Correlation between upper gastrointestinal endoscopic findings and Helicobacter pylori detection in gastric biopsy specimens: a retrospective study

¹Gastroenterology Unit, Department of Medicine, University College Hospital, Ibadan, Nigeria, ²Pathology Department, University College Hospital, Ibadan, Nigeria

^&Corresponding author
Morenike Adedoyin Osundina, Gastroenterology Unit, Department of Medicine, University College Hospital, Ibadan, Nigeria

Introduction    

Helicobacter pylori (Hp) is a non-sporulating, gram-negative bacterium. Morphologically, it can vary in shape from spiral, rod-shaped, curved-rod, coccoid, gull-winged, or U-shaped. It is microaerophilic and motile, the latter being made possible due to its embellishment with five to six sheathed lophotrichous flagellae [1]. It was initially thought that the acidic milieu of the stomach made it a sterile organ. However, in the 1980s, Marshall and Warren demonstrated that HP not only colonizes but also induces gastric inflammation [2]. H. pylori is considered a neutralophiles, that is, capable of surviving between pH levels of 4 and 8, while growth is possible between pH 6 and 8 [3]. It is with these and other virulence factors that allow this organism to earn the status of a true resident of the gastric mucosa [4].

The global prevalence of Hp infection amongst both children and adults has decreased from 52.6% before 1990 to 43.9% in the years between 2015 and 2022. Helicobacter pylori inhabit the gastrointestinal tracts of more than half of the inhabitants of Africa, which is the highest recorded for any region of the world [5]. The infection is believed to be acquired during childhood and remains asymptomatic in many individuals, with only about 20 -30% becoming symptomatic. The transmission remains unclear, but oral-oral, faeco-oral, and gastric-oral modes have been identified as routes of infection, particularly in low socioeconomic households where poor sanitation and overcrowding drives the spread [6]. The deleterious effects of this organism on the gastroduodenal mucosa are well established, ranging from chronic gastritis, duodenitis, gastroduodenal ulcers, mucosal associated lymphoid tissue lymphoma, and gastric cancer [7-9]. Additionally, it has been linked to a cascade of events that starts off as seemingly benign mucosal lesions culminating in a catastrophic malignant condition, in what is currently referred to in the literature as the Correa Cascade. While its effects have been well characterized in the stomach, research work have also been extended to evaluate potential links between H. pylori and other diseases in other organ systems such as cardiometabolic, autoimmune diseases and even nutrient deficiencies particularly Vitamin B12 and iron [10,11]. It therefore implies that elimination of the organism from the gastrointestinal tract is imperative to forestall the dysfunctional epithelial barrier it incites on its way to inducing carcinogenesis.

Diagnosis of Hp could be via non-invasive or invasive techniques. Non-invasive methods include the urea breath test (UBT), serology, and stool antigen test. The invasive methods include the use of an endoscopic biopsy for rapid urease test (RUT), histology, culture or polymerase chain reaction (PCR) [9,12-15]. Histology remains the gold standard for diagnosing H. pylori, providing direct insights into the presence and severity of the infection, as well as associated conditions like intestinal metaplasia, glandular atrophy, dysplasia, and neoplasia. Commonly used stains for detecting H. pylori include hematoxylin and eosin (HE), Giemsa, and Romanowsky. However, histology has its limitations, such as inter-observer variability in tissue assessment and the necessity of an endoscopy to obtain samples. Given the patchy distribution of H. pylori in the stomach, biopsies must be taken from multiple sites. The sensitivity and specificity of histology range from 53% to 90%, but these can be enhanced by increasing the number of biopsies and using specific stains [9]. Upper gastrointestinal endoscopy is a veritable tool used in obtaining gastric biopsies for the detection of Hp, either by culture, rapid urease test, or histological demonstration of the organism. Endoscopic features of Hp status have been documented in some studies, particularly in Caucasians and Asians. It is generally believed that regular arrangement of collecting venules (RAC) along the lesser curvature of the stomach is a marker of Hp-negative status, whereas findings such as diffuse redness, patchy/spotty redness, sticky mucus, antral nodularity, enlarged gastric folds, xanthoma, hyperplastic polyps, and gastric mucosa edema predict the presence of Hp infection [16].

Recent trends emphasize the direct diagnosis of Helicobacter pylori during endoscopy. Research has been conducted to ascertain whether mucosal changes can predict the presence of H. pylori infection or at least determine its absence, thereby guiding the decision to obtain biopsies for histological confirmation and avoiding delays in treatment. It has been demonstrated that enlarged gastric folds and antral nodularity predict H. pylori infection with a positive predictive value (PPV) of 100%, while fundic gland polyps and red streaks provide a 100% PPV for the exclusion of H. pylori infection using white light imaging during endoscopy. Furthermore, image-enhanced endoscopic techniques have shown superior sensitivity and specificity compared to white light imaging [16].

In the Kyoto classification, 5 endoscopic parameters (atrophy, enlarged folds, nodularity, diffused redness, and intestinal metaplasia) are scored with a minimum score of 0 and a maximum of 8. A score of 2 or more is associated with Hp positive status, while scores ≥ 4 are associated with higher gastric cancer risk [17]. Whether these findings are enough to immediately initiate treatment is debatable. In resource-poor settings like ours, where histopathological turnaround time is longer, we seek to understand endoscopic features of Hp status in a simplified manner and whether the endoscopic appearance is enough to commence Hp eradication therapy. Unfortunately, information on endoscopic predictors of Hp status is sparse in Africa, with the most recent work done in Tanzania, where the presence of duodenal and gastric ulcers was the only statistically significant predictor of Hp presence [18]. This study aimed to correlate gastroduodenal endoscopic findings with the detection of H. pylori in histological examinations

Methods     

Study design and setting: this was a retrospective, descriptive cross-sectional study conducted at the University College Hospital (UCH), Ibadan, Oyo State, Nigeria. The study covered the period from 1 January 2021 to 31 December 2023. University College Hospital (UCH) is a tertiary referral center where upper gastrointestinal endoscopy is routinely performed by consultant gastroenterologists and gastroenterology trainees under the supervision of the consultants.

Participants: all patients who underwent upper gastrointestinal endoscopy within the study period were included.

Variables: the primary outcome was Helicobacter pylori infection status, determined by histological examination of gastric biopsy specimens. Exposure variables included endoscopic findings in the stomach and duodenum, recorded at the time of endoscopy (erosions, ulcers, polyps, masses, nodularity, haemorrhagic changes, and others). Additional variables extracted were demographic characteristics (age, sex) and indications for endoscopy (dyspepsia, upper gastrointestinal bleeding, dysphagia, recurrent vomiting, suspected malignancy, unexplained weight loss, and others).

Data sources and measurement: data was obtained from the endoscopy unit register and matched with pathology department records to retrieve histology results. Histology was regarded as the gold standard for H. pylori diagnosis. Endoscopic findings were documented prospectively at the time of the procedure using standard reporting formats, and the data were collected based on the findings at the time the procedures were done. Where multiple endoscopic abnormalities were present, each was recorded separately.

Bias: potential sources of bias included inter-observer variability in endoscopic reporting and missing histology results. Biopsies were routinely taken from the antrum even in patients with seemingly normal findings. However, some patients didn´t have biopsies taken. All procedures were performed by trained gastroenterologists or supervised trainees using a standardized reporting template.

Study size: no formal sample size calculation was performed. Instead, all eligible endoscopies performed during the study period were included, yielding 1,147 procedures, of which 823 (71.7%) had corresponding histology reports.

Quantitative variables: age was recorded as a continuous variable and expressed as mean ± standard deviation. For analysis, categorical variables such as sex, indications for endoscopy, and endoscopic findings were expressed as frequencies and percentages.

Statistical methods: data was analyzed using SPSS version 20 (IBM Corp., Armonk, NY, USA). Categorical variables were compared between H. pylori-positive and negative groups using the chi-square (χ²) test. Sensitivity and specificity of selected endoscopic findings (e.g., gastric erosion, any abnormal stomach or duodenum finding) for predicting H. pylori infection were calculated. A p-value <0.05 was considered statistically significant. Missing data was excluded from the analysis, particularly patients without a histological diagnosis.

Ethical considerations: ethical approval was obtained from the Joint University of Ibadan/University College Hospital Ethics Committee (UI/EC/24/0845). Patient confidentiality was maintained by anonymizing all data during extraction and analysis.

Results     

Participants: a total of 1,147 upper gastrointestinal endoscopic procedures were performed during the study period. Histological reports were available for 823 cases in which antral biopsies were obtained. The remaining procedures had no histological results of gastric antral biopsy, which is routinely performed to identify Helicobacter pylori infection even in endoscopically normal mucosa.

Descriptive data: out of the 1,147 individuals who had the procedure done, 609 (53.2%) were females and 535 (46.8%) were males, with an average age of 51.4±16.7 years. The most prevalent indication for endoscopy was dyspepsia in 642 (56.1%), followed by upper gastrointestinal bleeding in 163(14.2%) and dysphagia in 65 (5.7%).

Outcome data: the commonest endoscopic finding that correlates with the presence of HP was gastric erosions. Other less common findings were antral nodularity and haemorrhagic gastritis, as shown in Figure 1, while Figure 2 is one of the antral biopsy histology showing HP-like organisms.

Main results: histology was traceable in 823 (71.7%) procedures. 377 (45.8%) had H. pylori detected (Hp+) at histology, and 446 (54.2%) without (Hp-). The mean age for those with Hp+ status was 51.1 ± 16.0, while it was 52.5 ± 15.9 for Hp- status. The gender distribution as well as the indication for upper GI endoscopy were similar in both groups. Indications and endoscopic findings based on HP status are shown in Table 1, Table 2,Table 3.

Other analysis: the sensitivity of gastric erosion was 75.1% while erosions in the duodenum had a sensitivity of 89.2%. Diagnostic performance of any abnormality in the stomach and duodenum, as shown in Table 4 while Table 5 shows the summary of histological reports of patients with HP negative status.

Discussion     

In this study, abnormal findings in the stomach and duodenum were strongly associated with Hp-positive status, while normal findings either in the gastric or duodenal mucosa were associated with Hp-negative status. The prevalence of H. pylori diagnosed histologically was 46%, which was higher than the 44.1% reported by Mao and colleagues [19], 26% reported in Japan [20] and 14.16% in Lebanon [21]. Only 6% of patients with Hp had normal gastric epithelium on white light endoscopy, which was considerably lower than the 20 percent reported in a different study [16]. This difference may be because the latter study reported a lower prevalence, despite using advanced image-enhanced endoscopy techniques. The sensitivity of gastric erosion observed in our study was 75%, whereas the specificity was significantly lower than 50%. This finding is in contrast to previous research that demonstrated low sensitivity but high specificity of gastric erosion for Hp infection, although those studies questioned the clinical utility of gastric erosion as a predictor of Hp status [19,20,22]. This discrepancy may be attributed, in part, to variations in the reporting methods of pathological gastric findings. Additionally, our study revealed a notably high prevalence of gastric erosions, even among Helicobacter pylori-negative patients, at a rate of 55.6%.

There were significant differences in endoscopic findings between patients with Hp-positive (Hp+) and Hp-negative (Hp-) histology results. In the stomach, erosions were more frequently observed at endoscopy in patients with histologically confirmed H. pylori infection (75.1% vs 56.1%), suggesting that the presence of this abnormality may be associated with H. pylori infection (χ²= 33.8, p < 0.001). This is consistent with previous research and aligns with the well-documented role of Hp in gastric mucosal pathology [19]. This underscores the potential for endoscopy to reveal mucosal abnormalities linked to Hp infection, thereby providing valuable diagnostic information [20,23].

In contrast to the findings in the stomach, there appears to be a similar pattern for both groups in their endoscopic duodenal findings. This is a less common finding and warrants further investigation [24]. These results suggest that while Hp infection is a key factor in gastric pathology, its role in duodenal conditions may be less pronounced or influenced by other factors. This is particularly evident when the sensitivity of any abnormality in the gastric mucosa (93.9%) and duodenal mucosa (21.0%) are compared. This finding is similar to what was found among the Lebanese cohort, where less than a third had inflammatory changes in the duodenum, though this reached statistical significance, unlike in our study [21]. Unlike this study where there was an attempt to characterize duodenal mucosa changes, some similar studies did not document the endoscopic appearance of the duodenal mucosa as an indicator of Hp status. [19,20].

Furthermore, there were more histologically confirmed gastric cancer cases in the Hp-negative group. Sampling of mass lesions without adjacent areas may explain this finding. Another plausible explanation is bias in reporting; histopathological findings of a cancer may lead to under-reporting of H. pylori. In addition, the hit-and-run hypothesis of Hp carcinogenesis has been well described in the literature [25-27]. The theory proposes that long-standing infection induced carcinogenesis, but continuous presence of the organism in the stomach is not required once gastric cancer is established. Nevertheless, 3.9% (45/1147) of all patients who had upper gastrointestinal endoscopy done had gastric masses, and about a third had Hp-associated gastritis. However, it has been shown that genomic techniques can increase the identification of Hp in gastric cancer tissue, with 89% reported for Latin America and 85% in Europe in the Legacy study, and we may benefit from this in our region [28]. Our findings highlight the importance of integrating endoscopic and histological evaluations for diagnosing H. pylori. Gastric erosions observed during endoscopy should raise suspicion of Hp infection, given their strong association with Hp infection in our study. Histological confirmation, however, remains essential due to the varied presentation of H. pylori infection.

Limitation: while the Kyoto classification was used in most studies assessing endoscopic predictors of Hp infection, this was not the case in this work. While our study provides significant insights, it is limited by the retrospective nature and the single-centre setting, which may affect the generalizability of the findings. Future research could make use of a multicentric approach and a prospective study design to validate our results.

Conclusion     

Gastric abnormalities were more prevalent at endoscopy, and gastric erosion was well correlated with the histological diagnosis of H. pylori. Therefore, empiric eradication therapy could be instituted in this category of patients, particularly in resource-poor settings like ours.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Conception and study design: Morenike Adedoyin Osundina, Hakeem Abiodun Alimi and Jesse Abiodun Otegbayo. Data collection: Morenike Adedoyin Osundina, Hakeem Abiodun Alimi, Tajudeen Adeolu Badmos, Modupe Kehinde Ijikoyejo, and Ifeanyichukwu Dupe Nwanji. Data analysis and interpretation: Hakeem Abiodun Alimi, Morenike Adedoyin Osundina, Jesse Abiodun Otegbayo, Tajudeen Adeolu Badmos, Modupe Kehinde Ijikoyejo and Ifeanyichukwu Dupe Nwanji. Manuscript drafting: Morenike Adedoyin Osundina, Hakeem Abiodun Alimi, Tajudeen Adeolu Badmos, and Modupe Kehinde Ijikoyejo. Manuscript revision: Hakeem Abiodun Alimi, Morenike Adedoyin Osundina, Jesse Abiodun Otegbayo, Tajudeen Adeolu Badmos, Modupe Kehinde Ijikoyejo and Ifeanyichukwu Dupe Nwanji. All the authors have read, and agreed to, the final version of this manuscript for publication.

Acknowledgments     

We acknowledge the contribution of all endoscopists as well as endoscopy nurses of the University College Hospital, Ibadan as well as the pathology residents of the same institution for providing us with the necessary support to get this work done.

Tables and figures     

Table 1: gender distribution and clinical indications by H. pylori status

Table 2: gastric endoscopic findings by H. pylori status

Table 3: duodenal findings and other endoscopic features by H. pylori status

Table 4: diagnostic performance of endoscopic findings in predicting H. pylori-positive status

Table 5: histological findings in gastric biopsies of H. pylori-negative patients

Figure 1: endoscopic images of some patients who had Helicobacter pylori confirmed by histology: A) revealed diffused nodularity while; B) had features of haemorrhagic gastritis (black arrows in A and B, respectively)

Figure 2: photomicrograph of the stomach antrum demonstrating florid colonization of the surface of the mucous layer by short rod-shaped bacilli (Giemsa x400)

References