The immunosuppressive effects of irradiation are well known; however, under certain circumstances irradiation also augments the local immune response by as yet undefined mechanisms. Because of the importance of HLA class I antigen in immune regulation and the fact that killing of tumor cells by cytotoxic T cells is HLA antigen-restricted, the authors studied HLA class I antigen expression in eight glioblastomas multiforme, four meningiomas, and four medulloblastomas. Twenty fragments of each tumor specimen were placed in short-term cultures immediately after resection. For each tumor, control Sample 1 was not irradiated. Sample 2 was irradiated on Day 1, and two groups of the remaining pieces of each tumor (specimens 3 to 10) were irradiated on two consecutive days. Escalating radiation doses were given, starting at 200 cGy/day for Sample 2 up to 1000 cGy/day for Sample 10. The total dose range was 200 to 2000 cGy. Corresponding nonirradiated tumor fragments served as controls. Four hours after irradiation, each sample was processed and stained for HLA class I antigen using the immunoperoxidase technique. The tumor cells were intensely stained in nonirradiated glioblastomas and meningiomas, whereas no staining was observed in medulloblastomas. In four of the eight glioblastomas and in all four meningiomas, irradiation augmented HLA class I antigen expression compared to controls. This effect was dose-dependent and was maximum in the 1200 cGy-treated specimens. No change was observed in the other four glioblastomas or in the medulloblastomas. The data suggest that irradiation does not decrease and may even induce HLA class I antigen expression in some brain tumors. This may be one of the mechanisms by which immunotherapy operates after irradiation. Further studies are required to elucidate optimum radiation doses and fractionation as well as optimum timing of immunotherapy.

The expression of HLA class I and II antigens was studied in 41 patients with gastric cancer and in 2 normal stomachs. The normal gastric epithelia were diffusely stained for class I antigens. No staining was observed for class II. In gastric cancer an overall reduction in staining was observed, which was related to the degree of tumor penetration through the gastric wall. In tumors which penetrated the mucosa and submucosa only, 80–90% of tumor cells were stained for class I, as well as the normal areas surrounding the tumor. On the other hand, in tumors which penetrated through the organ to the fat, only 10–20% of cells were stained. Staining for class II was observed in only 12 cases, all of them advanced, with penetration to muscle or fat. No staining was observed in the normal areas surrounding the tumor. The importance of HLA antigens in immune surveillance is discussed.

No abstract