Castle, Philip E. PhD, MPH; Gage, Julia C. PhD, MPH; Wheeler, Cosette M. PhD; Schiffman, Mark MD
Cervical intraepithelial neoplasia (CIN) grade 1 is the most common histologic biopsy diagnosis after referral for colposcopy for a positive cervical cancer screening test. Although CIN 1 is recognized as primarily the histologic manifestation of a human papillomavirus (HPV) infection, CIN 1 often is (incorrectly) grouped with more severe grades, CIN 2 and CIN 3, as “CIN” or cervical neoplasia, implying precancer. In part, this is the result of the diagnostic challenge of distinguishing CIN 1 from CIN 2 and from negative histology reproducibly.1,2
Nonetheless, the clinical implications of CIN 1 are not well understood because there are few contemporary prospective studies to explore the subsequent risk of cervical precancer. An earlier review suggested that 10% of women with CIN 1 develop CIN 3 in the next 10 years.3 In the atypical squamous cells of undetermined significance (ASCUS) and low-grade squamous intraepithelial lesions (LSIL) triage study (ALTS), the 2-year risk of CIN 3 was 9%.4 Other studies have found lower risks. For example, a randomized trial for the management of CIN 1 found that 4.4% of women with CIN 1 were diagnosed with CIN 2–3 in 18 months.5 An analysis of 1,001 CIN 1 found that 7% were diagnosed with CIN 2–3 at the 6-month follow-up.6 The widely variable risks between studies may be the result of many factors, including diagnostic errors, differences in diagnostic thresholds for each CIN category, population differences, the follow-up protocol, and the time of follow-up.
Now that HPV infection is known to be the causal agent underlying cervical carcinogenesis, we have become interested in the meaning of histologic CIN 1 diagnoses compared with not finding CIN 1 (ie, negative biopsies or no biopsies taken as a result of normal colposcopic impressions). Taking HPV infection status into account, does finding CIN 1 imply greater risk of CIN 3 than not finding it? We therefore conducted a retrospective study of primarily younger (younger than 30 years) women diagnosed with CIN 1 at enrollment in ALTS as a follow-up of a previous study.4
MATERIALS AND METHODS
The ALTS (1997–2001) was a multisite, randomized clinical trial comparing three management strategies (immediate colposcopy, HPV triage, or conservative management) for women referred for ASCUS (n=3,488) or LSIL (n=1,572) conventional cytology.7–11 “Atypical squamous cells of undetermined significance” under the 1991 Bethesda system12 was slightly more inclusive, particularly of probable reactive changes and atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H), than the “atypical squamous cells of undetermined significance” category of the 2001 Bethesda system.13 The National Cancer Institute and local institutional review boards approved the study and all participants provided written, informed consent.
At enrollment and follow-up visits over the 2-year duration, all women underwent a pelvic examination with collection of two cervical specimens. The first specimen was placed in PreservCyt for ThinPrep cytology and clinical HPV testing by Hybrid Capture 2 (HC2; Qiagen) and the second in specimen transport medium. Women in all three arms of the study were re-evaluated by cytology every 6 months during the 2 years and sent to colposcopy if cytology was high-grade squamous intraepithelial lesions (HSIL). An exit examination with colposcopy was scheduled for all women. We refer readers to other references for details on randomization, examination procedures, patient management, and laboratory and pathology methods.7–11
We used diagnoses by clinical center pathologists to establish the baseline histologic status (CIN 1, negative histology, or no biopsy). We restricted our analysis to women randomized to the immediate colposcopy or HPV triage arms and either referred to colposcopy for an ASCUS Pap test result and testing HC2-positive (HPV-positive ASCUS) or for an LSIL Pap test result to better reflect the current practice of referral to colposcopy.14 We note that in the HPV triage arm, a small proportion of women referred with a LSIL Pap test result were not included in this analysis because they tested HPV negative by HC2 (33 of 224 [14.7%] in the HPV triage arm) at enrollment. In contrast, current guidelines recommend that all women with LSIL, regardless of HPV status, undergo colposcopy.14 We excluded CIN 1 in the conservative management study arm because women in this arm were only referred to colposcopy if they had enrollment HSIL cytology.
There were 738 women with enrollment biopsies diagnosed as CIN 1 (including 16 women with missing data on whether a biopsy was taken and 10 who had a baseline diagnosis of CIN 1 but no biopsy data recorded), 346 diagnosed as negative by the clinical center pathologists, and 346 who went to colposcopy but had histologic outcome (including 17 women with missing data on whether a biopsy was taken). This analysis was further restricted to those women who were exited from the study, ie, those who had an exit visit at 24 months or were treated and censored from the study: 594 (80.5%) with CIN 1, 289 (83.5%) with negative histology, and 281 (81.2%) with no biopsy. There was no significant difference in the percent of women who exited from the study (P=.5, Fisher's exact test) by biopsy outcome. Over the 2-year duration of ALTS, the quality control pathology group diagnosed 100 women with CIN 3, which was the primary end point for this analysis. We also used an endpoint of CIN 2–3 (n=138) as diagnosed by the clinical center pathologists because it is relevant to clinical practice because these women would be treated according to current management guidelines.15
HC2, a DNA test that targets (with some crossreactivity for other types) a pool of 13 carcinogenic HPV genotypes (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68), was performed on residual PreservCyt specimens as previously described.16 STM specimens were tested for 27 or 38 HPV genotypes using a polymerase chain reaction assay, line blot assay as previously described.14,17 We used the HPV16 and HPV18 results from line blot assay among the HC2-positives to mimic currently available tests that offer HPV16 and HPV18 genotyping and classify women hierarchically according to established cancer risk (HPV risk group or status)18: HPV16-positive, else HPV16-negative and positive for HPV18, else negative for HPV16 and HPV18 but positive for the other carcinogenic HPV genotypes as measured by HC2, or negative for all carcinogenic HPV genotypes (HPV16>HPV18>other carcinogenic HPV>carcinogenic HPV-negative). Four women with no biopsy, six women with negative biopsy, and 20 women with CIN 1 were missing HPV results.
The enrollment characteristics for women with CIN 1, negative histology, or no biopsy were compared and tested for statistical significance using Pearson's χ2 tests. We calculated the 2-year risk of quality control pathology CIN 3 and clinical center pathology CIN 2–3, overall and stratified by baseline status (CIN 1, negative histology, or no biopsy), relevant screening results (eg, referral Pap and HC2 results), and HPV risk group. Fisher's exact or a Pearson's χ2 tests as appropriate were used to test for statistical differences in risk. Binomial exact 95% confidence intervals (CIs) were calculated for CIN 3 risks where noted.
Finally, we used logistic regression19 to calculate odds ratios (ORs) and 95% CI to model the association of risk factors with quality control pathology CIN 3 and clinical center pathology CIN 2–3. P values of <.05 were considered statistically significant. STATA 11.0 was used for analyses.
There were statistically significant but slight differences in the age at enrollment for women with CIN 1, negative histology, or no biopsy (median ages, 23 years, 24 years, and 23 years, respectively; P=.04). A comparison of the study and behavior characteristics among these three populations is shown in Table 1. There were significant differences among the three groups for some study characteristics: the enrolling clinical center and the referral Pap test. Women with CIN 1 were more likely to be classified as having riskier HPV genotypes (P=.02) and were more likely to have a smoking history (P=.02) than the other groups. There was no significant difference between groups by race, ever being pregnant, or oral contraceptive use (data not shown). Comparing negative histology with women without a biopsy, the only differences between the two groups were the enrolling clinical center (P<.001). There were significant differences (P<.001) in the distribution of colposcopic impressions among the three groups with 86.5% of those of CIN 1 and 90.0% of those with negative histology having a low-grade colposcopic impression or more severe, whereas 81.3% of women with no biopsy had a completely normal colposcopic impression (data not shown).
The 2-year risks of CIN 3 were 10.3% (95% CI 7.9–13.0) for women with CIN 1, 7.3% (95% CI 4.6–10.9) for negative histology, and 6.4% (95% CI 3.8–9.9) for women referred to colposcopy and no biopsies were taken (P=.1; Table 2). There were no obvious patterns or trends in the small differences in the CIN 3 risk between women referred for an ASCUS Pap test result and tested HPV-positive compared with a LSIL Pap test result. Most women referred for a LSIL Pap test result and being diagnosed with CIN 1, negative histology, or no biopsy taken tested HC2-positive (90.2%, 85.6%, or 80.9%, respectively).
We also present the risks after referral ASCUS and LSIL Pap test results for women randomized to the immediate colposcopy arm as an unbiased reference. We noted that there were no appreciable differences in the risks after a LSIL Pap test result for the immediate colposcopy and HPV arms compared with immediate colposcopy alone despite the per-protocol exclusion of the women with HPV-negative LSIL from undergoing colposcopy in the HPV study arm. Because there was no statistical difference in the overall 2-year risk of CIN 3 (P=.7) and CIN 2–3 (P=.7) between women with negative histology and no biopsy, for greater power, we combined the two groups in the subsequent analyses into a single group (“no CIN 1”) (n=570) with a 2-year risk of CIN 3 of 6.8% (95% CI 4.9–9.2).
In the initial analysis that did not take HPV status into account, women with a CIN 1 diagnosis were nonsignificantly more likely to have an enrollment high-grade cytology (HSIL or ASC-H) than women with no CIN 1, 8.0% compared with 5.8% (P=.2). Among women with CIN 1, those having an enrollment high-grade cytology were at twice the risk of CIN 3 compared with women without an enrollment high-grade cytology (19.2% compared with 9.6%, respectively; P=.05). Among women with no CIN 1, those having an enrollment high-grade cytology were also at approximately twice the risk of CIN 3 compared with women without an enrollment high-grade cytology (12.1% compared with 6.6%, respectively; P=.2), albeit at a lower absolute level.
Before HPV status was considered, the risk of CIN 3 (compared with no CIN 3) was greater (P=.04) for those with CIN 1 than those from those with no CIN 1 at enrollment. Stratified by HPV risk group status (Table 3), there was a gradient of risk for CIN 3 according to the HPV risk group (Ptrend<.00120): 19.1% (95% CI 13.1–25.3) for HPV16-positives, 13.9% (95% CI 7.2–23.5) for HPV18-positives, 5.7% (95% CI 4.2–7.6) for HC2-positives (HPV16 or HPV18-negative), and 5.2% (95% CI 1.1–14.4) for HC2-negatives. There was no statistically significant difference in the 2-year risk of CIN 3 by baseline histology status stratified by HPV risk group and relevant screening result (HC2-positive ASCUS or LSIL), although no CIN 1 tended to have slightly lower risks than CIN 1. Similar patterns were observed for the clinical center pathology end point of CIN 2–3.
Finally, we examined the relationship of enrollment characteristics, notably including HPV status, and the 2-year risk of CIN 3 using a logistic regression model that considered CIN 3 and CIN 2 as distinct end points (Table 4). Being HPV16-positive (OR 3.2, 95% CI 1.6–6.7) (compared with HC2-positive and HPV16 or HPV18-negative) and having an enrollment high-grade cytology (HSIL or ASC-H) (OR 2.4, 95% CI 1.2–4.9) (compared with less severe high-grade cytology) were independently associated with an elevated CIN 3 risk. Being HPV18-positive (OR 1.2, 95% CI 0.26–5.5) was not independently associated with CIN 3. Enrollment CIN 1 (compared with no CIN 1) (OR 0.99, 95% CI 0.54–18) was not associated with CIN 3. Similar associations were observed for the clinical center pathology end point of CIN 2–3 (data not shown).
Other enrollment variables, including the number of biopsies taken at enrollment and number of HPV genotypes, were not associated with CIN 3 or CIN 2 (data not shown). When added to the model, nonnormal colposcopic impression did not significantly predict CIN 3 or CIN 2 compared with a normal colposcopic impression (data not shown).
Excluding whose women with baseline histology was upgraded to CIN 3 by quality control pathology (18 cases from CIN 1 and four cases from no CIN 1), perhaps as a result of misclassification, the 2-year risks of CIN 3 were 7.5% and 6.2% (P=.2) after CIN 1 and no CIN 1 diagnosis, respectively. The results of the logistic regression did not appreciably change when we excluded the aforementioned cases (data not shown).
We examined whether having a diagnosis of CIN 1 represented an independent risk factor above HPV infection alone for the development of incident CIN 3. After controlling for HPV risk group and the presence of enrollment high-grade cytology, we found no evidence that having CIN 1 significantly elevated the 2-year risk of incident CIN 3 compared with negative histology or colposcopically normal women. Thus, differences in 2-year risk of CIN 3 between CIN 1 and negative histology or no biopsy were primarily the result of differences in the distribution in HPV genotypes and the fraction of women with enrollment CIN 3 missed by colposcopically directed biopsy but indicated by concurrent (enrollment) high-grade cytology. Thus, etiologically, HPV genotype is the primary cause of CIN 3 rather than the distinction between CIN 1 and no CIN 1.
From a clinical perspective, women with CIN 1 had a 10% 2-year risk of CIN 3 compared with 7% for negative histology and 6% for no biopsy, statistically significant but small absolute risk differences. Given the general similarities in risk, these outcomes of screening might be managed similarly.21 That is, given that there are a limited number of management options (routine screening, increased surveillance, or follow-up colposcopy) and the general similarities in risk, it might difficult to justify differential management by histologic status in this population, depending on the accepted thresholds of risk for each management option. We note that the risks of CIN 3 reported here were similar to those reported in a previous analysis on CIN 1 from ALTS.4
Having HPV16-positive or HPV18-positive CIN 1 or no CIN 1 elevated the 2-year risk threefold and twofold, respectively, compared with testing positive by HC2 for other carcinogenic HPV genotypes. Similar patterns were observed when using quality control pathology diagnosis of CIN 2–3 as the end point: women positive for HPV16 (19.9%) and HPV18 (16.1%) were at an elevated risk compared with those positive for other, non-HPV16 or HPV18 carcinogenic HPV genotypes (10.3%) (P=.002).
All 13 cases of CIN 3 in the HC2-negative group tested negative for any carcinogenic HPV genotype and 10 of 13 tested negative for all HPV genotypes by polymerase chain reaction. These data indicate that some of these CIN 3 diagnosed after a HC2-negative result developed from newly acquired HPV infections after enrollment or were misclassified by the quality control pathology group; fewer quality control pathology-diagnosed CIN 3 was confirmed by the clinical center pathologists as CIN 3 after a HC2-negative compared with a HC2-positive result (17.5% compared with 51.5%, respectively; P=.04).
We note several additional limitations in our analysis: 1) we had few cases of CIN 3 diagnosed, limiting our ability to provide more precise measures of CIN 3 risk; 2) our population was relatively young and therefore is not representative of older women who are diagnosed with CIN 1, negative, or no histology. We did not find that age was associated with risk of CIN 3 (data not shown) but we had only approximately 20% who were 30 years and older and 5% who were 40 and older. Therefore, these data do not inform about the appropriate management of women with CIN 1; 3) we did not have longer-term follow-up of these women to determine the outcomes of all infections and associated histopathology. Women who are HPV-positive remain at risk for CIN 3 or cancer for many years, at least until there is molecular evidence of clearance22; 4) although all women exiting ALTS received colposcopy, almost certainly some CIN 3 was missed as a result of limitations in sensitivity of colposcopy to find disease, especially when only a single colposcopically directed biopsy that was the common practice at the time of ALTS23,24; 5) women with a high-grade Pap test result were not referred into ALTS and therefore we cannot generalize our findings to all women with these colposcopy and biopsy results. It seems likely that women referred for a high-grade Pap test result would be a greater risk of CIN 3 than women with less severe cytologic abnormalities as a result of their a priori risk. Indeed, we observed that the women who had an enrollment HSIL cytology were more likely to develop CIN 3; and 6) a small fraction of women were referred into ALTS and had LSIL were not included in this analysis because they were randomized to the HPV arm and they tested HPV-negative. We would expect women with HPV-negative LSIL and a histologic diagnosis of CIN 1 or less to have a lower risk than the HPV-positive LSIL counterpart. Thus, we may have slightly overestimated the CIN 3 risk after a LSIL Pap test result and CIN 1 or less histology compared with women who would be referred to colposcopy with LSIL in current clinical practice, although we did not observe a qualitative difference in risk for the women with a LSIL Pap test result in the immediate colposcopy arm, which includes all HPV-negative and HPV-positive women, compared with those in the HPV.
The strength of the study is the use of consensus CIN 3 diagnosis as the primary end point rather than CIN 2–3. Previous studies have shown that CIN 2 is an equivocal diagnosis of precancer2 and more important in this context that CIN 2 has very poor interrater agreement1,2 and is much more likely to regress25,26 than CIN 3. Therefore, the risks of CIN 2–3 are more apt to vary depending on who is making the diagnosis and the follow-up protocol, respectively, making comparisons of risks between studies that included CIN 2 as an end point challenging to interpret. We did include the clinical center pathology diagnosis of CIN 2–3 to show that the general patterns of risk and determinants were nevertheless the same as for quality control pathology CIN 3.
In conclusion, we found nothing intrinsic in the diagnosis of CIN 1 that significantly elevated the 2-year risk of CIN 3 positivity above the molecular cause of the CIN 1, HPV infection. Women who were diagnosed with CIN 1 were more likely to have HPV16, which in turn increased the risk of CIN 3, than those who did not have CIN 1. This is perhaps expected because HPV16-positive cervical lesions are more apparent to colposcopists than lesions causes by other HPV genotypes,27 naturally leading to a bias as to where biopsies were taken and therefore identifying women with HPV16-positive CIN 1. Although the risk differences between HPV16-positive and HPV16-negative (or HPV18-positive or HPV18 negative) CIN 1 or no CIN 1 may not be sufficiently different from offer differential clinical management of patients, when valid testing of HPV16 results are presented in this context, doctors should be cognizant that HPV16-positive (and HPV18-positive) women are at an elevated risk of CIN 3 in the subsequent few years. Finally, categorizing histologic diagnoses as “CIN” is a gross misclassification of disease because CIN 1, CIN 2, and CIN 3 represent different biological entities and their associated cancer risks differ greatly.
1. Carreon JD, Sherman ME, Guillen D, Solomon D, Herrero R, Jeronimo J, et al.. CIN2 is a much less reproducible and less valid diagnosis than CIN3: results from a histological review of population-based cervical samples. Int J Gynecol Pathol 2007;26:441–6.
2. Castle PE, Stoler MH, Solomon D, Schiffman M. The relationship of community biopsy-diagnosed cervical intraepithelial neoplasia grade 2 to the quality control pathology-reviewed diagnoses: an ALTS report. Am J Clin Pathol 2007;127:805–15.
3. Ostor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 1993;12:186–92.
4. 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.
5. Elit L, Levine MN, Julian JA, Sellors JW, Lytwyn A, Chong S, et al.. Expectant management versus immediate treatment for low-grade cervical intraepithelial neoplasia : a randomized trial in Canada and Brazil. Cancer 2010;117:1438–45.
6. Bansal N, Wright JD, Cohen CJ, Herzog TJ. Natural history of established low grade cervical intraepithelial (CIN 1) lesions. Anticancer Res 2008;28:1763–6.
7. Schiffman M, Adrianza ME. ASCUS-LSIL Triage Study. Design, methods and characteristics of trial participants. Acta Cytol 2000;44:726–42.
8. Human papillomavirus testing for triage of women with cytologic evidence of low-grade squamous intraepithelial lesions: baseline data from a randomized trial. The Atypical Squamous Cells of Undetermined Significance/Low-Grade Squamous Intraepithelial Lesions Triage Study (ALTS) Group. J Natl Cancer Inst 2000;92:397–402.
9. 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;188:1393–400.
10. 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.
11. Solomon D, Schiffman M, Tarone R; ALTS Study group. Comparison of three management strategies for patients with atypical squamous cells of undetermined significance: baseline results from a randomized trial. J Natl Cancer Inst 2001;93:293–9.
12. Kurman RJ, Malkasian GD Jr, Sedlis A, Solomon D. From Papanicolaou to Bethesda: the rationale for a new cervical cytologic classification. Obstet Gynecol 1991;77:779–82.
13. Solomon D, Davey D, Kurman R, Moriarty A, O'Connor D, Prey M, et al.. The 2001 Bethesda System: terminology for reporting results of cervical cytology. JAMA 2002;287:2114–9.
14. Wright TC Jr, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D. 2006 consensus guidelines for the management of women with abnormal cervical screening tests [published erratum appears in J Low Genit Tract Dis 2008;12:255]. J Low Genit Tract Dis 2007;11:201–22.
15. Wright TC Jr, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D. 2006 consensus guidelines for the management of women with cervical intraepithelial neoplasia or adenocarcinoma in situ [published erratum appears in J Low Genit Tract Dis 2008;12:63]. J Low Genit Tract Dis 2007;11:223–39.
16. Castle PE, Wheeler CM, Solomon D, Schiffman M, Peyton CL. Interlaboratory reliability of Hybrid Capture 2. Am J Clin Pathol 2004;122:238–45.
17. Schiffman M, Wheeler CM, Dasgupta A, Solomon D, Castle PE; 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.
18. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, et al.. A review of human carcinogens—Part B: biological agents. Lancet Oncol 2009;10:321–2.
19. Long JS. Nominal outcomes: multinomial logit and related models. Regression models for categorical and limited dependent variables. Thousand Oaks (CA): Sage Publications; 1997. p. 148–86.
20. Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985;4:87–90.
21. Castle PE, Sideri M, Jeronimo J, Solomon D, Schiffman M. Risk assessment to guide the prevention of cervical cancer. J Low Genit Tract Dis 2008;12:1–7.
22. Kjaer SK, Frederiksen K, Munk C, Iftner T. Long-term absolute risk of cervical intraepithelial neoplasia grade 3 or worse following human papillomavirus infection: role of persistence. J Natl Cancer Inst 2010;102:1478–88.
23. Gage JC, Hanson VW, Abbey K, Dippery S, Gardner S, Kubota J, et al.. Number of cervical biopsies and sensitivity of colposcopy. Obstet Gynecol 2006;108:264–72.
24. Jeronimo J, Schiffman M. Colposcopy at a crossroads. Am J Obstet Gynecol 2006;195:349–53.
25. Trimble CL, Piantadosi S, Gravitt P, Ronnett B, Pizer E, Elko A, et al.. Spontaneous regression of high-grade cervical dysplasia: effects of human papillomavirus type and HLA phenotype. Clin Cancer Res 2005;11:4717–23.
26. Castle PE, Schiffman M, Wheeler CM, Solomon D. Evidence for frequent regression of cervical intraepithelial neoplasia-grade 2. Obstet Gynecol 2009;113:18–25.
27. 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.
© 2011 by The American College of Obstetricians and Gynecologists.