Aims This study was designed to establish the relative prevalence of intestinal-type and signet-ring carcinoma in gastric biopsy specimens from ambulatory patients, to determine the percentage of signet-ring carcinomas that could be expected based on the available clinical and endoscopic information, and to estimate the likelihood of missing a tumour.

Methods We extracted data of all patients with a diagnosis of primary gastric carcinoma from a national pathology database. We then reviewed clinical information and original slides, classified tumours as intestinal or signet-ring-type, and categorised the latter as ‘unexpected’ (no alarming symptoms, no mention of suspicious lesions) or ‘expected’ (clinical or endoscopic information suggestive of tumour). Unexpected signet-ring carcinomas were categorised as ‘obvious’ or ‘challenging’ (rare signet-ring cells; immunohistochemical stains used to confirm the nature of the infiltrates).

Results There were 310 109 patients with gastric biopsies; 615 patients had primary gastric carcinoma (359 intestinal and 256 signet-ring-type). Gastric cancer was more common in men (OR 2.54; 95% CI 2.05 to 3.14; p<.0001) for intestinal-type and (OR 1.90; 95% CI 1.48 to 2.42; p<0.0001) for signet-ring cell type). Intestinal-type carcinoma occurred in older patients than signet-ring-type (median age 74 vs 65 years, p<0.001). There were 196 expected and 60 unexpected signet-ring carcinomas; 47 of the 60 unexpected cases were histopathologically obvious. Thus, only 13 signet-ring carcinomas (1 in 25 000 gastric biopsy sets) were truly unexpected.

Conclusions Signet-ring carcinoma is a rare finding in gastric biopsy specimens from ambulatory patients; routine due diligence and the clinical/endoscopic information provided are usually adequate to raise pathologists’ index of suspicion.