Introduction: The challenging process of obtaining pure islet fractions from donor pancreata hampers the widespread use of allogenic islet transplantation for patients with complicated diabetes mellitus. Limited evidence suggests that transplantation of islet products with a higher percentage of non-islet cells is associated with improved long-term metabolic outcomes. However, conclusive evidence is lacking due to indirect metabolic outcome measurements, the small number of islet recipients, and the lack of an objective and reproducible assessment of islet purity. We aimed to retrospectively evaluate the effect of islet purity on long-term graft function using robust measurement methods to determine islet purity and graft function.

Method: In a cohort of islet recipients that underwent an allogenic islet transplantation procedure at Leiden University Medical Center, digitalized microscopic images of the dithizone-stained transplanted islet graft were analyzed using a reproducible computerized deep learning method (IsletNet, version 2020-01-20) to calculate islet purity (expressed as %), islet size index (estimate of the average islet size, index <1 indicates an average islet size <150 μm) and the percentage of embedded islets. The cohort was divided into tertiles based on the purity: low purity (0-38%), intermediate purity (39-58%) and high purity (59-100%). Short- and long-term graft function was evaluated by calculating the area-under-the-curve (AUC) of C-peptide measurements that were obtained with mixed meal tests three months after transplantation, and subsequently yearly up to 5 years.

Results: Forty-one islet transplantation patients were included. Twenty-eight grafts consisted of islets derived from 1 donor pancreas and 13 grafts from 2 donor pancreata. The purity (mean±sd) in the low, intermediate and high group was 26±10, 46±6 and 67±7%, respectively. A higher islet purity was positively associated with islet size index (R²=0.47, p<0.0001) and negatively associated with the percentage of embedded islets (R²=0.56, p<0.0001). The C-peptide AUC (mean±sd) at three months after transplantation was 125.4±74.0 (low purity), 107.7±57.1 (intermediate purity), and 172.4±73.0 nmol/L (high purity; high vs low purity p=0.11). Also at 5 years there was no clear difference in C-peptide AUC (mean±sd): 55.5± 65.7 (low purity), 115.2±62.3 (intermediate purity) and 96.0±84.2 nmol/L (high purity).

Conclusion: The use of higher purity islet grafts for transplant is not associated with better long-term graft function. Further investigations are required to elucidate the mechanisms underlying the effect of non-islet cells on long-term islet function.