for 37 HPV genotypes (excluding HPV 57) as previously described.17 Of note, among those with a single high-risk HPV genotype detected by Linear Array, increasing band intensity (categorized as 1 [low] to 4 [high]), as judged subjectively from the line blot, was strongly associated with increasing Hybrid Capture 2 signal strength (P<.001, nonparametric test of trend18) (data not shown), a semi-quantitative measure of viral load.19 We therefore considered band intensity a qualitative measure of HPV viral load.
HPV genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68 were considered the primary high-risk HPV genotypes. Line Blot Assay and Linear Array results were categorized into an HPV risk group according to a priori established cervical cancer risk: 1) positive for HPV 16; or 2) positive for any high-risk HPV genotype(s) excluding HPV 16; or 3) positive for any non–high-risk HPV genotypes and negative for all high-risk HPV genotypes; or 4) negative.
Clinical management was based on the clinical center pathologists’ cytologic and histologic diagnoses. In addition, referral smears, ThinPreps, and histology slides were sent to the Pathology Quality Control Group (Quality Control pathology) based at the Johns Hopkins Hospital for review and secondary diagnoses. Cervical intraepithelial neoplasia 2 or more severe (CIN-3, adenocarcinoma in situ, carcinoma in situ, adenocarcinoma, and squamous cell carcinoma) (≥CIN-2) histologic diagnoses based on the clinical center pathology or a CIN-3 or more severe (≥CIN-3) histologic diagnoses based on the Quality Control pathology review triggered treatment by loop electrosurgical excision procedure. In addition, women with persistent LSIL or high-risk HPV-positive ASC-US at the time of the exit from the study were offered loop electrosurgical excision procedure. There were 621 women who had a diagnosis of CIN-3 by either pathology group and this was the subset of women included in this analysis. (nb, seven cases of squamous cell cancer, all high-risk HPV positive, were included in this analysis. Exclusion of these cases did not appreciably change the results of the study.)
We conducted a blinded, retrospective analysis of the visual appearance of the cervix of high-risk HPV-negative and HPV-positive CIN-3 by reviewing Cervigrams (National Testing Laboratories Worldwide, Fenton, MO) of high-risk HPV-negative cases matched 1:2 to high-risk HPV-positive cases on study arm, clinical center, time of diagnosis (enrollment, follow-up, or exit), and referral Pap test. Cervigrams were digitized using standard features20 to obtain high-resolution images equivalent to 6-megapixel pictures. One of us (J.J.) reviewed both baseline and time-of-diagnosis Cervigrams for the cases diagnosed during follow-up. Baseline Cervigrams from 29 high-risk HPV negative and 57 high-risk HPV-positive cases (one high-risk HPV-negative case had only one matching high-risk HPV-positive case) and time-of-diagnosis Cervigrams (nb, some baseline Cervigrams were also time-of-diagnosis Cervigrams) from 27 high-risk HPV-negative and 50 high-risk HPV-positive cases (four high-risk HPV-negative cases had only one matching high-risk HPV-positive case) were evaluated; other high-risk HPV-negative cases were either missing Cervigrams or no match could be found. Evaluations were done using a Web-based software that presented the digitized Cervigram and masked the evaluator to any clinical data.21
We defined women with CIN-3 as high-risk HPV negative if Line Blot Assay and Hybrid Capture 2 were negative for high-risk HPV at baseline or if one test was negative for high-risk HPV and the other test result was missing (Table 1); we designated 588 cases (94.7%) as high-risk HPV positive and 33 cases (5.3%) as high-risk HPV negative. We considered women who were Hybrid Capture 2–positive, Line Blot Assay–positive for non–high-risk HPV genotypes but negative for high-risk HPV genotypes (n=11) as high-risk HPV-positive cases. However, we could not ascertain whether these results were the consequence of false-negative Line Blot Assay result or false-positive Hybrid Capture 2 result due to cross-reactivity with non–high-risk HPV genotypes (eg, HPV 53 and HPV 67).14,15 Exclusion of these cases did not appreciable alter our findings. We did not consider the Linear Array results in our primary definition of HPV-negative CIN-3 cases because we had Linear Array testing only on ASCUS referrals.
We first described our cases of CIN-3 by the distribution of paired diagnoses rendered by both pathology groups for all women and stratified by HPV status (positive compared with negative) as described above. We also compared the timing of diagnosis (enrollment, follow-up, or exit) of CIN-3 by HPV status for women enrolled in the immediate colposcopy and conservative management arms. Women enrolled in the HPV arm of ALTS were excluded from this analysis because of the inherent bias of referral based on HPV testing results by Hybrid Capture 2.
Standard contingency table methods, with Fisher exact tests, were used to assess possible univariate associations of categorical variables with HPV status. Odds ratios (ORs) and 95% confidence intervals (95% CIs) adjusted for relevant parameters (eg, identified as part of preliminary data analysis and stepwise modeling) were calculated using multivariate logistic regression.
Finally, we considered the results of Linear Array testing conducted in the ASC-US referral population. There are several studies to suggest that Linear Array may be more analytically sensitive than Line Blot Assay for the detection of HPV genotypes. We compared the distribution of HPV-negative and HPV-positive cases for HPV risk group status as determined by Linear Array.
Fisher exact test was used to test for statistical significance (P<.05) for categorical variables; Kruskal- Wallis test was used to test for statistical significance for continuous variables. Stata 8.2 (Stata Corporation, College Station, TX) was used for all statistical analyses.
In Table 2, we show the paired diagnoses by the two pathology groups at the time of the CIN-3 histologic diagnosis for anyone with at least one CIN-3 diagnosis during ALTS. Overall, 55% of women with a CIN-3 diagnosis had a CIN-3 diagnosis by both pathology groups, and 93% had CIN-3 diagnosis by one pathology group and CIN-2 diagnosis or worse by the other group. Hence, only 7% with a >CIN-3 diagnosis by one group were interpreted as normal or low grade (CIN-1) by the other group. The distribution of paired diagnoses was very different by HPV status (P<.001). High-risk HPV-negative CIN-3 were less likely to have a concordant CIN-3 diagnosis by both groups (24% compared with 56%) (P<.001) and were less likely to have a second diagnosis of at least CIN-2 (76% compared with 94%) (P =.001) than the high-risk HPV-positive CIN-3.
Table 3 shows the time of detection of CIN-3. In both the immediate colposcopy arm (P =.001) and the conservative management arm (P =.02), HPV-negative CIN-3 was found later than HPV-positive CIN-3 and predominantly at the exit visit. In the immediate colposcopy arm, in which all women underwent colposcopy at baseline, a greater proportion of high-risk HPV-negative CIN-3 (55.0%) was diagnosed at exit than the proportion of high-risk HPV-positive CIN-3 (17.8%). In the conservative management arm, in which women went to baseline colposcopy only for enrollment HSIL cytology, a greater proportion of high-risk HPV-negative CIN-3 (75.0%) were diagnosed at exit than the proportion of high-risk HPV-positive CIN-3 (31.5%).
We also examined what other factors might be associated with case HPV status using contingency tables and then multivariate modeling to mutually adjust for all relevant covariates (Table 4). Cases of high-risk HPV-negative CIN-3 (compared with high-risk HPV-positive CIN-3) were: 1) less likely to be referred into ALTS because of an LSIL Pap test (compared with ASC-US Pap test) (OR 0.44, 95% CI 0.18–1.1); 2) far less likely to have an enrollment cytology interpreted by the clinical center pathologist as HSIL (compared with <HSIL) (OR 0.10, 95% CI 0.046–0.22); 3) more likely to be enrolled at the clinical center 3 (OR 2.7, 95% CI 0.99–7.6) (nb, CIN-3 diagnosed by clinical center 3 were also less likely to be confirmed CIN-3 by Quality Control pathology [data not shown].); and 4) less likely to report having abnormal discharge in the last year (compared with no discharge) (OR 0.33, 95% CI 0.12–0.91).
Enrollment specimens from high-risk HPV-negative cases of CIN-3 were much more likely to again test negative for high-risk HPV by a third test, Linear Array, (64.0%; 28.0% negative for all genotypes and 36.0% for non–high-risk HPV genotypes) than high-risk HPV-positive cases (2.7%; 1.5% negative for all genotypes and 1.2% for non–high-risk HPV genotypes)(P<.001) (Table 5). Among those cases in which Linear Array was positive for only one high-risk HPV genotype (n=84), the band intensity was lower among the cases of CIN-3 originally identified as high-risk HPV negative (n=9) than high-risk HPV positive (n=75) (P<.001).
Cervigrams collected at the time of diagnosis from high-risk HPV-negative cases were marginally less likely to be called high-grade or cancerous than those from high-risk HPV-positive cases (52% compared with 70%, P =.1). Cervigrams from high-risk HPV-negative cases had marginally smaller areas of acetowhitening than those from high-risk HPV-positive cases (P =.08).
Considering all measurements of HPV throughout the duration of ALTS, only 82% of the CIN-3 initially high-risk HPV-negative cases at enrollment were positive for any HPV genotype on any occasion. Only 64% of these cases were ever positive for any high-risk HPV genotype, and 30% were ever positive for HPV 16. All percentages were lower than the percentages observed for cases of CIN-3 diagnosed in cross-sectional studies22,23 (see Appendix online at www.greenjournal.org/cgi/content/full/111/4/847/DC1). Among cases of CIN-3 diagnosed during follow-up that tested HPV-negative at enrollment (n=26), 19 (73%) were positive for high-risk HPV by either Line Blot Assay or Hybrid Capture 2 at the time of diagnosis, and only 10 of 19 of those that were positive by either test were positive by both assays (53% compared with 93% for the baseline high-risk HPV-positive cases, P<.001).
In ALTS, it seemed that the few high-risk HPV-negative CIN-3 cases diagnosed were less definitively “CIN-3.” Based on these data, we suggest four possible causes of HPV-negative CIN-3: 1) incident (new) cases of CIN-3 that developed during follow-up; 2) “non–high-risk HPV” HPV genotypes that rarely if ever cause cancer but may cause CIN-3; 3) histologic false-positives, classified incorrectly (misclassified) as CIN-3; and 4) false-negative high-risk HPV test results. To help illustrate examples of each, we have included case histories for all 33 HPV-negative CIN-3 (see the Appendix online at www.greenjournal.org/cgi/content/full/111/4/847/DC1). Although it is difficult to ascribe with certainty any one cause for these high-risk HPV-negative cases, upon independent review of the case histories by four of the authors (P.C., T.C., J.J., and M.S.), we suggest that of the 621 CIN-3 cases overall there were 12 incident cases (1.9%), five cases (0.8%) due to non–high-risk HPV, and eight cases (1.3%) due to incorrectly classified CIN-3, leaving only eight cases (1.3%) with false-negative high-risk HPV test result(s) (nb, based on using only Hybrid Capture 2 test results, the 37 Hybrid Capture 2–negative cases could be ascribed to 10 incident cases (1.7%), five cases (0.8%) due to non–high-risk HPV, and seven cases (1.2%) due to incorrectly classified CIN-3, leaving 15 cases (2.5%) with false-negative Hybrid Capture 2 test results.).
The patterns of some high-risk HPV-negative cases were suggestive of incident CIN-3 (nb, our a priori assumption was that cases diagnosed during follow-up that were high-risk HPV positive at baseline were missed prevalent disease. This assumption was based on the less-than-perfect sensitivity of colposcopy and the very similar number of cases of CIN-3 in the different arms of the trial despite the different sensitivity for detection of disease at baseline, which resulted in different timing of disease detection.9 Some CIN-3 likely developed from a baseline, prevalently detected high-risk HPV infection, but distinguishing between incident and missed prevalent CIN-3 is not possible.). Most compelling were 10 of 12 (83%) cases called incident in this analysis that were diagnosed at exit from ALTS—approximately 24 months after enrollment. All 12 of these cases had a second diagnosis of at least CIN-2.
There were clear examples of CIN-3 that seemed to be caused by genotypes other than those genotypes included in our definition of high-risk HPV. The CIN-3 in Participant 4 seemed to be caused by HPV 26, which has been described as possibly high-risk HPV.24 The CIN-3 in Participant 32 seemed to be caused by HPV 67, which is not considered a high-risk HPV genotype but shares the same phylogenetic grouping as HPV 16 and related high-risk HPV genotypes.25 Thus, some HPV genotypes closely related to high-risk HPV genotypes might occasionally cause CIN-3, a full epithelial thickness lesion indicative of precancer, but very rarely progressing to invasive cancer.
Some cases of high-risk HPV-negative CIN-3 were also attributable to misclassified (false-positive) histology, ie, morphologic changes that mimic CIN-3. One exemplary case was Participant 20, who tested high-risk HPV negative by both Hybrid Capture 2 and Line Blot Assay at all five visits. Yet, at the exit visit that included colposcopy, a biopsy was taken and diagnosed as CIN-3 by Quality Control pathology and CIN-2 by CC pathology. Interestingly, the colposcopic impression at that time was atypical metaplasia, which may be difficult to differentiate from CIN-3 histologically because both may display a full-thickness intraepithelial basaloid appearance.26
Finally, there were a few examples of false-negative high-risk HPV cases, ie, HPV testing errors, despite the use of dual testing. Participant 5 tested negative at baseline by Hybrid Capture 2 and Line Blot Assay but tested positive by Hybrid Capture 2 and positive for HPV 16, HPV 51, and HPV 59 by Line Blot Assay at the 12-month and exit visits (there were no test results at 6 and 18 months because of missing visits). Subsequently, HPV 51 was detected in the baseline specimen by Linear Array, consistent with persistent HPV 51 eventually leading to a CIN-3 diagnosis. Figure 1 shows the Cervigrams collected from this case at enrollment, 6-month, 12-month, and 24-month (exit) visits and shows that the lesion was distant from the cervical os, where cervical sampling occurs, at the enrollment visit, which may have resulted in poor sampling of the lesion and testing high-risk HPV negative by Line Blot Assay and Hybrid Capture 2 but positive by the more analytically sensitive Linear Array. Participant 33 (Fig. 2) might be either a false-negative high-risk HPV or incident case diagnosed at the 12-month visit, depending on whether a small, equivocal-appearing change in the epithelium at baseline that was unapparent at 6 months was a small focal lesion or not.
Eltoum and colleagues27 also evaluated cases of HPV-negative high-grade cervical precancer (≥CIN-2) in their population of women with ASC-US, determining that most lesions were small and shed few abnormal cells. Another possibility is poor exfoliation of cells due to altered adhesion molecule distribution.13 The authors concluded that these lesions were either early in development or were regressing, questioning their clinical significance. Another likely and clinically important explanation, as observed in this analysis, is that some false-negative tests are the result of poor sampling of the lesion.
So what do these cases mean clinically? First, the high-risk HPV-negative CIN-3 cases that are the result of misclassified histology do not truly represent CIN-3 and therefore do not present a risk to women. These are clinically relevant only with respect to possible overtreatment. Similarly, cases due to low-risk HPV genotypes are unlikely to develop into invasive cancer. Of the remaining high-risk HPV-negative CIN-3 cases, five were attributable to false-negative high-risk HPV and 12 to incident cases. Most women would have been diagnosed during follow-up with adherence to current guidelines1,2,28 for the management of abnormal cervical cytology.
The presence of these few false-negative high-risk HPV cases of CIN-3 emphasizes that no test, or even combinations of tests, will achieve 100% sensitivity for identifying women with cervical precancer. However, we caution against misguided attempts to achieve 100% clinical sensitivity for HPV assays by either increasing the analytic sensitivity and/or targeting additional HPV genotypes for detection, because such small gains in sensitivity will result in dramatic increases in the number of false positives (ie, high-risk HPV positive but no CIN-3).29,30 Changes in analytic sensitivity or targeting additional HPV genotypes for detection to improve clinical sensitivity require formal evaluations and validation, using receiver operating characteristic curve or other analytic approaches and thoughtful consideration of the balance between true and false positives, or more apropos to the clinical setting, negative and positive predictive values.
We conclude that in any sizeable population, even among women with evidence of cytologic abnormalities, there will be a few cases of cervical precancer that will test high-risk HPV negative for one or more of a number of possible reasons. In addition to understanding their meaning clinically, it is important to recognize that high-risk HPV-negative cases will occur in the clinical validation studies of new tests, which does not necessarily imply failure of the new test. On the whole, these high-risk HPV-negative CIN-3 cases were rare even over a 2-year period and differed substantially from the typical, HPV-positive CIN-3. About one half of the high-risk HPV-negative cases were the result of misclassification, either of enrollment high-risk HPV status or histopathologic diagnosis, and most of the rest were due to new incident HPV detection. As expected in a short window of observation time, there were a relatively small number of incident cases caused by a newly acquired infection leading to a characteristically small precancerous lesion within 24 months.
Although we took advantage of dual HPV testing to minimize testing error, we found very similar results relying only on Hybrid Capture 2 test results, naturally with a small increase in the false-negative test results. Eight of the 37 (22%) Hybrid Capture 2 negatives, or one half of the false Hybrid Capture 2 negatives, were called high-risk HPV positive by Line Blot Assay. Failure due to error by the test itself is rare, giving us confidence that a negative high-risk HPV result provides significant reassurance against clinically relevant CIN-3, and more importantly against cancer, over at least a couple of years. Of note, none of 1,796 (0%, 95% CI 0–0.2%) women who tested Hybrid Capture 2 negative at enrollment were diagnosed with cancer in 2 years, whereas all seven cases of cancer diagnosed during ALTS tested Hybrid Capture 2-positive at enrollment.31 Even when an HPV test misses a bona fide CIN-3, the lesion is most likely small or significantly peripheral to the cervical os; lacking the typical features associated with potential invasion, such as large size, gland duct involvement, and location proximal to or within the cervical canal.32 Additionally, fewer than one half of CIN-3 lesions evaluated in this study as high-risk HPV negative at enrollment remained negative in follow-up, further reducing the risk that a false-negative high-risk HPV test or an incident lesion will continue to go undetected in women managed according to clinical guidelines. Thus, these results confirm that HPV testing as a management option is significantly reassuring to be clinically useful in the follow-up of women initially referred for the evaluation of ASC-US or LSIL and not found to have CIN-2/3 at colposcopy.
1. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin. Clinical Management Guidelines for Obstetrician-Gynecologists. Number 61, April 2005. Human papillomavirus. Obstet Gynecol 2005;105:905–18.
2. Wright TC Jr, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D, et al. 2006 consensus guidelines for the management of women with abnormal cervical cancer screening tests. Am J Obstet Gynecol 2007;197:346–55.
3. Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, et al. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. CA Cancer J Clin 2002;52:342–62.
4. Cuzick J, Mayrand MH, Ronco G, Snijders P, Wardle J. Chapter 10: New dimensions in cervical cancer screening. Vaccine 2006;24 suppl 3:S90–7.
5. Wright TC Jr, Schiffman M. Adding a test for human papillomavirus DNA to cervical-cancer screening. N Engl J Med 2003;348:489–90.
6. ASCUS-LSIL Triage Study (ALTS) Group. A randomized trial on the management of low-grade squamous intraepithelial lesion cytology interpretations. Am J Obstet Gynecol 2003 Jun;188:1393–400.
7. ASCUS-LSIL Triage Study (ALTS) Group. Results of a randomized trial on the management of cytology interpretations of atypical squamous cells of undetermined significance. Am J Obstet Gynecol 2003;188:1383–92.
8. Cox JT, Schiffman M, Solomon D, ASCUS-LSIL Triage Study (ALTS) Group. Prospective follow-up suggests similar risk of subsequent cervical intraepithelial neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed biopsy. Am J Obstet Gynecol 2003;188:1406–12.
9. Guido R, Schiffman M, Solomon D, Burke L, ASCUS LSIL Triage Study (ALTS) Group. Postcolposcopy management strategies for women referred with low-grade squamous intraepithelial lesions or human papillomavirus DNA-positive atypical squamous cells of undetermined significance: a two-year prospective study. Am J Obstet Gynecol 2003;188:1401–5.
10. Kjaer S, Hogdall E, Frederiksen K, Munk C, van den Brule A, Svare E, et al. The absolute risk of cervical abnormalities in high-risk human papillomavirus-positive, cytologically normal women over a 10-year period. Cancer Res 2006;66:10630–6.
11. Safaeian M, Solomon D, Wacholder S, Schiffman M, Castle P. Risk of precancer and follow-up management strategies for women with human papillomavirus-negative atypical squamous cells of undetermined significance. Obstet Gynecol 2007;109:1325–31.
12. Lonky NM, Felix JC, Naidu YM, Wolde-Tsadik G. Triage of atypical squamous cells of undetermined significance with hybrid capture II: colposcopy and histologic human papillomavirus correlation. Obstet Gynecol 2003;101:481–9.
13. Lonky NM, Felix J, Tsadik GW, Lonky S. False-negative hybrid capture II results related to altered adhesion molecule distribution in women with atypical squamous cells pap smear results and tissue-based human papillomavirus-positive high-grade cervical intraepithelial neoplasia. J Low Genit Tract Dis 2004;8:285–91.
14. Castle PE, Schiffman M, Burk RD, Wacholder S, Hildesheim A, Herrero R, et al. Restricted cross-reactivity of hybrid capture 2 with nononcogenic human papillomavirus types. Cancer Epidemiol Biomarkers Prev 2002;11:1394–9.
15. Schiffman M, Wheeler CM, Dasgupta A, Solomon D, Castle PE, The ALTS Group. A comparison of a prototype PCR assay and hybrid capture 2 for detection of carcinogenic human papillomavirus DNA in women with equivocal or mildly abnormal papanicolaou smears. Am J Clin Pathol 2005;124:722–32.
16. Cogliano V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, et al. Carcinogenicity of human papillomaviruses. Lancet Oncol 2005;6:204.
17. Castle PE, Gravitt PE, Solomon D, Wheeler CM, Schiffman M. Comparison of linear array and line blot assay for detection of human papillomavirus and diagnosis of cervical precancer and cancer in the atypical squamous cell of undetermined significance and low-grade squamous intraepithelial lesion triage study. J Clin Microbiol 2008;46:109–17.
18. Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985;4:87–90.
19. Gravitt PE, Burk RD, Lorincz A, Herrero R, Hildesheim A, Sherman ME, et al. A comparison between real-time polymerase chain reaction and hybrid capture 2 for human papillomavirus DNA quantitation. Cancer Epidemiol Biomarkers Prev 2003;12:477–84.
20. Jeronimo J, Long R, Neve L, Ferris D, Noller K, Spitzer M, et al. Preparing digitized cervigrams for colposcopy research and education: determination of optimal resolution and compression parameters. J Low Genit Tract Dis 2006;10:39–44.
21. Jeronimo J, Massad LS, Schiffman M, National Institutes of Health/American Society for Colposcopy and Cervical Pathology (NIH/ASCCP) Research Group. Visual appearance of the uterine cervix: correlation with human papillomavirus detection and type. Am J Obstet Gynecol 2007;197:47.e1–8.
22. Herrero R, Castle PE, Schiffman M, Bratti MC, Hildesheim A, Morales J, et al. Epidemiologic profile of type-specific human papillomavirus infection and cervical neoplasia in Guanacaste, Costa Rica. J Infect Dis 2005;191:1796–807.
23. Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R, et al. Human papillomavirus type distribution in invasive cervical cancer and high-grade cervical lesions: a meta-analysis update. Int J Cancer 2007;121:621–32.
24. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518–27.
25. de Villiers EM, Fauquet C, Broker TR, Bernard HU, zur Hausen H. Classification of papillomaviruses. Virology 2004;324:17–27.
26. Regauer S, Reich O. CK17 and p16 expression patterns distinguish (atypical) immature squamous metaplasia from high-grade cervical intraepithelial neoplasia (CIN III). Histopathology 2007;50:629–35.
27. Eltoum IA, Chhieng DC, Crowe DR, Roberson J, Jin G, Broker TR. Significance and possible causes of false-negative results of reflex human Papillomavirus infection testing. Cancer 2007;111:154–9.
28. American College of Obstetricians and Gynecologists. ACOG Practice Bulletin number 66, September 2005. Management of abnormal cervical cytology and histology. Obstet Gynecol 2005;106:645–64.
29. Stoler MH, Castle PE, Solomon D, Schiffman M, American Society for Colposcopy and Cervical Pathology. The expanded use of HPV testing in gynecologic practice per ASCCP-guided management requires the use of well-validated assays. Am J Clin Pathol 2007;127:335–7.
30. Schiffman M, Khan MJ, Solomon D, Herrero R, Wacholder S, Hildesheim A, et al. A study of the impact of adding HPV types to cervical cancer screening and triage tests. J Natl Cancer Inst 2005;97:147–50.
31. Atkins KA, Jeronimo J, Stoler MH, ALTS Group. Description of patients with squamous cell carcinoma in the atypical squamous cells of undetermined significance/low-grade squamous intraepithelial lesion triage study. Cancer 2006;108:212–21.
32. Sherman ME, Wang SS, Tarone R, Rich L, Schiffman M. Histopathologic extent of cervical intraepithelial neoplasia 3 lesions in the atypical squamous cells of undetermined significance low-grade squamous intraepithelial lesion triage study: implications for subject safety and lead-time bias. Cancer Epidemiol Biomarkers Prev 2003;12:372–9.
Supplemental Digital Content
© 2008 by The American College of Obstetricians and Gynecologists.