Maternal Human Papillomavirus and Preterm Premature Rupture of Membranes: A Retrospective Cohort Study

Abstract

Introduction:

To determine whether maternal human papillomavirus (HPV) infection is associated with preterm premature rupture of membranes (PPROM).

Materials and Methods:

We conducted a retrospective cohort study of singleton deliveries at our institution from 2010 to 2015. Women, ages 16–49, with HPV genotyping or cervical cytology results 3 years before delivery were included. Chi-squared and logistic regression analyses were used.

Results:

In our cohort of 2153 women, 38.5% were HPV positive. PPROM was observed in 2.88% of women. HPV infection (p = 0.02), history of PPROM (p < 0.001), history of cervical conization or loop electrical excision procedure (LEEP) (p < 0.05), parity (p = 0.001), maternal body mass index at delivery (p < 0.001), drug use or smoking (p < 0.001), and ethnicity (p = 0.01) were associated with PPROM. HPV infection (odds ratio [OR] = 2.07, 95% confidence interval [CI]: 1.03–4.14) remained associated with PPROM when adjusting for history of PPROM, cervical conization, drug use or smoking, parity, ethnicity, and insurance. PPROM was associated with preterm delivery (OR = 105.50, 95% CI: 29.49–377.46) when adjusting for HPV infection, pregnancy-related hypertension, diabetes, placenta previa and abruption, cervical conization, smoking or drug use, ethnicity, and history of PPROM. HPV infection was associated with preterm delivery (p = 0.04) in univariate analysis, but not after adjusting for PPROM (p = 0.13). HPV infection had a univariate association with newborn septicemia (p = 0.02), respiratory distress syndrome (RDS) (p = 0.01), neonatal intensive care unit (NICU) admission (p = 0.001), and low birthweight (p = 0.03).

Conclusions:

HPV infection was associated with an increased risk of PPROM in this cohort. However, maternal HPV infection does not increase the risk of preterm delivery beyond those caused by PPROM. The observed association between maternal HPV infection and neonatal morbidity is likely due to the relationship between PPROM and preterm delivery.

Introduction

In the United States, human papillomavirus (HPV) is the primary cause of cervical dysplasia and the most common sexually transmitted infection, with 39.9% of women infected with any genital HPV.¹ HPV is a double-stranded DNA virus with more than 100 genotypes. Maternal HPV infection in the prepregnancy and early pregnancy period has been associated with adverse pregnancy outcomes, such as pregnancy-related hypertension, preterm delivery, and spontaneous abortion.^2

–6 Limited work also suggests that maternal infection with high-risk HPV may also increase the risk of premature rupture of membranes (PROM).^7,8

PROM affects ∼8% of term pregnancies and 3% of preterm pregnancies in the United States.^9
–11 Preterm premature rupture of membranes (PPROM) is associated with an increased risk of infection^12,13 and a myriad of adverse neonatal outcomes associated with prematurity.^14
–16 Among pregnancies complicated by PPROM, 50%–60% result in delivery within 1 week.¹⁷ A 2011 cross-sectional study (n = 311) conducted in Korea found that women with high-risk HPV, measured at 6 weeks postpartum, had a higher rate of any PROM.⁷ Similarly, a 2016 retrospective cohort study in the United States reported a univariate association between maternal HPV infection and any PROM, but did not conduct a full multivariable analysis.⁸

The primary aim of this study was to determine whether maternal prepregnancy infection with high-risk HPV, based on cervical cytology or HPV genotyping, increases the risk of PPROM. Secondarily, our study aimed to determine whether maternal infection with HPV increased the risk of preterm delivery due to other mechanisms beyond PPROM.

Materials and Methods

We conducted a retrospective cohort study of women with singleton deliveries that occurred at our institution from 2010 to 2015. Ethical approval was obtained from the Institutional Review Board for Health Sciences Research at the University of Virginia with a waiver of consent. For women with multiple deliveries between 2010 and 2015, only the most recent delivery was included. Those with a diagnosis of current or prior cervical cancer were excluded. Women with a history of treatment for cervical dysplasia, including cryotherapy, loop electrosurgical excision procedure, and cold knife conization, were not excluded, however, these procedures were adjusted for, as necessary, during statistical analysis. Women, ages 16–49, were included if cervical cytology and/or HPV genotyping results were available within 3 years before delivery.

The primary exposure of interest in this study was prepregnancy maternal HPV infection. HPV infection was based on the patient's most recent cervical cytology and/or HPV genotyping results. Both were used to diagnose HPV infection because it is not routine at our institution to conduct HPV genotyping for all women with abnormal Pap smears. Laboratory reports for cervical cytology and/or HPV genotyping were obtained from electronic medical records by a database administrator blinded to the study's hypothesis and objective. Cervical cytology and genotyping samples were obtained using a ThinPrep liquid-based Pap test. Women were considered HPV positive if their most recent HPV genotyping results were positive for high-risk types (types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68). For women without HPV genotyping results, women were considered positive if their most recent two consecutive cervical cytology results had high-grade or low-grade squamous intraepithelial lesions (HSIL and LSIL), atypical squamous cells cannot rule out high-grade lesions (ASC-H), or atypical squamous cells of undetermined significance (ASCUS). However, no women with two consecutive cervical cytology results with ASCUS lacked HPV genotyping. Remaining patients were considered HPV negative if their last two consecutive cervical cytology results were negative for HSIL, LSIL, ASC-H, and ASCUS. Women were also considered negative if their most recent HPV genotyping results were negative combined with at least one negative cervical cytology result. This diagnostic criteria was intended to optimize sensitivity, which was reported to be 94.6% for HPV DNA testing and 55.4% for single cervical cytology testing.¹⁸ In addition, the American Society for Colposcopy and Cervical Pathology's most recent screening guidelines for the prevention and early detection of cervical cancer included cervical cytology testing every 3 years in low-risk women, ages 21–65.¹⁹

The primary outcome was PPROM, which was confirmed by a chart review. Mode of delivery and gestational age at the time of delivery were abstracted from electronic medical records by a database administrator. Other secondary pregnancy outcomes were based on an International Classification of Diseases, Version 9 (ICD-9) diagnosis and included endometritis, chorioamnionitis, placental abruption, placenta previa, and postpartum hemorrhage. All neonatal outcomes were based on an ICD-9 diagnosis.

Maternal sociodemographic and insurance information, as well as smoking and drug use during pregnancy, was self-reported during regular prenatal visits and later abstracted from electronic medical records. Maternal body mass index (BMI) was calculated based on prenatal height and weight at the time of admission for delivery and categorized into normal/underweight (BMI <25.0), overweight (BMI 25–29.9), obese (BMI 30–39.9), and morbidly obese (BMI ≥40).

All potential risk factors assessed in our study, including maternal demographic information, concurrent genital infections, and maternal medical and obstetric history, were compared between women with and without PPROM using chi-squared tests. Potential risk factors that displayed a significant association with both the primary outcome and maternal HPV infection were considered potential confounders and adjusted for in multivariable analysis. Multivariable logistic regression was used to adjust for confounding factors. Multicollinearity was assessed during intermediate multivariable models. The association between secondary neonatal outcomes as well as secondary pregnancy outcomes and maternal HPV infection was assessed using chi-squared tests. Multivariable logistic regression was used to assess the association between preterm delivery and maternal HPV infection. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were reported for all logistic regression analyses. An estimated 2478 participants were needed based on a sample size calculation assuming a 4.0% and 2.0% prevalence of the primary outcome (PPROM) in HPV-positive and HPV-negative women, respectively, an alpha of 0.05, and beta of 0.20. Preterm delivery was a secondary outcome of our study that would have required an estimated sample size of 1864 based on a prevalence of 12.0% and 8.0% in women who are HPV-positive and HPV-negative, respectively. All statistical analyses were conducted using STATA 14.1 (College Station, TX).

Results

Of the 9471 patients who had a singleton delivery from 2010 to 2015, 38 were excluded due to current or prior cervical cancer, and 41 were excluded due to age <16 or >49 years. Of the remaining women in our cohort, 2153 women had available cervical cytology and/or HPV genotyping results within 3 years before delivery. In this final cohort, 1324 (61.5%) and 829 (39.5%) were HPV negative and HPV positive, respectively. Among the 1324 HPV-negative women, 486 (36.71%) women were negative based on HPV genotyping and the remaining 838 (63.29%) diagnoses were based on cervical cytology. Among the 829 HPV-positive women, 126 (15.20%) women were considered positive based on HPV genotyping and the remaining 703 (84.80%) women were based on cervical cytology results. PPROM was observed in 2.88% (n = 269) of women.

As displayed in Table 1, the majority of women were white, non-Hispanic, overweight or obese at delivery, multiparous, and had Medicaid. The mean maternal age was 28.8 years. The overall prevalence of smoking or drug use during pregnancy was 11.2%. The percentage of women with a history of PPROM was 2.31% (n = 48), and 16.41% (n = 348) of women had a history of cervical conization and/or loop electrical excision procedure (LEEP).

Table 1.

Demographics

Variable	(−) PPROM	(+) PPROM	P
Maternal age (%)			0.14
<25	495 (24.34)	11 (26.83)	506 (24.39)
25 to <30	635 (31.22)	10 (24.39)	645 (31.08)
30 to <35	615 (30.24)	13 (31.71)	628 (30.27)
35 to <40	234 (11.50)	5 (12.20)	239 (11.52)
≥40	55 (2.70)	2 (4.88)	57 (2.75)
Race and ethnicity (%)			0.01
White, non-Hispanic	1105 (53.13)	20 (48.78)	1125 (53.04)
Black, non-Hispanic	483 (23.22)	14 (34.15)	497 (23.43)
Hispanic	288 (13.85)	1 (2.44)	289 (13.63)
Asian	64 (3.08)	2 (4.88)	66 (3.11)
Other	140 (6.73)	4 (9.76)	144 (6.79)
Maternal BMI at delivery (%)			<0.001
Normal/under (<25)	128 (13.01)	4 (14.29)	132 (13.04)
Overweight (25 to <30)	280 (28.46)	6 (21.43)	286 (28.26)
Obese (30 to <40)	266 (27.03)	9 (32.14)	275 (27.17)
Morbidly obese (≥40)	310 (31.50)	9 (31.14)	319 (31.52)
Parity (%)			0.001
Multiparous	1415 (73.97)	31 (77.50)	1446 (74.04)
Nulliparous	498 (26.03)	9 (22.50)	507 (25.96)
Insurance type (%)			0.25
Private	658 (39.21)	15 (37.50)	673 (39.17)
Medicaid	924 (55.07)	21 (52.50)	945 (55.01)
Self-pay	18 (1.07)	1 (2.50)	19 (1.11)
M-Care	78 (4.65)	3 (7.50)	81 (4.71)
HPV status (%)			0.02
Negative	1290 (62.02)	18 (43.90)	1308 (61.67)
Positive	790 (37.98)	23 (56.10)	813 (38.33)
History of PROM (%)			<0.001
No	2002 (98.28)	28 (68.29)	2030 (97.69)
Yes	35 (1.72)	13 (31.71)	48 (2.31)
History of preterm labor (%)			0.58
No	2023 (99.31)	40 (97.56)	2063 (99.28)
Yes	14 (0.69)	1 (2.44)	15 (0.72)
History of cervical conization/LEEP (%)			<0.05
No	1741 (83.70)	32 (78.05)	1773 (83.59)
Yes	339 (16.30)	9 (21.95)	348 (16.41)
Other concurrent genital or sexually transmitted infections (%)			0.84
Negative	1708 (82.12)	33 (80.49)	1741 (82.08)
Positive	372 (17.88)	8 (19.51)	380 (17.92)
Concurrent bacterial vaginosis infection (%)			0.06
Negative	2073 (99.66)	40 (97.56)	2113 (99.62)
Positive	7 (0.34)	1 (2.44)	8 (0.38)
Smoking or drug use during pregnancy (%)			<0.001
No	1853 (89.09)	31 (75.61)	1884 (88.83)
Yes	227 (10.91)	10 (24.39)	237 (11.17)
Pregnancy-related hypertension (%)			0.06
No	1559 (92.41)	128 (99.22)	1687 (96.18)
Yes	66 (4.06)	1 (0.78)	67 (3.82)

BMI, body mass index; HPV, human papillomavirus; LEEP, loop electrical excision procedure; PPROM, preterm premature rupture of membranes.

Table 1 displays the results of our univariate analysis. HPV-positive women had a higher proportion of PPROM (p = 0.02). Maternal BMI (p < 0.001), maternal history of PPROM (p < 0.001), maternal history of cervical conization and/or LEEP (p < 0.05), smoking or drug use during pregnancy (p < 0.001), parity (p = 0.001), and maternal ethnicity (p = 0.05) were also associated with PPROM during univariate analysis. Insurance type (p = 0.25) was not associated with PPROM during univariate analysis but was included in multivariable analysis a priori as a marker of socioeconomic status. History of preterm labor, other concurrent genital and sexually transmitted infections, pregnancy-induced hypertension, and maternal age were not associated with PPROM (p > 0.05). We did not assess whether pregnancy-induced hypertension was associated with HPV infection in multivariable analysis.

Table 2 shows the results from multivariable analysis. Maternal HPV-positive status remained associated with PPROM after adjusting for drug use or smoking during pregnancy, history of PPROM, history of cervical conization, maternal ethnicity, parity, and insurance status. HPV-positive status increased the odds of PPROM by a factor of 2.07 (95% CI: 1.03–4.14).

Table 2.

Maternal Human Papillomavirus Infection and Preterm Premature Rupture of Membranes: Multivariable Logistic Regression

Variable	OR	95% CI	p
Positive HPV status	2.07	1.03–4.14	0.04
Drug use or smoking during pregnancy	1.50	0.62–3.64	0.37
History of PROM	24.91	10.66–58.20	<0.001
History of cervical conization	0.70	0.28–1.77	0.45
Nulliparous	1.34	0.58–3.08	0.50
Ethnicity			0.09
Black, non-Hispanic	1.50	0.70–3.19
Hispanic	0.25	0.03–1.98
Asian	3.49	0.75–16.15
Other	2.45	0.71–8.51
Insurance status			0.79
Medicaid	0.99	0.48–2.03
Medicare	1.11	0.11–11.71
Self-pay	2.06	0.52–8.06

CI, confidence interval; OR, odds ratio.

As displayed in Table 3, HPV-positive status was associated with preterm delivery in univariate analysis (p = 0.04). However, after adjusting for PPROM, pregnancy-induced hypertension, drug use or smoking during pregnancy, maternal ethnicity, and insurance status, no association was observed between HPV positive and preterm delivery beyond other factors included in the model (OR = 1.35, 95% CI: 0.89–2.04).

Table 3.

Association Between Maternal Human Papillomavirus Infection and Preterm Delivery: Univariate and Multivariable Logistic Regression

Variable	aOR	95% CI	p
HPV positive status^a	1.35	0.89–2.04	0.16
PPROM^a	102.05	29.46–353.48	<0.001
Pregnancy-related hypertension^a	5.76	2.62–12.67	<0.001
Drug/smoking during pregnancy^a	2.01	1.18–3.43	0.01
Ethnicity^a			0.45
White, non-Hispanic	Base	—	—
Black, non-Hispanic	0.98	0.59–1.61	0.93
Hispanic	0.63	0.31–1.29	0.21
Asian	0.91	0.28–2.95	0.87
Other	0.44	0.14–1.41	0.17
Insurance status^a			0.17
Private	Base	—	—
Medicaid	1.48	0.95–2.30	0.09
Medicare	2.9	0.61–13.66	0.18
Self-pay	2.13	0.78–5.84	0.14

aOR adjusted for HPV-positive status, PPROM, pregnancy-related hypertension, smoking or drug use during pregnancy, ethnicity, and insurance status.

aOR, adjusted odds ratio.

Maternal HPV-positive status was not associated with other adverse maternal outcomes in univariate analysis, including endometritis, chorioamnionitis, placental abruption, placenta previa, and postpartum hemorrhage (p > 0.05). Maternal HPV-positive status was associated with several adverse neonatal outcomes during univariate analysis, including septicemia, neonatal intensive care unit (NICU) admission in the first 28 days of life, respiratory distress syndrome, and low birthweight (p < 0.01). However, maternal HPV-positive status was not associated with oligohydramnios or polyhydramnios, neonatal bacteremia, necrotizing enterocolitis, nongenetic fetal abnormalities, stillbirth, and neonatal APGAR score <7 (p > 0.05). We did not conduct a multivariable analysis for other maternal or neonatal outcomes.

Discussion

Our study indicates that prepregnancy maternal infection with HPV was associated with an increased risk of PPROM in this cohort. HPV does not appear to have any additional effect on preterm delivery outside of preterm deliveries that result from PPROM. The observed association with PPROM is consistent with multivariable cross-sectional and univariate cohort studies (without multivariable results) reported in prior work in both Korea and the United States.^7,8 The increased risk for PPROM may result from differences in the rate of infection and inflammation among women at earlier gestational ages. Prior work indicates that gestational age may play an important role in PPROM due to choriodecidual infection and inflammation.²⁰

The pathophysiology of PROM is multifactorial.¹⁷ Although the mechanism through which HPV causes PROM is not clearly identified, one study reported a higher rate of HPV in the extravillous trophoblast region of placentas in women with preterm delivery.⁴ The same study reported that HPV leads to a higher rate of extravillous trophoblast apoptosis and inhibited placental invasion.⁴ A pathological study of the placentas of 268 women found that pregnancies complicated with PPROM had significantly more markers of acute inflammation compared with women with term labor, term PROM, and term cesarean delivery.²⁰ Intra-amniotic bacterial infection has been associated with PPROM.¹² Although PPROM increases the risk of infection, determining whether markers of infection and inflammation following pathological examination solely result from PPROM or precede PPROM is difficult. Nonetheless, these prior results suggest that choriodecidual infection with HPV occurs and could explain the resulting increased risk of PPROM.

Maternal HPV-positive status did not appear to increase the risk of preterm delivery outside of preterm deliveries caused by PPROM. As described earlier, two studies reported that HPV infection increases the risk of preterm delivery.^3,4 Although a 2014 study reported no association between abnormal cervical screening and any preterm delivery, the study also reported that an association between abnormal cervical screening and spontaneous preterm delivery, specifically, was observed.¹³ Our results suggest that the observed association between HPV infection and preterm delivery may be due to spontaneous PPROM, which is consistent with results from the 2014 study. Results from this study also expand on prior studies by suggesting that maternal HPV infection may be an upstream cause of preterm delivery that acts through a variety of downstream causes, including PPROM.

Maternal HPV-positive status may also increase the risk of other adverse neonatal outcomes, including septicemia, respiratory distress syndrome, low birthweight, and NICU admission, based on the results of our study. Prematurity is an established risk factor for these neonatal outcomes.^14
–16 Further studies could clarify whether maternal HPV-positive status increases the risk of these adverse neonatal outcomes beyond the baseline risks associated with prematurity.

Strengths of this study include a large sample size in a diverse population. Our study also relied on consistent diagnostic criteria, using both cervical cytology and HPV genotyping for maternal HPV infection, based on recommendations from the American Society for Colposcopy and Cervical Pathology.¹⁹ This study did not include women with low-risk HPV infection or perform subgroup analyses examining different genotypes, which remain areas of research. A major limitation of our study was the discrepancy between the number of deliveries and number of women ultimately included in the study who had HPV testing results available in their health record. Although this finding was expected because a significant proportion of women who deliver at our institution do not typically receive their prenatal or medical care here, this may have introduced selection bias if the women who receive prenatal care at our institution differ from the general population. Due to this limitation of our retrospective study design, this limits our ability to infer causality. Although we considered other genital infections, including sexually transmitted infections, as a potential confounder, this was based on reported diagnoses and not laboratory results. As a result, interaction between HPV and other pathogenic organisms was not assessed in our study. Given the high correlation between other genital infections and HPV infection, this means that HPV infection could serve as a marker for other genital infections. Furthermore, our study cohort was from a single institution and, thus, may not be generalizable to other patient populations. However, our cohort has similar demographics as found in the general population.

Conclusions

We found that HPV infection was associated with an increased risk of PPROM in this cohort, which in turn was associated with an increased risk of neonatal morbidity associated with preterm delivery. Although prior work has established an association between HPV infection and preterm delivery, we found no association between HPV infection and preterm delivery when accounting for PPROM, pregnancy-related hypertension, and other potential risk factors. This observation suggests that HPV infection may increase the risk of upstream causes of preterm delivery. Further studies may elucidate the mechanism of HPV infection contributing to PPROM allowing for additional mechanisms to prevent premature birth. We suggest that HPV infection may be a modifiable risk factor for preterm birth, and underscores the importance of widespread administration of the HPV vaccine. While the number of preterm births potentially attributable to HPV infection may be a relatively small proportion of all preterm births, a simple and cost-effective vaccine strategy could contribute to decreasing the preterm birth rate.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

McQuillan

, Kruszon-Moran

, Markowitz

, Unger

, Paulose-Ram

Prevalence of HPV in Adults Aged 18–69: United States, 2011–2014 [NCHS Data Brief, No. 280]. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics, April 2017.

McDonnold

, Dunn

, Hester

, et al. High risk human papillomavirus at entry to prenatal care and risk of preeclampsia. Am J Obstet Gynecol, 2014; 210:138.e1–e5.

Zuo

, Goel

, Carter

. Association of cervical cytology and HPV DNA status during pregnancy with placental abnormalities and preterm birth. Am J Clin Pathol, 2011; 136:260–265.

Gomez

, Ma

, Ho

, McGrath

, Nelson

, Parry

. Placental infection with human papillomavirus is associated with spontaneous preterm delivery. Hum Reprod, 2008; 23:709–715.

Hermonat

, Han

, Wendel

, et al. Human papillomavirus is more prevalent in first trimester spontaneously aborted products of conception compared to elective specimens. Virus Genes, 1997; 14:13–17.

Malhomme

, Armbruster-Moraes

, Dutheil

, Dupressoir

, Rabreau

, Schlehofer

. Human genital tissues containing DNA of adeno-associated virus lack DNA sequences of the helper viruses adenovirus, herpes simplex virus or cytomegalovirus but frequently contain human papillomavirus DNA. J Gen Virol, 1997; 78:1957–1962.

Cho

, Min

, Hong

, et al. High-risk human papillomavirus infection is associated with premature rupture of membranes. BMC Pregnancy Childbirth, 2013; 13:173.

Subramaniam

, Lees

, Becker

, Tang

, Khan

, Edwards

. Evaluation of human papillomavirus as a risk factor for preterm birth or pregnancy-related hypertension. Obstet Gynecol, 2016; 127:233–240.

Martin

, Hamilton

, Ventura

, Osterman

, Wilson

, Matthews

. Births: Final data for 2010. Natl Vital Stat Rep, 2012; 61:1–71.

10.

Matthews

, MacDorman

. Infant mortality statistics from the 2006 period linked birth/infant death data set. Natl Vital Stat Rep, 2010; 58:1–31.

11.

Waters

, Mercer

. PPROM: Prediction, prevention, principles. Clin Obstet Gynecol, 2011; 54:307–312.

12.

Garite

, Freeman

. Chorioamnionitis in the preterm gestation. Am J Obstet Gynecol, 1982; 59:539–545.

13.

Hong

, Berggren

, Campbell

, Smith

, Rahangdale

. Abnormal cervical cancer screening in pregnancy and preterm delivery. Paediatr Perinat Epidemiol, 2014; 28:297–301.

14.

Beydoun

, Yasin

. Premature rupture of the membranes before 28 weeks: Conservative management. Am J Obstet Gynecol, 1986; 155:471–479.

15.

Lemons

, Bauer

, Oh

, Korones

, Papile

, Stoll

. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996. Pediatrics, 2001; 107:E1.

16.

Spinillo

, Capuzzo

, Stronati

, Ometto

, Orcesi

, Fazzi

. Effect of preterm premature rupture of membranes on neurodevelopmental outcome: Follow up at two years of age. Br J Obstet Gynaecol, 1995; 102:882–887.

17.

Yoon

, Romero

, Park

, et al. Fetal exposure to an intra-amniotic inflammation and the development of cerebral palsy at the age of three years. Am J Obstet Gynecol, 2000; 182:675–681.

18.

Mercer

BM.

Preterm premature rupture of the membranes. Obstet Gynecol, 2003; 101:178–193.

19.

Mayrand

, Duarte-Franco

, Rodrigues

, et al. Human papillomavirus DNA versus papanicolaou screening tests for cervical cancer. N Engl J Med, 2007; 357:1579–1588.

20.

Saslow

, Solomon

, Lawson

, 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; 62:147–172.