RITA/Registry of Industrial Toxicology Animal Data: A comparative immunohistochemical study of 77 islet cell carcinomas in Sprague-Dawley and Wistar rats using antibodies against insulin, glucagon, somatostatin and gastrin

The objective of this study was to investigate spontaneous islet cell carci nomas with particular reference to possible existing strain differences bet ween Sprague Dawley (SD) and Wistar (W) rats in incidence and immunohistoch emical staining pattern. Secondly the occurrence of somatostatin and/or gas trin-positive islet cell tumors should be tested. Islet cell adenocarcinomas (34 from SD, 43 from W-rats) were selected from the RITA-data base [16] and company in-house data base out of an animal poo l of 3915 (1681 SD, 2234 W-rats). They were untreated or sham-treated (vehi cle) control animals from carcinogenicity studies and whole life-span exper iments. Islet cell carcinomas occurred in a higher incidence in male rats ( 2.98 % for SD, 3.23 % for W) than in female rats (1.07 % for SD, 0.63 % for W). All specimens were immunohistologically stained with antibodies agains t insulin, glucagon, somatostatin and gastrin and, selected specimens with additional antibodies (pancreatic polypeptide, lipase, chymotrypsin, S100-p rotein, actin and cytokeratin). 94 % (SD) and 93 % (W), respectively, were insulin-positive and the mean st aining intensity (on a scale ranging from 0 - 4) for insulin was 3.58 (SD) versus 3.37 (W). This high insulin staining incidence and intensity charact erized most islet cell carcinomas as malignant insulinomas. 24 % (SD) and 3 7 % (W), respectively, were glucagon-positive. Except two tumors in W-rats with a focal strong glucagon expression, the mean staining intensity for gl ucagon was low (0.38 SD, 0.72 W). 38 % (SD) and 44 % (W), respectively, wer e somatostatin-positive, but except for five cases having a focal to multif ocal, moderate to marked staining, only a few tumor cells were positive for somatostatin in the other cases and the mean staining intensity for somato statin was low (0.50 SD, 0.84 W). 6 % (SD) and 23 % (W), respectively, were gastrin-positive, but only one case of a male Wistar rat exhibited a focal strong staining in parts of the tumor. The other cases showed only a few t umor cells which were positive for gastrin. The mean staining intensity for gastrin was low (0.06 SD, 0.35 W). In all tumors with marked glucagon, som atostatin or gastrin expression, the immunostaining for insulin was still p redominating. Thus, insulin was the major hormone produced by most of the t umor cells. Five out of 77 tumors evaluated were immunohistologically negat ive with all applied antibodies. Conclusion: This study presents the first immunohistochemical survey on spo ntaneous islet cell carcinomas in SD and Wistar rats stained with antibodie s against the endocrine pancreas hormones insulin, glucagon, somatostatin a nd gastrin. No major differences in incidence or immunohistochemical staini ng pattern between SD and W-rats could be detected. In contrast to SD rats, Wistar rats had multihormonal coexpression in 16.3 %. The multihormonal ap pearance of the neoplasms is well comparable with the findings in other ani mal species and human insulinomas. Moreover, this is the first study in rat s which reports five cases with a marked co-expression of somatostatin and one case with marked focal co-expression of gastrin in malignant islet cell adenocarcinomas.