Because the etiology of insulin-dependent diabetes mellitus (IDDM) is thought to be autoimmune, several clinical trials have utilized immunosuppression to treat newly diagnosed diabetic patients. In the University of Miami trial, cyclosporine A (CyA) was used to treat one group (n = 10), while the other received placebo (n = 13). During the 1-year study, islet beta-cell function was better preserved in the CyA group compared to the placebo group, based on the response (C-peptide production) to a physiologic stimulus (meal challenge). Specifically, when measured by regression analysis, the slope defining the rate of decline of beta-cell function was significantly lower for the CyA-treated group (p < 0.05). Cytokine levels were analyzed retrospectively from frozen (-70 degrees C) stored sera from both groups. At time 0, tumor necrosis factor alpha (TNF alpha) levels were similar in the CyA (40.1 +/- 14.2 pg/mL) and placebo group (38.5 +/- 12.1 pg/mL) of IDDM subjects (normal 32.0 +/- 5.0 pg/mL). At 1 month, the level of TNF alpha in the CyA group was significantly lower than that observed in the placebo group (22.3 +/- 7.2 versus 53.3 +/- 8.9 pg/mL (P < .05). TNF alpha levels subsequently fell in the placebo group and were not significantly different between placebo and CyA groups. Soluble interleukin 2 receptor (IL-2R) levels in IDDM patients were significantly higher than in normal subjects at diagnosis of IDDM. For the next 6 months, these levels fell consistently in both the CyA and placebo groups.(ABSTRACT TRUNCATED AT 250 WORDS)