Introduction
Bariatric surgery is a leading treatment for severe obesity [1], with laparoscopic sleeve gastrectomy (SG) being the most common procedure worldwide [2]. The need for preoperative testing for Helicobacter pylori before bariatric surgery remains unclear. The 2020 International Federation for the Surgery of Obesity (IFSO) consensus statement made no recommendation on routine H. pylori eradication before surgery, though preoperative esophagogastroduodenoscopy (EGD) is often considered [3, 4]. In Poland, EGD with H. pylori testing and eradication therapy is standard practice [5].
Various H. pylori detection methods exist, including direct methods (rapid urease test, histology) and indirect ones (fecal antigen, urease breath test, serology), but none is a gold standard [6, 7]. Some suggest combining methods to improve accuracy [7, 8]. In bariatric patients, H. pylori testing is usually done during endoscopy, using rapid urease test (RUT), histology, or both, though multiple tests can be costly and time-consuming.
Few studies have focused on H. pylori detection methods in bariatric patients.
Aim
This study aimed to examine the concordance between RUT and histopathology results, the two most commonly used diagnostic methods in this population, while also collecting data on EGD and histology findings.
Materials and methods
Patient selection and methods
This retrospective cohort study included 287 patients with obesity eligible for bariatric treatment [9] operated on between January 2022 and December 2023 in the Tertiary Bariatric Center at University Hospital in Krakow, Poland. As part of the preoperative work-up, they were scheduled for EGD with a RUT for H. pylori detection and histopathology sampling.
Seven were excluded due to their medical history (oncological treatment, partial gastrectomy). All patients were advised to stop proton pump inhibitors (PPI) 14 days before EGD. None of the patients were administered antibiotic treatment 14 days before EGD [10].
Esophagogastroduodenoscopy
The procedure was performed according to a standard protocol [11] approximately 30 days before the surgery. An Olympus endoscope was used to perform the procedure. Tissue samples were obtained from the antrum and body for the RUT test and histology evaluation.
The diagnosis of macroscopic gastropathy and the extent of mucosal inflammation reported in the EGD results were collected in the database. The diagnosis of endoscopic gastritis was made based on the presence of diffuse redness with or without a regular arrangement of collecting venules, atrophy, intestinal metaplasia, enlarged folds, or nodularity [12].
RUT
Lencomm Gold Hp dry urease tests were used for quick diagnosis of H. pylori. The tests were stored and used as recommended by the manufacturer. Two specimen samples from the antrum and body were transferred to the reaction field on the test plate. If the reaction field became red or pink, the test for H. pylori presence was considered positive. The unchanged field was considered negative. Every sample was kept at room temperature for up to 1 h before the final result was confirmed.
Histopathology sampling
Specimens were dehydrated in a series of increasing concentrations of ethanol (Sakura Prisma), cleared in xylene (Sakura Prisma), and embedded in paraffin blocks (Sakura Prisma). The specimen was cut in a rotating microtome into 2.5 µm thick slides. After drying, samples were stained with hematoxylin and eosin (H&E) (Sakura Prisma) using the Sakura Prisma dying system. H&E-stained slides were examined by experienced histopathologists using light microscopy. A modified Giemsa stain was used to improve the sensitivity of H. pylori detection. The presence, type, and intensity of gastritis, intestinal metaplasia, dysplasia, and H. pylori colonization were evaluated in the specimen according to the Sydney System of Classification for Gastritis [13].
Statistical analysis
The study was designed according to STROBE guidelines [14] for observational studies. A descriptive analysis was performed. Categorical variables were described using frequency rate and percentage. Continuous variables were described with mean and range. The Cohen’s kappa inter-rater reliability test was used for concordance calculation [15]. Data were collected and analyzed using Microsoft Excel Spreadsheet Software.
Results
Clinical characteristics
Of the 280 patients included in the analysis, 204 (72.86%) were female, with a mean age of 40.11 years (ranging from 21 to 65 years). Forty (14.28%) patients suffered from type 2 diabetes mellitus, 33 (11.79%) had insulin resistance, 135 (48.21%) had hypertension, and 55(19.64%) had obstructive sleep apnea.
Forty-five (16.07%) patients experienced dyspeptic symptoms; of these, only 37 (82.22%) were on regular proton pump inhibitor therapy and 19 (42.22%) tested positive either on RUT or the histopathology test.
H. pylori in RUT vs. histopathology
Ninety-eight (35.00%) out of 280 patients were diagnosed with H. pylori infection by at least one diagnostic test used in the study. In 75 (26.79%) patients, both tests were positive; in 17 (6.07%) patients, H. pylori was detected only in histopathology; and in 6 (2.14%) patients, only the RUT results were positive. The overall Cohen’s kappa inter-rater agreement between the two diagnostic tests was 0.80. The details are presented in Table I. Macroscopic gastropathy was present in 76 (77.56%) of 98 patients. Microscopic evaluation showed inflammatory changes in 97 (98.99%) patients.
Table I
Cohen’s kappa inter-rater agreement between the RUT and histopathological H. pylori tests
| Parameter | Kappa | 95% CI | SE | |
|---|---|---|---|---|
| All cases | 0.80 | 0.73–0.88 | 0.03 | |
| Gastritis on EGD | ||||
| Yes | 0.73 | 0.62–0.83 | 0.05 | |
| No | 0.94 | 0.86–1.0 | 0.04 | |
| Gastritis on histopathology | ||||
| Yes | 0.72 | 0.61–0.82 | 0.05 | |
| No | 0.98 | 0.61–0.82 | 0.01 | |
Overall, H. pylori infection was detected in 23 (19.32%) out of 119 patients with normal gastric mucosa and in 75 (46.58%) out of 161 patients with gastritis in EGD.
Gastritis in EGD vs. gastritis in histology
One hundred sixty-one (57.50%) of the study group were diagnosed with macroscopic gastropathy during EGD evaluation. Among these, histopathological evaluation confirmed inflammation in 119 (73.91%) patients. Chronic gastritis was present in 112 (90.91%) patients, with active inflammation in 69 (57.98%) biopsy specimens. Atrophic changes were described in 22 (18.48%) cases, and intestinal metaplasia was detected in 11 (9.24%) specimens, including 9 (8.03%) cases of mixed complete and incomplete second-degree metaplasia, 1 (0.89%) case of complete metaplasia, 1 (0.89%) case of incomplete metaplasia, and 1 (0.89%) case of low-grade dysplasia.
Of the 119 patients with normal gastric mucosa on EGD, 41 (34.45%) had positive findings on microscopic evaluation. Chronic gastritis, either low-grade or moderate-grade, was detected in 38 of these cases (92.68%). Active inflammation was present in 21 (51.21%) out of the 41 cases, atrophic gastritis was identified in 5 (12.19%) cases, and complete metaplasia was observed in 4 (9.30%) cases.
Overall, 157 (56.07%) patients from the study cohort had abnormal histopathology reports, chronic gastritis was detected in 150 (53.57%) patients, atrophic gastritis in 27 (9.64%), low-grade dysplasia in 1 case, and metaplasia in 23 (8.21%) biopsies. Of these, 8 (354.78%) cases showed mixed complete and incomplete second-degree metaplasia (Table II).
Table II
Histopathology vs. EGD diagnosis. The percentage of EGD findings is calculated in relation to the cohort sample and the percentage of histopathology in relation to normal or abnormal EGD findings
The sensitivity of EGD evaluation of gastritis was 75.43%, and the specificity was 67.86%, with an accuracy of 72.47%. The details are presented in Table III.
Table III
Evaluation of gastroscopic diagnosis of gastric mucosal changes
Discussion
The need for preoperative identification and eradication of H. pylori infection in the bariatric population remains undetermined [3]. If performed, the most commonly employed methods during the preoperative assessment are RUT or histology sampling. In our study, we retrospectively analyzed the concordance between the two diagnostic techniques and the current prevalence of H. pylori infection and EGD findings in bariatric patients in Poland.
In our study, 34.84% of the bariatric patients were identified as having H. pylori infection based on at least one diagnostic tool. Previous epidemiological estimates on the general Polish population reported an H. pylori colonization prevalence ranging between 68% and 84% among adults [16, 17]. However, the prevalence of H. pylori infection is influenced by socioeconomic status and hygiene levels, and more recent data report decreasing prevalence in the selected Polish populations [18, 19]. Additionally, the mean age of our study group was 40 years, which could also explain the relatively low prevalence of H. pylori, as it increases with age, and is more common among older adults [16, 20–22].
According to our results, the overall Cohen’s kappa inter-rater agreement between the diagnostic tests was 0.80, which signifies a good concordance between the two methods [23]. Our findings are in keeping with the results reported by Valadares et al., who reported 87.1% accuracy of RUT when compared to histopathology as a reference point for H. pylori detection. Interestingly, the agreement was slightly lower (0.82) among patients with signs of gastritis diagnosed on either gastroscopy or histopathology.
Our research outcomes align with a recent systematic review that compared histology with the RUT, revealing similar high levels of sensitivity and specificity for histopathologic tests and RUT among patients experiencing dyspepsia [24].
Numerous commercially available RUT tests exist, including CLOtest, HpFast, PyloriTek, ProntoDry, UFT300, and EndoscHp, which have been commonly referenced in the literature [25, 26]. These tests possess slight variations in specificity, sensitivity, and reaction times for delivering results [25, 26]. In our investigation, we employed Lencomm Gold Hp dry urease tests, which, to the best of our knowledge, have not been investigated in comparative research of various rapid urease tests.
56% of the study cohort had abnormal histopathology reports, chronic gastritis was present in 54%, atrophic gastritis with complete second-degree metaplasia in 10%, and intestinal metaplasia in 8%. Similarly, in the study by García-Gómez-Heras, the prevalence of chronic gastritis, gastric atrophy, and gastric intestinal metaplasia among their bariatric population was 47%, 3.6%, and 7.7% [27].
The sensitivity of EGD evaluation of gastritis was 75.43%, and the specificity was 67.86%, with an accuracy of 72.47%. These results are in agreement with the previously published reports demonstrating an overall poor correlation between gastroscopy and histopathology results in diagnosing gastritis [28–30].
This is a retrospective analysis of the medical records of bariatric patients comparing the concordance of the two different H. pylori detection methods. Taking into consideration previously reported false positive and false negative results in both tests, the true prevalence of H . pylori infection among bariatric patients, as well as the positive and negative predictive value of the tests, could not be established. Additionally, no data on previous H. pylori eradication were obtained.
The study population included only patients who underwent laparoscopic sleeve gastrectomy. By definition, these were bariatric patients whose EGD results did not alter the surgical decision. Therefore, the applicability of the conclusions regarding the gastroscopic findings observed in the study is limited and should not be extrapolated to the entire population of bariatric candidates.
Conclusions
There is a good concordance between RUT and histopathological diagnoses of H. pylori among the bariatric population. Employing combined diagnostic methods may be unnecessary, considering the uncertain clinical benefit of H. pylori eradication before bariatric surgery. Additionally, there is a low accuracy rate in EGD assessments of gastric mucosa changes.
