Expression of eosinophil chemotactic factors in stomach cancer

We have occasionally experienced eosinophilic abscess of the liver in patients with gastric carcinoma, suggesting that some eosinophil mobilizing (chemotactic and proliferative) factors might be produced by carcinoma cells. The aim of this study was to determine whether or not gastric carcinoma expresses the well-known eosinophil chemotactic factors (ECFs) and whether or not the expression is related to the histologic subtypes. Seventeen consecutive surgically removed tumor-bearing stomachs were collected: 7 signet ring cell type, 7 poorly differentiated tubular adenocarcinoma, and 3 moderately differentiated tubular adenocarcinoma. Hematoxylin-eosin stained sections were re-evaluated for eosinophil and mast cell infiltration. The expression of IL-2, IL-5 and granulocyte-macrophage colony stimulating factor (GM-CSF) were examined by immunocytochemical stain. There was no available frozen tissue for IL-2 and IL-5 in one case. Gastric carcinoma expressed IL-2 in all 16 cases, IL-5 in 12 of 16 cases and GM-CSF in 10 of 17 cases. Of particular interest, 7 of 10 GM-CSF-expressing carcinomas were signet ring cell type. Even in the remaining 3 cases, most GM-CSF-positive cells were signet ring cells scattered within tubular adenocarcinoma. No correlation of ECF expression between either eosinophil/mast cell infiltration or peripheral blood eosinophilia was identified. In conclusion, most gastric carcinomas express the well-known ECFs and the expression of GM-CSF is specific for signet ring carcinoma cells.

Eosinophilic Liver Abscess in Patients with Gastric Carcinoma

Sixteen cases of heavy eosinophil infiltration or eosinophilic abscess of the liver in patients with gastric carcinoma were analyzed to draw attention to this interesting combination and to persue the pathogenetic mechanism. Peripheral blood eosinophilia and hepatic granuloma were found in only 5 and 4 cases, respectively. Neither the patients' stool nor the hepatic tissues disclosed any parasitic worms or eggs, although the skin tests for Clonorchis sinensis and Paragonimus westermani were positive in 2 cases. Among stomach carcinomas, early gastric cancer tended to have more eosinophils than advanced carcinoma, but was less frequently associated with the infiltration of mast cells. In the regional lymph nodes, there was no infiltration of eosinophils even in the presence of tumor metastasis. In the liver, none of the 16 cases had metastatic gastric carcinoma and mast cells were found in only 2 cases. The results suggest that heavy hepatic infiltration of eosinophils in gastric carcinoma patients is not of the parasitic or allergic cause, but of certain eosinophil chemotactic factor which may gain access to accumulate in the liver following released from the gastric carcinoma and transfered through the portal vein.