Letter: gastric atrophy as a precursor of gastric cancer – authors’ reply

Abstract

The clinico-pathological assessment (and natural history) of autoimmune gastritis (AG) is a matter of debate.1 In their appreciated letter,2 Barbosa and Botelho de Carvalho suggest starting with the histological evidence of corpus-restricted gastritis as a reliable marker of autoimmunity targeting the oxyntic cell population. However, if we do, we should be aware that these ‘putative-AG’ patients include cases whose corpus inflammation/atrophy is part of a multifocal atrophic gastritis due to Helicobacter pylori infection. In our trial,3 we ‘cleansed’ our study population by requiring serology-confirmed gastric autoimmunity. Consistent with this choice, our study sample included 52 patients with antiparietal/anti-intrinsic factor antibodies, with corpus-predominant gastritis, but no histological evidence of corpus-restricted gastric atrophy. Among the patients with positive serology, 116 were followed up for longer than 24 months. At the end of the follow-up, 2 of the 6 cases enrolled as having non-atrophic gastritis featured stage II gastritis, supporting the hypothesis that the natural history of AG includes two subsequent phenotypes: initial non-atrophic and advanced atrophic disease.3 As for the OLGA staging, here again – and also in the autoimmune setting – OLGA staging consistently stratified patients by their different cancer risk.4-7 In fact, the stage of gastritis expresses ‘how advanced’ the disease is, irrespective of its aetiology. In our AG series, all incidental cancers coexisted with high-risk stages, basically confirming both that gastric atrophy is the field in which gastric cancer develops, and the clinical usefulness of staging in gastric cancer secondary prevention strategies.5 As for carcinoids, the related risk coincides with the histological score for corpus atrophy (more than with the OLGA stage): 17 of 19 (89.5%) carcinoids coexisted with score 3 corpus atrophy.3 In conclusion, in the presence of a compatible clinical syndrome, serology does identify autoimmune gastric disease, with a histology that includes both non-atrophic and atrophic phenotypes. Irrespective of the aetiology of gastritis, OLGA staging can convey clinically important information on the gastritis-associated cancer risk. The authors’ declarations of personal and financial interests are unchanged from those in the original article.3