Signet ring cell cancer in a patient with Russell body gastritis – a possible diagnostic pitfall

ing colon. Endoscopic gastric biopsies were obtained on three different visits over a period of 3 years, and were named biopsy 1, biopsy 2, and biopsy 3. Helicobacter pylori phenotyping was performed by serological assay on all three visits. Light microscopy of biopsy 1 revealed mild chronic pangastritis, intestinal metaplasia (IM) and atrophy in the body and cardia, and no pseudolipomatosis foci. Biopsies 2 and 3 revealed moderate chronic gastritis and IM, and marked atrophy in the body and cardia. In both biopsies, empty-looking, clear, oval-shaped vacuoles, 50–100 lm in size, were observed in the cardiac mucosa (Figure 1A,B). Some of these vacuoles coalesced to form an occasional large submucosal vacuole. These vacuoles were associated with mild fibroblastic proliferation. In the second case, mild pangastritis, IM and atrophy were detected in the corpus mucosa, whereas, in the third case, no atrophy or IM was observed. These cases also showed pseudolipomatosis foci, like the index case. Follow-up biopsy of the second case did not reveal any pseudolipomatosis foci in the gastric mucosa, and the third case is still under follow-up. Immunohistochemically, the lining of these pseudoadipocytic vacuoles expressed CD34 (Figure 1C,D). However, S100, factor VIII and Pim-1 were not expressed. Previously, Alper et al. and Belinli et al. proposed that pseudolipomatosis in the stomach may be associated with H. pylori infection. However, Stebbing and Wyatt reported a case of pseudolipomatosis in the gastric body associated with severe atrophy and extensive IM, and negative for H. pylori infection. In our index case, the H. pylori phenotype was negative both histologically and serologically, and the lesion was associated with moderate to marked IM and atrophy, similar to the case of Stebbing and Wyatt. Immunohistochemically, the lining of the pseudolipocytic vacuoles was positive for CD34. CD34 is a transmembrane glycoprotein detected in haematopoietic progenitor cells, endothelium, endoneurium, and stromal cells. Its expression in our case suggests involvement of the residual fibroblasts. The lining of the vacuoles was negative for both S100 and Pim-1, ruling out any adipocytic nature of this condition. In addition, the lack of factor VIII expression argues against a possible vascular nature. To summarize, we suggest that pseudolipomatosis stomachalis is a reactive fibroblastic lesion observed on a background of metaplastic and atrophic gastric mucosa. This lesion is likely to be precipitated by repeated endoscopy in a previously thinned out atrophic gastric mucosa, leading to the migration of gas in the lamina propria, at least in some cases, as indicated by case 1 in our study.