Impact of aging on the in vivo potency of human islet preparation

R. Misawa¹, R.D. Molano², C. Fotino², E. Zahr-Akrawi², A. Miki², J. Molina², S. Barker², S. Villate², A. Khan², P. Witkowski³, L. Inverardi², H. Ichii², C. Ricordi², A. Pileggi^2
1 University of Chicago, University of Miami, Chicago, Miami, USA; ² University of Miami, Miami, USA; ³ University of Chicago, Chicago, USA

Objective: Islet transplantation into immunodeficient mice allows assessing cell viability and function, in the absence of allo-/auto-immunity. We sought to identify potential donor and isolation variables influencing in vivo islet functionality in the mouse bioassay.

Methods: For 175 human islet preparations, grafts of 2,000IEQ islets were implanted in the renal subcaspular space ofstreptozotocin-induced athymic nu/nu (nude) mice. Preparations were retrospectively divided into two groups: achieving or not diabetes reversal after transplant and univariate analysis was done for donor [age, gender, BMI, and cold ischemic time (CIT)] and isolation variables [yield (IEQ) and Islet Particle Number (IN), Average Particle Size (APS; IEQ/IN ratio) and digestion time]. Binary logistic regression identified significant predictors and carrying out partition analysis to optimize outcomes in mice. Conventional in vitro potency assays (viability by FDA/PI) and insulin stimulation index (SI) in glucose challenge were compared to the bioassay.

Results: From 175 mice transplanted with human islets, 137 achieved euglycemia. Univariate comparisons revealed a significant difference in donor age (reversal vs. non reversal: 38.92±1.12 vs.47.42±1.62, P<0.001), whereas no significant differences were observed in gender (M/F:85/43 vs. 29/18), BMI (28.64±0.46vs. 28.56±0.77) and CIT (644.2±22.0vs. 715.9±36.3). APS significantly differed (1.43±0.07 vs. 1.70±0.11, P<0.05). No significant differences were found in yield (283,476±13,930 vs. 264,217±23,792IEQ), IN (221,229±12,174 vs. 182,512±20,091) and Digestion Time (14.90±0.49 vs. 16.01±0.81). Binary logistic regression analysis identifiedsignificant differences in age (P<0.01; odds ratio=20.68) and CIT(P<0.05; OR=5.18) correlating with in vivo function. Partition analysis for bioassay and donorage demonstrated correlation with potency with diabetes reversal rates of 100% (age<23 years), 76% (23≦age<56), and 25% (age

Conclusions: Donor age may influence the outcome in vivo human islet function.