Multiple human papillomavirus infection with or without ... - Springer Link

Cancer Causes Control (2014) 25:1669–1676 DOI 10.1007/s10552-014-0471-6

ORIGINAL PAPER

Multiple human papillomavirus infection with or without type 16 and risk of cervical intraepithelial neoplasia among women with cervical cytological abnormalities Arsenio Spinillo • Barbara Gardella • Marianna Roccio • Paola Alberizzi Stefania Cesari • Morbini Patrizia • Enrico Silini

•

Received: 1 March 2014 / Accepted: 30 September 2014 / Published online: 9 October 2014 Ó Springer International Publishing Switzerland 2014

Abstract Purpose To evaluate the impact of multiple human papillomavirus (HPV) infections on the risk of cervical intraepithelial neoplasia grade 3 or worse (CIN3?) in subjects with cervical cytological abnormalities. Methods A cross-sectional study of 3,842 women attending a colposcopy service was carried out. Genotyping of 18 high-risk, seven low-risk, and two undefined-risk HPVs was carried out by the INNO-LiPA genotyping system. Results The final colposcopic/pathological diagnoses were as follows: 1,933 (50.3 %) subjects were negative; 1,041 (27.1 %) CIN1; 280 (7.3 %) CIN2; 520 (13.5 %) CIN3; and 68 (1.8 %) invasive cervical cancer. The prevalence of HPV infection was 75.8 % (2,911/3,842), whereas multiple HPVs were detected in 34.5 % of HPVpositive subjects (2,255/3,842). The adjusted risks of CIN3? in the group with multiple compared to the group with single infection were 2.31 (95 % CI = 1.54–3.47), among HPV16-positive women, and 3.25 (95 % CI = 2.29–4.61, p = 0.21 compared with HPV16-positive subjects), in HPV16-negative subjects. Out of a total of 1,285 subjects with mild lesions, followed up for a median A. Spinillo (&) B. Gardella M. Roccio Department of Obstetrics and Gynecology, IRCCS Fondazione Policlinico San Matteo, University of Pavia, P.le Golgi, 19, 27100 Pavia, Italy e-mail: [email protected] P. Alberizzi S. Cesari M. Patrizia Department of Pathology, IRCCS Fondazione Policlinico San Matteo, University of Pavia, Pavia, Italy E. Silini Department of Pathology, Azienda Ospedaliero-Universitaria, Universita` di Parma, Parma, Italy

of 16.1 months (interquartile range = 8.9–36.8), the rate of progression to CIN2–3 was 0.6 % (5/541) among subjects negative or with low-risk HPVs, 1.7 % (8/463) among those with single high-risk HPV, and 5 % (14/281, p \ 0.001 compared with HPV-negative/low-risk HPV and p = 0.038 compared with single high-risk HPV) among those with multiple high-risk HPVs. Conclusions Among women with cervical cytological abnormalities, infection by multiple high-risk HPVs increased the risk of CIN3? in both HPV16-positive and HPV16-negative subjects. These findings suggest a potential synergistic interaction between high-risk HPVs, favoring the progression of CIN lesions. Keywords Cervical cancer Cervical intraepithelial neoplasia Human papillomavirus infection Colposcopy

Introduction International studies have shown that infection by multiple HPV types in subjects with cervical intraepithelial neoplasia (CIN) is common, ranging from 17 to more than 55 % of CIN3 lesions [1–4]. Multiple HPV types have been linked to increased rates of CIN and to more severe and persistent HPV infection [3–6]. Although the rate of multiple HPV infections in subjects with invasive cervical cancer is low, the pattern of HPV coinfection, including causal attribution of cervical lesions to each HPV type, is considered important in the evaluation of future efficacy of vaccination programs [7, 8]. Studies on the clustering of HPV infections have shown both positive (synergistic) and negative (competitive) associations between the various HPV types identified [5, 9]. In multiple infections, the relative contribution of HPV16 and other low- and high-

123

1670

risk HPV to the occurrence of CIN3 is controversial [1, 5, 6]. Most studies have been carried out on a limited number of subjects and have failed to take into account the relative contribution of HPV16 and other high- or low-risk HPVs in case of simultaneous infection [1, 3, 6, 10]. Although molecular mapping suggests the etiological dominance of HPV16 as the causal agent of CIN3, in 18–20 % of CIN2–3 lesions associated with multiple HPV types, individual components of CIN3 lesions contain a different high-risk HPV [11, 12]. The purpose of this study was to evaluate the risk of CIN and cervical cancer associated with the pattern of coinfection between low- and high-risk HPVs and with the presence or absence of HPV16 in a consecutive series of women attending a colposcopy service because of cervical cytological abnormalities.

Patients and methods The initial study population included all women aged between 21 and 65 years who attended the colposcopy clinic of the Department of Obstetrics and Gynecology of the IRCCS Fondazione Policlinico San Matteo, Italy, from 2008 to 2012 because of abnormal Pap smear. Subjects were referred by the cytological screening service of our department, from private practice, and from screening services of external institutions. Exclusion criteria included pregnancy, HPV test or treatment for CIN in the last year, total hysterectomy, lack of a recent (1 month) Pap smear, and use of vaginal medication in the previous 2 days. Pap smears had been performed by referring physicians and were not institutionally reviewed. The Institutional Review Board of our Hospital approved the study, and informed consent was obtained from all subjects. All patients were treated according to an established protocol including HPV DNA detection and typing and colposcopy with targeted biopsies. Cervical samples for HPV typing were obtained immediately before colposcopy. After speculum examination, scrapes were taken with a cervix brush, suspended in ThinPrep-PreservCyt Solution (Cytic Corporation, Marlborough, MA, USA), and stored at 4 °C. DNA extraction was performed by lysis and digestion with proteinase K. HPV sequences from the L1 region were amplified by polymerase chain reaction (PCR) using SPF10 primers in a 50 ll final reaction volume for 40 cycles. Appropriate positive and negative controls were introduced for each set of reactions. Concurrent amplification of beta globin sequences was used as a control for DNA adequacy. HPV type-specific sequences were detected by the line probe assay INNO-LiPA HPV genotyping assay, version EXTRA (Innogenetics NV, Ghent, Belgium), according to the manufacturer’s instructions. The EXTRA version of the assay allows the simultaneous and separate detection of 18

123

Cancer Causes Control (2014) 25:1669–1676

high-risk HPV types (HPVs 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59, with proven carcinogenicity and Group I of the classification of International Agency for Research on Cancer (IARC); HPVs 26, 53, 66, 68, 73, and 82, with probable carcinogenicity and Group II IARC classification), seven low-risk HPV types (6, 11, 40, 43, 44, 54, and 70), and two unclassified-risk HPV types (69/71 and 74) [13]. Hybridization patterns were automatically analyzed by the LiRAS system and checked by two independent readers. Overall, 27 different viral types were identified. A standardized colposcopic examination was performed immediately after cervical brushing for HPV typing by two gynecology residents certified by the Italian Society of Colposcopy. Multiple targeted cervical biopsies were obtained in all cases where CIN was suspected on colposcopy and in all cases of high-grade squamous cervical lesions (HSIL) irrespective of colposcopic impression. Endocervical curettage was performed, according to the clinician’s judgment, when the extent of the lesion or the squamocolumnar junction was not entirely visible or in the case of atypical glandular cells (AGC) on Pap smear. In the analysis of data, we used either the histological diagnosis of punch biopsy or, when more severe, the diagnosis after cone biopsy obtained by loop electro-excision procedure (LEEP) or cold knife excision. After diagnostic workup and treatment, patients referred from the cytological screening service of our department were enrolled in a follow-up program including (a) observation, colposcopy, and/or cytology every 6–12 months for subjects with negative colposcopic impression and/or negative histological findings after an abnormal Pap smear; (b) observation, colposcopy, and cytology every 6 months for CIN1 lesions; and (c) HPV test coupled with colposcopy and cytology every 6 months for treated CIN2–3 lesions. In statistical analysis, outcome was defined as negative when colposcopy and/or cervical biopsy was negative, or as CIN1, CIN2, CIN3, and invasive squamous or adenocarcinoma of the cervix. HPV infections were categorized as negative, untypable, single/multiple low-risk, single high-risk, single high-risk/single low-risk/multiple lowrisk, multiple high-risk/single low-risk/multiple low-risk, multiple high-risk, and unclassified-risk types. Data were analyzed by Kruskal–Wallis analysis of variance and Bonferroni-corrected post hoc test to compare continuous and chi-square test to compare categorical variables, respectively. The Bonferroni method for multiple comparisons was used to adjust chi-square tests in multiway contingency tables. Multinomial logistic models were constructed to evaluate the association of category of HPV infection with abnormal colposcopic/pathological outcomes. Explanatory variables inserted in the models were age (continuous variable), previous treatment for squamous


intraepithelial lesions (SIL; yes, no), human immunodeficiency virus status (negative, unknown, positive), ancestry (Italian, foreign), parity (no, 1, [1), Pap smear results (Low-grade SIL, atypical squamous cells of undetermined significance, atypical squamous cells cannot exclude highgrade SIL, high-grade SIL-cancer, AGC), and category of HPV infection (negative, untypable, single/multiple lowrisk, single high-risk, single high-risk/single low-risk/ multiple low-risk, multiple high-risk/single low-risk/multiple low-risk, multiple high-risk infection). Data were analyzed with Stata/MP 10 for Windows (StataCorp LP, College Station, TX, USA). Significant differences between regression coefficients obtained from multinomial logistic equations were compared by Wald test [14].

Results During the study, 4,118 subjects with abnormal Pap smear were seen at our colposcopy clinic and 4,012 (97.4 %) agreed to participate in the study. A further 170 subjects were excluded because of missing or inadequate samples for HPV DNA testing (129 subjects), or inadequate colposcopic examination (41 subjects), leaving 3,842 subjects for analysis. The median age (interquartile range, IQR) of the entire cohort was 37 years (IQR = 29–45). The final colposcopic/pathological diagnoses were as follows: 1,933 (50.3 %) subjects were negative; 1,041 (27.1 %) CIN1; 280 (7.3 %) CIN2; 520 (13.5 %) CIN3; and 68 (1.8 %) invasive cervical cancer. Negative subjects included 915 (23.8 %) subjects with negative cervical biopsies and 1,018 (26.5 %) subjects with negative colposcopy in whom cervical biopsy was considered unnecessary. Pathological diagnoses C CIN1 were obtained either by cervical biopsies of LEEP or cold knife conization. Of the 1,041 CIN1 lesions, 706 (67.8 %) were diagnosed on cervical biopsies and 335 (32.2 %) on LEEP specimens. Diagnoses of CIN2–3 and invasive cervical cancer were obtained from either LEEP or cold knife conization in 581 (72.6 %) out of 800 subjects and in 57 (83.8 %) out of 68 subjects, respectively. According to the classification of the International Federation of Obstetrics and Gynecology, of the 68 subjects with a diagnosis of invasive cervical cancer, eight (11.8 %) were at stage IA, 47 (69.1 %) were at stage Ib– IIa, and the remaining 13 (19.1 %) were in more advanced stages of the disease. The relationship between final colposcopic/pathological classification, demographic data, and Pap smear results at enrolment are reported in Table 1. Age and parity differed across the categories of colposcopic/ pathological classification. Post hoc analysis and chisquare analysis showed that the median age was higher (p \ 0.01 for all comparisons) and parity [1 was more

1671

frequent (p \ 0.01 for all comparisons) in subjects with invasive cervical cancer than in all other categories of colposcopic/pathological classification. The 41 subjects with severe cytological abnormalities (28 ASC-H and 13 HSIL) and negative colposcopic/histological results were all biopsied, and histology was negative. The likelihood of CIN2? was 7.4 % after ASCUS (187/1,387), 18.6 % after LSIL (362/1,944), and 79.2 % (300/379) after ASC-H/ HSIL. Out of the 3,842 subjects enrolled in the study, 1,678 women (43.7 %) had a Pap smear outside our institution. A group of 211 (168 with ASCUS/LSIL, 36 with HSIL and seven with AGC) subjects repeated Pap smear at our institution before colposcopy. The overall positive agreement between readings was 81.5 % (95 % CI = 74.8–87.1) (137/168) for ASCUS-LSIL, and 86.1 % (95 % CI = 70.5–95.3) (31/36) for HSIL. The prevalence of HPV infection was 75.8 % (2,911/ 3,842); multiple HPV types were detected in 34.5 % of subjects (1,324/3,842) and in 45.5 % (1,324/2,911) of HPV-positive cases. The prevalence of the more common genotypes was as follows: HPV16 in 30 % of subjects (875/2,911); HPV31 in 14.3 % (416/2,911); HPV51 in 13.4 % (391/2,911); HPV52 in 17.6 % (513/2,911); and HPV53 in 7.5 % (217/2,911). The median age of subjects harboring multiple infections (36 years, IQR = 29–44) was consistently lower than that of subjects with a single infection (39 years, IQR = 30–48, p \ 0.001). Of the 2,911 HPV-positive subjects, 350 (9.1 %) had untypable HPV infection, 306 (8 %) were infected by single/multiple low-risk HPVs, 968 (25.2 %) by a single high-risk HPV, 243 (8.3 %) by single high-risk/single lowrisk/multiple low-risk HPVs, 382 (13.1 %) by multiple high-risk/single low-risk/multiple low-risk HPVs, and 655 (17 %) by multiple high-risk HPVs. In seven cases (two negative, four CIN1, and one CIN3), HPV infection was caused by HPV types (69/71 and 74) with an undefined risk of carcinogenicity. These cases were excluded from further analyses. The association between HPV infection classified on the basis of potential carcinogenicity and cytological and colposcopic/pathological outcome is reported in Table 2. The rates of high-risk multiple infection was 16.7 % (232/ 1,387) among ASCUS, 28.7 % (558) among LSIL (p \ 0.001 compared to ASCUS), and 59.3 % (224/378) among ASC-H/HSIL subjects (p \ 0.001 compared to ASCUS and LSIL). After Bonferroni correction, colposcopic/pathological outcomes were significantly different among subjects infected by a single high-risk HPV compared with those infected by multiple high-risk/single low-risk/multiple lowrisk (p \ 0.01) or multiple high-risk HPVs (p \ 0.01). The risks of CIN and invasive cervical cancer according to HPV classification are reported in Table 3. As expected,

123

1672


Table 1 Demographic variables and Pap smear results at enrolment

Age (median, IQR)

Foreigner

Negative n = 1,933

CIN1 n = 1,041

CIN2 n = 280

CIN3 n = 520

Invasive cancer n = 68

p value*

38 (30–36)

36 (27–44)

35 (28–43)

36 (30–43)

43 (37–51)

\.001

n (%)

n (%)

n (%)

n (%)

n (%)

207 (10.70)

95 (9.12)

35 (12.50)

84 (16.15)

13 (19.11)

935 (48.37)

507 (48.70)

149 (53.21)

212 (40.76)

10 (14.47)

\.001

Parity 0 1

715 (36.98)

353 (33.90)

73 (26.07)

202 (38.84)

20 (29.41)

[1

283 (14.64)

181 (17.38)

58 (20.71)

106 (20.38)

38 (55.88)

Previous SIL n.

63 (3.25)

19 (1.82)

9 (3.21)

14 (2.69)

4 (5.88)

0.2

HIV seropositive

44 (2.27)

9 (0.86)

7 (2.50)

12 (2.30)

2 (2.94)

0.07

LSIL

947 (48.99)

640 (61.47)

169 (60.35)

190 (36.53)

3 (4.41)

ASCUS

864 (44.69)

337 (32.37)

66 (23.57)

109 (20.96)

12 (17.64)

41 (2.12)

38 (3.65)

40 (14.28)

212 (40.76)

48 (70.58)

81 (4.19)

26 (2.49)

5 (1.78)

9 (1.73)

5 (7.35)

ASC-H, HSIL, invasive cancer AGC

\.001

\.001

IQR interquartile range, CIN cervical intraepithelial neoplasia, HIV human immunodeficiency virus, LSIL low-grade squamous intraepithelial neoplasia, HSIL high-grade squamous intraepithelial lesions, ASCUS atypical squamous cells of undetermined significant, ASC-H atypical squamous cells cannot exclude HSIL, AGC atypical glandular cells * As obtained by Kruskal–Wallis analysis of variance or chi-square test

among subjects with high-risk HPV, either as a single infection or multiple infections, the adjusted risk of CIN and invasive cervical cancer was significantly higher compared with that of uninfected subjects. In logistic regression analysis, adjusted odds ratios of CIN2 and of CIN3 and invasive cervical cancer associated with a single high-risk HPV infection were lower compared with those associated with multiple high-risk/single lowrisk/multiple low-risk HPV infections (p \ 0.001 for comparisons of CIN2 and CIN3, and p = 0.013 for comparison of invasive cervical cancer odds ratios) and with multiple high-risk HPV infections (p \ 0.001 for comparisons of CIN2 and CIN3, and p = 0.01 for comparison of invasive cervical cancer odds ratios). There were no significant differences between odds ratios of CIN2, CIN3, and cancer between subjects infected by a single high-risk HPV and those with multiple infections caused by a single high-risk/single low-risk/multiple low-risk types (p = 0.31, 0.36, and 0.4, respectively), suggesting that the risk of CIN or cancer associated with a single high-risk HPV was not influenced by simultaneous coinfection by low-risk types. When the categories of HPV infection were collapsed in two groups (single infection vs multiple infections), the adjusted risk of CIN3? was 40.3 (95 % CI = 19.5–83.5) for a single high-risk infection and 118.2 (95 % CI = 57–244.9, p \ 0.001 in comparison with a single infection) for multiple high-risk infections compared with uninfected subjects.

123

To separate the effect of HPV16 from that of other infections, we did a stratified analysis including only highrisk HPV infections (two categories of single infection and two categories of multiple infections; Table 4). Although the effect was less marked among HPV16-positive subjects, multiple infections were associated with a significantly higher risk of CIN2 and CIN3 compared with a single HPV infection. The adjusted risk of CIN2 and CIN3? in the group of multiple infections compared with the group of single infection were 2.14 (95 % CI = 1.44–3.18) and 2.31 (95 % CI = 1.54–3.47), respectively, among HPV16-positive subjects and 2.19 (95 % CI = 1.3–3.68) and 3.25 (95 % CI = 2.29–4.61, p = 0.21 compared with HPV16-positive subjects) in HPV16-negative subjects. Out of a total of 800 subjects with CIN2–3 at enrolment and treated by LEEP or cold knife conization, 440 (55 %) were enrolled in our follow-up program for a median time of 15.4 months (IQR = 8.5–33.7). Biopsy-proven recurrence/persistence of CIN2–3 was diagnosed in 5.4 % (2/ 37) of subjects with negative, untypable, or low-risk HPV, in 12.3 % (18/146) of patients with a single high-risk HPV, and in 8.9 % (23/257) of patients with multiple high-risk HPV infections (p = 0.4). Among the 199 HPV16-positive subjects followed up, the rate of CIN2–3 persistence/ recurrence was similar between those with a single (6/58) high-risk infection and those with multiple high-risk HPV infections (11/141, p = 0.58). Out of a total of 2,974


1673

Table 2 Association between type of human papillomavirus (HPV) infection, cytological, and colposcopic/pathological classification Type of HPV infection

ASCUS n = 1,387 n (%)

LSIL n = 1,944 n (%)

HSIL, ASC-H, Invasive cancer n = 378 n (%)

Negative (n = 931)

515 (37.13)

365 (18.77)

13 (3.44)

38 (30.16)

Untypable (n = 350) Single/multiple low risk (n = 306)

169 (12.18) 125 (9.01)

158 (8.13) 162 (8.33)

8 (2.12) 7 (1.85)

15 (11.90) 12 (9.52)

Single high risk (n = 968)

AGC n = 126 n (%)

284 (20.47)

552 (28.40)

98 (25.92)

34 (26.98)

Single high risk/single low risk/multiple low risk (n = 243)

62 (4.47)

149 (7.66)

28 (7.41)

4 (3.17)

Multiple high risk/single low risk/multiple low risk (n = 382)

92 (6.63)

184 (9.46)

99 (26.19)

7 (5.55)

140 (10.09)

374 (19.23)

125 (33.07)

16 (12.70)

Type of HPV infection

Negative n = 1,931 n (%)

CIN1 n = 1,037 n (%)

CIN2 n = 280 n (%)

Negative (n = 931)

882 (45.67)

19 (1.83)

22 (7.86)

5 (0.96)

Untypable (n = 350)

230 (11.91)

93 (8.97)

16 (5.71)

10 (1.93)

Multiple high risk (n = 655)

CIN3 n = 519 n (%)

Single/multiple low risk (n = 306)

154 (7.97)

125 (12.05)

17 (6.07)

9 (1.73)

Single high risk (n = 968)

352 (18.23)

375 (36.16)

81 (28.93)

140 (26.97)

Single high risk/single low risk/multiple low risk (n = 243)

82 (4.25)

103 (9.93)

18 (6.43)

36 (6.94)

Multiple high risk/single low risk/multiple low risk (n = 382)

80 (4.14)

98 (9.45)

49 (17.50)

139 (26.78)

151 (7.82)

224 (21.60)

77 (27.50)

180 (34.68)

Multiple high risk (n = 655)

LSIL low-grade squamous intraepithelial neoplasia, HSIL high-grade squamous intraepithelial lesions, ASCUS atypical squamous cells of undetermined significant, ASC-H Atypical squamous cells cannot exclude HSIL, AGC atypical glandular cells

subjects with negative colposcopy/biopsy or CIN1, 1,285 (43.2 %) were followed up for a median period of 16.1 months (IQR = 8.9–36.8). Progression to biopsyproven CIN2–3 occurred in 0.6 % (5/541) of subjects with negative, untypable, and low-risk HPV infection, 1.7 % (8/ 463) of those with a single high-risk HPV infection, and 5 % (14/281, p \ 0.001 compared with negative/untypable/low-risk infection, and p = 0.038 compared with single high-risk infection) of those with multiple high-risk HPV infections. Among high-risk infections, the rates of progression were similar among subjects infected by HPV16 at enrolment (6/233, 2.6 %) or by other HPV types (16/511, 3.1 %, p = 0.82).

Discussion The results of this study have shown that multiple high-risk HPV infections are associated with an increased risk of CIN2 and CIN3? compared with a single high-risk HPV infection among women with cervical cytological abnormalities. These results were replicated in both HPV16positive and HPV16-negative subjects, suggesting that multiple high-risk HPVs can interact synergistically, increasing the risk of CIN. Follow-up data of subjects with either negative colposcopy/biopsy or CIN1 at enrolment confirmed that infection by multiple high-risk HPV types

was associated with an increased risk of progression to CIN2–3 compared with a single high-risk infection. Multiple infections by one high-risk HPV and infection by one or more low-risk HPVs carry a risk of CIN similar to that of infection by a single high-risk HPV. This suggests that coinfection with low-risk HPV types does not modify the risk of precancerous cervical lesions associated with highrisk HPVs. Several cross-sectional and longitudinal studies have shown that multiple infections with high-risk HPV types increase the risk of CIN [1–6]. The main differences of the present study from others involve the type of population and the number of subjects studied. Previous cross-sectional studies on this topic have been carried out on archived cervical biopsies or on a limited number of women attending colposcopic services and analyzed by univariate methods [1, 3, 4, 6, 10]. Age, Pap smear results, parity, immunodepression, and other factors could influence the rate of multiple infections and CIN [1, 4, 6]. The high number of CIN cases in a homogeneous population of consecutive subjects with abnormal cytological findings and the multivariate methods used to correct for potential confounders are the main strengths of our study compared with others. On the other hand, one of the potential limitations of this study is the lack of institutional review of Pap smears, which meant that cytological diagnoses were not checked for consistency. In addition, subjects who were

123

1674


Table 3 Odds Ratios of cervical intraepithelial neoplasia (CIN) and invasive cervical cancer according to the category of human papillomavirus (HPV) infection Type of HPV infection

Negative

Negative

Baseline

CIN1 OR (95 % CI) Reference



Invasive cancer OR (95 % CI) Reference

Untypable

19.4 (11.6–32.5)

2.8 (1.4–5.4)

7.6 (2.5–22.6)

1.1 (0.1–11.2)

Single/multiple low risk

38.5 (23–64.4)

4.3 (2.2–8.3)

10.2 (3.3–31.2)

1.9 (0.2–19.5)

Single high risk

49.1 (30.4–79.3)

7.9 (4.8–13)

56.4 (22.7–140)

11.6 (3.3–41.3)

Single high risk/single low risk/multiple low risk

58.1 (33.7–100)

7.4 (3.8–14.5)

59.3 (22.4–159

Multiple high risk/single low risk/multiple low risk

55.1 (32–95.1)

19.3 (11–33.8)

201 (79.1–511)

29.1 (7.8–109)

Multiple high risk

66.4 (40.2–109.8)

16.4 (9.8–27.5)

154 (61.6–385)

28.4 (7.9–101)

8.6 (1.7–42.4)

Odds Ratios (OR) and 95 % confidence intervals (CI) as obtained by multinomial logistic equations including colposcopic/pathological outcome as outcome variable, and type of HPV infection, age (continuous variable), previous SIL treatment (yes, no), human immunodeficiency virus status (negative, unknown, positive), ancestry (Italian, foreign), parity (no, 1,[1), Pap smear results LSIL, ASCUS, ASC-H, HSIL-cancer, AGC, and category of HPV infection (negative, untypable, single/multiple low-risk, single high-risk, single high-risk/single low-risk/multiple low-risk, multiple high-risk/single low-risk/multiple low-risk, multiple high-risk infection) as explanatory variables

followed up were only a part of the entire cohort, and this could have introduced a potential bias in the evaluation of the role of multiple infections on the incidence and recurrence of CIN among at-risk women. Consolidate literature data suggest that the immediate risk of CIN2? is 6–12 % after referral for ASCUS, 15–30 % after LSIL, and at least 60 % after ASC-H and HSIL [15, 16]. In our study, the immediate risk of CIN2? was 7.4 % after ASCUS, 18.6 % after LSIL, and 79.2 % after ASC-H/HSIL, suggesting that cytological readings were appropriate. We found that coinfection with low-risk HPV types did not increase the risk of CIN3? associated with high-risk HPV types among women with cytological abnormalities. This finding confirms older serological data and recent epidemiological studies suggesting that in case of multiple HPV infections, only high-risk HPVs play a causal role in cervical oncogenesis and that low-risk HPVs could interact with high-risk ones without increasing the risk of CIN and invasive cervical cancer [17, 18]. Although the rate of multiple infections in invasive cervical cancer is lower than 10 % [7], worldwide epidemiological studies have shown that the prevalence of multiple infections among women with cervical cytological abnormalities or CIN is higher than 50 % of subjects [1–6, 19]. The 45.5 % prevalence of multiple infection in our study is similar to that reported by other authors [1, 6, 9] in cervical screening studies. The association between multiple high-risk HPV infections and risk of CIN could be attributed to several causative factors including methodological, sociodemographic, and biological reasons. The rate of multiple high-risk infections is consistently lower in biopsy compared with cytological studies [3, 6]. Cytological sampling collects material from the entire cervix and vaginal areas, allowing the identification of HPV types infecting other CIN zones and also normal and metaplastic epithelium [20]. The simultaneous presence of multiple

123

HPV types in normal and metaplastic epithelium and in surrounding areas of CIN1 has been demonstrated by laser capture method in histological specimens of CIN3 [11, 12]. Young age, increased susceptibility of cervical transformation zone among adolescents, and sexual behavior have been cited as explanations for the association between CIN and multiple HPV infections [4, 8, 9]. However, none of these methodological or sociodemographic factors explain the correlation between multiple HPV infections and increasing severity of CIN as reported in this and other studies [1–6]. Some authors have attributed the correlation between multiple infections and severity of CIN to the increasing duration of HPV infection associated with multiple HPV types [6]. However, persistence of HPV infection, which is considered a key risk factor in cervical carcinogenesis, has been found to be associated with multiple HPV infections in some studies [21, 22] but has not been confirmed in others [23]. Multiple incident and prevalent HPV infections could have a promoting effect on CIN independently of persistence of infection. Genomic studies have shown that multiple HPV infections are associated with overexpression of E6/E7 mRNA which, in turn, is a well-known marker of oncogenic activity [24]. Longitudinal studies have found that multiple HPV types and HPV metatyping (infection with different types of HPVs during follow-up) are strongly associated with HSIL and progression of cervical disease [6, 25]. Studies using laser capture micro-dissection have shown that in histological samples, single CIN3 lesions contain distinct viruses, mostly HPV16 [11, 12]. The authors also found that two out of 11 (18.2 %) CIN3 and four out of 11 (36.4 %) CIN2 samples contained different lesions of similar severity caused by different HPV types [12], suggesting that at least in some cases, multiple high-risk HPVs can be associated with simultaneous and distinct CIN areas.

96 (46.60) 27 (13.11) 20 (9.71) 63 (30.58)

Negative n = 458 n (%) reference 256 (55.90) 54 (11.79 %) 60 (13.10 %) 88 (19.21 %)

Single high risk n = 342

Single high risk/single low risk/multiple low risk n = 78

Multiple high risk/single low risk/multiple low risk n = 136

Multiple high risk n = 319

HPV16 negative

Single high risk n = 626 Single high risk/single low risk/multiple low risk n = 164

Multiple high risk/single low risk/multiple low risk n = 246

Multiple high risk n = 336

2.1 (1.3–3.4)

1.6 (1.2–2.2)

2 (1.2–3.5) 41 (30.83)

137 (26.10)

1.1 (0.8–1.7)

28 (21.05)

0.8 (0.4–1.8)

1.5 (1–2.2) 66 (12.57)

54 (40.60) 10 (7.52)

CIN2 n = 133 n (%) OR (95 % CI)

1.8 (1–3.3)

36 (39.13)

245 (46.67) 77 (14.67)

CIN1 n = 525 n (%) OR (95 % CI)

1 (0.66–1.5)

87 (31.64)

21 (22.83) 3.4 (1.6–7.2)

32 (11.64)

1.1 (0.4–2.8)

8 (8.70)

27 (29.35)

CIN2 n = 92 n (%) OR (95 % CI)

1.1 (0.6–2.1)

0.7 (0.4–1.3)

26 (9.45)

130 (47.27)

CIN1 n = 275 n (%) OR (95 % CI)

2.8 (1.8–4.3)

61 (26.40)

4.2 (2.6–6.6)

83 (35.93)

1.4 (0.8–2.6)

65 (28.14) 22 (9.52)

CIN3 n = 231 n (%) OR (95 % CI)

2 (1.3–3.2)

119 (45.08)

3.2 (1.7–6)

56 (21.21)

0.7 (0.3–1.6)

14 (5.30)

75 (28.41)

CIN3 n = 264 n (%) OR (95 % CI)

5.4 (1.7–16.7)

9 (36.00)

4.2 (1.3–13.1)

9 (36.00)

0.7 (0.07–6.0)

6 (24.00) 1 (4.00)

Invasive cancer n = 25 n (%) OR (95 % CI)

1.1 (0.45–2.8)

14 (36.84)

1.9 (0.6–5.9)

7 (18.42)

0.8 (0.2–3.4)

3 (7.89)

14 (36.84)

Invasive cancer n = 38 n (%) OR (95 % CI)

Odds Ratios (OR) and 95 % confidence intervals (CI) as obtained by multinomial logistic equations including colposcopic/pathological outcome as outcome variable, and type of HPV infection, age (continuous variable), previous SIL treatment (yes, no), human immunodeficiency virus status (negative, unknown, positive), ancestry (Italian, foreign), parity (no, 1, [1), Pap smear results LSIL, ASCUS, ASC-H, HSIL-cancer, AGC, and category of HPV infection (negative, untypable, single/multiple low-risk, single high-risk, single high-risk/single low-risk/ multiple low-risk, multiple high-risk/single low-risk/multiple low-risk, multiple high-risk infection) as explanatory variables

Negative n = 206 n (%) reference

HPV16 positive

Table 4 Odds Ratios of cervical intraepithelial neoplasia (CIN) and invasive cervical cancer according to the category of human papillomavirus (HPV) infection among HPV16-positive and HPV16-negative women

Cancer Causes Control (2014) 25:1669–1676 1675

123

1676


In summary, the present study showed that coinfection by multiple high-risk HPVs increased the risk of CIN2 and CIN3? in both HPV16-positive and HPV16-negative samples. Multiple infections were also associated with an increased progression to more severe CIN during follow-up of mild lesions, suggesting a potential synergistic mechanism between highrisk HPVs that favors the progression of CIN lesions. Conflict of interest interest.

The authors state that they have no conflict of

References 1. Cuschieri KS, Cubie HA, Whitley MW et al (2004) Multiple high risk HPV infections are common in cervical neoplasia and young women in a cervical screening population. J Clin Pathol 57:68–72 2. Dickson EL, Vogel RI, Bliss RL, Downs LS Jr (2013) Multiple type human papillomavirus (HPV) infections: a cross-sectional analysis of the prevalence of specific types in 309,000 women referred for HPV testing at the time of cervical cytology. Int J Gynecol Cancer 23:1295–1302 3. Kirschner B, Schledermann D, Holl K et al (2013) HPV-genotypes in high-grade intraepithelial cervical lesions in Danish women. Acta Obstet Gynecol Scand 92:1032–1040 4. Figueiredo Alves RR, Turchi MD, Santos LE et al (2013) Prevalence, genotype profile and risk factors for multiple human papillomavirus cervical infection in unimmunized female adolescents in Goiaˆnia, Brazil: a community-based study. BMC Public Health 13:1041 5. Chaturvedi AK, Katki HA, Hildesheim A et al (2011) Human papillomavirus infection with multiple types: pattern of coinfection and risk of cervical disease. J Infect Dis 203:910–920 6. Trottier H, Mahmud S, Costa MC et al (2006) Human papillomavirus infection with multiple types and risk of cervical neoplasia. Cancer Epidemiol Biomark Prev 15:1274–1280 7. De Sanjose S, Quint WGV, Alemany L et al (2010) Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study. Lancet Oncol 11:1048–1056 8. Tota J, Ramanakumar AV, Jiang M et al (2013) Epidemiologic approaches to evaluating the potential for human papillomavirus type replacement postvaccination. Am J Epidemiol 178:625–634 9. Mejlhede N, Pedersen BV, Frisch M, Fomsgaard A (2010) Multiple human papilloma virus types in cervical infections: competition or synergy? APMIS 118:346–352 10. Pista A, de Oliveira CF, Lopes C, Cunha MJ, CLEOPATRE Portugal Study Group (2013) Human papillomavirus type distribution in cervical intraepithelial neoplasia grade 2/3 and cervical cancer in Portugal: a CLEOPATRE II study. Int J Gynecol Cancer 23:500–506 11. van der Marel J, Quint WGV, Schiffman M et al (2012) Molecular mapping of high grade cervical intraepithelial neoplasia shows etiological dominance of HPV 16. Int J Cancer 131:E946–E953

123

12. Quint W, Jenkins D, Molijn A et al (2012) One virus, one lesion. Individual components of CIN lesions contain a specific HPV type. J Pathol 227:62–71 13. Bouvard V, Baan R, Straif K et al (2009) A review of human carcinogens—part B. Biological agents. Lancet Oncol 10:321– 322 14. Scott JL, Freese J (2006) Regression models for categorical dependent variables using Stata, 2nd edn. Stata Press, College Station, TX, USA 15. Massad LS, Einstein MH, Huh WK et al (2013) 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors. Obstet Gynecol 121:829–845 16. Arbyn M, Buntinx F, Van Ranst M, Paraskevaidis E, Martin-Hirsch P, Dillner J (2004) Virologic versus cytologic triage of women with equivocal Pap smears: a meta-analysis of the accuracy to detect high-grade intraepithelial neoplasia. J Natl Cancer Inst 96:280–293 17. Luostarinen T, af Geijersstam V, Bjørge T et al (1999) No excess risk of cervical carcinoma among women seropositive for both HPV16 and HPV6/11. Int J Cancer 80:818–822 18. Castle PE, Hunt WC, Langsfeld E, Wheeler CM, New Mexico HPV Pap Registry Steering Committee (2014) Three-year risk of cervical precancer and cancer after the detection of low-risk human papillomavirus genotypes targeted by a commercial test. Obstet Gynecol 123:49–56 19. Spinillo A, Dal Bello B, Gardella B, Roccio M, Dacco’ MD, Silini EM (2009) Multiple human papillomavirus infection and high grade cervical intraepithelial neoplasia among women with cytological diagnosis of atypical squamous cells of undetermined significance or low grade squamous intraepithelial lesions. Gynecol Oncol 113:115–119 ¨ et al (2011) Prevalence and 20. Bayram A, Erkılıc¸ S, Balat O genotype distribution of human papillomavirus in non-neoplastic cervical tissue lesion: cervical erosion. J Med Virol 83:1997– 2003 21. Matsumoto K, Oki A, Furuta R et al (2011) Predicting the progression of cervical precursor lesions by human papillomavirus genotyping: a prospective cohort study. Int J Cancer 128:2898–2910 22. Clarke M, Schiffman M, Wacholder S, Rodriguez AC, Hildesheim A, Quint W, Costa Rican Vaccine Trial Group (2013) A prospective study of absolute risk and determinants of human papillomavirus incidence among young women in Costa Rica. BMC Infect Dis 13:308 23. Campos NG, Rodriguez AC, Castle PE et al (2011) Persistence of concurrent infections with multiple human papillomavirus types: a population-based cohort study. J Infect Dis 203:823–827 24. Andersson E, Ka¨rrberg C, Ra˚dberg T et al (2012) Type-dependent E6/E7 mRNA expression of single and multiple high-risk human papillomavirus infections in cervical neoplasia. J Clin Virol 54:61–65 25. An HJ, Sung JM, Park AR et al (2011) Prospective evaluation of longitudinal changes in human papillomavirus genotype and phylogenetic clade associated with cervical disease progression. Gynecol Oncol 120:284–290