Gastric Intestinal Metaplasia (GIM). Natural History and Surveillance

Abstract

Purpose: Determine the prevalence of GIM in GI subspecialty patients, assess the diagnostic value of histological grading, including gastritis and presence of Helicobacter pylori (HP) relative to the risk of adenocarcinoma (GC) distal to the cardia. Methods: In the course of 19 years, there were 11,600 EGDs that diagnosed 354 male patients with GIM (one GIM in 33 EGDs). Of these, 185 patients had at least one follow-up EGD (53%) and 11% had four or more EGDs. The interval between the first and last EGD was 39.4 (34.5 SD) months (1-204). Results: Compared with 350 non-gastric randomly sampled GI controls, patients with GIM tended to be older, with 55.7% being over the age of 70. Fifty percent were African-American and 44.6% were Caucasian, almost identical with the controls. GIM was graded 1 - 5 with a mean grade of 2.2 on initial EGD in non GC patients, while in the 21 patients with GC (5.9% of all GIM cases), it was 4.8. Of the patients with GC, 67% were African-Americans, higher than their distribution among other GIM patients. GIM was labeled as focal in 58%, mild to moderate in 30%, extensive in 8% and malignant in 4%. Severity of GIM did not differ among the age groups, ethnic groups, 153 HP positive and 201 HP negative patients. Of 90 HP positive patients who had repeated EGD, HP was diagnosed only at the initial EGD in 56% (effect of therapy?). Histological severity of gastritis and the number of mucosal inflammatory nodules did not differ among the subgroups. Conclusion: GIM contributes to gastric cancer in a large segment of patients at risk. Patients older than 70 and African-Americans are at higher risk. HP and gastritis were not reliable predictors of neoplasia (therapy? sampling error?). Histologic grading of GIM is relatively non sensitive for staging and progression. Patients who are diagnosed with GIM for the first time should undergo endoscopic surveillance irrespective of the initial grade of GIM. There should be biopsies from multiple gastric sites to increase the diagnostic yield. A biomarker for GIM would increase the effectiveness of screening for gastric cancer.