Human islet graft function in NOD-SCID mice predicts clinical response in islet transplant recipients.

The purpose of this study was to evaluate the utility of nondiabetic immune-deficient NOD-SCID mouse model in assessing the functional capacity of isolated human islets. We transplanted 2000 islet equivalents obtained from six preparations used for human islet transplantation in three patients under the kidney capsule of groups of 10 mice. Human (Hu) C-peptide and insulin levels were determined following intraperitoneal (i.p.) glucose challenge at days 0, 7, 14, 21, 30, 60, 90, and 120. The Hu C-peptide level >1.5 ng/mL was the threshold for islet function in this model. The first patient did not achieve insulin independence and had minimal (0.5 ng/mL) fasting C-peptide levels that mirrored the low C-peptide levels observed in the mice. After the first infusion, the insulin requirements were reduced by 50% in the second patient. She became insulin free 10 days after her second infusion with a C-peptide level of 3.0 ng/mL, which corresponded to the peak C-peptide level (3.9 ng/mL) observed in the mice. By 150 days' posttransplant, the decline in C-peptide level paralleled the decline observed in mice. Within 2 weeks after the first transplant, insulin dose was reduced by 75% in the third patient, which corresponded to the robust C-peptide production in mice (7.3 ng/mL). Both patient and mice had a delay in islet function following the second infusion. She remained with a C-peptide level of 1.8 ng/mL and insulin free until suffering a rejection episode 3 months later. We observed that human islet graft function in NOD-SCID mice correlated with clinical response in islet transplant recipients.