Secondary Logo

Journal Logo

Original Research

Rate of and Risks for Regression of Cervical Intraepithelial Neoplasia 2 in Adolescents and Young Women

Moscicki, Anna-Barbara MD; Ma, Yifei MS; Wibbelsman, Charles MD; Darragh, Teresa M. MD; Powers, Adaleen NP; Farhat, Sepideh MS; Shiboski, Stephen PhD

Author Information
doi: 10.1097/AOG.0b013e3181fe777f
  • Free

Cervical intraepithelial neoplasia (CIN) and cervical cancers are caused by human papillomavirus (HPV). Cervical intraepithelial neoplasia 1 is a histologic diagnosis associated with benign viral replication and, in most cases, spontaneously regresses.1–3 Studies in adult women show regression rates of 70% to 80%, whereas in adolescents and young women, more than 90% show regression.1–4 Because of these high regression rates, it is recommended in the United States that clinicians manage conservatively with observation, rather than treat, CIN 1 in adolescents.5

In contrast, CIN 3 is considered a true precancer with the potential to progress to invasive cancer at the rate of 0.2% to 4% within 12 months.6 The biologic behavior of CIN 2 is more controversial. Many clinicians consider CIN 2 a precancerous lesion and, therefore, routinely treat these lesions.5 The annual regression rate of CIN 2 in adult women is estimated to range from 15% to 23%, with up to 55% regressing by 4 to 6 years.3,7,8 As with CIN 1, data in adolescents suggest that CIN 2 has a much higher likelihood of regression. In a chart review of 23 patients, Moore et al9 observed a regression rate of 65%, whereas 13% progressed to CIN 3. In a database review, Fuchs et al10 reported that 39% of adolescents with untreated CIN 2 showed regression to normal, with 92% showing CIN 1 or less after 3 years. Only 8% had CIN 2 persistence or progression. Certainly, rates of cervical cancer are low in adolescents and young women, supporting that progression of CIN 2 to cancer in this age group is extremely rare.11

The purpose of this article is to describe the natural history of CIN 2 in a prospective study of adolescents and young women aged 13 to 24 years, and to examine behavioral and biologic factors associated with regression and progression.


Females 13 to 24 years of age who had abnormal cervical cytologic screening showing atypical squamous cells of undetermined significance, low-grade squamous intraepithelial lesion, or high-grade squamous intra-epithelial lesion while attending one of the 12 participating clinics within Kaiser Permanente, Northern California, were eligible for recruitment. Details of the recruitment have been reported elsewhere.12 Exclusion criteria included previous treatment for CIN, immunosuppression, pregnancy, or planning to leave the area within 3 years. This study was approved by the Institutional Review Boards of the University of California, San Francisco, and Kaiser Permanente, Northern California. At the time this study was initiated in 2002, atypical squamous cells of undetermined significance/high-risk (HPV-positive), repeat atypical squamous cells of undetermined significance, low-grade squamous intraepithelial lesion, and high-grade squamous intraepithelial lesion were all immediately referred for colposcopy. Recruitment was completed in 2007. Eighty percent of those who were contacted agreed to participate. Data are unavailable for those who were never contacted or refused to participate. At baseline and each 4-month follow-up visit, charts were reviewed to verify all reported sexually transmitted infections and a face-to-face interview was conducted to obtain behavioral data. The following were obtained at each visit: a vaginal sample for bacterial vaginosis13 and wet mount for yeast and Trichomonas vaginalis, and cervical sample for cytology and HPV and for Neisseria gonorrhoeae and Chlamydia trachomatis at the annual visit or if symptomatic. Biopsy samples obtained at baseline were sent to each of the respective Kaiser Permanente, Northern California, site pathology laboratories and then released for a second review by a single pathologist (T. Darragh, University of California, San Francisco). Cervical intraepithelial neoplasia 3 diagnosis were confirmed by a third pathologist.12 Patients with CIN 2 diagnosed by either University of California, San Francisco, or Kaiser Permanente, Northern California, were allowed into the follow-up study. If CIN 3 was diagnosed at either site, the patient was excluded.12 All study cytology and histology samples during follow-up were sent to the centralized laboratory to be reviewed by a single pathologist (T. Darragh). During follow-up, patients underwent biopsy if the colposcopist was concerned about progression or if the cytology on any visit suggested progression. At the end of the study, all participants were asked to consent to a cervical biopsy, regardless of cytologic diagnosis or colposcopic impression. If no colposcopic abnormality was noted on this final visit, then the site of initial CIN 2 was targeted for biopsy. Not all patients underwent biopsy at the end because they either refused or did not attend the final visit.

Samples for cytology and HPV were immediately placed into liquid-based media (PreservCyt; Hologic, Marlborough, MA) and were sent to the University of California, San Francisco, laboratory, where 9 mL was removed for amplification and genotyping.12,14 All HPV testing used the Roche Linear Array Assay (Roche Molecular Diagnostics, Alameda, CA) testing for HPV types 6, 11, 16,18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73, 81, 82, 83, 84, and 89.

Only those patients with at least two visits (baseline and at least one follow-up visit) were considered for this analysis. Definition of progression was biopsy-proven CIN 3 at any visit after the baseline determination of CIN 2. Definition of regression was based on having three consecutive visits with negative cytology and negative biopsy on any of these visits, if available. If there was insufficient follow-up (ie, only one visit with negative cytology), then the analysis was censored at the last visit with abnormal cytology or histology. If the patient continued to have low-grade squamous intraepithelial lesion or high-grade squamous intraepithelial lesion on cytology or CIN 1 or CIN 2 on histology at end of follow-up, then she was considered to be a nonregressor (ie, persistent). Histologic diagnosis, if available, was used for the analysis over a negative cytologic diagnosis at any visit.

To compare the sociodemographics between participants and those with no follow-up data (nonparticipants), we used t tests for continuous variables and χ2 test or two-tailed Fisher exact test for categorical variables, when appropriate. Kaplan-Meier estimates of time to CIN 2 regression or progression were based on the date of CIN 2 detection and the first of three consecutive visits with normal cytology or the first CIN 3 diagnosis. Once an event was reached, the remaining visits were censored from the analysis. The comparisons between Kaplan-Meier curves were based on log-rank test. The independent variables included fixed and time-dependent covariates, and their effects on CIN regression or progression were estimated in Cox proportional hazards regression models. When variables of similar characteristics were significant in univariable models (ie, HPV persistence), we entered each of them in a separate multivariable model to avoid potential colinearity problems. For the CIN 2 regression analysis, the variables that were significant at P≤.05 in univariable models were considered in the multivariable regression models. No significant violations of the proportional hazards assumption were detected for fitted regression models. We also performed a sensitivity analysis for those CIN 2 cases confirmed by a second pathologist. For the analysis for progression to CIN 3, the multivariable analysis was severely limited because of the small number of patients whose CIN progressed. We focused on the most biologically plausible variable, HPV status. We performed a regression analysis for each of the four HPV variables (HPV persistence of any type, HPV 16/18 persistence, HPV 16/18 status at entry, and HPV 16/18 status at last visit). Each analysis adjusted for the other variables found to have P<.1 in the univariable analysis.


Of the 715 patients screened, 120 met criteria for CIN 2 follow-up at the baseline visit. Twenty-five patients had no follow-up visit because they either verbally refused to enter the CIN 2 follow-up study or did not return for follow-up. Consequently, 95 agreed to participate and had at least two visits (baseline and first follow-up). The demographics of these 95 patients are given in Table 1. There were no statistical differences for these demographic characteristics between those who participated and those who did not return (Table 1). Forty-eight patients had CIN 2 diagnosed by University of California, San Francisco, and Kaiser Permanente, Northern California, pathologists. Forty-two had discordant diagnosis (ie, one of the diagnosis was less than CIN 2). Sixteen had CIN 2 determined by Kaiser Permanente, Northern California, and 26 had CIN 2 determined by University of California, San Francisco. Five patients with CIN 2 were evaluated only by University of California, San Francisco. Mean period of observation was 27.4 months (SD, 11.6), with a range of 3.8 to 46.8 months. Sixty-five (68%) patients had an exit biopsy. Median number of biopsies during follow-up was one (range, 0–6).

Table 1
Table 1:
Demographics and Behavioral Characteristics of the Population at Baseline

Time to clearance for the 95 patients with CIN 2 is shown in Figure 1. Thirty-eight percent (95% confidence interval [CI] 29%–49%) of patients with CIN 2 had regression by 1 year, 63% (95% CI 53%–74%) had regression by 2 years, and 68% (95% CI 57%–78%) had regression by 3 years. Time to clearance by HPV 16/18 baseline status is shown in Figure 2. Of those with HPV 16/18 (n=42), 31.6% (95% CI 20%–48%) showed clearance by 1 year, 44.1% (95% CI 30%–62%) showed clearance by 2 years, and 55.1% (95% CI 39%–72%) showed clearance by 3 years. In comparison, among individuals with non-HPV 16/18 CIN 2, 43.7% (95% CI 31%–59%) showed clearance at 1 year (P=.27), 78% (95% CI 65%–89%) showed clearance at 2 years (P=.01), and 78% (95% CI 65%–89%) showed clearance at 3 years (P=.03). No difference in time to clearance was found if the CIN 2 diagnosis was concordant between the Kaiser Permanente, Northern California, and University of California, San Francisco, pathologists compared with discordant diagnoses (Fig. 3; P=.75).

Fig. 1.Moscicki. Regression of CIN 2 in Young Females. Obstet Gynecol 2010.
Fig. 2.Moscicki. Regression of CIN 2 in Young Females. Obstet Gynecol 2010.
Fig. 3.Moscicki. Regression of CIN 2 in Young Females. Obstet Gynecol 2010.

Closer examination of the 42 patients with HPV 16/18 showed that 20 with HPV 16/18 immediately had clearance of their HPV 16 (ie, they had a single positive HPV 16 DNA test). Of these, 18 (90%; 95% CI 76.9%–100%) had clearance of their CIN 2 and 2 had CIN 2 persistence. These two patients also had infection with other HPV types: one had persistently positive results for HPV 52 and the other had positive results for HPV 18 and 31 at baseline, at second visit had HPV 18 and 61, at third visit had HPV 61, and at fourth and last visits had HPV 51. Of the 22 patients with HPV 16/18 persistence, defined by at least two positive HPV 16 tests over the observed period (persistence ranged from 4 months to 26 months), eight (36.4%; 95% CI 16.3%–56.5%) had progression to CIN 3, 11 (50%; 95% CI 29.1%–70.9%) had regression, and three (13.6%; 95% CI 2.9%–34.9%) had regression to CIN 1.

The appearance of new HPV types over the course of the observed period was extremely common, with 84 (88.4%; 95% CI 82%–94.9%) patients exhibiting new HPV types. In addition to the 15 patients who had progression or persistence, 17 (44.7%; 95% CI 28.6%–61.7%) were considered nonregressors because they had CIN 1 at the end. Of these 17 patients, eight (47.1%; 95% CI 23.3%–70.8%) had HPV persistence of the type found at the baseline CIN 2 visit, and nine (52.9%; 95% CI 29.2%–76.7%) had a new HPV type with clearance of the initial HPV type. Six patients were found to have only two consecutive visits with normal cytology before being lost to follow-up. Of these six, four (67%; 95% CI 22.3%–95.7%) had clearance of the HPV observed at the CIN 2 visit and two (33%; 95% CI 4.3%–77.7%) had persistence of that type.

Table 2 shows the univariable associations with CIN 2 regression. Factors associated with regression at P<.05 included older age at first intercourse, a recent reported infection with N. gonorrhoeae, and months using medroxyprogestereone acetate. Factors associated with nonregression included greater number of months of combined hormonal contraception use, HPV persistence, HPV 16/18 persistence, HPV 16/18 infection status at baseline, and HPV 16/18 infection status at last visit. Because the HPV variables were highly correlated, multivariable models were performed for each HPV variable separately. In the four multivariable models, all the variables remained significant except for HPV 16/18 status at entry, which became marginally significant. All the biologic and behavioral variables had similar hazard ratios, as presented in Table 3. The results are summarized in Table 3. The sensitivity analysis for the CIN 2 cases that were confirmed by a second pathologist showed similar results as presented in Tables 2 and 3 (data not shown).

Table 2
Table 2:
Univariable Analysis for Risk of Cervical Intraepithelial Neoplasia 2 Regression
Table 3
Table 3:
Multivariable Analysis for Risk of Cervical Intraepithelial Neoplasia 2 Regression

Time to progression to CIN 3 is shown in Figure 4. Two percent (95% CI 1%–9%) of patients showed progression to CIN 3 by year 1, 12% (95% CI 8%–22%) showed progression by year 2, and 15% (95% CI 9%–26%) showed progression by year 3. No cases of cervical cancer occurred during follow-up. Of the 11 patients with progression, eight were positive for HPV 16, one was positive for HPV 51, one was positive for HPV 31, and one was positive for HPV 51, 52, and 58 at the CIN 3 visit.

Fig. 4.Moscicki. Regression of CIN 2 in Young Females. Obstet Gynecol 2010.

In the univariable analysis, factors associated with progression to CIN 3 at P<.1 are summarized in Table 4. All the HPV variables (HPV persistence of any type, HPV persistence of HPV 16/18, HPV 16/18 status at entry, and HPV 16/18 status at last visit) were significantly associated with progression. In addition, young age of menarche, history of reported douching since the last visit, weekly alcohol use, engaging in anal sex, and having a history of genital warts were associated with progression. The small number of cases that progressed precluded us from performing any meaningful full multivariable model. Hence, we focused the analysis on the HPV variables. After adjusting for the five potentially confounding variables, HPV 16/18 persistence (hazard ratio 25.27; 95% CI 2.65–241.2; P=.005) and HPV 16/18 status at last visit (hazard ratio 7.25; 95% CI 1.07–49.36; P<.05) remained significant, whereas HPV persistence of any type (hazard ratio was not calculable; P=.99) and HPV 16/18 status at entry (hazard ratio 2.87; 95% CI 0.48–17.06; P=.25) were no longer significant. Of note, young age of menarche remained significantly associated (P≤.05) with progression in all four models. The lowest hazard ratio was for the model with HPV 16/18 persistence (hazard ratio 0.36; 95% CI 0.13–1.0; P =.05), and the highest hazard ratio was for the model with HPV persistence (hazard ratio 0.44; 95% CI 0.021–0.92; P=.03). All the other variables were either not significant or inconsistently significant.

Table 4
Table 4:
Univariable Analysis for Risk of Cervical Intraepithelial Neoplasia 2 Progression to Cervical Intraepithelial Neoplasia 3


In this prospective study of CIN 2 in adolescents and young women, regression of CIN 2 was common, with almost 70% having regression to normal within 3 years. To be conservative, we classified those with CIN 1 at the last visit as nonregressors. Because many of the CIN 1 lesions appeared to be associated with new HPV types, the regression rate of CIN 2 is likely even higher. However, we note that few showed regression after 2 years of follow-up. In the similar vein, progression to CIN 3 was not common in the 1 to 2 years after the CIN 2 diagnosis. These data support the 2006 American Society for Colposcopy and Cervical Pathology Consensus Guidelines recommendation for observation for up to 2 years for adolescents and young adult women with CIN 2.5 These rates of regression are higher than those reported in previous prospective studies and are likely attributable to the younger age of our cohort. Younger age likely reflects a shorter time of HPV persistence at entry into study. The association found in our study between older age of first intercourse and regression underscores this premise. Syrjanen et al3 reported a CIN 2 regression rate of 53% and a progression rate of 21%. Their study had several differences in that patients were older and not all study participants had confirmation of CIN 2 by biopsy. This group also reported regression and progression rates for those with CIN 2 confirmation and reported that 24 of 70 (34%) with CIN 2 had regression and 14 (20%) had progression.15 Another older study by Nasiell et al2 also showed a lower regression rate of 54%, with a progression rate of 30%. This study used cytology as the sole entry diagnosis and included older women, ranging from 15 to 72 years. The relevance of most of these studies to young women today is unclear because sexual behavior, contraceptive use, and smoking habits all have changed since most of these cohort studies were performed. Although not prospective studies, two more recent studies by Moore et al9 and Fuchs et al10 had estimates similar to ours.

Although the rate of regression was high, it was lower than that reported for CIN 14 and higher than that reported for CIN 3.6 This suggests that CIN 2 may have some distinct biologic characteristics, or the morphologic classification may not accurately reflect the biologic potential. Some have argued that CIN 2 as a lesion does not exist. Most agree that the reproducibility of a histologic diagnosis of CIN 2 is poor, as shown in our study.7,16 Interestingly, concordance of diagnosis did not influence the regression rate.

Not surprisingly, HPV persistence was a factor associated with nonregression, as found in many other studies.17 In comparison, HPV 16/18 status at entry was not significant, underscoring the high rate of HPV 16/18 clearance seen in most young women. These observations warrant further clinical studies that examine the potential use of HPV DNA testing in follow-up of CIN 2. The finding associated with hormonal contraception and CIN 2 persistence was also not surprising, because its use has been associated with the development of CIN 3 and invasive cervical cancer.18 We previously reported on risk factors from the baseline visit of this study associated with CIN 3.12 In that report, one of the risks was time using combined oral contraceptives.12 This finding may suggest that the mechanism by which oral contraceptives play a role is by assisting in the transcription of viral oncoproteins that result in the histologic changes of CIN 2.19 Interestingly, progesterone-only contraceptives seem to assist in clearance. This finding is incongruous with the literature that show medroxyprogesterone acetate and combined oral contraceptives are associated with cancer.20

Interpreting the observation that infections with N. gonorrhoeae assisted in regression is limited because of the few cases that occurred. We hypothesize that the intense inflammatory response induced by N. gonorrhoeae infections may have serendipitously assisted in viral clearance.

Because few participants in this study had progression to CIN 3, as expected in this young population, our findings regarding risks for progression have serious limitations. The wide CIs underscore the potential instability of the variables. As expected, we found that HPV 16/18 persistence was a risk for progression, with an almost 20-fold risk of progression over the observed period. Although there were wide CIs, the association remained relatively stable even after adjusting for numerous potential confounders. It is difficult to comment on the persistence of other HPV types because HPV 16/18 was predominant in this analysis. In comparison, single-point HPV 16/18 testing appeared less informative. Although we focused the analysis on the HPV variables, we observed that the association with young age of menarche appeared relatively stable in all the models. This is interesting because other gynecologic cancers, including ovarian and endometrial, have been associated with young age at menarche.21–23 Young age of menarche is thought to reflect a higher cumulative exposure to estrogen. This would be consistent with our findings associated with oral contraceptive use and CIN 2 persistence and oral contraceptive use and CIN 3 in the cross-sectional baseline analysis. The lack of finding an association with oral contraceptive use in this study may have been attributable to the small number of CIN 3 cases or that the influence associated with oral contraceptive use targets viral persistence and not mutational events that result in CIN 3.

In summary, our data show that CIN 2 commonly regresses spontaneously in adolescents and young women, supporting the conservative approach in observing young patients with CIN 2. Factors associated with CIN 2 regression and progression to CIN 3 were correlated with HPV persistence, specifically HPV 16/18 infections. The findings with combined hormonal contraceptive use and age at menarche support the premise that reproductive hormones are important influences on persistence and progression.


1. 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.
2. Nassiel K, Nassiel M, Vaclavinkova V. Behavior of moderate cervical dysplasia during long term follow-up. PMCID: PMC2735396. 1983;61:609–14.
3. Syrjanen K, Kataja V, Yliskoski M, Chang F, Syrjanen S. Natural history of cervical human papillomavirus lesions does not substantiate the biologic relevance of the Bethesda System. PMCID: PMC2735396. 1992;79(5 Pt 1):675–82.
4. Moscicki AB, Shiboski S, Hills NK, Powell KJ, Jay N, Hanson EN, et al. Regression of low-grade squamous intra-epithelial lesions in young women. Lancet 2004;364:1678–83.
5. 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. Am J Obstet Gynecol 2007;197:340–5.
6. Goldie SJ, Kohli M, Grima D, Weinstein MC, Wright TC, Bosch FX, et al. Projected clinical benefits and cost-effectiveness of a human papillomavirus 16/18 vaccine. J Natl Cancer Inst 2004;96:604–15.
7. Castle PE, Schiffman M, Wheeler CM, Solomon D. Evidence for frequent regression of cervical intraepithelial neoplasia-grade 2. PMCID: PMC2735396. 2009;113:18–25.
8. Insinga RP, Dasbach EJ, Elbasha EH. Epidemiologic natural history and clinical management of human papillomavirus (HPV) disease: a critical and systematic review of the literature in the development of an HPV dynamic transmission model. BMC Infect Dis 2009;9:119.
9. Moore K, Cofer A, Elliot L, Lanneau G, Walker J, Gold MA. Adolescent cervical dysplasia: histologic evaluation, treatment, and outcomes. Am J Obstet Gynecol 2007;197:141 e1–6.
10. Fuchs K, Weitzen S, Wu L, Phipps MG, Boardman LA. Management of cervical intraepithelial neoplasia 2 in adolescent and young women. J Pediatr Adolesc Gynecol 2007;20:269–74.
11. Ries LAG, Melbert D, Krapcho M, Mariotto A, Miller BA, Feuer EJ, et al. SEER Cancer Statistics Review, 1975–2004. Bethesda (MD): National Cancer Institute; 2007.
12. Moscicki AB, Ma Y, Wibbelsman C, Powers A, Darragh TM, Farhat S, et al. Risks for cervical intraepithelial neoplasia 3 among adolescents and young women with abnormal cytology. PMCID: PMC2735396. 2008;112:1335–42.
13. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991;29:297–301.
14. Moscicki AB, Widdice L, Ma Y, Farhat S, Miller-Benningfield S, Jonte J, et al. Comparison of natural histories of human papillomavirus (HPV) detected by clinician- and self- sampling. Int J Cancer 2010;127:1882–92.
15. Kataja V, Syrjanen K, Mantyjarvi R, Vayrynen M, Syrjanen S, Saarikoski S, et al. Prospective follow-up of cervical HPV infections: life table analysis of histopathological, cytological and colposcopic data. Eur J Epidemiol 1989;5:1–7.
16. Dalla Palma P, Giorgi Rossi P, Collina G, Buccoliero AM, Ghiringhello B, Gilioli E, et al. The reproducibility of CIN diagnoses among different pathologists: data from histology reviews from a multicenter randomized study. Am J Clin Pathol 2009;132:125–32.
17. Koshiol J, Lindsay L, Pimenta JM, Poole C, Jenkins D, Smith JS. Persistent human papillomavirus infection and cervical neoplasia: a systematic review and meta-analysis. Am J Epidemiol 2008;168:123–37.
18. Appleby P, Beral V, Berrington de Gonzalez A, Colin D, Franceschi S, Goodhill A, et al. Cervical cancer and hormonal contraceptives: collaborative reanalysis of individual data for 16,573 women with cervical cancer and 35,509 women without cervical cancer from 24 epidemiological studies. Lancet 2007;370:1609–21.
19. Ruutu M, Wahlroos N, Syrjänen K, Johansson B, Syrjänen S. Effects of 17beta-estradiol and progesterone on transcription of human papillomavirus 16 E6/E7 oncogenes in CaSki and SiHa cell lines. Int J Gynecol Cancer 2006;16:1261–8.
20. Moodley M, Moodley J, Chetty R, Herrington CS. The role of steroid contraceptive hormones in the pathogenesis of invasive cervical cancer: a review. Int J Gynecol Cancer 2003;13:103–10.
21. Dossus L, Allen N, Kaaks R, Bakken K, Lund E, Tjonneland A, et al. Reproductive risk factors and endometrial cancer: the European Prospective Investigation into Cancer and Nutrition. Int J Cancer 2010;127:442–51.
22. Fujita M, Tase T, Kakugawa Y, Hoshi S, Nishino Y, Nagase S, et al. Smoking, earlier menarche and low parity as independent risk factors for gynecologic cancers in Japanese: a case-control study. Tohoku J Exp Med 2008;216:297–307.
23. Moorman PG, Palmieri RT, Akushevich L, Berchuck A, Schildkraut JM. Ovarian cancer risk factors in African-American and white women. Am J Epidemiol 2009;170:598–606.
© 2010 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.