Frequency of HPV detection in chagasic megaesophagus associated or not with esophageal squamous cell carcinoma

Background: Chagasic megaesophagus (clinical manifestation of chagasic disease) has been reported as an etiological factor for squamous cell carcinoma of the esophagus, as well as the presence of human papillomavirus (HPV). Objective: We accessed the prevalence of HPV DNA in a series of squamous cell carcinomas of the esophagus associated or not with the chagasic megaesophagus, and within samples of chagasic megaesophagus without cancer. Data obtained was further correlated to the pathological clinical data of affected individuals. Methods: Retrospective study that used a total 92 samples tissue/biopsy specimens of formalin xed and paran embedded tissues were retrospectively collected from the southeast region of Brazil from patients treated in three hospitals: Barretos Cancer Hospital, Barretos, São Paulo; Federal University of Triângulo (UFTM), Uberaba, Gerais; and São Paulo State University (UNESP), Botucatu, São Paulo. Cases were divided in three groups: i) 24 patients with chagasic megaesophagus associated with esophageal ESCC (CM/ESCC); ii) 37 patients with esophageal ESCC without chagasic megaesophagus (ESCC); iii) 31 patients with chagasic megaesophagus without esophageal ESCC (CM). Results: We detected a higher prevalence of high-risk HPVs in patients from both CM (12/31, 38.8%) and CM/ESCC groups (8/24, 33.3%), as compared to individuals of the ESCC group (6/37, 16.3%), although data was not statistically signicant. We further observed that HPV-16 was more prevalent in patients of the ESCC (4/9, 44.5%) and CM/ESCC groups (2/8, 25.0%). In addition, some of these samples presented infection by multiple HPV types. High-risk HPVs detected were HPV-31, 45, 51, 53, 56, 66, and 73, of which the majority was identied in patients from the CM group. Furthermore, low-risk HPV-11 and HPV70 were identied in individuals from both ESCC and CM groups. Conclusion: This is the rst report regarding the presence of HPV DNA in megaesophagus associated with esophageal squamous cell carcinoma. In the present study, HPV infection appears to be directly related to the development of esophageal squamous cell carcinoma in patients with chagasic megaesophagus. Further studies are warrantee to conrm and better understand the role of oncogenic HPV persistent infection in these patients. positive individuals have a longer survival in relation to HPV negative individuals. These results are in line with previous reports on esophageal cancer, that shown that HPV infection is not associated with improved survival, suggesting that the presence of HPV may not be useful for a possible prognostic assessment related to fact or that contribute to cancer of the esophagus known that infection by the protozoan Trypanosoma cruzi in the esophageal mucosa leads to several physiological changes that result in the development of the chagasic megaesophagus. The altered microenvironment of chagasic megaesophagus has been shown to be a favorable environment for HPV. In the present study, HPV was associated with esophageal squamous cell carcinoma in patients with chagasic megaesophagus. Further studies are needed to conrm and better understand the potential role of oncogenic HPV persistent infection in patients with chagasic megaesophagus.


Introduction
Esophageal cancer is ranked as the 8 th most frequent type of cancer and the 6th most lethal cancer worldwide [1][2][3] . Most esophageal cancers occur in developing countries, such as in Brazil, and, according to the Brazilian National Cancer Institute (INCA), they are considered the 10th most common type of cancer (6th in men and 15th in women) with 10,000 new cases estimated for 2020 3 .
Among the histological subtypes of esophageal cancer, squamous cell carcinoma (ESCC) is the most frequent, accounting for 90% of the cases 4 ; this is a very aggressive malignant neoplasm with high incidence and high mortality rates even in the scenario of recent advances in treatment and diagnosis 1,2 . The main risk factors for ESCC development are the consumption of alcohol and tobacco (particularly in combination), food and hot drinks, as well as high consumption of red meat [5][6][7] . Moreover, chronic diseases, such as chagasic megaesophagus, and infectious agents, such as human Papillomavirus (HPV), have been suggested to in uence esophageal carcinogenesis [8][9][10] .
Chagasic megaesophagus occurs due to Chagas' disease (American Trypanosomiasis), caused by the parasitic infection of the protozoan Trypanosoma cruzi 11 . It is an endemic disease in South and Central America; Brazil is one of the main endemic countries where about 4 million people now are infected with the parasite 12,13 . However, due to the high number of Latin American immigrants, non-endemic regions, such as the United States, Canada, Europe, Australia, and Japan, may present cases of the disease, and thus this may prove to be an important public health problem worldwide [13][14][15][16] . Chagas' disease is divided into two forms: acute (initial) and chronic (late) 11 . During the acute phase, which lasts from 4 to 8 weeks, a high number of parasites circulate in the blood, but in most cases, the symptoms are absent or mild and unspeci c, thus making early diagnosis di cult 11 . The chronic form occurs in 30% of the individuals infected with the parasites, and clinical manifestations include the dilation of some organs such as the colon (megacolon), esophagus (megaesophagus), and heart (cardiomegaly) 11 .
Chagasic megaesophagus affects about 3% of individuals chronically infected with Chagas' disease 11 . It occurs when the amastigote forms of the protozoa destroy the myenteric plexus, which is part of the enteric nervous system composed of a chain of neurons responsible for contractions of the gastrointestinal tract. This destruction occurs when a agellar antigen of the parasite (similar to a protein released by these neurons) leads to cross-immunoreactivity, attracting immune cells into the ganglia, causing in ammation and leading to the deposition of dense connective tissue in interstitial brosis between muscle bers 17,18 . Consequently, uncoordinated contractions of peristalsis occur, reducing the esophageal body and altering the functioning of the lower esophageal sphincter. These changes lead to the buildup of food, resulting in the development of megaesophagus 17 . Importantly, about 2-10% of individuals with chagasic megaesophagus will develop esophageal squamous cell carcinoma 19, 20 ; however, due to the scarcity of studies regarding the carcinogenic environment, the neoplastic mechanism leading to ESCC remains uncertain 21 . Nevertheless, it is believed that in ammation and chronic esophagitis may play an important role in the development of epithelial dysplasia and, subsequently in cancer 22,23 .
HPV (Human Papillomavirus) belongs to the Papillomaviridae family and has a tropism towards lining tissues such as the parenchyma of the epidermis and mucous membranes 24,25 . Over 200 types of HPV (http://pave.niaid.nih.gov/) have been described and clustered within ve genera: alpha-, beta-, gamma-, muand nu-HPV. High-risk alpha(α)-HPVs are associated with the development of several tumors, including the uterine cervix and the head and neck 26 . Its oncogenic potential occurs due to the expression of viral oncoproteins that interact and block the activities of host cell cycle regulatory proteins 24,25 . Brie y, the E6 viral oncoprotein interacts with the p53 tumor suppressor protein leading to its degradation and consequently the inhibition of DNA repair and apoptosis; additionally, the E7 oncoprotein interacts with the retinoblastoma protein (pRB) leading to the release of the E2F transcription factor, leading to a deregulated cell proliferation and consequently cancer 24, 25, 27 . In the early 1980s, esophageal tissues (benign and malignant) with cytopathological alterations characteristic to those found in HPV-infected uterine cervix tumors were rst observed 9 , suggesting that HPV could be involved in esophageal carcinogenesis, since the oral mucosa extends to the squamous epithelium of the esophagus being exposed to viruses such as HPV through oral transmission 10 . This relationship is still a matter of debate since HPV DNA prevalence in ESCC varies substantially within studies, ranging from 0 to 100% [28][29][30] . Moreover, in the megaesophagus context, solely a single study from Crema and colleagues identi ed a high frequency of HPV (63.3%) in individuals with chagasic megaesophagus when compared to healthy individuals without megaesophagus (13.8%) 31 . Nevertheless, to the best of our knowledge, currently there are no studies that have analyzed the presence of HPV within individuals with chagasic megaesophagus associated with esophageal cancer.
In summary, esophageal squamous cell carcinoma has as etiologic factors, the chagasic megaesophagus and possibly HPV infection. Facing the evidence of the presence of HPV in individuals with chagasic megaesophagus (whose carcinogenic mechanism is still uncertain), we hypothesized if the presence of HPV in the chagasic megaesophagus may increase the chance of these individuals to develop esophageal cancer.

Study population
The casuistic of this retrospective study was composed of patients examined between 1990 and 2016 in the Upper Digestive Tract Department of three hospitals in Southeastern Brazil: Barretos Cancer Hospital (BCH), Barretos, São Paulo; Federal University of Triângulo Mineiro (UFTM), Uberaba, Minas Gerais; and São Paulo State University (UNESP), Botucatu, São Paulo.

Inclusion criteria
We included only patients diagnosed with an image or histopathological proven chagasic megaesophagus with or without a positive serologic test for Chagas' disease; we also included patients with ESCC without chagasic megaesophagus serologic negative for Chagas' disease whose exams (image and histopathology) con rmed the malignant disease. Clinical-pathological variables of patients, as well as the molecular status of TP53 and PI3KCA hotspot mutations and MSI phenotype, were previously reported from these individuals 21,32,33 . All clinical and pathological information was obtained through medical record´s review.

DNA isolation
A slide containing a 4µm para n section of each sample was stained with hematoxylin-eosin (HE, Merck KGaA, GE) and evaluated by a specialist pathologist to delineate necrosis free areas. DNA was isolated from FFPE tissues representative of the tumor lesions in ESCC and CM/ESCC groups and esophageal tissues in the CM group, as previously described 34 . Sections were depara nized by heating at 80°C for 20 min, followed by sequential washing in xylol (5 min) and decreasing concentrations of ethanol (1 min -100%, 70%, and 50%) and water free of nucleases (1 min). DNA extraction was performed using the commercial kit QIAamp DNA Micro Kit (Qiagen, USA) following the recommended protocol.

HPV detection and genotyping
Detection of HPV DNA was assessed in all samples by a genotyping assay combining multiplex PCR (TS-MPG) and bead-based Luminex technology (Luminex Corp., USA), as previously described 35 . This technique is able to simultaneously identify 20 types of α-HPV: 3 low risk: HPV-6, 11, 70; 6 probably high risk: HPV-26, 53, 66, 68, 73, 82; and 11 high risk: HPV- 16, 18, 31, 33, 35, 39, 45, 52, 56, 58, 59 36, 37 . As a positive control for the quality of the template DNA, primers for the β-globin gene were included in these reactions. PCRs were performed with 10µl of template DNA in a 96-well format in 25-μl/well nal reaction volume. HPV multiplex PCR was performed with Qiagen Multiplex PCR Kit (Qiagen, Germany), following the manufacturer protocol. Each reaction consisted of 45 cycles: 94°C for 30 seconds, 63°C for 3 minutes, and 72°C for 90 seconds. The rst cycle was preceded by incubation at 95°C for 15 minutes, and the last cycle was extended for 10 minutes at 72°C. Hybridization reactions were performed according to Schmitt et al (2006) 38 . For each HPV type-speci c probe, the mean uorescence intensity (MFI) values obtained when no PCR product was added to the hybridization mixture were considered as background. The cutoffs were calculated by adding 5 MFI to 1.1 times the value of the median background. MFI values > 20 were considered positive.

Statistical analysis
Characterization of the study population was analyzed through frequency tables for qualitative variables, and measures of central tendency and dispersion (mean, standard deviation, minimum, and maximum) for the quantitative variables, comparing the different groups. To assess the association of HPV DNA detection, clinical-pathological characteristics, and the mutation status of TP53 and PIK3CA between study groups, we used Chi-square or Fisher's exact tests. We performed an overall survival analysis using the Kaplan-Meier limit estimator and we applied the Logrank test to compare overall survival curves in the groups with cancer.

Characterization of the study population
Clinical-pathological and molecular features of the patients studied are described in Table 1 21 . As previously reported by our group 32 , most of the patients were males and under 60 years old. In relation to the principal risk factors for the development of esophageal cancer, patients from groups CM/ESCC and ESCC were statistically associated with higher consumption of tobacco and alcohol (87.5% and 86.5%; 66.7% and 81.0%, respectively). In addition, most patients had advanced stage esophageal cancer and chagasic megaesophagus ( Table 1).

Detection of HPV DNA
Using a very sensitive HPV detection and typing protocol based on multiplex PCR (Luminex), we observed a high prevalence of high-risk HPVs DNA in patients of both CM (12/31, 38.8%) and CM/ESCC groups (8/24, 33.3%), when compared to patients of the ESCC group (6/37, 16.3%), although data was not statistically signi cant (Table 2).
We analyzed the frequency of individual HPV types within each group (Table 3). We observed that HPV-16 was more frequently detected in patients from both ESCC (4/9, 44.5%) and CM/ESCC groups (2/8, 25.0%). In addition, for some samples, HPV co-infection by more than one viral type was observed. We additionally detected several high-risk HPVs (HPV-31, 45, 51, 53, 56, 66, and 73), of which the majority was identi ed in patients from the CM group. Finally, low-risk HPV-11 and HPV-70 were identi ed in individuals from both ESCC and CM groups (Table 3).
We next thought to analyze the association between HPV prevalence and clinical-pathological features of patients independent or not of the study group. The data obtained are depicted in tables 4 and 5. Nevertheless, no signi cant associations were observed for any of the variables studied (Table 4 and 5).
Finally, in order to better understand the role of HPV in these patients, we analyzed if there was any correlation between the presence of high-risk HPV DNA and the mutation status of TP53, which was reported previously in our studies 32 . Despite the lack of signi cant associations, we observed that patients who tested positive for HPV were more prone to lack mutation within the TP53 gene (18/25, 72.0%) ( Table 4). However, when disease groups were analyzed independently, only HPV positive samples from individuals from the CM group did not present mutations within TP53 (12/12, 100.0%), as opposed to patients of the CM/ESCC and ESCC groups, among which HPV frequencies were similar in patients with or without TP53 mutation (Table 5).

Overall survival analysis
Finally, we compared the Kaplan-Meier survival curves and the Log Rank test among all groups. As expected, we observed that individuals in the group with only megaesophagus (CM) had a signi cant better survival when compared with the other groups with cancer ( Figure 1). Regarding the HPV status, HPV positive cases exhibited a better outcome; however, it was not statistically signi cant (Figure 2). We performed the Cox regression test to assess the survival ratio between the group ESCC/CM and the positive HPV (Table 6). We observed that the hazard ratio is signi cantly associated with the ESCC/MEC group.

Discussion
There are several well-known risk factors for the development of esophageal squamous cell carcinoma, but two remain controversial: chagasic megaesophagus caused by neglected Chagas disease, endemic in Latin and South America regions and considered a signi cant public health problem 16,17,39 ; and persistent HPV infection, which has been suggested since the early 1980s as a potential risk factor 9, 10 . However, no studies have until now investigated the possible association between HPV detected in patients with chagasic megaesophagus associated with esophageal cancer. In this sense, the present study assessed, for the rst time. the presence of HPV in patients with chagasic megaesophagus associated with esophageal squamous cell carcinoma (CM/ESCC) and compared with patients with esophageal squamous cell carcinoma without chagasic megaesophagus (ESCC) and chagasic megaesophagus without esophageal squamous cell carcinoma (CM).
The results obtained in this study showed a higher prevalence of high-risk HPVs in the context of the chagasic megaesophagus without (12/31, 38.8%) or associated with ESCC (8/24, 33.3%). These ndings are partially in concordance with the previous Brazilian study that reported an HPV frequency of 63.3% in individuals with chagasic megaesophagus without esophageal cancer 31 . Moreover, the high frequency of HPV is similar to that reported in esophageal tumors in regions in China considered to be at high risk (~32-63,6%) 40 . In addition, we observed low HPV prevalence in patients with esophageal cancer of the ESCC group (6/37, 16.3%). These results agree with those reported by Brazilian studies in which frequencies reached 15% [41][42][43][44] , and with a recent study from our group that identi ed a frequency of 13.8% using the same methodology, in patients with esophageal squamous cell carcinoma 34,45 .
The HPV type more commonly detected within samples analyzed in the current study was HPV-16, present within the three groups, but more frequently detected in the ESCC group (4/9, 44.5%) followed by CM/ESCC (2/8, 25.0%) and CM groups (2/14, 14.3%). Notably, persistent HPV infection, mainly HPV-16, is associated with carcinogenesis in several anatomical sites such as cervical cancer worldwide [46][47][48] . In addition, we identi ed a low frequency of several other high-risk HPV types, including HPV- 31,39,45,51,52,53,56,59,66,68, and 73 ( Table 3). Some of the HPV types detected in our analysis coincide with that reported in our previous study, also using the Luminex technique 34 . These types identi ed with low frequency are little studied in the literature 49 .
Importantly, the oncogenic potential of high-risk HPV is well-established and related to the expression of the E6 and E7 oncoproteins that interact and cause degradation of p53 24,25 . Therefore, in an attempt to better understand the role of HPV in the groups, we evaluated the presence of HPV and the mutation status of the TP53 gene previously reported in a study of our group 31 . We observed that most HPV positive patients did not have TP53 mutations. What caught our attention was the fact that the HPV positive CM group of patients was 100% TP53 wild-type, while in the CM/ESCC and ESCC groups HPV frequencies were similar in the mutated or wild-type patients for the TP53 gene. According to the literature, it is expected that HPV associated cancers do not present TP53 mutations, being mutually exclusive events. 50,51 . Some studies report that in cervical tumors, HPV infection and TP53 mutations are mutually exclusive, evidencing that such infection leads to the development of this tumor type, but in head and neck tumors both HPV infection and TP53 mutations can occur simultaneously. In these cases, the study shows that HPV did not play a causal role in tumor development, but it was only casually present [52][53][54][55][56][57] .
We also assessed the relationship between survival between study groups and HPV status. We observed that HPV status does not in uence patient's survival, although HPV positive individuals have a longer survival in relation to HPV negative individuals. These results are in line with previous reports on esophageal cancer, that shown that HPV infection is not associated with improved survival, suggesting that the presence of HPV may not be useful for a possible prognostic assessment related to fact or that contribute to cancer of the esophagus 58, 59 .
It is known that infection by the protozoan Trypanosoma cruzi in the esophageal mucosa leads to several physiological changes that result in the development of the chagasic megaesophagus. The altered microenvironment of chagasic megaesophagus has been shown to be a favorable environment for HPV. In the present study, HPV was associated with esophageal squamous cell carcinoma in patients with chagasic megaesophagus. Further studies are needed to con rm and better understand the potential role of oncogenic HPV persistent infection in patients with chagasic megaesophagus.        Figure 1 Kaplan-Meier curve for assessing the estimated overall survival probability of follow-up time among patients from diagnosis to death per group. ESCC: group with squamous cell carcinoma of the esophagus; MEC/ESCC: chagasic megaesophagus group associated with cancer.