Agradecimentos

Comparing Pap smear cytology, aided visual inspection, screening colposcopy, cervicography and HPV testing as optional screening tools in Latin Study design and baseline data of the LAMS study. Anticancer Research 2005, 25(5): 3469-80, interval. Women with positive and negative HPV test were compared regarding sociodemographic and reproductive factors using relative risk (RR) and stepwise logistic regression analysis calculated with 95% confidence interval (CI). Also the incidence rate and RR of developing any cytological or histological abnormality during the follow-up were calculated within 95% confidence limits. Colposcopy was considered as the gold standard. When colposcopy result was normal or biopsy result was cervicitis it was considered as negative diagnosis. Women with histologic diagnosis of cervical intraepithelial neoplasia (CIN) 1 or higher were considered as positive diagnosis. Results : Incidence of low and high-grade cytological lesion was higher in women with positive HPV testing than in women with negative HPV testing after 12 and 24 months of follow-up. In up to 12 months of follow-up, women with baseline positive HPV test had a significantly higher proportion of low-grade (1.4; 95% CI 1.1-1.7) and high -grade (1.5; 95%CI 1.4-1.7) cytological lesion. The RR for high-grade lesion increased to 1.7 (95%CI 1.5-1.9) for those followed-up in up to 24 months. For histological outcomes, the incidence of CIN 1, 2 or 3 was also higher in women with positive HPV testing than in women with negative HPV testing after 12 and 24 months of follow-up. Women with positive HPV test had a higher RR of CIN 2 and 3 (1.5; 95%CI 1.4-1.6) during the follow-up in up to 12 months and the RR increased to 1.7 (95%CI 1.5-1.9) for those followed-up in up to 24 months. Conclusions : HPV test is useful in addition to cytology to select from women with normal cytology those who are at highest risk for underlying cervical lesion. Objective: To investigate the incidence of cytological and histological cervical lesions in a 24 months follow-up, according to HPV detection among women with baseline normal cytology, in a subgroup of women included in the Latin American Screening study (LAMS). Study design: A group of 365 women with normal Pap smear with negative and positive high-risk Hybrid Capture II test were prospectively followed in up to 24 months at Campinas e São Paulo (Brazil). The relative risk (RR) of developing any cytological or histological abnormality during the follow-up was calculated with 95% confidence interval (CI). Results: Until 12 months of follow-up, women with baseline positive HPV test had a significantly higher proportion of low-grade (1.4; 95%CI 1.1-1.7) and high-grade (1.5; 95%CI 1.4-1.7) cytological lesion. The RR for high-grade lesion increased to 1.7 (95%CI 1.5-1.9) for those followed in up to 24 months. For histological outcomes, women with positive HPV test had a higher proportion of cervical intraepithelial neoplasia 2 and 3 (1.5; 95%CI 1.4-1.6) during the follow-up in up to 12 months and the RR increased to 1.7 (95%CI 1.5-1.9) for those followed in up to 24 months. Conclusions: HPV test is useful in addition to cytology to select from women with normal cytology those who are at highest risk for underlying cervical lesion.

Palavras-chave: citologia, estudo de coorte, papilomavírus humano, neoplasia intra-epitelial cervical, seguimento, colo uterino, câncer. for high-grade lesion increased to 1.7 (95%CI 1.5-1.9) for those followed-up in up to 24 months. For histological outcomes, the incidence of CIN 1, 2 or 3 was also higher in women with positive HPV testing than in women with negative HPV testing after 12 and 24 months of follow-up. Women with positive HPV test had a higher RR of CIN 2 and 3 (1.5; during the follow-up in up to 12 months and the RR increased to 1.7 (95%CI 1.5-1.9) for those followed-up in up to 24 months.
Conclusions: HPV test is useful in addition to cytology to select from women with normal cytology those who are at highest risk for underlying cervical lesion.

Introduction
Several epidemiological studies in the last decade have established a strong relation between high-risk types of human papillomavirus (HR-HPV) and the development of cervical cancer and precancerous lesions [1,2]. Indeed, it has been reported that 99.7% of all cervical squamous cells carcinomas contain HR-HPV [1].
Insight into this relation between HR-HPVs and the natural history of squamous intraepithelial lesions (SIL), it has been proposed that testing for the presence of these viruses in cervical scrapes as an adjunct of cervical cytology could be of great value in a variety of clinical settings [2]. The commercial available Hybrid Capture (HC) II, is a non radioactive, rapid, reproductive and sensitive assay, designed to detect 18 HPV types divided into high-risk (types 16,18,31,33,35,39,45,51,52,56,58, 59 e 68) and low-risk (types 6,11,42, 43 e 44) groups [3,4,5].
HC II test is being proposed as a general population screening test in conjunction with the Papanicolaou (Pap) test for women 30 years of age and older, as reassurance of the absence of high-grade SIL (HSIL) or invasive carcinoma [6,7]. However, most HPV DNA positive women are cytologically negative, considering that most HPV infections are transient and not associated with detectable cytological abnormality. Many authors consider that a positive HPV testing selects a population of highest risk of developing a HSIL, while a concurrent negative Pap smear and negative HC II test have a substantially decreased risk of cervical lesion, providing strong reassurance that incident disease will be absent [6,8,9,10,11]. Consequently, the knowledge about clinical significance of an HPV-positive, cytologic-negative result with regard to prediction of subsequent cervical lesions is still needed.
Thus, we carried out a prospective study to investigate the role of HR-HPV detected at enrollment visit in the development of SIL in women who had no previous diagnosis of cytological abnormalities, during a 24-month follow-up.

Study design
The subjects of this study represent a group of women derived from a major  Figure 1 shows the different exclusions and follow-up outcomes for the entire cohort.

Follow-up Visits
Follow-up consultations comprised clinical history and gynecological examination with Pap smear and colposcopy in a six-month interval. Targeted biopsies of the cervix were taken when abnormal epithelium was found. Accordingly, the same routine was adopted at following visits. Altogether 287 women completed the 6 and 12month follow-up visit and 234 completed the 18 and 24-month follow-up visit.

Cervical cytology
Screening cervical cytology was tested in two modes: conventional and liquidbased cytology (LBC) techniques. Conventional Pap smear was taken by Campinas centre in 285 women, following conventional procedures of smear taking, fixation and staining. The system of LBC was tested in 80 women at Sao Paulo centre using DNA-Cytoliq® (Digene Brazil, Sao Paulo Brazil). The samples were collected using the brush of DNA-Cytoliq system and immersed in the Universal Collecting Medium vials. The sample processing followed the manufacturer's instructions [13]. At the follow-up visits, only conventional cytology was performed. Slides were fixed in absolute ethanol and stained by the modified Papanicolaou method. Final cytological diagnoses were rendered using the Bethesda System [14] and were classified as normal/inflammatory, ASC, LSIL or HSIL. For statistical purposes, normal/inflammatory exams were grouped as negative and ASC, LSIL or HSIL as positive.

Colposcopy
Colposcopy was performed for all women who comprised follow-up visits.
Experienced and certified colposcopists performed all examinations. The classification was according to the terminology of the International Federation of Cervical Pathology and Colposcopy (IFCPC) [15].

Histology
Biopsy specimens were fixed in 10% phosphate buffered formalin, embedded in paraffin, and stained with hematoxylin and eosin (HE). Pathological diagnoses were rendered according to the World Health Organization [16] criteria's and classified as negative, CIN 1, CIN 2 or CIN 3. For statistical analysis, results of cervical biopsies were grouped as negative when normal/cervicitis was present and positive when showing CIN 1, CIN 2 or CIN 3.

Statistical analysis
We calculate the association between HPV-positive test and some reproductive and sociodemographic factors using relative risk (RR) with stepwise logistic regression analysis. We also calculated the incidence rate and RR of developing any cytological or histological abnormality during the follow-up among women with HPV-positive or HPV-negative test at enrollment. Colposcopy was considered as the gold standard. When colposcopy result was normal or biopsy result was cervicitis it was considered as negative diagnosis. Women with histological diagnosis of CIN 1 or higher were considered as positive diagnosis. All calculations were performed with EpiInfo V3.0 (Centers for Disease Control, Atlanta, USA) within 95% confidence limits.

Results
The women in the study group ranged between 15 and 65 years of age with the This risk was not associated with women who never smoke, who were former smoker or who smoked in the past. After the logistic regression analysis, only women who had more than two lifetime sexual partners (RR=1.5 95%CI 1.0-2.2) remained significantly associated with a positive HPV testing (Table 1).
We examined the incidence rate and the relative increase in risk of developing any cytological abnormality during a 24-month follow-up among those women who were cytologically negative and either HPV-positive or HPV-negative at enrollment. In up to 12 months of follow-up, the incidences of LSIL and HSIL were 4.6% and 1.0%  (Table 2).
We repeated the analysis separately for histological outcomes. In up to 12 months of follow-up, the incidences of CIN 1 and CIN 2/3 were 6.2% and 2.6% respectively for women with positive HPV test, compared to 3.2% and 0% for those with negative HPV test result. In up to 24 months, the incidences of CIN 1 and CIN 2/3 decreased to 2.1% and 0.7% respectively for women with positive HPV test. Women with baseline normal cytology and positive HPV test had a significantly higher RR of CIN 2 and 3 (1.5; 95%CI 1.4-1.6) during the follow-up in up to 12 months. At the follow-up in up to 24 months, the RR of CIN 1 (1.7; 95%CI 1.5-1.8) was also remarkably higher. There was observed a similar trend of increasing RR of CIN 3 over time, from 1.5 (95%CI 1.4-1.6) for women followed-up in up to 12 months to 1.7 (95%CI 1.5-1.8) for those followed-up in up to 24 months. Similarly to cytological analysis, in HPV-negative women followed-up in up to 12 months, two cases of CIN 1 were observed and there was no CIN 3 diagnosed over time (Table 3).

Discussion
To date, the detection of cervical cancer and its precursors by Pap test is widely recognized as the most effective method for preventing carcinoma of the cervix [2,10,17]. However, cytological screening suffers from limited reproducibility and high rate of false negative and false positive results [18,19]. This limitation is further compounded by a substantial number of Pap smears being reported as negative that may represent an unrecognized precursor to atypia in many cases [20].
To overcome this problem, with the knowledge about the role of HPV in natural history of cervical lesions, it has been suggested that testing for HPV as an adjunct of cytology could dramatically improve the detection of cervical disease [2,11].
Furthermore, the fact the HPV test has high sensitivity and low specificity and Pap smear low sensitivity and high specificity could be used to support each other.
In the present study, the overall prevalence of HR-HPV in women with negative Pap smear was 14%, whereas others have observed HPV prevalence ranging from 10% [21] to 21% [22]. Risk for HPV infection was associated with single women, age at first intercourse, number of lifetime sexual partners and partner with previous sexual transmitted disease. It suggests that sexual behavior increases the probability of contact with an infected partner and reflects the important role of sexually transmitted agents in the etiology of cervical lesions [21,23,24,25]. However, hormonal contraceptive use, multiparity and smoking, factors that are thought to influence the likelihood of HPV infection [26,27] did not show this association in the population studied. After logistic regression analysis, only the number of lifetime sexual partners remained associated with having an HPV infection.
The incidence rates for HSIL cytology and CIN 2/3 in women with positive HPV testing were respectively 1.0% and 2.6% in women followed-up in up to 12 months. For those followed-up in up to 24 months, the incidence rate for HSIL cytology increased to 2.1% and the incidence rate for CIN 2/3 decreased to 0.7%.
This study also indicates that over the entire duration of follow-up, women with baseline negative Pap smear and positive HPV test are more likely to develop HSIL cytology and CIN 3 histology result as those with both tests negative. This is consonant with earlier articles that provide confirmation that HPV infection precedes the development of CIN [8,9,11,21,27,28]. In a study carried out by Rozendaal et al [8], among 1622 women followed for a mean of 40 months, HPV detection was associated with 116-fold increased risk for the development of CIN 3 in contrast to women without HR-HPV. Evaluating 496 women attending a family planning clinic, Moscicki et al [28] reported that approximately 20% of participants with an HPV infection were subsequently found to have a Pap test interpreted as LSIL during a follow-up of 50 months. Bory et al [22] evaluated 3091 women with normal smears and found a relative risk of 96.7 of incident HSIL when HPV was detected at enrollment.
For instance, Kjaer et al [9] followed 10177 cytologically normal women and observed that HPV status at enrollment predicted future development of high-grade lesions. Castle et al [6] also described that about 17% of 2511 women attending screening clinic with a negative Pap test and a positive HPV DNA test will develop an abnormal Pap within approximately five years. Relying merely on simple HPV detection as a biomarker of increased risk of subsequent cervical lesion suffers from some practical limitations. We have observed a small proportion of LSIL and ASC cytology in HPV-negative women during the follow-up. It could be explained for three reasons: 1) fluctuations on viral load below the detection threshold of screening tests leading to misclassification of some infected women as false negative [29], 2) HPV infection could be due to low-risk types, which were not evaluated in this study, 3) it is likely that some subsequent abnormal Pap tests called ASC, and even a few LSIL, were not due to any HPV infection, particularly in older women. They were a consequence of age-related atrophic changes in the cervical epithelia that mimic equivocal cytologic changes [6].
However, regarding histological outcome, as well as Dalstein et al [30] and Sherman et al [11] described, we found no high-grade SIL among women who tested negative for HPV during the entire follow-up. The low risk of lesion among women with both negative tests reflects mainly the high sensitivity and negative predictive value of HPV testing. In consequence, as previous suggested several studies, in such conditions, the use of HPV testing would allow the screening interval to be safely lenghetened [8,10,22,30,31]. In the other hand, a single positive test for oncogenic HPV predicts increased risk for disease [11] and women with positive HPV test and a negative Pap test should be targeted for more frequent surveillance.
A limitation of our trial is that only the screening HC II test result was used for predicting the RR of disease during the follow-up. Bory et al. [22] have shown that an HPV test remaining positive at two controls is significantly associated with an increased risk of incident HSIL. Moreover, a second negative HR-HPV test one or two years after the initial test can reach a negative predictive value of 100% [32], reassuring that women with double negative tests really represent a low risk population of incident disease.
Taking this into account, this ongoing study has extended follow-up and we have been collecting another samples for HPV after one, two and three years of follow-up. This approach will provide more adequate analysis of the duration of the risk after a positive HPV test, which might lead to better understanding of the natural history of squamous intraepithelial lesions.
Results from this study and others [6,10,11,21,22,28] highlight the central etiologic role of HPV in cervical lesions by establishing that viral infection precedes the development of disease. It is, therefore, suggested that while Pap test should continue to be the routine screening method for detection of cervical lesions, HPV testing can be utilized as an adjunct to cytology for effective screening, especially for those women with negative Pap smear [16,21].
In summary, our study using HC II assay confirms the usefulness of this test in addition to cytology to select from a population of women with normal cytology those who are at highest risk for underlying cervical lesion.