Histopathological prevalence of Helicobacter pylori in gastric and duodenal biopsies in the central region of Burkina Faso: a retrospective multicenter cross-sectional study

¹Laboratory of Molecular Biology, Epidemiology and Surveillance of Foodborne Bacteria and Viruses, Doctoral School of Science and Technology, Joseph Ki-Zerbo University, Ouagadougou, Burkina Faso, ²Tengandogo University Hospital Centre, Ouagadougou, Burkina Faso, ³Ziniaré University Centre, Ouagadougou, Burkina Faso, ⁴Higher Institute of Health Sciences/Yembila Abdoulaye Toguyeni University Fada-Ngourma, Fada-Ngourma, Burkina Faso, ⁵Bogodogo University Hospital Centre, Ouagadougou, Burkina Faso

^&Corresponding author
Saydou Beogo, Laboratory of Molecular Biology, Epidemiology and Surveillance of Foodborne Bacteria and Viruses, Doctoral School of Science and Technology, Joseph Ki-Zerbo University, Ouagadougou, Burkina Faso

Introduction    

Helicobacter pylori (H. pylori) is a gram-negative, spiral-shaped, microaerophilic bacterium [1] that can progress to severe gastroduodenal pathologies, including peptic ulcer, gastric cancer and gastric mucosa-associated lymphoid tissue lymphoma [2]. Gastric ulcer is a chronic condition characterized by spontaneous flare-ups and an ulcerative crater, a deep, irregularly shaped mucosal defect with well-defined edges that affects the muscularis layer, which is replaced by sclerotic tissue [3].

Transmission occurs through oral-oral or fecal-oral routes, often facilitated by poor hygiene, low socioeconomic conditions, and limited knowledge about the mode of transmission. Most infected individuals remain asymptomatic [4]. However, if untreated and undiagnosed, the infection can develop into chronic gastritis and peptic ulcer disease [5]. It is also associated with distal gastric adenocarcinoma, mucosa-associated lymphoid tissue (MALT) lymphoma, primary non-Hodgkin's lymphoma, and various extra-gastric diseases [6,7]. About 10% of infected individuals develop peptic ulcers, 1-3% develop gastric cancer, and 0.1% develop MALT lymphoma [8]. The prevalence of Helicobacter pylori infection varies according to geography, socioeconomic status, education, living conditions, and occupation [8]. The World Health Organization (WHO) reports that over half of the global population is infected [4]. In industrialized countries, the prevalence ranges from 20% to 40%, whereas in developing nations, it can reach 70% to 90%, posing a significant public health concern [9]. In West Africa, prevalence rates vary significantly: 75% in Ghana [10]; 62-97% in Senegal [11-13], 54-91% in Côte d'Ivoire [14-17], and 34.7-72% in Benin [18-20]. In Burkina Faso, estimated prevalence rates range from 47.60% to 91.43% among children and adults [3,21-23]. Based on the data found in the literature, there is considerable variation in the prevalence and clinical manifestations of Helicobacter pylori infection. Most of the available data come from single-centre studies, and there is a limited amount of multicentre histopathological data showing the distribution across different diagnostic laboratories within the same region.

This study aimed to determine the histopathological prevalence of Helicobacter pylori infection in gastric and duodenal biopsies from patients presenting with gastroduodenal symptoms in the pathological anatomy and cytology laboratories of the Central region of Burkina Faso and to be able to describe the associated gastric lesions according to the Sydney system.

Methods     

Study design and setting: this is a descriptive cross-sectional study based on a retrospective analysis of pathological data, conducted from 1 January to 31 December 2023. It was carried out in ten public and private pathological anatomy and cytology laboratories located in the Centre Region of Burkina Faso.

Study population: the study included all symptomatic patients who underwent histopathological examination of a gastric or duodenal biopsy due to gastroduodenal symptoms during the study period. We included all histopathological reports of gastric or duodenal biopsies providing information on the patient's age and sex, the biopsy site, the type of chronic gastritis according to the revised Sydney classification, and the presence or absence of Helicobacter pylori. We excluded reports that did not conform to the revised Sydney classification and those that did not specify the status of Helicobacter pylori. Missing data, such as the site of the biopsy and clinical symptoms, did not constitute grounds for exclusion; this information was treated as missing data in the corresponding analyses. No formal sample size calculation was performed, as the study included all available records meeting the inclusion criteria (exhaustive sampling).

Data collection: a standardized data collection form was utilized to record sociodemographic details (age and sex), biopsy indications, anatomical biopsy sites, types of chronic gastritis, and the presence of Helicobacter pylori. Reported precancerous lesions (atrophy, intestinal metaplasia, dysplasia) and malignant tumors (adenocarcinoma, GIST, lymphoma, neuroendocrine tumors) were also documented.

Laboratory procedures: all the laboratories in which data were collected followed the same analytical protocol: after digestive fibroscopy, tissue fragments were placed in 10% buffered formalin; following this fixation step, the tissue underwent dehydration, paraffin embedding, microtome sectioning at 3 µm, and haematoxylin-eosin and Giemsa staining. All diagnoses were established by qualified pathologists according to the histological criteria of the Sydney system.

Definitions: patients are categorised according to gender and age groups as defined in the study database. Chronic gastritis is defined and classified according to the revised Sydney system (taking into account the intensity of inflammation, activity, atrophy, intestinal metaplasia and the presence of Helicobacter pylori). Helicobacter pylori infection was considered positive when the bacterium was visualised on histopathological examination of gastric or duodenal biopsies. Precancerous lesions included gastric atrophy, intestinal metaplasia and dysplasia; malignant lesions corresponded to the presence of gastric adenocarcinoma, lymphomas, gastrointestinal stromal tumours and neuroendocrine tumours.

Statistical analysis: the data were exported and analysed using Stata® version 17 (StataCorp, College Station, TX, USA). Qualitative variables were described in terms of their frequencies and proportions, whilst quantitative variables were summarised by their mean and standard deviation or by their median and percentiles where appropriate. The age distribution was described using measures of central tendency and dispersion. The prevalence of histological Helicobacter pylori colonisation was estimated along with the corresponding number of cases. Comparisons of proportions between groups were carried out using Pearson's X² test or Fisher's exact test when the sample sizes were insufficient. Comparisons of quantitative variables between two independent groups were carried out using the Mann-Whitney test when the conditions for applying parametric tests were not met.

In order to identify factors associated with Helicobacter pylori carriage, bivariate logistic regression analyses were initially performed. Variables showing clinical relevance or a statistical association in the bivariate analyses were then included in a multivariate logistic regression model to estimate adjusted odds ratios (aOR) and their 95% confidence intervals (95% CI). The results were expressed as odds ratios with their 95% CIs. Given the high proportion of missing data observed for the ‘sampling site' variable, two multivariate models were evaluated: a primary model excluding this variable in order to maximise the size of the sample analysed, and a secondary model including the sampling site as part of a sensitivity analysis. The main model was selected for interpreting the results due to its larger sample size and its better representativeness of the study population.

The goodness of fit of the multivariate model was assessed using the Hosmer-Lemeshow test. We verified model specification using the Link test and assessed discriminatory power using the area under the ROC curve (AUC). All analyses were conducted as two-sided tests, and the threshold for statistical significance was set at 5 per cent (p < 0.05).

Ethical considerations: we maintained patient confidentiality and anonymity throughout the study. Data were collected with the consent of the laboratory managers and used exclusively for research purposes. This study was conducted in accordance with the ethical principles of health research. Ethical approval was obtained from the relevant ethics committee under decision no. 2026-02-0091. Due to the retrospective nature of the study and the use of anonymised data, individual patient consent was not required.

Results     

The study included 3,699 patients who had undergone a gastric histopathological examination. The mean age was 44.4 ± 16.3 (median: 43 years), ranging from 10 to 98 years. Women accounted for 53.1% of the study population. Epigastric pain was the main clinical presentation observed (53.2%), followed by gastro-oesophageal reflux (5.2%) and peptic ulcer disease (4.4%). Biopsies from the antrum and fundus (45.8%) and the antrum alone (44.5%) accounted for the majority of samples taken. Chronic gastritis was the predominant histological diagnosis (96.8%). Among the associated histological lesions, lymphoid follicles were the most common (13.1%), followed by glandular atrophy (6.6%), intestinal metaplasia (4.1%), and dysplasia (0.5%) (Table 1).

The prevalence of Helicobacter pylori was 80.3%. No significant difference was observed between the sexes. However, Helicobacter pylori colonisation varied significantly depending on the site from which the sample was taken for histological examination (p < 0.001). Biopsies from the antrofundic region were associated with a higher prevalence of carriage (ORb = 1.60), whilst those from the antroduodenal and fundic sites showed no association with carriage (ORb = 0.34 and 0.37). Inflammatory activity (ORb = 14.74; 95% CI: 9.86-22.03), the presence of intestinal metaplasia (ORb = 2.32) and lymphoid follicles (ORb = 3.81) were factors associated with carriage. Glandular atrophy (ORb = 0.37) was associated with a reduction in carrier status. No statistically significant association was observed with dysplasia (ORb = 1.84; P = 0.410) (Table 2).

After adjusting for age, sex, clinical symptoms, and histological characteristics, several factors remained significantly associated with Helicobacter pylori carriage. The presence of intestinal metaplasia (OR = 3.06; 95% CI: 1.72-5.43; P < 0.001) and that of lymphoid follicles (OR = 3.78; 95% CI: 2.58-5.55; P < 0.001) were independently associated with Helicobacter pylori colonisation. Conversely, the presence of glandular atrophy was associated with a significant reduction in the odds of bacterial colonisation (OR = 0.32; 95% CI: 0.23-0.43; P < 0.001). Of the symptoms studied, only ulcerative syndrome remained significantly associated with Helicobacter pylori carriage after adjustment (OR = 0.63; 95% CI: 0.43-0.93; P = 0.019). No statistically significant association was observed with age or sex. The model fit was satisfactory, as indicated by the Hosmer-Lemeshow test (χ² = 9.01; P = 0.342). The specification test (link test) revealed no major specification errors in the model (_hatsq: p = 0.856). The model's discriminatory power was moderate, with an area under the ROC curve of 0.639 (Table 3).

A total of 40 patients (1.1 per cent) had a neoplastic or precancerous lesion. The mean age of the patients was 46.9 ± 16.9 years, with no statistically significant difference (P = 0.412). Similarly, no significant association was observed with sex (P = 0.942) or the clinical symptoms reported (P = 0.324). However, a statistically significant association was observed between the presence of a neoplastic lesion and the site of the histological sample (P < 0.001). Lesions were proportionally more common in duodenal samples (4.76 per cent) than in samples from other sites (Table 4).

Discussion     

The aim of this study was to determine the prevalence of histological Helicobacter pylori colonisation and to identify the associated clinical and histological factors in patients who underwent an anatomical and cytopathological examination of gastric and duodenal biopsies in the Central region of Burkina Faso. A total of 3,699 patients who had undergone an anatomical pathology examination were included. The prevalence of histological Helicobacter pylori colonisation was 80.3% of cases. The study population consisted mainly of adults aged between 20 and 59 years (74.44%), with a predominance of women (53.1%). Epigastric pain (53.2%) was the main reason for the patients' consultation. Age and sex were not significantly associated with Helicobacter pylori colonisation. However, inflammatory activity was the factor most strongly associated with Helicobacter pylori colonisation. After adjustment, lymphoid follicles and intestinal metaplasia remained independently associated with Helicobacter pylori colonisation. Glandular atrophy and ulcerative syndrome, on the other hand, were associated with a lower prevalence of colonisation.

Helicobacter pylori infection remains the most common infection in developing countries. Its prevalence in our study was 80.31%. This figure falls within the range reported by numerous African studies (56.4% to 91.3%), whereas European data remain below 50% [24]. Our prevalence is lower than that reported by Bounder in Morocco (89.6%) and by Werme in Burkina Faso (91.43%) [23,25], and higher than that observed by Darko in Ghana (45.2%) and by Joutei in Morocco (69%) [24,26]. This high infection rate should be interpreted with caution; the high prevalence of Helicobacter pylori could be associated with several factors, such as hygiene conditions, socio-economic status, or environmental exposures, which are considered risk factors [27-30].

The study population was mainly composed of young adults aged 20 to 59 years (74.44%). Several studies report a concentration of cases in the 19-59-year age group [24,27,31,32]. A bimodal distribution has also been described, suggesting two peaks of prevalence, one among younger and one among older individuals [33]. Some studies, however, show an increased prevalence among patients over 50 years of age [28]. The prevalence observed in the 20-39 and 40-59-year age groups could be associated with infection acquired during childhood. Research conducted by Hazel M. and Mitchell shows that the prevalence of Helicobacter pylori infection in children under 10 years of age appears to increase by 0.5 to 2% per year [34]. Furthermore, socio-economic factors such as poor living conditions, household overcrowding (OR = 1.38) and precarious employment (OR = 1.23) could be associated with the increased risk observed in several population subgroups [35].

Demographic analysis revealed that women accounted for 53.1% of the study population. With regard to gender, the relationship with Helicobacter pylori infection appears ambiguous in the literature. Some studies report significantly higher infection rates among women [24,29-33], as observed in our study, whereas others report higher rates among men [34-37]. The higher number of cases among women in our study could be due to the larger proportion of women in the population of the Centre Region [18]. Statistical analysis showed no significant difference according to sex (P = 0.469) or age category (P = 0.162) in relation to infection carriage. By contrast, the work of Joutei et al. 2010, found that age alone was significantly associated with the prevalence of Helicobacter pylori infection (X² = 38.22; P = 0.000003 < 0.05) [24]. This situation could be explained by household crowding in developing countries and open-air defecation by children who are sometimes already infected [33,38]. This near-universal exposure to Helicobacter pylori from an early age may complicate the identification of subgroups with varying risk profiles.

Epigastric pain was the most frequent reason for upper digestive endoscopy (53.2%), a result higher than those reported in Algiers in 2019 (26.9%) [27] and in Togo in 2014 (48.8%) [39]. This predominance is probably consistent given that Helicobacter pylori infection is closely linked to inflammation of the gastric mucosa. Helicobacter pylori infection can be associated with acute gastritis, which may progress to chronic gastritis, a source of epigastric pain in the absence of treatment.

Multivariate analysis showed that the presence of intestinal metaplasia (aOR = 3.06; 95% CI: 1.72-5.43; P < 0.001) and of lymphoid follicles (aOR = 3.78; 95% CI: 2.58-5.55; P < 0.001) were independently associated with Helicobacter pylori carriage. Our data are consistent with those of Chen et al. and Genta et al. [40,41]. From a pathophysiological standpoint, following colonisation of the gastric mucosa by Helicobacter pylori, a local immune reaction may occur, involving proliferation of T lymphocytes and macrophages with the release of cytokines (interleukin-2 (IL-2) and IL-6), which could lead to B-cell proliferation and the development of follicles and lymphoid aggregates in the infected gastric mucosa [40,42]. Regarding intestinal metaplasia, our results corroborate those of Jmaa et al. Ohkuma et al. and Aumpan et al. [38,43,44]. Earlier studies show that intestinal metaplasia is more prevalent among patients with Helicobacter pylori infection (43.1%) than among those without (6.2%). The prevalence of intestinal metaplasia also appears to vary according to geographic context, being higher in countries with a high incidence of gastric cancer such as Japan and China, compared with sub-Saharan African countries [38].

The significant reduction in the risk of bacterial colonisation by Helicobacter pylori could be explained by the rarefaction of gastric glands that characterises gastric atrophy [28,34]. This finding is consistent with several histopathological studies confirming that as gastric atrophy progresses, bacterial density decreases and Helicobacter pylori becomes more difficult to detect [28,45]. The inverse association observed in multivariate analysis (OR = 0.63; 95% CI: 0.43-0.93; P = 0.019) may be related to confounding factors (advanced gastric lesions, prior treatments, selection bias among patients who underwent endoscopy), since numerous earlier findings show that dyspeptic symptoms do not reliably distinguish patients infected with Helicobacter pylori from those who are not [46,47].

This study has several methodological limitations that may significantly affect the interpretation of the results. First, this is a retrospective study in which several factors (socio-economic factors, hygiene conditions and environmental exposures, antibiotic or PPI use, alcohol or tobacco consumption) were not available. This limits the consideration of these factors in the conclusions and precludes establishing a causal relationship between the variables studied. Second, the study included only patients with gastric symptoms. This introduces selection bias and limits the representativeness of the results for the general population. In addition, the study was conducted in the Centre Region of Burkina Faso, which also reduces external validity and prevents generalisation of the results to the country as a whole. Finally, data collection was carried out in ten of the eleven pathological anatomy and cytology laboratories in the Centre Region; the exclusion of one laboratory could lead to a slight underestimation or overestimation of the true prevalence. From a histological standpoint, the Sydney system was not systematically applied at all levels, particularly regarding the presence of tumour lesions. This approach does not allow determination of whether these tumour lesions were directly or indirectly related to Helicobacter pylori.

In addition, certain variables had a non-negligible proportion of missing data, such as sampling site (33.2%) and symptoms (24.7%), which led to their exclusion from the main model. Excluded patients could indeed present clinical or histological characteristics different from those included, which may affect the estimates of the associations studied. Furthermore, histological interpretation based on the Sydney system may be subject to inter-observer variability. The absence of a concordance assessment in diagnosis between pathologists, or of a centralised re-reading of specimens, could constitute a limitation likely to influence the classification of gastric lesions or the detection of Helicobacter pylori. Finally, the identification of Helicobacter pylori was based solely on histopathological examination, without confirmation by molecular and/or bacteriological techniques. Despite these limitations, this study has several strengths, some of which can be highlighted: first, the majority of pathological anatomy and cytology laboratories (ten of eleven) were included in this retrospective study, which made it possible to capture the full population of patients who underwent upper digestive fibroscopy with biopsy in the Centre Region of Burkina Faso. Second, the use of the histological method made it possible to determine not only the prevalence of Helicobacter pylori but also the associated gastric lesions according to the Sydney system. Third, the multivariate analysis allowed identification of the factors independently associated with Helicobacter pylori infection. Finally, with 3,699 patients included, this study provides updated data on the histopathological epidemiology of Helicobacter pylori in Burkina Faso, thereby contributing to the body of available evidence in this context of Helicobacter pylori endemicity.

Conclusion     

This study confirms that the prevalence of Helicobacter pylori remains high in Burkina Faso, particularly in the Centre Region. In our study, lymphoid follicles and intestinal metaplasia were independently associated with Helicobacter pylori colonisation. Tumour lesions were also identified. In light of these findings, early screening for these lesions appears essential to enable appropriate treatment and close follow-up of individuals at high risk of developing gastric cancer. Moreover, improved sanitation, the promotion of hygiene education, and better access to diagnostic and treatment services could help slow the spread of Helicobacter pylori and reduce the associated morbidity and mortality. However, for these strategies to be effective, they must be adapted to the socio-economic and cultural contexts of the populations concerned. As a continuation of this work, we plan to investigate the aetiological factors of Helicobacter pylori infection, to analyse its antibiotic resistance profiles from biopsy samples, and to explore potential reservoirs via food sources.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Saydou Beogo, as lead author, contributed to study conception and design, data analysis, and manuscript drafting. Hassan Koanda performed the statistical analysis and contributed to data interpretation. Touwendsida Serge Bagre and Souleymane Ouattara critically reviewed the manuscript and contributed to its revision. Nicolas Barro supervised the study and contributed to critical revision of the manuscript. All the authors read and approved the final version of this manuscript.

Acknowledgments     

We would like to express our gratitude and appreciation to all public and private healthcare facilities that kindly agreed to allow the collection to take place on their premises. To all those who supported us in carrying out this study.

Tables     

Table 1: sociodemographic and clinical characteristics of the patients included in the study
Table 2: distribution of Helicobacter pylori carriage according to patients' sociodemographic, clinical and histopathological characteristics
Table 3: factors independently associated with histological Helicobacter pylori colonisation
Table 4: bivariate analysis of factors associated with neoplastic or precancerous lesions

References