For predicting Helicobacter pylori (H pylori) status, a simple endoscopic program and biopsy collection may be sufficient. These findings were published in BMJ Open Gastroenterology.
Between 2018 and 2020, patients (N=153) were prospectively recruited at Imperial College London for this diagnostic performance and validation study. All patients reported abdominal pain or dyspepsia and were candidates for an upper gastrointestinal tract white-light endoscopy and narrow-band imaging with updated Sydney Protocol biopsy collection. Features of the examination were used as input for a multinominal logistic regression model which predicted H pylori status.
Patients were aged mean 54.23±16.80 years, 56.9% were women, and 58.8% were of European ethnicity. A total of 72.5% were found to be H pylori naïve, 18.3% had active infection, and 9.2% had previous H pylori eradication.
According to the literature, regular arrangement of collecting venules (RAC) is the best predictor of H pylori. RAC were identified among 70.6% of all patients, 78.4% of naïve or eradicated patients, and 35.7% of positive patients.
RAC present in the lower part of the lesser curvature of the stomach was a strong predictor of negative or eradicated status (odds ratio [OR], 6.53; 95% CI, 2.70-15.80; P <.001), especially among patients aged <55 years (negative predictive value, 93.2%; 95% CI, 83.1%-97.3%).
On the basis of these data, the full diagnostic model took into account patient history of H pylori, RAC presence, diffuse redness or edema of the stomach, and map-like redness of the gastric antrum. This model had a goodness-of-fit of 21.8 (P =.005).
A validation cohort of 30 patients was used to assess the predictive model. The diagnostic accuracy was 80.0%, in which no patient with an active infection was misclassified as naïve or eradicated.
This study may have been limited in certain ways, as the proportion of patients with an active infection was substantially lower than anticipated (18.3% vs 35%, respectively). It remains unclear whether these findings may be generalizable.
The study authors concluded that they generated a simple algorithm for predicting H pylori status among patients with possible symptoms using a simple endoscopic program.
Glover B, Teare J, Patel N. Assessment of Helicobacter pylori status by examination of gastric mucosal patterns: diagnostic accuracy of white-light endoscopy and narrow-band imaging. BMJ Open Gastroenterol. 2021;8(1):e000608. doi:10.1136/bmjgast-2021-000608