Gastric Cancer and Associated Pathogens: Is There Any Association in Moroccan Region?

Purpose: Helicobacter pylori, Epstein-Barr virus and human papillomavirus are three pathogens associated with various human cancers. This study aimed to investigate the role of these pathogens in gastric cancer in Moroccan population Methods: For this, a retrospective study has been conducted on participants attending the gastroenterology department of Hassan II University Hospital of Fez. A total of 279 participants were enrolled. H. pylori, EBV and HPV were detected and genotyped by PCR. Results: A signicant association has been established between H. pylori, EBV and gastric cancer. 93.4% and 43.3% of gastric cancer cases are related to H. pylori and EBV respectively (p ≤ 0.01). H. pylori-EBV co-infection is responsible of 31.6% of gastric cancer cases (p<0.01). Correlation between pathogens genotypes and gastric cancer shows 55.6% of GC EBV positives are carrying the 30bp deletion in LMP1gene, while 16% of gastric cancers cases are carrying high-risk genotypes of HPV (p=0.21). Conclusion: The obtained results highlight the possible role of co-infection in gastric cancer development.


Introduction
Cancer is a complex and multifactorial disease related to genetic, environmental, lifestyle factors and sometimes infectious agents [1][2][3]. In fact, the discovery of infectious agents-related to carcinogenesis helped to develop cancer prevention strategies based on vaccination, infection prevention and earlier diagnosis before cell transformation.
Gastric cancer is the fth most frequently diagnosed cancer and the third leading cause of cancer-related mortality in the world [4]. Its association with Helicobacter pylori (H. pylori) has been well documented and this bacteria is responsible for a six fold increased risk for gastric non-cardia cancer [5,6]. Like other cancer-associated infectious agents, H. pylori is widespread and harbored by around 75% of people aged 40 years old and above. However, not all carriers will develop gastric cancer [7], thus other factors are required for cell transformation [8]. Among them, carcinogenic viruses could be with major importance. In fact, in the last years, the focus has shifted to the role of Epstein-Barr virus (EBV) on GC, and recent studies identi ed that relatively 10% of patients with gastric cancer exhibited EBV infection [9,10]. Thus, EBV antigenic epitopes have been detected speci cally in gastric cancer cells, demonstrating that EBV may be a virulence factor in gastric cancer [11]. This data was considered by the Cancer Genome Atlas in the novel molecular pathological classi cation of gastric cancer and EBVassociated gastric cancer (EBVaGC) was categorized as a distinct type of gastric cancer [9]. However, only few studies were interested on and documented the simultaneous presence of EBV and H. pylori in gastric cancer specimens. The determined co-infection rate range from 6-12% and the signi cance of this co-infection was assessed in few meta-analysis [12][13][14]. Also, EBV and H. pylori co-infection with the human immunode ciency virus (HIV) were found to be associated to about 10% of gastric cancer and increased risk of this cancer cases respectively [15][16][17]. In addition to EBV and HIV, the implication of Human Papillomavirus (HPV) in gastric cancer development has been recently reported [18] despite the nonconclusive results of several studies [19,20]. In Morocco, even if the association of H. pylori and gastric cancer was well documented, no data on the occurrence and implication of EBV or HPV in gastric cancer cases is available.
The aim of the present study were i) to determine the prevalence of EBV and HPV and the rate of their co-infection with H. pylori in gastric cancer; ii) to correlate their co-presence with gastric cancer in the studied population, iii) to study the correlation between genotypes of infectious pathogens and gastric cancer.

Methods
From May 2009 to May 2014, 1134 patients aged 18 years and more, attending Gastroenterology Department of Hassan II University Hospital (CHU) of Fez and undergoing endoscopy for diagnosis of abdominal pain or discomfort were enrolled in a prospective study to determine the H. pylori association with gastric pathologies. Patients aged below 18 years or who were on medications (antibiotics, proton pump inhibitors) for the last 3 months, as well as pregnant or nursing women were excluded from this study. Consenting participants had a personal interview to ll a questionnaire before undergoing endoscopy. For illiterate or semi-literate participants, the written consent was read to them by the principal investigator. This study was approved by the CHU Hassan II Ethical Committee.
From the 1134 participants initially enrolled [21], 279 samples were selected to be included in the present study. The selection criteria were based on histological and endoscopic results. Hence, all samples of patients with gastric cancer (GC) and precancerous lesions (pk lesions: metaplasia and /or dysplasia) were used as cases and samples of patients with duodenal ulcer (DU) (only cases whose histological exam didn't show cancerous or pk lesions) were used as control group. Gastritis cases were not included in this study as their evolution to GC or DU still possible. The GC and pk lesions were classi ed according to the Sydney system [22].
The DNA directly extracted from gastric biopsies was used for H. pylori, EBV and HPV molecular diagnosis using polymerase chain reaction. For all reactions, positive and negative controls were included. The PCR products were resolved in a 2% agarose gel, stained with ethidium bromide and visualized under UV source.

H. pylori molecular diagnosis and genotyping
All specimens have already been analyzed as a part of a study previously conducted by our team [21,23]. Brie y, PCR using glmM primers to detect H. pylori as previously described [24] and positive specimens for H. pylori were genotyped to determine CagA status and vacA subtypes by multiplex PCR using speci c primers and conditions previously described [25,26].

EBV molecular diagnosis and genotyping
The specimens previously analyzed for H. pylori diagnosis have been subjected to PCR using TC60 primers to detect BamW gene of EBV as previously described [27] and LMP1 genes was characterized for all EBV positive specimens using speci c primers [28, 29].

HPV molecular diagnosis and genotyping
HPV DNA was detected by PCR using the primers GP5+/GP6+ as previously described [30]. HPV positive samples were then analyzed to identify the low and high-risk genotypes and to determine the speci c high-risk genotypes using a recently developed genotyping method. In fact. Two separate multiplex PCR were setup for the speci c identi cation of 16 high risk-HPV genotypes (16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 68, 73 and 82).

Statistical analysis
All collected data were entered on an Excel spread sheet; statistical analysis was then performed using SPSS (Statistical product and services solutions, version20, SPSS Inc. Chicago, Illinois, USA) software. The studied population was described and then the association's pathogens-histological lesions and pathogens genotypes-histological lesions were studied. Chi-square or Fisher's exact tests were applied to establish all statistically signi cant associations. All Chi2 test results with p-values ≤0.05 were considered as statistically signi cant.

Results
A total of 279 consenting participants were enrolled in this study. They were 188 (67.4%) men and 91 (32.6%) women with an average age of 52 years, ranging from 18 to 90 years, and only 88/272 (32.4%) and 42/231 (15.4%) of them are smokers and alcoholics, respectively. Most of participants (124/279) were with pk lesions and 76 were with GC. In this study, DU cases (79/279) were considered as control group.
Molecular diagnosis was performed to determine the prevalence of H. pylori, HPV and EBV infections on the studied population. The results con rm the presence of H. pylori, EBV and HPV infection in 57.3% (160/279), 28% (78/279) and 26.2% (73/279) cases respectively.
HPV positives samples genotyping revealed that 87.7% cases were with low risk strains; while 12.3% were with high risk ones, which correspond to HPV 35, 45, 51 and 53. HPV 53 was the most prevalent genotype in the high risk strains.

Correlation between infection and the studied groups
Infections status has been correlated to the pathologic groups in this series and the results show a signi cant association between H. pylori, EBV infections and GC (p≤0.01) separately. In fact, 93.4% and 43.4% of GC cases were related to H. pylori and EBV infection respectively, while no association has been obtained with HPV infection (Table 1).  Table 2) and 33% of Pk lesions cases didn't show an ampli cation of any of those micro-organisms.

Discussion
H. pylori infection is strongly associated to GC development which is one of the most common cancers, contributing to 6.8% of total cancer cases and 8.8% of total cancer-associated mortality worldwide [31][32][33][34][35]. In a previous study, conducted by our team, in the central north region of Morocco [36], we came to the evidence that all our GC cases are not H. pylori dependent. In fact, recent studies reported the implication of other factors in the gastric carcinogenesis. EBV and HPV has been described as potentially responsible of some of these cases [7,15,18,37] and EBVassociated gastric cancer (EBVaGC) was categorized as a distinct type of gastric cancer according to the last Cancer Genome Atlas classi cation.
To date, H. pylori co-infection with EBV and especially with HPV remains little explored in GC [18,[38][39][40], and most of conducted study studies on this eld were concentrated in Latin America except one that were conducted in India. In our knowledge, this is the rst study conducted in Morocco and North Africa with the aims to investigate the implication of these pathogens and their co-infection with H. pylori mainly in GC cases and to determine the genotypes implicated in the gastric carcinogenesis. This will help to understand GC causes and establish adapted preventing strategy.
In this study, the prevalence of H. pylori, EBV and HPV infection in all the studied groups, namely GC, pk lesions and DU groups was determined. The choice of the two rst groups was based on the histological lesions evolution, while DU is considered as control group. Patients with DU were used as control group because gastritis can evolved towards pk lesions, which predispose to GC disease, besides DU and GC seem to be mutually exclusive outcomes of H. pylori infection [41].
The obtained rates of H. pylori, EBV and HPV infections were 57.3%, 28% and 26.2% respectively. These rates increase clearly in GC cases to reach 93.4%, (p<0.01) for H. pylori, 43.4% for EBV (p=0.01) and 31.6% for HPV (p=0.42). The prevalence of H. pylori infection in GC corroborate our previous results [42] and the results reported in north Brazil where H. pylori infection has been detected in 88% of GC cases, while EBV infection has only been detected in 9.6% of GC cases which is lower than the rate detected in our population [43]. The studied population and the EBV diagnosis methods can explain this difference. In fact, in this Brazilian study the RNA in situ hybridization was used for EBV diagnosis. Also, the HPV prevalence detected in our GC group (31.6%) is comparable to that recently reported in a meta-analysis conducted in Asia (28%) [18].
The H. pylori-EBV or HPV co-infection has been correlated to the studied groups. These correlations show that most GC cases are associated to the H. pylori-EBV co-infection (31.6%) compared to the control group (10.1%). H. pylori single infection and the H. pylori -HPV co-infection are responsible of 30.3% and 21.1% of GC cases respectively (p<0.01). These signi cant associations con rm that H. pylori -EBV coinfections are risk factors for GC development. However, the fact that no cases of HPV single infection was observed in GC cases while 21.1% of GC cases presented co-infection versus 3,8 in control group leads to suppose that HPV may be a cofactor of GC development in H. pylori positive patients. Also, it's interesting to highlight that no case of GC has been related to HPV-EBV co-infection and that 5.3% of our GC cases are pathogens independent (p<0.01). In these GC cases, other risk factors can be incriminated such as nutritional, environmental and genetic predisposition. This also can be explained by the fact that H. pylori can likely disappear spontaneously during malignant transformation of gastric epithelia as reported by several studies [44,45]. The H. pylori co-infection with EBV and HPV in the studied groups has also been investigated according to the participants age (younger and older than 40 years  [43]. Therefore, EBV and HPV genes have been well studied and identi ed mostly in nasopharyngeal carcinoma and cervical cancer respectively but limited in GC. For this reason, we proposed to investigate the correlation between EBV and HPV genotypes with the studied groups. The 30-bp LMP1gene deletion was observed in 54.5% of our EBV+ GC cases that is comparable to the results of Brazilian study (50%) but lower than that of Japanese study (88.5% and 91.7%) [46,47].
In our study, women are more likely than men to develop EBV-associated gastric cancer. In fact, a signi cant association was found between gender and EBV infection in GC cases. Thus, 59% of women with GC were EBV positives vs 34% of man. However, 51.5% of EBV + GC+ were man. These data corroborate the data of meta-analysis conducted by Tavakoli  H. pylori cagAis more co-present with EBV. This tendency is observed in all cases but is signi cant in Pk lesion cases. Supposing that H. pylori play role in EBV reactivation, the last data lets suggest that some H. pylori strains and specially Cag-strains are more likely applied in this reactivation.
The implication of HPV in GC has long been controversial [19,37,49]. When a recent Brazilian study reported a lower prevalence of HPV in GC cases (3%) [50], a meta-analysis conducted on 1917 cases reported a higher rates (a pooled prevalence : 28%) and that HPV prevalence in GC cases is 7 fold higher than in normal gastric tissues [18]. In our series, 31.6% of GC cases are HPV infected which is concordant with the results of the meta-analysis. The genotyping analysis shows that only 16% of GC-HPV positives cases were with high-risks HPV vs 0% in DU cases (p=0.21). However, only the HPV 16 and 18 strains were detected in Brazilian GC cases [50].

Conclusion
The obtained results show that H. pylori and EBV infections are signi cantly associated to GC and that the majority of GC cases are associated to H. pylori-EBV co-infection. The most recognized virulent genotypes of different pathogens seem to be those associated to gastric cancer. So, participants with coinfection and /or carrying the most virulent genotypes may need more attention to prevent gastric cancer development.

Declarations
Funding: This work was supported by the National Centre for Scienti c and Technical Research (CNRST) and the "Ministère de l'éducation nationale, de la formation professionnelle, de l'enseignement supérieur et de la Availability of data: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Author contributions: All authors contributed equally to this work