A step forward for improving the endoscopic diagnosis of autoimmune gastritis

We read with interest the paper by Chen et al. [11], who performed a retrospective study in a cohort of Chinese patients with autoimmune gastritis (AIG) to evaluate the performance of an endoscopic artificial intelligence (AI) system in assisting the diagnosis of AIG. The performance of the AI system was comparable to that of expert endoscopists and significantly outperformed nonexperts. We greatly appreciate such a study, as AIG is a largely underdiagnosed and misdiagnosed condition [22]. However, some considerations are warranted.

According to the inclusion criteria, only seropositive AIG patients (i.e. positive for either anti-parietal cell or anti-intrinsic factor antibodies) were included, with any grade of corpus atrophy. Nonetheless, based on the published endoscopic images, it appears that most of the patients included actually exhibited moderate to severe atrophy, which is the most commonly reported form in a large AIG series [33]. Unfortunately, histopathological data on the severity of atrophy, as well as data on dysplasia, neuroendocrine neoplasms, and overt adenocarcinoma, are not provided. It is known that milder, or very early forms of AIG may show no obvious endoscopic signs of atrophy (e.g. disappearance of gastric folds, prominence of vessels) [44], and these patients could benefit more from an early diagnosis. Thus, the AI system described by Chen et al. may fail to detect the disease in early AIG phases. Additionally, we believe the ideal setting for application of this AI technique is in cases of unsuspected AIG, such as patients undergoing upper gastrointestinal endoscopy for reasons other than AIG. In such scenarios, especially when overt endoscopic signs of AIG are absent, diagnosis may be significantly delayed if biopsies are not taken, as has been previously reported [22].

In conclusion, while we appreciate the utility of this novel AI system, further prospective studies in unselected patients undergoing an upper gastrointestinal endoscopy are warranted.

Publication History

Article published online:
25 March 2025

• References

• 1 Chen S, Xu L, Yan L. et al. A novel endoscopic artificial intelligence system to assist in the diagnosis of autoimmune gastritis: a multicenter study. Endoscopy 2024;
• 2 Lenti MV, Miceli E, Cococcia S. et al. Determinants of diagnostic delay in autoimmune atrophic gastritis. Aliment Pharmacol Ther 2019; 50: 167-175
• 3 Lenti MV, Miceli E, Lahner E. et al. Distinguishing features of autoimmune gastritis depending on previous Helicobacter pylori infection or positivity to anti-parietal cell antibodies: results from the autoimmune gastRitis Italian netwOrk Study grOup (ARIOSO). Am J Gastroenterol 2024;
• 4 Kotera T, Ayaki M, Sumi N. et al. Characteristic endoscopic findings in early-stage autoimmune gastritis. Endosc Int Open 2024; 12: E332-E338