Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Apr 1.
Published in final edited form as: J Low Genit Tract Dis. 2022 Apr 1;26(2):127–134. doi: 10.1097/LGT.0000000000000667

Contribution of etiologic cofactors to CIN3+ risk among women with HPV-positive screening test results

Maria Demarco 1, Didem Egemen 1, Noorie Hyun 1, Xiaojian Chen 1, Anna-Barbara Moscicki 2, Li Cheung 1, Olivia Carter-Pokras 1, Anne Hammer 3, Julia C Gage 1, Megan A Clarke 1, Philip E Castle 1, Brian Befano 1, Jie Chen 1, Cher Dallal 1, Xin He 1, Kanan Desai 1, Thomas Lorey 4, Nancy Poitras 4, Tina R Raine-Bennett 4, Rebecca B Perkins 5, Nicolas Wentzensen 1, Mark Schiffman 1
PMCID: PMC8940696  NIHMSID: NIHMS1776333  PMID: 35249974

Abstract

Objective:

U.S. screening and management guidelines for cervical cancer are based on the absolute risk of precancer estimated from large clinical cohorts and trials. Given the widespread transition toward screening with HPV testing, it is important to assess which additional factors to include in clinical risk assessment to optimize management of HPV-infected women.

Methods:

We analyzed data from HPV-infected women, ages 30 to 65 years, in the NCI-Kaiser Permanente Northern California Persistence and Progression study. We estimated the influence of HPV risk group, cytology result, and selected cofactors on immediate risk of CIN3+ among 16,094 HPV-positive women. Cofactors considered included, age, race/ethnicity, income, smoking, and hormonal contraceptive use.

Results:

HPV risk group and cytology test result were strongly correlated with CIN3+ risk. After considering cytology and HPV risk group, other cofactors (age, race/ethnicity, income, smoking, and hormonal contraceptive use) had minimal impact on CIN3+ risk and did not change recommended management based on accepted risk thresholds. We had insufficient data to assess the impact of long-duration heavy smoking, parity, history of sexually transmitted infection, or immunosuppression.

Conclusions:

In our study at KPNC, the risk of CIN3+ was determined mainly by HPV risk group and cytology results, with other cofactors having limited impact in adjusted analyses. This supports the use of HPV and cytology results in risk-based management guidelines.

Keywords: cervical cancer screening, risk estimation, guidelines, smoking, hormonal contraception

Precis:

HPV risk group and cytology are major predictors of CIN3+ risk. Etiologic cofactors, such as smoking and contraception, do not alter risk estimates enough to change clinical management.

Introduction

Cervical cancer screening is shifting from cytology to the more sensitive approach of human papillomavirus (HPV) testing, either as HPV and cytology co-testing, or as HPV testing alone.1,2 Testing HPV negative permits safe extension of screening intervals. However, testing HPV positive alone does not necessarily require colposcopy referral and/or treatment; additional tests or predictive factors are needed to estimate precancer risk and determine the next step in clinical management. The 2019 ASCCP Risk-Based Management Consensus Guidelines assess an individual’s risk of precancer using combinations of past and current screening results and indicate recommended management accordingly, with the goal of maximizing benefits (i.e., cervical cancer prevention) and minimizing harms (e.g., unnecessary referrals to colposcopy and treatment).1 The clinical actions range from intensified surveillance for women at a slightly elevated 5-year risk of Cervical Intraepithelial Neoplasia grade 3 or higher (CIN3+), to colposcopic referral, to possible treatment, or even preferred treatment for very high immediate risk.

Data from large observational studies and regulatory trials were used to generate precise precancer risk estimates for the many possible combinations of test results and cofactors. An essential step was to assess the importance of each risk predictor considered for risk estimation. Specifically, management guidelines for HPV-positive women needed to address which factors have sufficient impact on the estimated risk of precancer to warrant routine consideration in management for each clinical encounter. 3 HPV infection is the necessary cause of cervical cancer, and both HPV infection and abnormal cytology results are a strong predictors of elevated immediate risk of precancer and cancer. Several factors have also been found to be associated with cervical cancer, including HPV genotype, genetic and immunologic host factors, and other factors such as smoking status, long term contraceptive use, parity, and sexually transmitted infections. Some factors may elevate the risk of cancer up to approximately two-fold.1, 4-13 In the context of the 2019 guidelines, this analysis aimed to assess whether, in the setting of known HPV infection and cytology results, some of the previously evaluated cofactors and potential risk modifiers for cervical precancer warrant routine consideration in clinical management of patients with HPV-positive screening test results.

Methods

Study design and population

This was a retrospective cohort analysis using data from HPV co-testing of a large set of residual specimens following HPV testing conducted at Kaiser Permanente Northern California (KPNC), with supplemental HPV typing directed by the National Cancer Institute. This study was approved by institutional review boards at the National Cancer Institute and Kaiser Permanente Northern California. Written consent was obtained from participants. At KPNC, women were tested by Hybrid Capture 2 (HC2) for the group of high-risk genotypes of HPV (as a pool without genotyping) as a co-test with cytology in women ages 30 years and older beginning in 2003.14.15 Women younger than 30 years, who had HC2 results mainly for ASC-US triage, were not comparable to women aged 30 years and older undergoing routine screening with HPV/cytology co-testing and were not considered for this analysis.

The HPV Persistence and Progression Cohort (PaP Cohort) was created by banking residual, discarded cervical specimens, collected into specimen transport medium (STM; Qiagen), from women tested by HC2. Women were contacted, and 92% agreed to specimen storage and research testing. The study collection from the enrollment phase of the PaP Cohort (January 2007- January 2011) consisted of a group of 45,000 HPV-positive residual specimens (approximately 80% of HC2 positive specimens from KPNC co-testing during that time). Women were followed prospectively through 2017 or until they left the KPNC system. The core collection used for the present analysis was drawn from nearly 30,000 specimens from women 30 years or older who tested positive by HC2 during routine screening, with a small number of HC2-negative specimens included as negative controls. Given the interest in evaluating progression of type-specific infection to cervical intraepithelial neoplasia grade 3, adenocarcinoma in situ, or cancers (CIN3+), we restricted this analysis to women with a positive HPV test result by HC2 test at baseline and information on HPV typing. Specifically, this analysis used data from 16,094 women with positive screening HC2 results who were chosen by stratified random sampling in previously published investigations for masked HPV typing to distinguish the individual carcinogenic HPV risk group(s) present.14,15 Specimens were processed at BD Diagnostics (Sparks MD) by the Onclarity assay, or at Roche Molecular Systems (Pleasanton CA) by cobas and/or Linear Array (the latter intended for research and not FDA approved for clinical use). Fewer than 10% were HPV typed using research-assays, either MY09-MY11 with dot-blot typing (R. Burk MY09-MY11 PCR) or Linear Array (F. Coutlee lab).

Variables

This analysis studied the immediate risk of CIN3+ following the baseline screening visit to determine whether, after controlling for HPV risk group and cytology results, sociodemographic cofactors changed risk estimates sufficiently to merit different clinical actions. Infection with a specific HPV genotype was defined hierarchically to match currently available HPV DNA assays (HPV16, else HPV18, else 45, else 31/33/52/58, or 35/39/51/56/59/66/68). Patients were assigned to the HPV risk group associated with their first detected infection.

Cytology results included NILM (negative for intraepithelial lesions or malignancy), ASC-US (atypical squamous cells of undetermined significance), LSIL (low-grade squamous intraepithelial lesion), ASC-H (atypical squamous cells of undetermined significance - cannot exclude high-grade squamous intraepithelial lesion (HSIL), AGC (atypical glandular cells), HSIL, adenocarcinoma in situ (AIS), and cancer. Cytology results were then grouped by precancer risk into four categories for analysis: (1) NILM, (2) ASC-US or LSIL, (3) ASC-H or AGC, and (4) HSIL, AIS, or Cancer (HSIL+) based on prior studies in this population.14,15

Potential sociodemographic and behavioral cofactors selected were based on previous literature. Age was categorized in five-year categories from 30 to 65 years. Race/ethnicity was categorized using medical record data as non-Hispanic White, Hispanic, non-Hispanic Black, non-Hispanic Asian/Pacific Islander, multiracial/other. Neighborhood income was categorized based on census tract: high (≥80% in Census tract were >200% of poverty level) low (≥20% households in Census tract were below poverty level), and middle (incomes between high and low). Smoking was categorized as never, former, and current. Hormonal contraceptive use was determined by the total number of oral contraceptive pill (OCP) packs or depot medroxyprogesterone acetate (DMPA) injections dispensed in the years prior to and including enrollment date, categorized as a proxy for duration of use: Never (0 packs or injections prescribed), 1-6 years (1-83 1-month packs or 1-28 3-month injections), 7-9 years (83-120 1-month packs or ≥28 3-month injections), and 10+ years (≥120 1-month packs; no DMPA users had >10 years of use recorded).

Statistical analysis

We compiled descriptive statistics for the study population as a whole and for women with CIN3+. Sample-weighted logistic-Cox models were used to combine risk of prevalent CIN3+ estimated by odds ratios (OR) with risk of incident CIN3+ estimated by hazard ratios (HR). To assess the impact of each factor independently, univariate cumulative risks combining OR at baseline and HR were shown for years 1, 3, and 5 (Table 3). Bi-covariate cumulative risks combined the main predictive variables of HPV genotype and cytology (Table 4). To consider the clinical utility of additional predictor variables, multivariate analyses considered as third variables those cofactors found to be significant in univariate models. Cofactors were dichotomized by clinical relevance for inclusion in the final model (shown in Table 5). We chose immediate risk among HPV+ women for the focus of this analysis as the primary outcome of interest was whether sociodemographic variables would change indications for colposcopic referrals. We report 95% confidence intervals when informative. Multiple comparisons corrections were not employed in the variable selection because the analysis is explorative rather than confirmative.

Table 3.

Univariate cumulative risk of CIN3+ among HPV+ women by sociodemographic and clinical characteristics

Cumulative Risk
1-year 95% CI 3-years 95% CI 5-years 95% CI
Age 30 to 34 10.9 10.2 11.6 13 12.2 13.8 13.8 12.9 14.7
35 to 39 10.4 9.6 11.3 12 11 12.9 12.8 11.8 13.9
40 to 44 10.4 9.4 11.3 12.3 11.2 13.4 13.2 12 14.5
45 to 49 8.7 7.8 9.7 10.1 9 11.2 10.9 9.7 12.2
50 to 54 6.7 5.7 7.6 7.9 6.8 9 9 7.8 10.3
55 to 59 5.7 4.7 6.6 6.8 5.7 8 7.9 6.6 9.3
60 to 65 7.4 6.1 8.6 8.4 6.9 9.9 9.3 7.7 10.9
Race/ethnicity Non-Hispanic White 10.3 9.8 10.8 12.2 11.6 12.9 13.6 12.8 14.4
Hispanic 9.3 8.6 10.1 10.6 9.8 11.5 11.5 10.5 12.4
Non-Hispanic African American 6.1 5.2 7 6.7 5.7 7.6 7.3 6.3 8.4
Non-Hispanic Asian/Pacific Islander 9.9 9 10.7 12.1 11 13.1 13.1 11.9 14.2
Multiracial 6.8 4.1 9.5 7.2 4.3 10.1 7.2 4.3 10.1
Income1 Low income 9.5 8.4 10.5 10.8 9.6 12 11.6 10.3 12.9
Bottom income 9.5 8.9 10.1 10.9 10.2 11.7 11.8 11 12.6
High income 9.5 9 9.9 11.2 10.7 11.8 12.6 11.9 13.3
Smoking Never 9.2 8.8 9.6 10.7 10.2 11.1 12.1 11.5 12.7
Former 9.1 8.1 10 11 9.8 12.1 12.1 10.8 13.4
Current 12.1 11 13.3 13.8 12.4 15.1 14.5 13.1 15.9
Duration of OCP Use (years)2 0 9.6 9.1 10.1 11.1 10.6 11.7 12.3 11.7 13
1-6 9.2 8.6 9.8 10.8 10.2 11.5 12 11.2 12.8
7-10 9.6 7.8 11.3 12.4 10.2 14.7 13.4 11 15.8
10+ 11.2 8 14.4 13.1 9.4 16.8 13.5 9.7 17.4
Duration of DMPA Use (years)2 0 9.5 9.1 9.8 11.0 10.6 11.4 12.4 11.9 13.0
1-6 10.0 8.1 11.8 12.2 9.9 14.4 13.4 10.9 15.9
7+ 9.3 7.6 11.0 10.5 8.6 12.5 11.6 9.5 13.7
Cytology3 NILM 2.9 2.6 3.2 4.7 4.3 5.1 6.0 5.4 6.6
ASC-US/LSIL 11.0 10.4 11.6 12.5 11.9 13.2 13.4 12.6 14.2
ASC-H/AGC 27.6 24.9 30.3 30.6 27.5 33.7 31.3 28.2 34.4
HSIL+ 64.3 60.4 68.1 65.2 61.3 69.0 66.0 62.2 69.9
HPV Risk Group 35/39/51/56/59/66/68 4.2 3.7 4.9 4.2 5.0 4.4 5.2 4.6 6.0
31/33/52/58 7.2 6.4 8.1 7.2 9.0 8.1 9.4 8.5 10.4
45 5.3 3.7 7.6 5.3 6.9 5.3 7.3 5.6 9.5
18 9.2 7.6 11.1 9.2 10.7 9.0 11.5 9.6 13.6
16 16.9 15.5 18.5 16.9 20.8 19.2 21.6 19.9 23.4
Open in a new tab
1

High income: >=80% in Census Tract are >200% of Poverty Level. Low income: >=20% Households in Census Tract are Below Poverty Level

2

Oral Contraceptive Pill (OCP) packs and Depot-medroxyprogesterone acetate (DMPA) injections prescribed were categorized with their equivalent in number of years used (never users: 0 packs or injections, 1-6 years: 1-83 packs or 1-28 injections, 7-9 years: 83-120 packs or ≥28 injections, and 10+ years: ≥120 packs).

3

NILM= negative for intra-epithelial lesion or malignancy, ASC-US= Atypical Squamous Cells of Undetermined Significance. LSIL= Low-grade Squamous Intraepithelial lesion, ASC-H = Atypical Squamous Cells, cannot exclude High-grade, AGC = Atypical Glandular Cells, HSIL+ = High-grade squamous intraepithelial lesion or higher including adenocarcinoma in situ or cancer

Table 4.

Bi-covariate cumulative risk of CIN3+ by combinations of HPV risk group and cytology

HPV Risk Group
(hierarchical)		 cytology N % Immediate
CIN3+
risk		 95% CI
35/39/51/56/59/66/68 NILM 5343 17 1.1 0.66 1.5
ASC-US/LSIL 4175 13 3.1 2.5 3.6
ASC-H/ AGC 322 1 11 7.1 14
HSIL+ 259 0.82 31 25 37
31/33/52/58 NILM 5307 17 2.5 1.9 3.1
ASC-US/LSIL 4015 13 5 4.3 5.8
ASC-H/ AGC 509 1.6 16 13 20
HSIL+ 449 1.4 41 36 46
45 NILM 1163 3.7 0 0 0.002
ASC-US/LSIL 621 2 3.6 2 5.2
ASC-H/ AGC 124 0.39 17 10 25
HSIL+ 86 0.27 42 29 54
18 NILM 1309 4.1 3.8 2.4 5.2
ASC-US/LSIL 954 3 5 3.5 6.6
ASC-H/ AGC 202 0.64 23 17 30
HSIL+ 125 0.39 37 27 47
16 NILM 3149 9.9 6 4.9 7.2
ASC-US/LSIL 2675 8.4 11 9.7 12
ASC-H/ AGC 462 1.5 30 25 35
HSIL+ 476 1.5 62 57 67
Open in a new tab

NILM= negative for intra-epithelial lesion or malignancy, ASC-US= Atypical Squamous Cells of Undetermined Significance. LSIL= Low-grade Squamous Intraepithelial lesion, ASC-H = Atypical Squamous Cells, cannot exclude High-grade, AGC = Atypical Glandular Cells, HSIL+ = High-grade squamous intraepithelial lesion or higher including adenocarcinoma in situ or cancer

Table 5.

Multi-covariate immediate risk of CIN3+ by HPV risk group, cytology, and selected cofactors

HPV Risk Group
(hierarchical)		 Cytology Age Race/ethnicity Smoking OCP (years)
Age
Group		 % CIN3+
Immediate
Risk		 Race/
ethnicity
Group		 % CIN3+
Immediate
Risk		 Smoking
Status		 % Immediate
CIN3+
Risk		 OCP
(years)		 % CIN3+
Immediate
Risk
35/39/51/56/59/66/68 NILM 30-44 11 1.3 (0.73, 1.8) White/ Hispanic/ Asian/Other 16 1.1 (0.67, 1.6) Never/ former 15 1.1 (0.61, 1.5) 0-6 16 1 (0.6, 1.5)
45-65 6.4 0.76 (0.39, 1.1) Black2 1.5 0.84 (0.28, 1.4) Current 1.8 1.2 (0.56, 1.9) 7+ 0.97 1.7 (0.65, 2.8)
ASC-US/ LSIL 30-44 8.9 3.5 (2.8, 4.3) White/ Hispanic/ Asian/Other 12 3.1 (2.5, 3.7) Never/ former 12 3 (2.4, 3.6) 0-6 12 3 (2.4, 3.5)
45-65 4.3 2.1 (1.5, 2.8) Black 0.97 2.3 (1.1, 3.6) Current 1.5 3.5 (2.1, 4.8) 7+ 0.77 4.9 (2.4, 7.3)
ASC-H/ AGC 30-44 0.6 12 (7.6, 17) White/ Hispanic/ Asian/Other 0.95 10 (6.6, 14) Never/ former 0.92 11 (6.7, 15) 0-6 0.89 9.8 (6, 14)
45-65 0.4 7.6 (4.3, 11) Black 0.04 7.9 (3, 13) Current 0.09 12 (5.7, 18) 7+ 0.1 15 (7.6, 23)
HSIL+ 30-44 0.55 34 (26, 41) White/ Hispanic/ Asian/Other 0.71 31 (24, 38) Never/ former 0.7 31 (24, 38) 0-6 0.81 30 (23, 37)
45-65 0.28 23 (16, 30) Black 0.12 25 (14, 36) Current 0.11 34 (23, 45) 7+ 0.02 42 (27, 57)
31/33/52/58 NILM 30-44 11 2.7 (2, 3.5) White/ Hispanic/ Asian/Other 15 2.4 (1.7, 3.1) Never/ former 15 2.5 (1.9, 3.2) 0-6 16 2.4 (1.8, 3)
45-65 5.9 2 (1.4, 2.5) Black 1.8 1.9 (1.1, 2.6) Current 2 2.6 (1.7, 3.5) 7+ 1 3.1 (1.6, 4.5)
ASC-US/LSIL 30-44 9 5.5 (4.6, 6.4) White/ Hispanic/ Asian/Other 12 5.1 (4.4, 5.9) Never/ former 11 5 (4.2, 5.8) 0-6 12 5 (4.2, 5.7)
45-65 3.5 4 (3, 4.9) Black 1 4 (2.4, 5.6) Current 1.6 5.2 (3.6, 6.8) 7+ 0.79 6.3 (3.9, 8.7)
ASC-H/ AGC 30-44 1.1 18 (14, 22) White/ Hispanic/ Asian/Other 1.5 17 (13, 21) Never/ former 1.3 16 (12, 20) 0-6 1.6 16 (12, 20)
45-65 0.53 13 (9.6, 17) Black 0.14 13 (7.7, 19) Current 0.26 17 (11, 22) 7+ 0.07 20 (12, 28)
HSIL+ 30-44 0.98 42 (36, 48) White/ Hispanic/ Asian/Other 1.3 41 (35, 46) Never/ former 1.2 41 (35, 47) 0-6 1.4 40 (34, 45)
45-65 0.42 34 (28, 41) Black 0.15 34 (25, 44) Current 0.25 41 (33, 50) 7+ 0.05 46 (34, 57)
45 NILM 30-44 2.4 0.26 (0, 1.5) White/ Hispanic/ Asian/Other 3.3 0 (0, 0.01) Never/ former 3.3 0 (0, 0.01) 0-6 3.5 0.02 (0, 0.71)
45-65 1.2 0.22 (0, 1.3) Black 0.38 0 (0, 0.01) Current 0.41 0 (0, 0.02) 7+ 0.17 0.01 (0, 0.24)
ASC-US/LSIL 30-44 1.3 3.8 (1.9, 5.7) White/ Hispanic/ Asian/Other 1.8 3.7 (2, 5.4) Never/ former 1.7 3.3 (1.6, 5) 0-6 1.9 3.7 (2, 5.5)
45-65 0.6 3.2 (1.1, 5.4) Black 0.16 4 (0, 8.1) Current 0.25 5.3 (1.1, 9.4) 7+ 0.07 1.3 (0, 4)
ASC-H/ AGC 30-44 0.21 18 (8.2, 28) White/ Hispanic/ Asian/Other 0.34 17 (8.4, 25) Never/ former 0.35 17 (8.5, 25) 0-6 0.35 18 (9.3, 26)
45-65 0.16 15 (6.8, 24) African 0.04 18 (3.3, 33) Current 0.04 24 (7.1, 42) 7+ 0.03 6.8 (0, 21)
HSIL+ 30-44 0.22 41 (27, 55) White/ Hispanic/ Asian/Other 0.25 40 (26, 54) Never/ former 0.23 40 (26, 54) 0-6 0.28 41 (27, 54)
45-65 0.07 37 (19, 55) Black 0.04 42 (17, 67) Current 0.04 52 (29, 74) 7+ 0 19 (0, 51)
18 NILM 30-44 2.7 4.1 (2.1, 6.1) White/ Hispanic/ Asian/Other 3.7 4.1 (2.5, 5.6) Never/ former 3.5 4 (2.4, 5.6) 0-6 3.9 3.8 (2.3, 5.3)
45-65 1.4 2.8 1.5, 4.1) Black 0.41 2.3 (0.18, 4.5) Current 0.61 2.6 (0.95, 4.3) 7+ 0.2 5.8 (1.2, 10)
ASC-US/LSIL 30-44 2.1 5.3 (3.5, 7.1) White/ Hispanic/ Asian/Other 2.7 5 (3.4, 6.6) Never/ former 2.6 5.3 (3.5, 7) 0-6 2.8 4.7 (3.1, 6.2)
45-65 0.84 3.6 (1.9, 5.3) Black 0.25 2.9 (0.25, 5.5) Current 0.37 3.5 (1.4, 5.6) 7+ 0.14 7.1 (2, 12)
ASC-H/ AGC 30-44 0.4 25 (17, 34) White/ Hispanic/ Asian/Other 0.6 23 (16, 31) Never/ former 0.55 24 (17, 32) 0-6 0.58 23 (15, 30)
45-65 0.21 18 (10, 27) Black 0.01 15 (1.9, 27) Current 0.09 17 (7.3, 27) 7+ 0.03 31 (13, 49)
HSIL+ 30-44 0.23 41 (29, 54) White/ Hispanic/ Asian/Other 0.34 40 (28, 51) Never/ former 0.29 40 (27, 52) 0-6 0.37 38 (27, 49)
45-65 0.15 32 (20, 44) Black 0.04 27 (8.5, 46) Current 0.1 30 (16, 44) 7+ 0.01 49 (27, 71)
16 NILM 30-44 7.1 6.4 (5.1, 7.7) White/ Hispanic/ Asian/Other 9.3 6.3 (5.1, 7.6) Never/ former 9 5.9 (4.7, 7.1) 0-6 9.4 6.1 (4.9, 7.3)
45-65 2.9 5.6 (4.3, 6.9) Black 0.72 4.2 (2.3, 6.1) Current 0.94 7.5 (5.2, 9.8) 7+ 0.6 7 (4, 10)
ASC-US/LSIL 30-44 6.5 11 (9.7, 13) White/ Hispanic/ Asian/Other 7.9 11 (9.7, 13) Never/ former 7.3 11 (9.2, 12) 0-6 7.8 11 (9.4, 12)
45-65 1.8 9.9 (7.9, 12) Black 0.45 7.6 (4.4, 11) Current 1.2 13 (10, 17) 7+ 0.52 12 (8, 17)
ASC-H/ AGC 30-44 1 31 (25, 36) White/ Hispanic/ Asian/Other 1.3 31 (25, 36) Never/ former 1.2 29 (24, 34) 0-6 1.3 30 (25, 35)
45-65 0.39 28 (22, 34) Black 0.1 22 (14, 31) Current 0.26 35 (27, 42) 7+ 0.09 33 (23, 43)
HSIL+ 30-44 1.2 61 (55, 67) White/ Hispanic/ Asian/Other 1.4 61 (55, 67) Never/ former 1.1 60 (54, 66) 0-6 1.5 60 (54, 66)
45-65 0.34 58 (50, 65) Black 0.14 50 (39, 62) Current 0.36 66 (59, 74) 7+ 0.05 64 (53, 74)
Open in a new tab
1

Distribution refers to the % of column total that is represented in each cell

2

non-Hispanic Black

NILM= negative for intra-epithelial lesion or malignancy, ASC-US= Atypical Squamous Cells of Undetermined Significance. LSIL= Low-grade Squamous Intraepithelial lesion, ASC-H = Atypical Squamous Cells, cannot exclude High-grade, AGC = Atypical Glandular Cells, HSIL+ = High-grade squamous intraepithelial lesion or higher including adenocarcinoma in situ or cancer

Results

As shown in Table 1, women studied in this subset of HPV-positive KPNC members had a median age of 37 years. Nearly half (49.4%) were non-Hispanic White, 21.5% were Hispanic, 18.2% were Asian, and 9.8% were non-Hispanic Black. Most (88.7%) lived in middle or high-income neighborhoods, and 88.9% were non-smokers. Most women (55.9%) had no documented prescriptions for oral contraceptive pills, and 91.6% had no documented prescriptions for depo-medroxyprogesterone acetate. Most (53.5%) of this sample of HPV-positive women had concurrently abnormal cytology; approximately 80% of cytology abnormalities were low-grade (ASC-US or LSIL) and 20% were high-grade (ASC-H, AGC, HSIL or higher).

Table 1.

Sociodemographic, behavioral and clinical characteristics of HPV infected (or HPV positive) women in the sample

Total CIN3+
n Column % n Column %
Age in years 30 to 44 10390 64.7 1440 9.0
45 to 54 3341 20.8 340 2.1
55 to 65 2318 14.4 206 8.9
Race/ethnicity Non-Hispanic White 8268 49.4 1224 7.6
Hispanic 3590 21.5 468 0.3
Non-Hispanic African American 1638 9.8 144 0.1
Non-Hispanic Asian/Pacific Islander 3047 18.2 437 0.3
Multiracial 189 1.1 15 0.0
Neighborhood Income1 Low 1969 11.3 258 0.2
Middle 5559 32.0 745 0.5
High 9858 56.7 1313 8.2
Smoking Never 12408 74.0 1593 9.9
Former 2356 14.1 322 0.2
Current 2000 11.9 321 0.2
Duration of OCP Use (years)2 0 9864 55.9 1268 7.9
1-6 6837 38.7 932 5.8
7-10 697 3.9 117 4.3
10+ 260 1.5 40 0.0
Duration of DMPA Use (years)2 0 16382 91.6 2177 13.6
1-6 1466 8.2 202 1.3
7+ 40 0.2 4 0.0
Cytology NILM 8077 46.5 613 3.8
ASC-US/LSIL 7403 42.6 922 5.7
ASC-H/ AGC 992 5.7 329 2.0
HSIL+ 894 5.1 465 2.9
HPV Risk Group 35/39/51/56/59/66/68 5269 34.2 226 1.4
31/33/52/58 4557 29.6 634 4.0
18/45 1732 11.3 227 1.4
16 3832 24.9 1198 7.5
Open in a new tab
1

High income: >=80% in Census Tract are >200% of Poverty Level. Low income: >=20% Households in Census Tract are Below Poverty Level

2

Oral Contraceptive Pill (OCP) packs and Depot-medroxyprogesterone acetate (DMPA) injections prescribed were categorized with their equivalent in number of years used (never users: 0 packs or injections, 1-6 years: 1-83 packs or 1-28 injections, 7-9 years: 83-120 packs or ≥28 injections, and 10+ years: ≥120 packs).

*

Note, CIN3+ cases refers to cases diagnosed at baseline or during follow up

Table 2 describes univariate measures of association for prevalent CIN3+. In this well-screened population, ages 50-65 were associated with lower CIN3+ prevalences than age 30-34 (ORs 0.5-0.7 (95% CIs 0.5-0.8). Non-Hispanic Black and Multiracial women had slightly lower CIN3+ prevalence than non-Hispanic White women (NH Black OR 0.6, 95% CI 0.5-0.7; multiracial OR 0.4 95% CI 0.4-0.8). Income, smoking status, and use of hormonal contraception were not associated with prevalent CIN3+ in univariate analyses. Cytology results and HPV risk group were most strongly associated with CIN3+, with a current HSIL cytology having a 67.6-fold increased risk of prevalent CIN3+ compared to a normal cytology (95% CI 54.6-80.7), and HPV16 having an 8.2-fold increased risk of prevalent CIN3+ compared to the lowest risk subtypes (95% CI 9.7-17.2). Table 3 illustrates the cumulative risk of developing CIN3+ over 5 years by different characteristics. Most CIN3+ was diagnosed within the first year, indicating prevalent disease. For most factors, cumulative absolute CIN3+ risks increased by approximately 2-4% between years 1 and 5. The greatest increase in cumulative CIN3+ risk was seen for HPV16 infections, increasing from 16.9% at year 1 to 21.6% at year 5.

Table 2.

Univariate measures of association [odds ratios (OR)] for prevalent cervical intraepithelial neoplasia grade 3 or more severe diagnoses (CIN3+) between sociodemographic and clinical characteristics

Prevalent CIN3+
OR 95% CI
Age (Years) 30 to 34 1.0 - -
35 to 39 0.9 0.8 1.1
40 to 44 1.0 0.8 1.1
45 to 49 0.8 0.7 0.9
50 to 54 0.6 0.5 0.7
55 to 59 0.5 0.4 0.6
60 to 65 0.7 0.6 0.8
Race/ethnicity Non-Hispanic White 1.0 - -
Hispanic 0.9 0.8 1.0
Non-Hispanic African American 0.6 0.5 0.7
Non-Hispanic Asian/Pacific Islander 0.9 0.8 1.0
Multiracial 0.6 0.4 0.8
Income1 High income 1.0 - -
Low income 1.0 0.9 1.1
Middle income 1.0 0.9 1.1
Smoking Never 1.0 - -
Former 1.0 0.9 1.1
Current 1.4 1.3 1.6
Duration of OCP Use (years)2 0 1.0 - -
1-6 1.1 1.0 1.2
7-10 1.1 0.9 1.3
10+ 1.3 0.9 1.6
Duration of DMPA Use (years)2 0 1.0 - -
1-6 1.2 1.0 1.3
7+ 1.0 0.2 1.9
Cytology NILM 1.0 - -
ASC-US/LSIL 5.3 4.6 6.0
ASC-H 20.5 16.7 24.2
AGC 12.2 8.9 15.5
HSIL+ 67.6 54.6 80.7
HPV risk group risk group 35/39/51/56/59/66/68 1.0 - -
31/33/52/58 3.0 2.7 3.4
18/45 2.7 2.3 3.1
16 8.2 7.3 9.0
Open in a new tab
1

High income: >=80% in Census Tract are >200% of Poverty Level. Low income: >=20% Households in Census Tract are Below Poverty Level

2

Oral Contraceptive Pill (OCP) packs and Depot-medroxyprogesterone acetate (DMPA) injections prescribed were categorized with their equivalent in number of years used (never users: 0 packs or injections, 1-6 years: 1-83 packs or 1-28 injections, 7-9 years: 83-120 packs or ≥28 injections, and 10+ years: ≥120 packs).

Table 4 describes the immediate risk of CIN3+ by combinations of HPV risk group and cytology. For the lowest risk types (35/39/51/56/59/66/68) immediate risk ranged from 1.1% for NILM cytology to 31% for HSIL+ cytology. For the second risk group (31/33/52/58) immediate risk ranged from 2.5% for NILM cytology to 41.0% for HSIL+ cytology. For HPV 18 and 45, immediate risks ranged from <3% for NILM cytology to approximately 40% for HSIL+ cytology. Again, HPV16 was associated with the highest risks: 6% for NILM cytology and 62% for HSIL+ cytology.

Table 5 describes the multi-covariate analysis of the cumulative risk of CIN3+ by HPV risk group, cytology, and selected sociodemographic cofactors, considering one additional sociodemographic cofactor at a time. The goal of the analysis was to determine whether these factors, when considered in the context of HPV and cytology results, would change management recommendations as defined by clinical action thresholds [1]. The 2019 ASCCP Risk-Based Management Consensus Guidelines defined clinical action thresholds that determined management recommendations. Specifically, an immediate CIN3+ risk of 4-24% leads to a recommendation of colposcopy, a risk of 25-59% leads to a recommendation of colposcopy or treatment, and a risk of 60% or higher leads to a recommendation of treatment. After accounting for HPV risk group and cytology status, sociodemographic variables did not substantially impact risk estimates. In no case did age, race, smoking status, or oral contraceptive use lead to a different clinical recommendation than would be recommended by HPV type and cytology result alone.

Discussion

As previously shown, the risk of prevalent or incident CIN 3+ was mainly determined by HPV typing and cytology results. Also, as previously observed, age, race/ethnicity, long-term oral contraceptive use, and current smoking were significant albeit much weaker stratifiers of risk. This study adds to prior literature by demonstrating the residual effects of demographic factors after considering the effects of HPV and cytology results. The 2019 Risk-Based Management Consensus Guidelines describe six possible clinical actions (return in 5 years, return in 3 years, return in 1 year, colposcopy recommended, expedited treatment or colposcopy acceptable, or expedited treatment preferred) with defined risk thresholds determined by a national consensus.1 These data demonstrate that HPV risk group and cytology results were clinically important risk factors. High-grade cytology increased the immediate risk of CIN3+ by up to tenfold, and high-risk HPV types increased the immediate risk of CIN3+ up to fivefold. However, after considering cytology result and HPV risk group, the additional sociodemographic factors considered in this study did not raise risk sufficiently to exceed any of these risk thresholds and lead to a change in management. These findings support the inclusion of current and past HPV and cytology test results, and the exclusion of sociodemographic factors in existing risk estimates used in the 2019 Risk-Based Management Consensus Guidelines.1, 28

As with the ASCCP 2019 Risk-Based Management Consensus Guidelines, we chose CIN3+ as the surrogate endpoint for screening.1,16,17 Analyses used in the 2019 ASCCP Risk-Based Management Consensus Guidelines showed that using CIN2+ yielded similar conclusions related to the limited clinical significance of sociodemographic factors on risk estimates.18,28 Specifically, ancillary analyses using the same methods to estimate risk of CIN2+ suggest that the choice of the surrogate endpoint may change the magnitude of the effects and the hierarchy of HPV risk groups but not the direction of the associations of cofactors studied in this paper (data not shown).

This study has several limitations. The study population consisted of women receiving care at KPNC, and typically have high rates of screening and follow-up of abnormal results. Previous literature showed that screened populations in other settings have similar patterns of risk stratification when considering cytology and HPV test results; unscreened populations have higher risks.22 Past history of abnormal screening is important as noted in the guidelines and as further analyzed in separate articles focused on previous screening and history of high-grade results.23 HPV vaccination is also relevant in lowering the risk of CIN3+ and will need to be considered in triage strategies as individuals vaccinated prior to the onset of sexual activity age into cervical cancer screening [24]. In addition, age of first sexual intercourse and number of partners have also been associated with cervical cancer. These were not assessed as they often are not reliably elicited or documented during medical care. We acknowledge that this analysis included few women with a history of smoking, long-term hormonal contraceptive use (>10 years), high parity (>4 births), and sexually transmitted infections, which limits our ability to draw meaningful conclusions related to these factors. 8-10,13,19,24-26 Because smoking is associated not only with cervical cancer, but with a variety of other negative health consequences, assessing for smoking and encouraging smoking cessation is warranted. Variables related to medications (OCP and DMPA) were based on prescriptions issued at KPNC, regardless of whether or when women took them, and do not include prescriptions obtained in other health systems. We also did not assess Body Mass Index (BMI) as a cofactor in this analysis, as a previous large study within the full KPNC population indicated a lower risk of precancer, but higher risk of cancer among overweight and obese women compared to those with normal BMI, likely due to underdiagnosis of precancer.27 Of note, HPV typing in the 2019 guidelines is limited to HPV 16 and 18. These results indicate that further risk stratification by HPV risk group may be helpful, and the inclusion of additional types will be addressed in future guideline revision.

The 2019 ASCCP Risk-Based Management Consensus Guidelines currently use risk estimation that include current HPV and cytology test results, as well as past history.1 This study aimed to assess whether certain sociodemographic cofactors, including age, race/ethnicity, smoking status, and oral contraceptive use meaningfully changed risk estimates that would affect clinical management based on the clinical action thresholds of the 2019 ASCCP Risk-Based Management Consensus Guidelines. Our results suggest that including these variables would add complexity without additional risk discrimination. Therefore, these results support continued use of the current risk estimates based on screening history.28

Acknowledgements

The field effort was a collaboration of the NCI and KPNC and was supported in part by the intramural program of the NCI. The NCI has received HPV and cytology test results at no cost from Roche Molecular Systems, Qiagen, and BD Diagnostics for independent evaluations of these technologies.

Abbreviations used

ASCCP

American Society for Colposcopy and Cervical Pathology

ASC-US

Atypical squamous cells of undetermined significance

ASC-H+

Atypical squamous cells of undetermined significance - cannot exclude HSIL

CIN

Cervical intraepithelial neoplasia

DMPA

Depot medroxyprogesterone acetate

HPV

Human papillomavirus

HR

Hazard ratio

KPNC

Kaiser Permanente Northern California

LSIL

Low-grade squamous intraepithelial lesion

NILM

Negative for intraepithelial lesion or malignancy

NH

Non-Hispanic

OR

Odds ratio

PaP

HPV Persistence and Progression Study

STM

Specimen transport medium

Footnotes

Conflict of interest statement:

The National Cancer Institute (incl. all NCI-affiliated authors) receives cervical screening results at reduced or no cost from commercial research partners (Qiagen, Roche, BD, MobileODT, Arbor Vita) for independent evaluations of screening methods and strategies, Dr. Moscicki is a Merck and GSK, Advisory Board member. Dr. Castle has received HPV tests and assays at a reduced or no cost from Roche, Becton Dickinson, Arbor Vita Corporation, and Cepheid for research. Dr. Hammer has received reagents for free from Roche, Denmark for an unrelated study. The remaining authors have no conflicts of interest to disclose.

Disclaimer: The conclusions, findings, and opinions expressed by authors contributing to this journal do not necessarily reflect the official position of the National Cancer Institute.

Ethics: This study was approved by institutional review boards at the National Cancer Institute and Kaiser Permanente Northern California.

References

  • 1.Perkins RB, Guido RS, Castle PE, et al. 2019 ASCCP Risk-Based Management Consensus Guidelines for Abnormal Cervical Cancer Screening Tests and Cancer Precursors. J Low Genit Tract Dis. 2020;24(2):102–131. doi: 10.1097/LGT.0000000000000525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Saslow D, et al. , American Cancer Society, American Society for Colposcopy and Cervical Pathology, and American Society for Clinical Pathology screening guidelines for the prevention and early detection of cervical cancer. CA Cancer J Clin, 2012. 62(3): p. 147–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Schiffman M, et al. , A cohort study of cervical screening using partial HPV typing and cytology triage. Int J Cancer, 2016. 139(11): p. 2606–15. [DOI] [PubMed] [Google Scholar]
  • 4.Maucort-Boulch D, et al. , Predictors of human papillomavirus persistence among women with equivocal or mildly abnormal cytology. Int J Cancer, 2010. 126(3): p. 684–91. [DOI] [PubMed] [Google Scholar]
  • 5.Xi LF, et al. , Relationship between cigarette smoking and human papilloma virus types 16 and 18 DNA load. Cancer Epidemiol Biomarkers Prev, 2009. 18(12): p. 3490–6 years. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Pratt MM, et al. , Polycyclic aromatic hydrocarbon-DNA adducts in cervix of women infected with carcinogenic human papillomavirus types: an immunohistochemistry study. Mutat Res, 2007. 624(1-2): p. 114–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Scott ME, et al. , Covariates of cervical cytokine mRNA expression by real-time PCR in adolescents and young women: effects of Chlamydia trachomatis infection, hormonal contraception, and smoking. J Clin Immunol, 2006. 26(3): p. 222–32. [DOI] [PubMed] [Google Scholar]
  • 8.Appleby P, et al. , Carcinoma of the cervix and tobacco smoking: collaborative reanalysis of individual data on 13,541 women with carcinoma of the cervix and 23,017 women without carcinoma of the cervix from 23 epidemiological studies. Int J Cancer, 2006. 118(6): p. 1481–95. [DOI] [PubMed] [Google Scholar]
  • 9.McIntyre-Seltman K, et al. , Smoking is a risk factor for cervical intraepithelial neoplasia grade 3 among oncogenic human papillomavirus DNA-positive women with equivocal or mildly abnormal cytology. Cancer Epidemiol Biomarkers Prev, 2005. 14(5): p. 1165–70. [DOI] [PubMed] [Google Scholar]
  • 10.Castle PE, Beyond human papillomavirus: the cervix, exogenous secondary factors, and the development of cervical precancer and cancer. J Low Genit Tract Dis, 2004. 8(3): p. 224–30. [DOI] [PubMed] [Google Scholar]
  • 11.Smith JS, et al. , Cervical cancer and use of hormonal contraceptives: a systematic review. Lancet, 2003. 361(9364): p. 1159–67. [DOI] [PubMed] [Google Scholar]
  • 12.Ho GY, et al. , Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med, 1998. 338(7): p. 423–8. [DOI] [PubMed] [Google Scholar]
  • 13.Poppe WA, et al. , Tobacco smoking impairs the local immunosurveillance in the uterine cervix. An immunohistochemical study. Gynecol Obstet Invest, 1995. 39(1): p. 34–8. [DOI] [PubMed] [Google Scholar]
  • 14.Schiffman M, et al. , A long-term prospective study of type-specific human papillomavirus infection and risk of cervical neoplasia among 20,000 women in the Portland Kaiser Cohort Study. Cancer Epidemiol Biomarkers Prev, 2011. 20(7): p. 1398–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Castle PE, et al. , Human papillomavirus (HPV) genotypes in women with cervical precancer and cancer at Kaiser Permanente Northern California. Cancer Epidemiol Biomarkers Prev, 2011. 20(5): p. 946–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Demarco M, et al. , A study of type-specific HPV natural history and implications for contemporary cervical cancer screening programs. EClinicalMedicine, 2020. 22: p. 100293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Stoler MH, et al. , The accuracy of colposcopic biopsy: analyses from the placebo arm of the Gardasil clinical trials. Int J Cancer, 2011. 128(6): p. 1354–62. [DOI] [PubMed] [Google Scholar]
  • 18.Massad LS, et al. , 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors. Obstet Gynecol, 2013. 121(4): p. 829–46. [DOI] [PubMed] [Google Scholar]
  • 19.Castle PE, et al. , Hormonal contraceptive use, pregnancy and parity, and the risk of cervical intraepithelial neoplasia 3 among oncogenic HPV DNA-positive women with equivocal or mildly abnormal cytology. Int J Cancer, 2005. 117(6): p. 1007–12. [DOI] [PubMed] [Google Scholar]
  • 20.Chung SH, Franceschi S, and Lambert PF, Estrogen and ERalpha: culprits in cervical cancer? Trends Endocrinol Metab, 2010. 21(8): p. 504–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Castle PE, How does tobacco smoke contribute to cervical carcinogenesis? J Virol, 2008. 82(12): p. 6084–5; author reply 6085-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Saraiya M, Cheung LC, Soman A, Mix J, Kenney K, Chen X, Perkins RB, Schiffman M, Wentzensen N, Miller J. Risk of cervical precancer and cancer among uninsured and underserved women from 2009 to 2017. Am J Obstet Gynecol. 2021. Apr;224(4):366.e1–366.e32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Castle PE, et al. , Role of Screening History in Clinical Meaning and Optimal Management of Positive Cervical Screening Results. J Natl Cancer Inst, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Castle PE, et al. , Impact of human papillomavirus vaccination on the clinical meaning of cervical screening results. Prev Med, 2018. 118: p. 44–50. [DOI] [PubMed] [Google Scholar]
  • 25.Hildesheim A, et al. , HPV cofactors related to the development of cervical cancer: results from a population-based study in Costa Rica. Br J Cancer, 2001. 84(9): p. 1219–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Robbins HA, et al. , Cervical cancer screening intervals and management for women living with HIV: a risk benchmarking approach. Aids, 2017. 31(7): p. 1035–1044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Clarke MA, et al. , Epidemiologic Evidence That Excess Body Weight Increases Risk of Cervical Cancer by Decreased Detection of Precancer. J Clin Oncol, 2018. 36(12): p. 1184–1191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Egemen DC, L.C.; Chen X; Demarco M; Kinney W; Poitras N; Befano B; Locke A; Perkins RB; Guido R; Wiser AL; Gage JC; Katki HA; Wentzensen N; Castle P; Schiffman M; Lorey T, Risk Estimates Supporting the 2019 ASCCP Risk-Based Management Consensus Guidelines. J Low Genit Tract Dis, 2020. 24: p. xx-xx. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES