Frequency of Human papillomavirus in women attending cervical cancer screening program in Chile

Background Human papillomavirus (HPV) is the etiological factor for cervical cancer and its precursor lesions. The characterization of HPV genotypes in preneoplastic lesions and cervical cancer could establishes the effectiveness of vaccination plan in Chilean population. The aim of this study was to determine HPV frequency in a group of women including in a cervical screening program in the public health care system in Chile. Methods We analyzed 985 cervical smears samples from women with different histological diagnosis, attending to public health care in Temuco-Chile between 2004 and 2012, to detect HPV genotypes, through PCR followed by reverse line blotting assay. Results HPV was found present in 80.8% (n = 796) of samples. Only a 5.6% of 985 samples were infected with a low-risk HPV, considering multiple infections. 10.5% (n = 8/76) of normal cervical epithelia, 83.5% (n = 208/249) and 87.6% (n = 557/636) of low and high grade squamous intraepithelial lesions, respectively, and 95.8% (n = 23/24) of squamous cervical carcinomas tested positive for HPV. HPV 16 was the most frequent genotype found (Overall 44.9%, n = 442/985; SCC: 62.5%, n = 15/24). A high variability of HPV types was also found in preneoplastic lesions, whereas there was a selection of genotypes in neoplasia. Also, there was a higher risk of infection with HPV 16 in women ≤26 years and 34–41 years old (p < 0.05), meanwhile infections with HPV 16 or HPV 18 have related with cancer development (p < 0.01). Conclusions These data provide further information about the frequency of HPV genotypes in women with cervical lesions in Chile, and the introduction of new targeted vaccines against a wider spectrum of HPV is suggested. Electronic supplementary material The online version of this article (doi:10.1186/s12885-017-3496-x) contains supplementary material, which is available to authorized users.

Worldwide, Papanicoulau smear is the screening technique to detect cervical lesions in women and the detection of HPV genotypes is considered complementary to the clinical diagnosis of patients affected with preneoplastic and neoplastic lesions [13]. HPV is present in different percentages in precancerous lesions: it is found in between 59 and 82% of LSIL, while in HSIL the frequency is generally more than 80% [3,[14][15][16]. The principal methods for HPV detection are PCR (polymerase chain reaction)-based HPV detection systems, where a broad spectrum of HPV types is amplified by consensus primers, followed by detection with type-specific probes [17]. Also, methodologies based in real-time PCR for detection of DNA and mRNA of HPV, are the most accepted detection methodologies in the last years [18].
The frequency of HPV in Chile has been described in only a few studies, in normal [6,19], precancerous lesions [20,21], and cervical cancer samples [22,23]. The present study illustrates the frequency of HPV in a large number of precancerous lesions in light of the importance of finding the principal HPV genotypes infecting preneoplastic lesions in the Chilean population in order to evaluate the effectiveness of HPV vaccination plan in this country to prevent cervical neoplasia.

Collection of clinical samples
The study corresponded to a cross-sectional type. All women attending to public health care to gynecological control were invited to participate and they signed an informed consent, previously approved by Araucanía Sur Health Service Ethics Committee. A total of 985 samples from cervical scrapes (cytobrush) were collected and evaluated for HPV infection. The median age of enrollment was 33 years old (Interquartile range 15 years). The age of participants was divided into four groups: ≤ 26 years old, between 27 and 33 years old, between 34 and 41 years old and ≥42 years old. Normal epithelium samples (n = 76) were collected from women who regularly attending to Miraflores public Clinic for gynecological control, and the cervical cytology (Papanicoulau) was found without alterations in current and past controls. Preneoplastic and neoplastic samples were separated into groups according the histological diagnoses (The 1991 Bethesda System [12]) as follows: 249 LSIL, 636 HSIL and 24 squamous cervical carcinomas (SCC). These samples were collected at the Doctor Hernán Henríquez Aravena Hospital in Temuco, Chile (public hospital), between 2004 and 2012, during gynecological examination. The diagnoses of preneoplastic and neoplastic lesions were confirmed by colposcopy-guided biopsy in the Pathology Anatomy and Citology Unit of Hernán Henríquez Aravena Hospital. Molecular test was performed at Molecular Pathology Laboratory, School of Medicine, Universidad de La Frontera.

Positive controls
HPV viral types were used for positive controls. HPV 16, 18, 31 and 33 corresponded to commercial plasmid clones (American Type Culture Collection, Manassas, VA, USA), the remaining HPV types were provided by Dr. Peter Snijders (VU University Medical Center, Amsterdam, The Netherlands). Negative controls consisted of commercial genomic DNA (Promega, Madison, WI, US) and deionized water.

Definitions for single and multiple HPV infections
The detection of only one positive HPV genotype signal in the reverse line blot was defined as a single HPV infection. A multiple infection was defined by the presence of a positive signal for two or more genotypes; in these cases, the predominant signals were considered for further analysis.

Statistical analysis
All data was analyzed by the software IBM SPSS Statistics 20.0 (IBM Corporation, NY, USA).
Age was categorized in four groups (<26, 26-33, 34-41 and ≥42 years old), due to quartiles determination. In order to improve statistical analysis, due to the low frequency of some HPV genotypes, the less frequent genotypes were grouped as HR-HPV (including HPVs 26, 33, 35, 39, 51, 52, 53, 56, 59, 66 and 68) and LR-HPV. The frequency of HPV genotypes was performed through descriptive statistics using as reference group women ≥42 years old, due to the highest frequencies of HPV infection usually are produced in young women (15-25 years old).
To establish the association between age range or lesion grade and HPV single infection we used multinomial logistic regression. For the analysis of multiple infections and their association with lesion degree, binary logistic regression was used considering only LSIL and HSIL, due the absence of multiple infections in normal and SCC cases. Age and lesion degree were considered as the independent variables in each case. In multinomial and binomial logistic regression, we determine the statistical significance, odd ratio and confidence interval. P-values <0.05 were considered significant, with 95% of confidence.

Discussion
Studies around the world have indicated that virtually all cervical cancers contain HPV DNA [2,22]. HPV positivity varies in different studies according the HPV  detection method and type of sample (biopsy or cervical smear) [7,25,26]. Due the importance of highrisk HPV infection to the development of cervical cancer, it is necessary to determine HPV genotype distribution in women with or without cervical lesions. In our study, a total of 985 cytobrush samples (normal, preneoplastic and neoplastics lesions) from women of south of Chile were evaluated for HPV infection. None of these women were vaccinated against HPV. Women with normal cytology have a frequency of 10.5% of infection by HPV. In 2004, Ferreccio et al. in a population-based prevalence study recorded a 14% of HPV infection in women. However, they analyzed general population and some participants had a preneoplastic lesion, therefore they found a higher frequency of HPV infection [6]. Nevertheless, in the same study, Ferreccio found 11.2% of HPV positive in women with normal cytology in Santiago de Chile, results similar to our study [6]. In other hand, we found that majority of HPV positive normal epithelia participants have a HR-HPV infecting. This data also is similar to Ferrecio study in Santiago [6].
The frequency of HPV was elevated in LSIL and HSIL. These percentages are higher than in other studies of HPV frequency in preneoplastic lesions [3,27]. However, most of HPV frequency investigations, used biopsy samples for HPV detection, in contrast to our study where we used cytobrush samples. It is known that the process of fixation with formalin and paraffin embedding involves DNA fragmentation [28], increasing the number of false-negatives in DNA-based HPV detection. Furthermore, the reverse line blot technique for HPV detection and typing is highly sensitive [22]. About the genotypes found in precancerous samples, studies from Latin-American countries show high frequencies of HPV 16,18,45,31 and 58 in preneoplastics lesions [3,20,21,26,27,29], which are correlated with our results.
In cervical cancer, we found a high percentage of HPV infection. Moreover, almost all SCC were positive for HPV and only one sample were tested HPVnegative. The DNA of this sample was tested twice to verify the result. These findings are similar to another study developed by our group, where 94.2% of SCC had the presence of HPV DNA [22].
Almost 30 HPV genotypes were detected in the studied group. The most frequent HPV genotype found was HPV 16, which is usually detected in half of cervical cancer cases and is the most frequent genotype worldwide [7,22]. Preneoplastic lesions showed the greatest variability in HPV types, while the number of genotypes found in SCC decrease considerably. This could be explained by the immune system clearance in order to eliminate HPV and infected cervical cells. There is a natural regression of preneoplastic lesions in a great number of women infected with HPV [9,30]. Also, it seems to be a selection of viral genotypes and only the most oncogenic and those with the ability to integrate their genome are capable of cell transformation. This hypothesis is supported by the decrease of the frequency of some HR-HPV genotypes in HSIL and the disappearance of low-risk HPV in SCC [9,31].
There is agreement regarding the principal HPV genotype found in SCC around the world, namely HPV 16 [9,22,32], followed by HPV 18. In our study we found a strong association between these two genotypes and the grade of lesion and SCC. Besides, we found a that HPV 31 is related with a major risk of developing HSIL, which has been observed in several studies [6,7,[33][34][35]. More importantly, HPV 16 and/ or 18 were found in 56.9% of analyzed samples and genotypes 31, 33, 52 and/or 58 were present in the 16.3%, including single and multiple infection. In 2015, a study performed in Chilean women, showed that 38.8% of women tested positive for HPV were infected with HPV 16 and or 18 [36]. Our results shown a higher frequency of HPV 16 and HPV 18 infections, which is related to the analyzed population, due our study was performed in the Region of La Araucanía, the poorest region of Chile, with the highest native population of the country, and also one of the region with the highest rates of mortality by cervical cancer in Chile [21,22].
In 2014, Chilean government included, in the national program of immunization, quadrivalent HPV vaccine (HPV 6, 11, 16, 18) for 9 years old girls. This vaccine also has cross-protective efficacy for another HPV genotypes, such as 31, 33, 52 and 58 [37]. Our results indicate that the coverture of HPV vaccine in Chilean population will be wide (approximately 60% of genotypes), and it could prevent a great number of preneoplastics lesion and cervical cancer, diminishing incidence of this pathology. However, these results indicate that almost 40% of SCC could not be prevented by current vaccines.
Otherwise, we found that young women had more risk of being infected by HPV 16 by itself or in association with another HR-HPV including HPV 18. This data suggests an early initiation of sexual relationships and also an early infection by a HR-HPV which could lead to carcinogenesis. Meanwhile, women over 42 years old have more risk of been infected with HPV16 associated with HPV 18 and others types of HR-HPV. Both results could be associated with the description made in others studies, where is observed a high frequency of infection by HPV in young women and a resurgence of the infection in older women [6,38]. This phenomenon could be explaining due to the beginning of the sexual intercourses in young women, characterized by several sexual partners, followed by period of stable relationship. These results are similar to others shown by Ferreccio et al. [6].
In other hand, PCR-reverse line blotting assay used to genotype HPV was able to detect single and multiple infections easily, because several genotypes can be detected in a single assay, while others methodologies, such as Hybrid Capture 2 assay, only can discriminate between LR and HR genotypes, but it is not capable of indicate the specific genotypes of HPV [39]. Also, there is evidence, that this technique has a high sensitivity, been capable of detect 0.1 fg of HPV [24]. Among the limitations of the study, there was a low collection of normal cytology samples, because the Health care center chose for the participant recruitment was attending mainly women with preneoplastics and neoplastics lesions.

Conclusions
It is clear that HPV infection continues to be a significant public health problem, particularly in developing countries. There is a wide spectrum of HPV genotypes infecting women around the world and the frequencies of each vary according to the geographic region. Therefore, it is important that genotypes causing cancer in every area be defined. Our results show the distribution of HPV infection through the different alterations in the cervix, finding a genotype selection as the lesion progresses in malignity. The incorporation of quadrivalent HPV vaccine in Chile will potentially diminish HPV frequency in this country, but a great percentage of genotypes will not be covered, evidencing the necessity of new vaccine covering a higher spectrum of HPV genotypes, and given more relevance to HPV detection and genotyping even when patients have been vaccinated for the prevention of cervical cancer.

Additional files
Additional file 1: Data S1. Specific HPV genotype infecting cervical smears samples from normal epithelium and preneoplastic and neoplastics lesions of the cervix. This data describes for each sample the infecting HPV genotype, according the histology diagnosis.