RÉSUMÉ
Evaluating the quality of isolated human islets before transplantation is crucial for predicting the success in treating Type 1 diabetes. The current gold standard involves time-intensive in vivo transplantation into diabetic immunodeficient mice. Given the susceptibility of isolated islets to hypoxia, we hypothesized that hypoxia present in islets before transplantation could indicate compromised islet quality, potentially leading to unfavorable outcomes. To test this hypothesis, we analyzed expression of 39 hypoxia-related genes in human islets from 85 deceased donors. We correlated gene expression profiles with transplantation outcomes in 327 diabetic mice, each receiving 1200 islet equivalents grafted into the kidney capsule. Transplantation outcome was post-transplant glycemic control based on area under the curve of blood glucose over 4 weeks. In linear regression analysis, DDIT4 (R = 0.4971, P < 0.0001), SLC2A8 (R = 0.3531, P = 0.0009) and HK1 (R = 0.3444, P = 0.0012) had the highest correlation with transplantation outcome. A multiple regression model of 11 genes increased the correlation (R = 0.6117, P < 0.0001). We conclude that assessing pre-transplant hypoxia in human islets via gene expression analysis is a rapid, viable alternative to conventional in vivo assessments. This approach also underscores the importance of mitigating pre-transplant hypoxia in isolated islets to improve the success rate of islet transplantation.