Association between high-risk human papillomavirus DNA load and precursor lesions of cervical cancer in Mexican women

Available online at www.sciencedirect.com R

Gynecologic Oncology 90 (2003) 310 –317

www.elsevier.com/locate/ygyno

Association between high-risk human papillomavirus DNA load and precursor lesions of cervical cancer in Mexican women Dulce M. Herna´ndez-Herna´ndez,a,* Laura Ornelas-Bernal,a Miriam Guido-Jime´nez,b,c Teresa Apresa-Garcia,a Isabel Alvarado-Cabrero,a Mauricio Salcedo-Vargas,a Alejandro Mohar-Betancourt,b,c and Alejandro Garcia-Carrancab,c a

Medical Research Unit in Oncology Diseases, Pathology Department, Oncology Hospital, Centro Me´dico Siglo XXI, Instituto Mexicano del Seguro Social (IMSS), Mexico City, Mexico b Division of Research, Instituto Nacional de Cancerologı´a (INCan), Mexico City, Mexico c Instituto de Investigaciones Biome´dicas, Universidad Nacional Auto´noma de Me´xico (UNAM), Mexico City, Mexico Received 18 September 2002

Abstract Objective. Our objective was to determine the association between viral load of high risk human papilloma virus (HPV) using the Hybrid Capture II (HC II) system and cervical intraepithelial neoplasia (CIN) lesion stage. Methods. A total of 182 consecutive women with confirmed diagnoses of CIN 1–3 and 182 healthy women with negative Pap were included. All subjects underwent structured interviews focused on socioeconomic and reproductive factors. HC II testing was used to detect human papilloma virus (HPV) DNA. Viral load was measured by light measurements expressed as relative lights unit (RLU) ratio (specimens/control). Log10RLU ratios were categorized for analysis into four groups: negative (ⱕ0); low viral load (0.01–1.0), middle viral load (1.01–2.0), and high viral load (2.0 –3.6). Frequencies and association measurement odds ratio (OR) adjusted by unconditional multinomial regression (UMR) were used in analysis. Results. A total of 75 of 80 (93.7%) patients with CIN 2–3, 82 of 101 (79.4%) with CIN 1, and 36 of 182 (19.8%) controls were positive for HPV DNA. The higher the viral load of HPV DNA infection observed, the higher the probability of being associated with stage of CIN (P ⬍0.001). Association between low viral load HPV and CIN 1 was 16.8 (7.2–39) compared with the highest association observed with high viral load and CIN 2–3 (ORa ⫽ 365.8, 94.7–1412). Both control and cases in the oldest women presented the highest viral load. Conclusions. We found high frequencies of HPV DNA in CIN 1 and in CIN 2–3 patients. A clear association between viral load of HPV DNA was determined by HC II assay and CIN stage. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Cervical intraepithelial lesions; Cervical cancer; Human papilloma virus; Viral load; Hybrid Capture II

Introduction Carcinoma of the cervix (CC) is second only to breast cancer as the most common malignancy affecting women worldwide. Each year, CC accounts for approximately * Corresponding author: Unidad de Investigacio´n Me´dica en Enfermedades Oncolo´gicas, Hospital de Oncologia, Centro Me´dico Siglo XXI, IMSS, Av. Cuauhte´moc #330, Col. Doctores, 06720 Me´xico, D.F., Me´xico. Fax: ⫹52-55-5761-0952. E-mail address: [email protected] (D.M. Herna´ndez-Herna´ndez).

470,600 new cases and 233,400 deaths [1]. In Mexico, this neoplasia occupies first place in incidence among carcinoma cases and there has been an increase in incidence and mortality rates during the past two decades [2]. Infection with oncogenic genital human papillomaviruses (HPV) has been established as the main etiologic agent for CC and of cervical intraepithelial neoplasia (CIN) [3]. Genital HPV types have been subdivided into low risk, found mainly in genital warts, and high risk, frequently associated with invasive cervical cancer. Epidemiologic classification based on HPV type-specific odds ratios (ORs)

0090-8258/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0090-8258(03)00320-2

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

and HPV prevalence among patients and controls identified 15 HPV types as high risk (16,18,31,33,35,39,45,51,52, 56,58,59,68,73,82), and 12 types (6,11,40,42,43,44,54,61, 70,72,81, and CP6108) were categorized as low risk [4]. Epidemiologic association between HPV infection and development of cervical neoplasia is strong independently of other risk factors and consistent in several countries [5]. Nevertheless, there is a marked discrepancy between number of women positive for HPV DNA and incidence of cervical neoplasia [6,7]. The vast majority of HPV infections are transient, with approximately 70% of infections cleared within 1 year [8]. Many epidemiologic studies have provided strong and consistent evidence that persistence infection with specific HPV types, in particular highly oncogenic types HPV 16 and 18, is responsible for development, maintenance, and progression of CIN and CC [9,10]. Persistence may depend on certain characteristics such as high viral load as a result of viral replication. The effect of high-risk types of HPV DNA infection on CIN development is highly influenced by viral number copies [11,12]. It was observed that women with high viral loads of HPV had at least 30 times greater risk for CC than HPV 16negative women in one decade prior to diagnosis. These authors found an increase in relative risk over time, and approximately 25% of women infected with high viral load prior to the age of 25 years developed carcinoma in situ within 15 years [13]. Hybrid Capture test II generation (HC II) for HPV DNA validated its sensitivity and showed to be an accurate reproducible test [14]. Compared with PCR-ELISA, the HC II showed a sensitivity of 91.7% and a specificity of 95.4%. Both tests did not differ in recognizing HPV DNA-positive patients with either low- or high-grade precursor lesions of CC [15]. In addition, its capability to determine quantitative viral load in cervical specimen provided a means of increasing test specificity for disease [16]. In Mexico, the National Screening Program does not consider the HPV molecular test and many controversies exist concerning its cost effectiveness as a screening test in the population worldwide [2,17,18]. For these reasons, it is important to have specific criteria to identify women at high risk of CC. The purpose of this study was to evaluate the association between viral load of high-risk HPV and precursor cervical lesions stage in a group of Mexican women.

Methods Patients were recruited from two outpatient clinics of the Department of Colposcopy at the Mexican Institute of Social Security (IMSS) and Hospital General of Mexico, both in Mexico City. These patients were referred by their general practitioner owing to abnormal cervical cytology report (CIN 1–3, mild, moderate, or severe dysplasia, in situ car-

311

cinoma), or cervical cytological abnormality discovered during gynecologic examination. Between March 1998 and December 2000, 225 cases and 230 controls were recruited. CIN 1–3 diagnosis was carried out in accordance with the World Health Organization Histopathological Classification [19]. Biopsy or cone specimens were evaluated by two pathologists who were blind to colposcopy or clinical information to classify lesion stage. All cases with disagreement were reanalyzed by an experienced pathologist in oncologic and gynecologic lesions (Alvarado CI); to improve agreement, we decided to classify patients into two groups: CIN 1 and CIN 2–3. A total of 43 of 225 (19.1%) cases were eliminated; 35 had cervical biopsies without histopathologic alterations, 2 cases were diagnosed as invasor cervical cancer, 4 cases did not have sufficient tissue for diagnosis, and 3 additional biopsies were not available. Women with negative Pap were selected as controls at primary-level medical units from the same influence area of cases. They were frequency-matched by age (⫾5 years) of cases. They had negative history of abnormal Pap smear and no cervical–vaginal symptoms. Cytologic samples for Papanicolaou (Pap) and human papilloma virus (HPV) tests were taken by trained nurses with a standardized technique. For persons to remain in the study, we required that Pap smear on day of study entry be negative. Only one subject included in this group had an abnormal result in Pap study (CIN 2) and in 13 cases there was not sufficient cervical material for a good reading on the slide; thus, they were eliminated. Cases and healthy women were residents of Mexico City and they were invited to participate as an additional part of their clinic visit. Women were asked to sign informed consent, to participate in an interview, and to provide cervical material for the HPV test. Social, clinical, and reproductive information was obtained from each patient at a standard interview. The final study sample included 182 cases with the following CIN diagnosis distribution: 102 (56.1%) were CIN 1, 31 (17%) CIN 2, and 49 (26.9%) CIN 3. Cases and controls were adjusted in sample size (n0,n1 ⫽ 182). The study protocol was approved by institutional ethical and research review boards of the participating institutions in Mexico (17-98-HO; 000161CI/98). HPV sampling procedure. A sample was taken from endocervix using a rotary motion with a Cytobrush (Digene Cervical Sampler, Digene Corp., Gaithersburg, MD, USA). Specimens for HPV DNA testing were stored at ⫺4°C and sent to the Molecular Biology Laboratory for masked highrisk HPV test using the Hybrid Capture II method [20]. Specimens were denatured and liberated single-stranded DNA was hybridized in solution with a bionucleic acid (RNA) probe mix consisting of high-risk HPV types (16,18,31,33,35,39,45,51,52,56,58,59,68). Each mixture reaction containing RNA/DNA hybrids formed was trans-

312

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

Table 1 Quantitative characteristics of healthy women (controls) and with CIN 1 and CIN 2–3 in Mexico City Variables

Controls X ⫾ (SD)a n ⫽ 182

CIN 1 X ⫾ (SD)a n ⫽ 102

CIN 2–3 X ⫾ (SD)a n ⫽ 80

P

Age (years) Education (years) Income budget (m) Age at first sexual intercourse (years) Sexual partners (n) Deliveries (n) Cesarean (n) Oral contraceptives (m)b Condom use (m) Age of onset of smoking (years)b Cigarettes/dayb HPV ratio light units (RLU)c

38.0 (10.9) 9.6 (3.8) 2,920 (2545) 20.0 (4.0) 1.5 (1.0) 1.7 (2.1) 0.6 (1.0) 32.7 (45.4) 7.6 (26.0) 20.2 (5.4) 3.1 (3.0) 55.8 (114)

37.0 (10.5) 9.0 (4.4) 1,635 (2774) 19.3 (3.8) 1.6 (1.0) 2.3 (2.1) 0.5 (1.1) 54.2 (76.6) 4.4 (20.5) 18.6 (4.9) 6.0 (7.9) 165.4 (546)

38.4 (9.6) 7.2 (4.0) 2,021 (1,904) 18.1 (3.4) 1.7 (1.0) 3.3 (2.4) 0.2 (0.5) 43.5 (38.5) 2.5 (8.2) 19.2 (7.3) 6.3 (7.5) 529.2 (700)

0.7 ⬍0.001 ⬍0.001 ⬍0.001 0.1 ⬍0.001 0.01 0.05 0.02 0.08 0.03 ⬍0.001

Median ⫾ standard deviation (SD). Only users. c Relative light units in positive HPV DNA; n ⫽ frequency. a

b

ferred to a capture tube coated with antibodies against RNA/DNA hybrids. Unreacted material was removed by washing and dioxetane-based chemiluminescent substrate was added, which binds to alkaline phosphatase. Light produced by the ensuing reaction was measured by a luminometer (DML 2000, Digene Corp.). Light measurements were expressed as relative lights units (RLU). Solutions of highrisk HPV at 10 pg/ml served as positive controls. All RLU measurements of specimens were divided by RLU of appropriate positive controls to provide a ratio. Specimen ratio of ⱖ1.0 was regarded as positive for HPV DNA, while a ratio ⬍1.0 was regarded as negative. Because the amount of light produced by hybrid capture assay is proportional to the amount of target DNA in each specimen, results are quantitative: the higher the ratio, the greater the amount of target HPV DNA in specimen, this considering equal distribution of virology-infected cells and number of cells obtained [20]. Data analysis. Data were validated by frequencies and range checks and analyzed with a software package (STATA PC) [21]. A nonparametric test (Kruskal–Wallis test X2) was used to compare quantitative variables. HPV test results and final histologic diagnosis from each subject were compiled and analyzed by a standard contingency table approach (X2 or Fisher exact test). Associations between various study parameters and risks of CIN were determined by comparison with control subjects in univariate analysis. Risk was estimated by calculating the odds ratio (OR) as an approximation of relative risk (RR) with 95% confidence interval (CI 95%) [22]. Viral load HPV DNA was measured as RLUs and agrees with previous specifications in the hybrid capture assay. Viral quantification data in RLUs were initially continuous measurements but were categorized for analysis. Positive HPV cases were separated and agreed with their percentile distribution in three equal groups. RLUs were transformed

into their logarithm (Log10) to provide graphics. Finally, four groups were considered as follows: negative HPV DNA (RLUⱕ0); low viral load (Log10RLU ⫽ 0.01–1.00); middle viral load (Log10RLU ⫽ 1.01–2.0), and high viral load (Log10RLU ⫽ 2.01–3.60). Unconditional multinomial regression was used to calculate adjusted risks to CIN 1/CIN 2–3 [21].

Results Many social, economic, and reproductive risk factors associated with increased risk for cervical carcinoma were identified in women with CIN 1 and CIN 2–3 compared with healthy women. Differences in risk factors among these three groups are shown in Table 1. Healthy women presented higher years of education and family income than cases (P ⬍ 0.001). Women with diagnosis of CIN 1/CIN 2–3 began sexual activity earlier than controls and had a similar number of sexual partners (P ⬍ 0.001, P ⫽ 0.1). We observed differences in number of deliveries and frequency of cesarean sections in cases and controls (P ⫽ ⬍ 0.001, P ⫽ 0.01). On average, time of condom use was higher in the control group than in the other two groups of cases, but great variability also was observed. Use time of oral contraceptives was not different among groups (P ⫽ 0.05), nor was age of onset of smoking (P ⫽ 0.08); however, amount of cigarettes smoked per day was different (P ⫽ 0.03). An important difference was observed in viral load among CIN 1, CIN 2–3, and controls (P ⬍ 0.001) (Table 1). The most important risk factors associated with precursor lesions of CC adjusting by HPV infection are shown in Table 2. A higher frequency (41.2%) of CIN 1 was observed in the younger group of women (42/102), while 45% of CIN 2–3 cases (36/80) occurred in the 35– 44 years of age group. Statistical differences were identified in prevalence to

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

313

Table 2 Social and reproductive factors associated to CIN stage adjusted by HPV DNA Variable Reference group Age (years) ⬍35 35–44 45–70 Place of birth Mexico City Other School grade (years) 7–23 0–3 4–6 Income budget ⱖ3 ⬍3 Smoking Negative Yes Familial history Negative Cervical cancer Other First sexual intercourse ⱖ18 ⬍18 Deliveries (n) None 1–2 ⱖ3 Sexual partners (his/her) 1 2–3 ⱖ4 Oral contraceptives Negative Positive Condom use Ever Never

Controls n0

CIN 1 n1

CIN 2–3

ORa

CI 95%

n2

ORa

CI 95%

68 71 43

42 36 24

1 0.6 0.3

0.3–1.1 0.1–0.7*

26 36 18

1 0.8 0.3

1 0.4–1.8 0.1–0.8*

131 51

58 44

1 2.3

1.2–4.4*

43 37

1 2.7

1 1.3–5.6*

135 10 37

69 14 19

1 1.9 0.9

0.7–5.4 0.4–2

40 17 23

1 3.8 2.0

1 1.2–11.3* 0.9–4.6

63 119

25 77

1 2.2

1.1–4.3*

13 67

1 3.9

1 1.7–8.8*

92 90

55 47

1 0.97

0.5–1.7

55 25

1 0.5

1 0.2–1.2

125 15 40

64 17 20

1 2.3 0.7

0.9–5.9 0.3–1.6

51 12 15

1 2.0 0.7

1 0.7–5.9 0.3–1.5

133 49

65 37

1 1.6

0.8–3

39 41

1 3.0

1 1.5–6.1*

69 60 53

23 37 42

1 1.9 2.1

0.7–4 0.9–4.4

7 27 46

1 4.5 7.2

1 1.6–12.8* 2.6–19.9*

97 51 34

32 35 35

1 2.1 3.3

1–4.4 1.5–7.2*

28 30 22

1 2.1 2.4

1 0.9–4.6 1.0–5.8

110 72

73 29/28.4

1 0.4

0.2–0.9*

60 20

1 0.4

1 0.2–0.7*

38 144

11 91

1 1.3

0.5–3.1

9 71

1 1.1

0.4–2.9

Note. OR, odds ratio adjusted by high risk HPV (16,18,31,33,35,39,45,51,52,56,58,59,68) in multinomial analysis. n ⫽ frequency.

HPV infection in controls and cases of CIN 1–3 by age groups (P ⫽ 0.001). Prevalence of HPV in healthy women was increasing with age group, while CIN 1–3 patients showed similar frequencies between age groups and higher frequencies than groups of control women (⬎78%) (Fig. 1). All positive HPV control women had lower viral load than cases; however, in both cases and controls older women presented highest viral load (Fig. 2). Place of birth different from Mexico City, low educational level, and low family income had higher probabilities for association with CIN 2–3 lesions. It is possible that these results were a reflection of fewer opportunities to access timely medical care attention for these women. Reproductive risk factors had higher association with late stage of cervical lesions (CIN 2–3). We did not observe association among smoking, positive family cancer history, and con-

dom use. However, condom use was a protective variable to HPV infection (ORa ⫽ 0.42, 0.2– 0.9) independently of CIN stage. Use of oral contraceptives showed a protector factor to CIN independently of HPV infection, but when other confounding variables were considered, association was lost (Table 2). Frequency of HPV DNA infection in control group was 19.8% (36/182), 79.4% in CIN 1 (81/102), and 93.7% in the CIN 2–3 women’s group (75/80). Adjusted risk ratio for HPV infection in CIN 1 women was 26.6 (12.9 –54.7), while for CIN 2–3 it was 111.4 (35.2–352) (Table 3). Association among CIN 1, CIN 2–3, and viral load is depicted in Table 3. We observed a significantly increased risk of CIN stage with higher exposure in viral load. Frequency of low viral load observed in the control group was 9.9% (18/182) and 26.5% of CIN 1 women, obtaining an

314

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

statistically significant trend (X2 trend ⫽ 140.7, P ⬍ 0.001) (Fig. 3).

Discussion

Fig. 1. Frequency distribution of positive HPV infection in controls and cases by age groups.

OR of 16.8 (7.2–39). Association between low viral load and CIN 2–3 was 23.6 (6.4 – 88). While percentage of high viral load in controls was 3.3% (6/182), 63.8% (51/80) of CIN 2–3 was observed, demonstrating an important increase of risk ratio ⫽ 365.8 (94.7–1412). Intermediate risk values were observed in association of middle viral load and both groups of cases, CIN 1 and CIN 2–3. Adjusted for age, sexual partners, and deliveries, association between viral load and CIN had a strong positive and

Findings of this study suggest that the amount of HPV DNA detected by hybrid capture is highly related with CIN stage of lesion. Several authors have reported a positive association between viral burden of HPV with high-grade intraepithelial lesion [23,24]. Consistent results were observed when using either the hybrid capture method or the low-stringency PCR method in follow-up studies [23–26]; nonetheless, not all researchers found this condition [27,28]. In this study, only 3% of healthy women presented high viral load, while 63.8% of women with CIN 2–3 showed highest frequency in high viral load. On the other hand, frequency of women with a low amount of HPV was 9.9%, and more than twice the number of CIN 1 cases (26.5%) and CIN 2–3 (23.6%) presented this condition. The increasing amount of HPV in higher CIN stage led us to observe a significant trend in the odds ratio, supporting a hypothesis of dose–response association. Results of this study were obtained with a single measurement of exposure; therefore, it was not possible to know whether an ordinal measure of HPV DNA could be a predictor of progression. However, results were consistent with some findings observed in follow-up study designs that studied viral load of HPV and development of CC. These studies demonstrated that the effect of longer-term persis-

Fig. 2. Human papilloma virus DNA load in control and women with CIN by age group.

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

315

Table 3 Multinomial analysis in HPV DNA load and women with cervical intraepithelial neoplasia (CIN) stage Viral load (Log10RLU)

Negative Low (0.01–0.99) Middle (1.0–1.99) High (2.0–3.6) HPV positive

Controls

CIN 1

CIN 2,3

n0 ⫽ 182

n1 ⫽ 102

ORa

CI 95%

n2 ⫽ 80

ORa

CI 95%

146 18 12 6 36

21 27 36 18 81

1.0 16.8 35.7 29.9 26.6

7.2–39 14.1–91 9.7–92 12.9–54.7

5 8 16 51 75

1.0 23.6 71.2 365.8 111.4

6.4–8.8 19.4–261 94.7–1412 35.2–352

Note. Control was comparison group; odds ratios were adjusted by age, deliveries, and sexual partners.

tence and viral load, especially in high-risk HPV types, had a much greater risk of incidence of high-grade squamous intraepithelial lesion (HSIL) and CC than women negative to HPV [9,24,25]. A consistently increased load of HPV 16 already ⬎10 years prior to diagnosis has been reported even when many smears were cytologically normal [12,13]. However, these authors had the same opinion concerning low predictive value to viral load HPV in cervical cancer diagnoses. Variation of amount of HPV in women could be explained by both environmental factors and host genetic factors. Certain environmental factors such as smoking, use of oral contraceptives, and number of pregnancies have shown a relationship with CC and precursor lesions, taking into consideration HPV infection [29 –31]. Although smoking has been identified as a cofactor in development of CC and has been related with higher genital HPV infection rate [29,32], we did not find a significant association with HPV

infection and/or viral load with CIN, as several other studies have shown [33,34]. We observed that use of oral contraceptives had a significant protective effect with CIN 1/CIN 2–3 even when we adjusted analysis by HPV infection; however, this association lost significance when other cofactors were considered for measuring association between viral load and CIN 1/CIN 2–3 in multinomial analyses. Studies concerning oral contraceptive use and cervical cancer have not been consistent in their findings [35,36]. Some have identified strong evidence of this association with adenocarcinoma of cervix but not with squamous cervical cancer [30,37]. However, other authors consider that studies not restricted to women positive for HPV could have underestimated the effect of oral contraceptives in squamous-cell carcinomas [31]. Significant associations between high parity and risk of invasive and in situ carcinomas of the cervix have been

Fig. 3. Human papilloma virus DNA load in CIN lesion degree. Median viral load shows significative differences between them (Kruskal–Wallis test X2 ⫽ 52.2, P ⬍ 0.001).

316

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

reported even in women who were all positive to HPV infection [38,39]. In accordance with these studies, we observed that number of deliveries was the most important cofactor in association of HPV infection and CIN 1/CIN 2–3; this association was maintained when the amount of HPV DNA was considered in the analysis. Higher exposure of a transformation zone observed on exocervix during the last trimester of pregnancy could be a condition that facilitates direct exposure to HPV infection [38,40]. However, additional cofactors such as immune response of women must be considered, assuming that genetic susceptibility to HPV infection has been reported [41]. In our data, we observed a higher frequency of CIN 2–3 in the 36 – 45 year of age group as the expected prevalence [6,34]. Frequency of HPV infection in healthy women has been reported to be relatively high in younger women and to decline with increasing age [42]. Contrary to these results, we observed higher frequencies of HPV in women in the oldest age as compared with controls, while similar frequencies were observed in cases independently of age and CIN stage. These findings are similar to those observed in a rural Costa Rican population except for frequency observed in younger group. In this study, authors identified HPV infections that peaked first in women ⬍25 years of age and noted increased prevalence after menopause among women with normal cytology [43]. These differences observed in prevalence of HPV infection in young women must be explained by a possible biased selection of participants in our study in that one criteria to be invited to participate was that the women did not have cervico–vaginal symptoms. This is a frequent condition in women with higher exposure to sexually transmitted diseases [42]. Another limitation of this study was the possibility of misclassifying cases, especially with CIN 2 diagnosis, in which we observed the lowest agreement between the two pathologists. For that reason, we decided to consider women with CIN 2 and CIN 3 in one group, so as to improve agreement in case classification. Cervical swabs were tested by HPV DNA by Probe B of Hybrid Capture II, an approved assay for HPVs. This method detects 13 cancer-associated HPV types and two other recently classified high-risk HPV types were not identified in the study (HPV 73 and HPV 82); however, there is evidence that ⬎95% of CC are associated with these genotypes [4]. On the other hand, it has been reported that women infected with persistent infection of HPV 16 or 18 had considerably higher relative risks (RR ⫽ 8.6) for any squamous intraepithelial lesions compared with women negative to high-risk HPV [12]. In our study, we did not determine genotypes of HPV and it was not possible to relate viral load of HPV 16 or 18 with CIN 1/CIN 2–3. Several studies have evaluated the relative utility of both cytology test and HPV DNA testing compared with the cytology test alone as a primary cervical cancer-screening tool [24,44]. Reported results showed that concurrent normal cytology test and negative HC II result have substan-

tially decreased risk of high-grade lesions on colposcopy relative to those for whom the sole screening information is a normal conventional cytology result [45]. On the other hand, a positive HC II result is not an absolute indicator that high-grade lesions exist or will develop. The prognostic value of a positive test result, especially in the absence of cytologic abnormality, has not been fully validated in prospective studies. Therefore, we think that if additional markers were used in women with abnormal Pap, such as genotypes of HPV and viral load, selection of high-risk women to develop cervical cancer could improve. In conclusion, we observed high association between a quantitative method of high-risk HPV and precursor lesions of cervical cancer in this population. But due to inconsistent results observed with its use, the utility of this ordinal measure must be studied.

Acknowledgments This work was partially supported by grants from IMSS, FOFOI/0038/437 and 0038/439.

References [1] Parkin MD, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden: Globocan 2000. Int J Cancer 2001;94:153– 6. [2] Secretaria de Salud. Programa de prevencio´ n y control del ca´ ncer cervico uterino 1998 –2000. Me´ xico: Secretaria de Salud 1998; p.5. [3] Kjaer SK, van den Brule AJC, Bock JE, et al. Human papillomavirus—the most significant risk determinant of cervical intraepithelial neoplasia. Int J Cancer 1996;65:601– 6. [4] Mun˜ oz N, Bosch FX, de Sanjose´ S, Herrero R, Castellsague X, Shah KV, Snijders PJF, Meijer CJLM. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 2003;348:518 –27. [5] Mun˜ oz N, Bosch FX. HPV and cervical neoplasia: review of casecontrol and cohort studies. IARC Sci Publ 1992;119:251– 61. [6] Bosch FX, Manos M, Mun˜ oz N, Sherman M, Jansen AM, Peto J, Schifman MH, Moreno V, Kurman R, Shah KV. International biological study on cervical cancer (IBSCC) study group. Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. J Natl Cancer Inst 1995;87:796 – 802. [7] zur Hausen H. Papillomavirus infections—a major cause of human cancers. Biochim Biophys Acta 1996;1288:F55–F78. [8] Ho GYF, Bierman R, Beardsley L, Chang CJ, Burk RD. Natural history of. cervicovaginal papillomavirus infection in young women. N Engl J Med 1998;338:423– 8. [9] Schlecht FN, Kulaga S, Robitaille J, Ferreira S, Santos M, Miyamura AR, Duarte-Franco E, Rohan ET, Ferenczy A, Villa LL, Franco LE. Persistent human papillomavirus infection as a predictor of cervical intraepithelial neoplasia. JAMA 2001;286:3106 –14. [10] Woodman CB, Collins S, Winter H, et al. Natural history of cervical human papillomavirus infection in young women: a longitudinal cohort study. Lancet 2001;357:1831– 6. [11] Sun CA, Lai HC, Chang CC, Neih S, Yu CP, Chu TY. The significance of human papillomavirus viral load in prediction of histologic severity and size of squamous intraepithelial lesions of uterine cervix. Gynecol Oncol 2001;83:95–9. [12] Ylitalo N, Sorensen P, Josefsson AM, Magnusson PKE, Kragh AP, Ponten J, Hans-Olov A, Gyllensten UB, Melbye M. Consistent high

D.M. Herna´ ndez-Herna´ ndez et al. / Gynecologic Oncology 90 (2003) 310 –317

[13]

[14]

[15]

[16]

[17]

[18] [19]

[20]

[21] [22] [23]

[24]

[25]

[26]

[27]

[28]

[29]

viral load of human papillomavirus 16 and risk of cervical carcinoma in situ: a nested case-control study. Lancet 2000;255:2194 – 8. Josefsson AM, Magnusson PKE, Ylitalo N, Per S␾rensen, Qwarforth-Tubbin P, Per Krag A, Melbye M, Hans-Olov A, Gyllensten UB. Viral load of human papilloma virus 16 as a determinant for development of cervical carcinoma in situ: a nested case-control study. Lancet 2000;355:2189 –93. Castle PE, Lorincz AT, Mielzynska-Lohnas I, Scott DR, Glass AG, Sherman ME, Schussler JE, Schiffman M. Results of human papillomavirus DNA testing with the hybrid capture 2 assay are reproducible. J Clin Microbiol 2002;40:1088 –90. Venturoli S, Cricca M, Bonvicini F, Giosa F, Pulvirenti FR, Galli C, Musiani M, Zerbini M. Human papillomavirus DNA testing by PCRELISA and hybrid capture II from a single cytological specimen: concordance and correlation with cytological results. J Clin Virol 2002;25:177– 85. Howard M, Sellors J, Kaczorowski J. Optimizing the hybrid capture II human papillomavirus test to detect cervical intraepithelial neoplasia. Obstet Gynecol 2002;100:972– 80. Sato S, Matunaga G, Tsuji I, Yajima A, Sasaki H. Determining the cost-effectiveness of mass screening for cervical cancer using common analytic models. Acta Cytol 1999;43:1006 –14. Davey DD, Armenti CA. HPV primary screening for cervical cancer: more pain than protection. Diagn Cytopathol 2000;22:333–5. Scully RE, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkerson EJ. Histological typing of female genital tract tumours. In: World Health Organization International Histological Classification of Tumours. Berlin: Springer-Verlag; 1994, p. 39 –54. Digene Corporation. Digene HPV test hybrid capture II test. In vitro test. A signal amplified hybridization microplate assay for the chemiluminescent detection of human papillomavirus, 1998. STATA. Statistics data analysis. Stata Corporation, 1984 –2001. Schsselmann JJ. Case-control studies. Design, conduct, analysis. New York: Oxford Univ. Press, 1982, p. 124 – 43. Abba MC, Mouron SA, Gomez MA, Dulout FN, Golijow CD. Association of human papillomavirus viral load with HPV16 and highgrade intraepithelial lesion. Int J Gynecol Cancer 2003;13:154 – 8. Castle PE, Wacholder S, Sherman ME, Lorincz AT, Glass AG, Scott DR, Rush BB, Demuth F, Schiffman M. Absolute risk of a subsequent abnormal pap among oncogenic human papillomavirus DNApositive, cytologically negative women. Cancer 2002;95:2145–51. Sun CA, Liu JF, Wu DM, Nieh S, Yu CP, Chu TY. Viral load of high-risk human papillomavirus in cervical squamous intraepithelial lesions. Int J Gynaecol Obstet 2002;76:41–7. Schlecht NF, Trevisan A, Duarte-Franco E, Rohan TE, Ferenczy A, Villa LL, Franco EL. Viral load as a predictor of the risk of cervical intraepithelial neoplasia. Int J Cancer 2003;103:519 –24. Lorincz AT, Castle PE, Sherman ME, Scott DR, Glass AG, Wacholder S, Rush BB, Gravitt PE, Schussler JE, Schiffman M. Viral load of human papillomavirus and risk of CIN3 or cervical cancer. Lancet 2002;360(9328):228 –9. Bory JP, Cucherousset J, Lorenzato M, Gabriel R, Quereux C, Birembaut P, Clavel C. Recurrent human papillomavirus infection detected with the hybrid capture II assay selects women with normal cervical smears at risk for developing high grade cervical lesions: a longitudinal study of 3,091 women. Int J Cancer 2002;102:519 –25. Prokopczyk B, Cox JE, Hoffmann D, Waggoner SE. Identification of tobacco-specific carcinogen in the cervical mucus of smokers and nonsmokers. J Natl Cancer Inst 1997;89:868 –73.

317

[30] Lacey JV Jr., Brinton LA, Abbas FM, et al. Oral contraceptives as risk factors for cervical adenocarcinomas and squamous cell carcinomas. Cancer Epidemiol Biomarkers Prev 1999;8:1079 – 85. [31] Mun˜ oz N, Franceschi S, Bosetti C, Moreno V, Herrero R, Smith JS, Shah KKV, Meijer CJL, Bosch XF. Role of parity and human papillomavirus in cervical cancer: the IARC multicentric case-control study. Lancet 2002;359:1093–101. [32] Eppel W, Worda C, Frigo P, Manavi M, Czerwenka K. The influence of cotinine on interleukin 6 expression in smokers with cervical preneoplasia. Acta Obstet Gynecol Scand 2000;79:1105–11. [33] Olsen AO, Gjoen K, Sauer T, Orstavik I, Naess O, Kierulf K, Sponland G, Magnus P. Human papillomavirus and cervical intraepithelial neoplasia grade II–III: a population-based control-study. Int J Cancer 1995;61:312–5. [34] Yoshikawa H, Nagata C, Noda K, Nozawa S, Yajima A, Sekiya S, Sugimori H, Hirai Y, Kanazawa K, Sugase M, Shimizu H, Kawana T. Human papillomavirus infection and other risk factors for cervical intraepithelial neoplasia in Japan. Br J Cancer 1999;80:621– 4. [35] Rolon PA, Smith JS, Mun˜ oz N, et al. Human papillomavirus infection and invasive cervical cancer in Paraguay. Int J Cancer 2000;85:486 – 91. [36] Bosch FX, Mun˜ oz N, de Sanjose´ S, et al. Human papillomavirus and cervical intraepithelial neoplasia grade III/carcinoma in situ: a casecontrol study in Spain and Colombia. Cancer Epidemiol Biomarkers Prev 1993;2:415–22. [37] Brinton AL, Tashima TK, Lehman HF, Levine RS, Mallin K, Savitz DA, Stolley PD, Fraumeni JF. Epidemiology of cervical cancer by cell type. Cancer Res 1987;47:1706 –11. [38] Parazzini F, Catenoud L, La Vecchia C, Negri E, Franceschi S, Bolis G. Determinants of risk of invasive cervical cancer in young women. Br J Cancer 1998;77:838 – 41. [39] Mun˜ oz N, Francheschi S, Bosetti C, Moreno V, Herrero R, Smith JS, Shah KV, Meijer CJLM, Bosch XF. Role of parity and human papillomavirus in cervical cancer: the IARC multicentric case-control study. Lancet 2002;359:1093–101. [40] Palle C, Bangsboll S, Andreasson B. Cervical intraepithelial neoplasia in pregnancy. Acta Obstet Gynecol Scand 2000;79:306 –10. [41] Odunsi K, Terry G, Ho L, Bell J, Cuzick J, Ganesan TS. Susceptibility to human papillomavirus-associated cervical intra-epithelial neoplasia is determined by specific HLA DR-DQ alleles. Int J Cancer 1996;67:595– 602. [42] Figueroa JP, Ward DE, Luthi TE, Vermund SH, Brathwaite AR, Burk RD. Prevalence of human papillomavirus among STD clinic in Jamaica; association of younger age and increased sexual activity. Sex Transm Dis 1995;22:114 – 8. [43] Herrero R, Hildesheim A, Bratti C, Sherman ME, Hutchinson M, Morales J, Balmaceda I, Greenberg MD, Alfaro M, Burk RD, Wacholder S, Plummer M, Schiffman M. Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst 2000;92:464 –74. [44] Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, Cohen C. American Cancer Society guideline for the early detection of cervical neoplasia and cancer. J Lower Genital Tract Dis 2003;7:67– 86. [45] Wright Jr TC, Schiffman M. Adding a test for human papillomavirus DNA to cervical cancer screening. N Engl J Med 2003;348: 489 –90.