Clinical Outcomes after Conservative Management of Cervical Intraepithelial Neoplasia Grade 2 (CIN2) in Women Ages 21–39 Years

Cervical cancer incidence rates have declined dramatically where screening programs have been successfully implemented, as precancerous high-grade cervical lesions are detected before the onset of cervical cancer and treated (1–3). There is evidence that about one third of large lesions graded cervical intraepithelial neoplasia grade 3 (CIN3), among older women, will eventually invade (4, 5). On the other hand, lesions graded CIN grade 2 (CIN2), a less reproducible diagnosis, are more biologically heterogeneous, ranging from benign human papillomavirus (HPV) infection to evolving CIN3 (6–8); CIN2 is more likely to regress than CIN3 (8, 9). Although exact figures are lacking, several hundred thousand cases of CIN2 are diagnosed in the United States per year.

Accordingly, the American Society for Colposcopy and Cervical Pathology (ASCCP) guidelines specify that observation is preferred for young women with a diagnosis of CIN2, acceptable for young women with CIN2/3, and not recommended for young women with unambiguous CIN3 (3, 10). Such conservative management (intensified surveillance with cytology and colposcopy instead of immediate treatment) is sometimes preferred among women still desiring childbearing because of the increased risk of future obstetric complications associated with excisional treatment (11, 12).

Only small studies have evaluated the clinical outcomes for women diagnosed with CIN2 or CIN2/3 and followed with conservative management (13–19), none in routine practice, and few studies included women over age 21 years (16, 17, 19) or more than 50 women (17). We determined clinical outcomes among 2,417 initially untreated women between the ages of 21 and 39 years in the Kaiser Permanente Northern California (KPNC) Cohort Study of over 1.5 million women undergoing cervical cancer screening from 2003 to 2015.

We conducted a retrospective cohort analysis nested within the larger KPNC study of women receiving cervical cancer screening with HPV and cytology cotesting between 2003 and 2015 (20). Eligibility criteria are summarized in Fig. 1. We initially identified 24,515 women with a first biopsy diagnosis of equivocal high-grade lesion or greater [CIN1/2, CIN2, CIN2/3, or CIN3 or more severe diagnosis (CIN3+)]. We restricted this population to younger women, defined here as ages 21 to 39 years, as they are the target for conservative management guidelines due to the potential for obstetric complications of treatment, leading to the exclusion of 7,871 women aged <21 or >39 years. We further restricted our analysis to lesions that would be considered eligible for conservative management (CIN1/2, CIN2, and CIN2/3), excluding 6,015 women with CIN3+. We then excluded women treated within 4 months of their biopsy diagnosis (N = 376) or >4 months after their biopsy diagnosis but without an intervening biopsy, HPV, or cytology test (N = 6,660), as these women received immediate treatment rather than conservative management. There were 3 cancers diagnosed among these excluded women. Finally, women were required to have at least 22 months of follow-up after biopsy diagnosis [the practical minimum time that would permit return to routine screening (i.e., at least 4 months until first follow-up cytology/colposcopy, then 6 months to the second, followed by a cotest 12 months later)], resulting in another 1,176 women being excluded (there were 89 high-grade lesions and no cancers among the 1,176 women excluded for insufficient follow-up time; the rest persisted with low-grade lesions). This resulted in a final cohort of 2,417 women conservatively managed with intensified surveillance (cytology and colposcopy) instead of immediate treatment for equivocal high-grade lesions (CIN1/2, CIN2, or CIN2/3).

Women were categorized into one of five mutually exclusive hierarchical groups according to their clinical outcomes as described in Table 1. Women were categorized as developing cancer if any future biopsy or treatment histology resulted in a cancer diagnosis. If a woman did not have cancer but underwent treatment including loop electrosurgical excision procedure, cold knife cone (cone biopsy), cryotherapy, and/or hysterectomy, they were considered treated. A woman was classified as “exited from colposcopy surveillance” and returned to routine KPNC screening with a cotest every 3 years if she met the KPNC criteria of two or more sequential negative cytologies and colposcopies at least 6 months apart, followed by a negative cotest 12 months later, and did not have cancer, a high-grade lesion, or receive treatment. Untreated women with at least one subsequent biopsy finding of CIN2, CIN2/3, CIN3, or adenocarcinoma in situ (AIS) that did not regress were categorized as having a “persisting high-grade lesion.” Finally, a woman was categorized as having a “low-grade lesion” if she did not fall into the other 4 categories, but had a CIN1 or low-grade squamous intraepithelial lesion result from either biopsy or cytology or tested HPV positive. HPV and cytology cotesting were done as described previously (20). Women were referred to colposcopy following local KPNC guidelines, which were largely in accordance with national standards (3).

The KPNC Institutional Review Board (IRB) approved use of the data. The Albert Einstein College of Medicine IRB and National Institutes of Health Office of Human Subjects Research deemed this study exempt from IRB review because there was no protected health information exchanged. The study was conducted in accordance with recognized ethical guidelines (e.g., Declaration of Helsinki, CIOMS, Belmont Report, U.S. Common Rule).

Of the 2,417 women included in this cohort, 428 (17.7%), 1,670 (69.1%), and 319 (13.2%) had an index diagnosis of CIN1/2, CIN2, and CIN2/3, respectively. We included all 3 diagnoses as CIN2. Overall, women had a median follow-up of 48 months [interquartile range (IQR), 31–71 months) from their initial CIN2 histology result. Table 2 shows the reason for the initial colposcopy referral, available for 2,189 of the 2,417 women (90.6%), leading to the index diagnosis. Most were referred (72.3%) due to HPV-positive ASC-US or LSIL cytology, whereas approximately 25% were referred because of a high-grade cytology result (ASC-H, HSIL, AIS, AGC), and <3% were referred for repeat HPV positivity. Women referred to colposcopy for high-grade cytology were most likely to receive treatment during their follow-up, regardless of their baseline histology. Conversely, fewer women referred for ASCUS/HPV⁺ or LSIL results had treatment, and approximately 45% of these women referred for less than high-grade cytology continued to have low-grade lesions diagnosed during follow-up.

We identified 6 (0.2%; 95% confidence interval, 0.1–0.5) cancer cases, all of which were among women with persistent high-grade lesions and/or HR HPV. Their clinical histories are described in Supplementary Table S1. Furthermore, half of these cases also occurred after a 2- to 3-year gap in follow-up, deviating from the recommended management protocols. All cases had a recent history of a high-grade cytology prior to the cancer diagnosis. Importantly, none of the cases occurred after a negative cotest in follow-up of the index CIN2 diagnosis.

Through the duration of follow-up, almost a third of women eventually received some form of treatment (Table 2), with a median time to treatment of 12 months (IQR, 7–20 months). At the time of the treatment, almost 30% had a worst biopsy of <CIN2/HSIL histology, 41% had CIN2/HSIL histology, and 31% had CIN3+ (Table 3). This pattern was similar across all baseline histology diagnoses, although women with baseline CIN1/2 had a slightly higher proportion of CIN2/HSIL histologies and slightly lower proportion of CIN3 during follow-up compared with those with higher grade baseline lesions.

Only 20% of women met the criteria to exit colposcopic follow-up and return to routine screening. Thus, half of the population remained in the surveillance/colposcopy protocol due to persistent “abnormal” results. Of these, only 7% of women had high-grade lesions identified on colposcopy in follow-up; the other 1,048 women (43%) had low-grade histology or HC2 positivity that neither progressed nor cleared during the period of observation, preventing them from returning to routine screening. With each increasing age group, we saw an increase in the proportion of women who were either treated or exited colposcopy and a decrease in those with both persistent high- and low-grade lesions (Supplementary Table S2). When further stratified by initial histology result, similar patterns by age group remained for those initially called CIN2 or CIN2/3 versus CIN1/2.

CIN2 is common, its natural history uncertain, and some women chose not to be treated immediately. Over an average of 4 years of follow-up of 2,417 women with an index CIN2 diagnosis, only 6 cancers were diagnosed. In comparison, approximately 30% were eventually treated and 20% met the criteria to exit surveillance and return to routine screening at 3-year intervals. Forty-three percent of women with an initially untreated CIN2 never received treatment during our follow-up period but remained by protocol in colposcopy clinic follow-up for 2 or more years in the absence of persisting CIN2 or more severe diagnoses (CIN2+). This may be due to persistence without disease or the acquisition of new HPV infections causing low-grade abnormal results that are common in younger sexually active women and unrelated to the initial CIN2. Longer follow-up is required to determine whether these women would eventually progress, return to normal, or continue to persist with low-grade abnormalities.

It is important to note that this is a retrospective analysis based on medical record data. As a result, we lack information on factors related to the decision for conservative management versus immediate treatment, such as whether this was a decision guided by other risk factors, provider recommendation, patient preference, or other unknowns. The initial CIN2 lesion was also not tested for p16 by IHC, so we are unable to assess whether that triage marker would have assisted in predicting the outcome.

Previous case series containing 40 to 385 women (mostly age <25 years) with CIN2 found regression rates ranging from 39% to 88% (13, 17, 18, 21) with wide variability in study design and definition of outcomes like “regression” and “persistence.” In our slightly older population, we found less evidence of complete return to normalcy, instead finding a large proportion of women with persistent but nonprogressing low-grade lesions. This failure to “clear” or return to routine screening in the absence of progression resulted in a long-term cycle of continued frequent follow-up visits, many including repeat colposcopies and/or biopsy procedures. This cycle comes at a considerable cost to the patient in terms of time, finances, and emotional distress or anxiety.

These findings warrant a reconsideration of the current guidelines requiring a return to complete normalcy to exit increased surveillance in colposcopy clinic and return to routine screening. Further research is needed to determine whether there is a way to identify which women with an initial CIN2 who are under increased surveillance can be safely returned to routine screening after fewer follow-ups without inordinately increasing their cancer risk. Perhaps HPV genotyping, a negative cotest, and/or lack of high-grade cytology could serve as further means to distinguish which women are at higher cancer risk and need continued follow-up and which are lower cancer risk and could be released safely from such intensive surveillance back to routine screening. Circumstances in which younger women with an untreated, transient CIN2 diagnosis but persisting low-grade abnormalities can return safely to less intensive screening deserve further study.

No potential conflicts of interest were disclosed.

Conception and design: M.I. Silver, J.C. Gage, M. Schiffman, T. Lorey, W.K. Kinney, P.E. Castle

Development of methodology: J.C. Gage, M. Schiffman, L.C. Cheung

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Schiffman, B. Fetterman, T. Lorey, A. Locke, W.K. Kinney, P.E. Castle

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.I. Silver, J.C. Gage, M. Schiffman, N.E. Poitras, L.C. Cheung, P.E. Castle

Writing, review, and/or revision of the manuscript: M.I. Silver, J.C. Gage, M. Schiffman, T. Lorey, L.C. Cheung, H.A. Katki, W.K. Kinney, P.E. Castle

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B. Fetterman, N.E. Poitras, L.C. Cheung, A. Locke

Study supervision: M. Schiffman, P.E. Castle

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.