BACKGROUND Pancreatic transplantation results in denervation and loss of splanchnic venous drainage and inflicts numerous metabolic abnormalities. However, it is unclear whether denervation or loss of splanchnic venous drainage is responsible for the observed metabolic abnormalities. METHODS To discern denervation's role in these abnormalities, four mongrel dogs underwent extrinsic pancreatic denervation with preservation of splanchnic venous drainage. These animals, as well as four innervated control subjects, underwent standardized enteral and intravenous glucose tolerance testing. In addition, hyperglycemic clamps that maintained stable serum glucose elevations at either 2.8 or 8.3 mmol/L above basal were also performed. RESULTS Prestimulated glucose (90.4 +/- 2.7 vs 92.6 +/- 4.9 mg/dl) and insulin levels (6.8 +/- 1.7 vs 8.5 +/- 1.4 muU/ml) did not differ between innervated and denervated groups. Integrated incremental enteral glucose (5320 +/- 1900 vs 7790 +/- 2000 mg/dl) and insulin (2565 +/- 350 vs 2836 +/- 598 muU/ml) levels did not differ between groups. Integrated incremental intravenous glucose (3680 +/- 400 vs 3950 +/- 1000 mg/dl) and insulin (741 +/- 70 vs 1053 +/- 326 muU/ml) levels also did not differ. During glucose clamp studies, time-weighted 60 to 120-minute insulin levels (2.8 mmol/L, 30 +/- 5.0 vs 24 +/- 4.8 muU/ml; 8.3 mmol/L, 57 +/- 5.9 vs 50 +/- 9.8 muU/ml) did not differ between groups. In addition, glucose disposal, cyclic insulin release, and insulin sensitivity indexes were unchanged by denervation. CONCLUSIONS Extrinsic pancreatic neural elements are not necessary for cyclic insulin release in response to enteral or parenteral glucose challenge or physiologic and pharmacologic hyperglycemia. These findings suggest that the previously described posttransplantation glucose and insulin abnormalities are not attributable to denervation.