Women missing follow-up data and those with baseline positive co-test results and/or other positive follow-up screening results without a colposcopy visit were considered indeterminate for the presence or absence of disease (n = 121). DS indicates p16/Ki-67 dual-stain testing; CIN2, cervical intraepithelial neoplasia grade of 2; CIN3, CIN grade of 3; Pap, Papanicolaou cytology testing.
The 5-year cumulative risk curves were generated using the logistic Weibull model. The dashed line corresponds to the threshold for a 1-year return, and the dotted line corresponds to the threshold for immediate colposcopy referral in this study population. Plus sign indicates positive; minus sign, negative; ASC-US atypical squamous cells of undetermined significance; CIN2, cervical intraepithelial neoplasia grade of 2; CIN3, CIN grade of 3; NILM, negative for intraepithelial lesion or malignancy.
The 5-year cumulative risk curves were generated using the logistic Weibull model. The dashed line corresponds to the threshold for a 1-year return, and the dotted line corresponds to the threshold for immediate colposcopy referral in this study population. Plus sign indicates positive; minus sign, negative; ASC-US, atypical squamous cells of undetermined significance; CIN2, cervical intraepithelial neoplasia grade of 2; CIN3, CIN grade of 3; NILM, negative for intraepithelial lesion or malignancy.
eMethods. Outcome definitions and prevalence-incidence mixture models
eFigure 1. Example screening data from electronic health records (EHRs) for 9 women
eTable 1. p16/Ki-67 dual stain positivity by cytology and worst histology results
eTable2. Cumulative Risk of CIN2+ and CIN3+ by Enrollment p16/Ki-67 Dual Stain and Cytology Results
eTable 3. Cumulative Risk of CIN2+ and CIN3+ by Number of p16/Ki-67 Dual Stain-Positive Cells and Cytology Categories
eFigure 2. Cumulative detection and risk of CIN2+ and CIN3+ in women with NILM cytology at enrollment.
eFigure 3. Cumulative detection and risk of CIN2+ and CIN3+ in women with ASC-US+ cytology at enrollment.
eFigure 4. Cumulative detection and risk of CIN2+ and CIN3+ in women p16/Ki-67 -negative at enrollment.
eFigure 5. Cumulative detection and risk of CIN2+ and CIN3+ in women p16/Ki-67-positive at enrollment.
eFigure 6. Cumulative risk of CIN2+ and CIN3+ by number of dual stain-positive cells and Pap cytology categories.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Clarke MA, Cheung LC, Castle PE, et al. Five-Year Risk of Cervical Precancer Following p16/Ki-67 Dual-Stain Triage of HPV-Positive Women. JAMA Oncol. 2019;5(2):181–186. doi:10.1001/jamaoncol.2018.4270
What is the long-term risk of cervical precancer in human papillomavirus (HPV)-positive women with positive and negative p16/Ki-67 dual-stain results in relation to clinical management thresholds?
Among the 1549 HPV-positive women in this prospective cohort study, p16/Ki-67 positivity was associated with significantly higher cumulative 5-year risks of cervical precancer compared with abnormal Papanicolaou test results. In p16/Ki-67–negative women, the risk of precancer remained below the colposcopy referral threshold for all 5 years, crossing the 1-year return threshold at 3 years.
The low risk of cervical precancer in p16/Ki-67–negative women permits safe extension of follow-up intervals for 3 years.
As cervical cancer screening transitions to primary human papillomavirus (HPV) testing, effective triage and management of HPV-positive women is critical to avoid unnecessary colposcopy referral and associated harms while maintaining high sensitivity for cervical precancer. Triage with p16/Ki-67 dual-stain (DS) testing has shown high sensitivity and specificity for detection of cervical precancers; however, longitudinal studies are needed to determine the long-term risk of precancer following a negative DS result.
To evaluate the longitudinal performance of p16/Ki-67 DS triage for detection of cervical precancer in HPV-positive women over 5 years of follow-up in the context of clinical management thresholds.
Design, Setting, and Participants
Prospective cohort study of HPV-positive women 30 years or older undergoing routine cervical cancer screening in 2012 with HPV and Papanicolaou (hereinafter “cytology”) co-testing within the Kaiser Permanente Northern California health care system. Follow-up of medical records was conducted through 2017.
All p16/Ki-67 DS testing was performed on residual SurePath material, and slides were evaluated for p16/Ki-67 positivity.
Main Outcomes and Measures
Histological end points were ascertained from the clinical database through 2017. We estimated 5-year cumulative risks of cervical intraepithelial neoplasia grades of 2 or worse (≥CIN2) or grades 3 or worse (≥CIN3) by baseline DS and cytology at yearly intervals using Logistic Weibull models. Risks were compared with clinical management thresholds for colposcopy referral and a 1-year return interval.
Among the 1549 HPV-positive women in this study, the mean age at enrollment was 42.2 years, and the median follow-up time was 3.7 years (range, 0.2-5.4 years). Positive DS results were associated with significantly higher cumulative 5-year risks of ≥CIN2 compared with abnormal cytology (31.0%; 95% CI, 27.2%-35.3% vs 25.0%; 95% CI, 21.7%-28.7%; P = .03). Women with DS-negative findings had significantly lower 5-year risks of ≥CIN2 compared with women with normal cytology (8.5%; 95% CI, 6.5%-11.1% vs 12.3%; 95% CI, 9.8%-15.4%; P = .04). In DS-negative women, the risks of both ≥CIN2 and ≥CIN3 remained below the colposcopy referral threshold for all 5 years, crossing the 1-year return threshold at 3 years.
Conclusions and Relevance
Triage with p16/Ki-67 DS provides better long-term risk stratification than cytology over 5 years. The low risk of cervical precancer in p16/Ki-67 DS–negative women permits safe extension of follow-up intervals for 3 years.
Human papillomavirus (HPV) testing has been approved for primary cervical cancer screening. Efficient management of HPV-positive women requires triage markers that distinguish those at high risk who need colposcopy from those who can safely return to routine screening.1,2 Currently, Papanicolaou testing (hereinafter “cytology”) alone or in combination with HPV genotyping is recommended for triage of HPV-positive women in some settings3; however, cytology requires frequent testing at shorter intervals because the risk of precancer in cytology-negative women is not low enough to provide long-term reassurance.4 To overcome these limitations, alternative molecular-based markers are currently being evaluated, including p16/Ki-67 dual stain (DS) testing, which detects coexpression of p16, a tumor suppressor protein upregulated by HPV oncogene activity, and the cell proliferation marker Ki-67.5-12 We and others have shown that p16 and DS testing have better performance than cytology for detection of precancers in HPV-positive women.6,8,10,11,13
Longitudinal studies are needed to establish optimal testing intervals for DS-negative women. To our knowledge, only 1 study has evaluated the longitudinal accuracy of p16, showing high sensitivity of p16 for cervical precancer for up to 3 years.13 However, prospective studies evaluating DS for HPV triage are lacking. Herein, we evaluate the 5-year risk of cervical precancer following DS triage of HPV-positive women undergoing screening.
This is a prospective cohort study of women undergoing cervical cancer screening with cytology and HPV co-testing at Kaiser Permanente Northern California (KPNC).6 Between January and May of 2012, discard SurePath cytology specimens were collected at KPNC from 2364 HPV-positive women 30 years and older. Disease end points were abstracted from electronic medical records through May 2017. We excluded 544 women with a previous abnormal cytology within 2 years of enrollment, leaving 1820 HPV-positive women. We further excluded 150 women who did not have an evaluable DS result, 81 missing follow-up, and 40 women who had indeterminate outcomes. These 271 excluded women were similar in age, but had a higher proportion of normal cytology results compared with the 1549 (85.1% of the 1820) women included in the analysis (70% vs 49%, P < .001). The study was approved by the KPNC institutional review board and a waiver of informed consent was granted. This study was exempt from institutional review at the National Cancer Institute.
SurePath slides (Becton Dickinson) were prepared, stained, and processed on the FocalPoint Slide Profiler (Becton, Dickinson and Company). Cytology was categorized per the 2001 Bethesda System as negative for intraepithelial lesion or malignancy (NILM), atypical squamous cells of undetermined significance (ASC-US), low-grade squamous intraepithelial lesion (LSIL), and high-grade squamous intraepithelial lesion (HSIL).14 The HPV testing was performed using Hybrid Capture 2 (HC2; Qiagen Inc) in specimen transport medium at the KPNC regional laboratory per the manufacturer’s instructions. All HPV-positive women were evaluated by guided screening–assisted screening and full manual review with knowledge of HPV status. All abnormal slides were sent for pathology review. In addition, all negative cytology results from HPV-positive women were rescreened manually. Management was not based on DS results.
All p16/Ki-67 DS testing was performed on the residual SurePath cell pellet, supplemented with CytoRich Fluid (Becton, Dickinson and Company) within 1 to 4 months of sample collection. Slides were produced at the manufacturer’s laboratory using the CINtec PLUS Cytology kit (Roche) according to instructions, with each staining run including 2 control specimens. Staining was performed on a Dako Autostainer using the program for SurePath slides, followed by hematoxylin counterstaining. Slides were evaluated by an expert cytotechnologist blinded to the associated cervical histology; the number of DS-positive cells was assessed semiquantitatively (0, 1, 2-5, 6-50, or >50).
At KPNC, women who tested HPV-positive with ASC-US or more severe cytologic abnormalities (≥ASC-US) were referred for colposcopy, while women who tested HPV-positive with NILM cytology underwent repeated co-testing after 12 months and were referred for colposcopy if either test result was positive. Histological diagnoses are based on the cervical intraepithelial neoplasia (CIN) classification.
Primary outcomes included risk of histologically confirmed CIN grades of 2 or worse (≥CIN2) and, alternatively, CIN grades of 3 or worse (≥CIN3) including adenocarcinoma in situ. We evaluated the distribution of DS results by enrollment cytology and worst histologic diagnosis using contingency tables and χ2 trend tests. We estimated 5-year cumulative risks with 95% confidence intervals (95% CI) of ≥CIN2 and ≥CIN3 by enrollment DS (positive vs negative) and cytology (≥ASC-US vs NILM) individually, and as combined categories, and fit separate models stratified by age (<45 years vs ≥45 years). We also evaluated cumulative risks of ≥CIN2 and ≥CIN3 by number of DS-positive cells (1, 2-5, 6-50, or >50 vs 0) and cytology categories (ASC-US, LSIL, or ≥HSIL vs NILM). Risks were estimated using a logistic Weibull model (eMethods and eFigure 1 in the Supplement).15,16P values were calculated using t tests. For comparison, we also report cumulative detection using the Kaplan-Meier method (eFigures 2-5 in the Supplement). Risks were compared with internal benchmarks in our study population, estimated as the baseline (1-year) risk of ≥CIN2 and ≥CIN3 in HPV-positive women with ASC-US (threshold for immediate colposcopy referral, 11.1% and 5.7%, respectively) and in HPV-positive women with NILM (threshold for a 1-year return, 6.2% and 3.1%, respectively). Analyses were performed with R software (version 3.3.1) and logistic Weibull models were estimated using the PIMixture package.16 All statistical tests were 2-sided with P < .05 considered significant.
Among the 1549 HPV-positive women included, the mean age at enrollment was 42.2 years, and the median follow-up time was 3.7 years (range, 0.2-5.4 years). A total of 11 women were diagnosed with cancer, 110 with CIN3, 12 with adenocarcinoma in situ, 108 with CIN2, and the remaining 1308 were <CIN2 (Figure 1 and eTable 1 in the Supplement). Among those with ≥CIN2, 137 were prevalent cases with a median time to diagnosis of 1.6 months, and 71 were incident cases with a median time to diagnosis of 34.8 months. Among the remaining 33 cases with unknown prevalent or incident ≥CIN2, the median time to diagnosis was 20.4 months.
A total of 705 women (45.5%) were DS positive, and 785 (50.7%) had ≥ASC-US cytology at baseline. The DS positivity increased with worsening cytologic severity from 32% in women with NILM to 83% in women with >HSIL (P < .001 for trend). Similarly, DS positivity increased with increasing severity of worst histologic diagnosis from 22% in women without biopsy to 77% in women with CIN3, and 91% in women with cancer (P < .001 for trend) (eTable 1 in the Supplement).
Cumulative risks of ≥CIN2 by DS and cytology are detailed in eTable 2 in the Supplement with corresponding cumulative risk curves shown in Figure 2. Women testing DS positive had significantly higher baseline and 5-year risks of ≥CIN2 (22.0%; 95% CI, 19.2%-25.2% and 31.0%; 95% CI, 27.2%-35.3%, respectively) vs women with ≥ASC-US cytology (18.0%; 95% CI, 15.5%-20.8% and 25.0%; 95% CI, 21.7%-28.7%, respectively). Women testing DS negative had significantly lower baseline and 5-year risks of ≥CIN2 (4.0%; 95% CI, 3.0%-5.5% and 8.5%; 95% CI, 6.5%-11.1%, respectively) vs women with NILM cytology (6.2%; 95% CI, 4.7%-8.1% and 12.3%; 95% CI, 9.8%-15.4%, respectively). We observed similar patterns for risk of ≥CIN3 (eTable 2 in the Supplement). We did not observe significant differences by age (data not shown).
Furthermore, we evaluated risk in combined categories of enrollment DS and cytology (eTable 2 in the Supplement and Figure 3). The baseline and 5-year risks of ≥CIN2 in women with DS-negative/≥ASC-US findings (5.3%; 95% CI, 3.4%-8.3% and 10.0%; 95% CI, 6.7%-14.8%, respectively) were similar to those with DS-negative/NILM cytology (3.2%; 95% CI, 2.0%-5.2% and 7.6%; 95% CI, 5.2%-11.0%, respectively). We observed similar patterns for risk of ≥CIN3 (eTable 2 in the Supplement) and did not observe any significant differences by age (data not shown). The baseline and cumulative 5-year risks of ≥CIN2 and ≥CIN3 increased with increasing number of DS-positive cells, providing additional levels of risk stratification compared with categories of cytology (eTable 3 and eFigure 6 in the Supplement).
Cumulative risks of ≥CIN2 and ≥CIN3 in DS-positive women and in women with ≥ASC-US were much higher than internal risk thresholds for colposcopy referral for all 5 years. In DS-negative women, cumulative risks of ≥CIN2 and ≥CIN3 remained below the colposcopy referral threshold for up to 5 years and crossed the 1-year return referral threshold at year 3, respectively. In contrast, cumulative risks of ≥CIN2 and ≥CIN3 in women with NILM cytology crossed the threshold for immediate colposcopy referral after year 4 and the 1-year return risk threshold at year 1, respectively (Figure 2). The combination of DS and cytology provided long-term risk stratification: risks of ≥CIN2 and ≥CIN3 in women with ≥ASC-US who were DS negative were lower than the colposcopy referral threshold for all 5 years. In contrast, the risks of ≥CIN2 and ≥CIN3 in women with NILM cytology who were DS positive were higher than the colposcopy referral threshold at baseline (Figure 3).
As cervical cancer screening transitions to primary HPV testing, finding effective triage and management for HPV-positive women remains a critical issue.1,2 Risk thresholds can guide whether a woman should return to routine screening or undergo repeated testing, colposcopy, or immediate treatment.2,17,18 Longitudinal studies evaluating the long-term negative predictive value of a test indicate how long test intervals can be safely extended. In this large prospective study of HPV-positive women, we evaluated the longitudinal 5-year performance of DS triage for detection of cervical precancer. Our findings support that HPV-positive, DS-negative women can avoid immediate colposcopy referral and be safely retested after 3 years.6,13 Additionally, we found that DS has lower positivity, and can provide superior long-term risk stratification than cytology and HPV co-testing, which may reduce costs associated with screening and management.
Results from our analysis of DS positivity thresholds suggest that additional risk stratification can be obtained from quantitative assessment of DS-positive cells. The risk of precancer in women with more than 50 positive cells approached the risk of HSIL, and it is possible that automated evaluation may enable even more refined risk stratification compared with the current semiquantitative, manual approach.19
To our knowledge, ours is the first study to evaluate the long-term performance of DS for triage of HPV-positive women. Previously, Carozzi et al13 assessed the 3-year performance of p16 staining in a prospective study nested within the New Technologies for Cervical Cancer Screening (NTCC) trial. In their study, p16 cytology provided long-term risk stratification and had high longitudinal sensitivity for cervical precancer over 3 years. Compared with their estimates, the 3-year risks of ≥CIN2 and ≥CIN3 found in the present study were slightly higher. In the NTCC trial, HPV-positive women were referred for colposcopy, with annual follow-up until clearance of HPV was observed. Thus, differences in underlying population characteristics, frequency of disease ascertainment, and methodological approaches for estimating risk may account for these differences.
Strengths of our analysis include a large population, providing precise risk estimates particularly for the long-term reassurance against ≥CIN2 of DS-negative results. Further, the uniform screening and management procedures assured excellent outcome ascertainment with little loss to follow-up over 5 years. Our study reflects real-world clinical practice in an organized setting, which is essential for informing screening guidelines.
A potential limitation of our study for evaluating baseline risk was the differential management of HPV-positive women with ≥ASC-US vs those with NILM cytology, and the fact that referral was not based on DS. To address this, we used logistic Weibull models, which accounted for undiagnosed prevalent disease that may have occurred in women testing HPV positive and NILM who were not immediately referred for colposcopy.15,16 This approach provided more accurate risk estimates compared with other commonly used methods such as Kaplan-Meier and enabled better estimation of prevalent precancer risk.
In conclusion, our prospective study of HPV-positive women undergoing cervical cancer screening shows that DS provides long-term risk stratification compared with cytology triage for up to 5 years. Since the risk of precancer in HPV-positive/DS-negative women is identical at 3 years to the risk in women who test HPV positive and cytology NILM at 1 year, repeated testing can be safely extended to 3-year intervals in these women.
Corresponding Author: Megan Clarke, PhD, MHS, Division of Cancer Epidemiology and Genetics, National Cancer Institute, 9609 Medical Center Dr, Room 6E552, Rockville, MD 20892 (firstname.lastname@example.org).
Accepted for Publication: July 6, 2018.
Published Online: October 11, 2018. doi:10.1001/jamaoncol.2018.4270
Author Contributions: Drs Clarke and Wentzensen had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Castle, Schiffman, Lorey, Wentzensen.
Acquisition, analysis, or interpretation of data: Clarke, Cheung, Castle, Tokugawa, Poitras, Lorey, Kinney, Wentzensen.
Drafting of the manuscript: Clarke, Wentzensen.
Critical revision of the manuscript for important intellectual content: Cheung, Castle, Schiffman, Tokugawa, Poitras, Lorey, Kinney, Wentzensen.
Statistical analysis: Clarke, Cheung, Poitras, Wentzensen.
Obtained funding: Wentzensen.
Administrative, technical, or material support: Schiffman, Tokugawa, Lorey, Kinney.
Study supervision: Lorey, Kinney, Wentzensen.
Conflict of Interest Disclosures: Drs Schiffman and Wentzensen are employed by the National Cancer Institute (NCI), and the NCI has received cervical cancer screening assays in kind or at reduced cost from Becton, Dickinson and Company and Roche for studies that Drs Schiffman and Wentzensen are working on. Dr Castle has received commercial HPV tests for research at reduced or no cost from Roche, Arbor Vita Corporation, and Cepheid. No other disclosures are reported.
Funding/Support: This research was supported by the Intramural Research Program of the National Institutes of Health and the NCI.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Information: This manuscript is dedicated to the memory of Barbara Fetterman.
Create a personal account or sign in to: