Long-term outcomes of pancreatic islet transplantation alone in type 1 diabetes: a 20-year single-centre study in Italy

Background: Islet transplantation has the potential to cure type 1 diabetes by restoring endogenous insulin production. However, its success relies on balancing improved glycaemic control with the risks of immunosuppressive therapy. This study aimed to evaluate long-term outcomes of islet transplantation alone for type 1 diabetes, focusing on the effects of islet mass and immunosuppressive regimens on graft survival and insulin independence, and weighing glycaemic control benefits against the risks of immunosuppressive therapy. Methods: This cohort study retrospectively analysed individuals aged 18–67 years with type 1 diabetes who received intraportal islet transplantation alone at IRCCS Ospedale San Raffaele, Milan, Italy. Inclusion criteria comprised adults with type 1 diabetes diagnosed before the age of 55 years with severe recurrent hypoglycaemia or glycaemic instability. Major exclusion criteria included a HbA1c of more than 12·5%, a BMI of more than 30 kg/m2, and insulin requirements exceeding 1·2 IU/kg per day, along with contraindications to immunosuppressive therapy. Participants were recruited from the hospital's islet transplant registry. Follow-up was conducted through regular clinical visits, with data collected retrospectively. Outcomes assessed included patient survival, graft survival, insulin independence, glycaemic control, and adverse events. Data were analysed using an intention-to-treat method, mixed-effects models, Kaplan–Meier estimates, and Cox and logistic regression to identify factors linked to metabolic success and reduced risks. Findings: 79 patients underwent intrahepatic or intraportal islet transplantation alone between Feb 16, 2001, and June 1, 2023, and received a total of 159 islet infusions, with a median total islet mass of 9637 islet equivalents (IEQ) per kg. Complications were infrequent and mostly involved minor bleeding, with only 3% (two of 79) of patients requiring surgical intervention. Glycaemic control improved significantly after infusion, with a reduction of HbA1c by –10·04 mmol/mol (–13·63 to –6·46), and a decrease in daily insulin requirements by –13·35 units per day (–17·04 to –9·65). The intention-to-treat analysis showed a median graft survival (fasting C peptide ≥0·3 ng/mL) of 3·9 years (95% CI 1·6 to 6·2) and 44% (35/79) insulin independence for a median of 6 years (95% CI 2·88 to 9·08). Patients receiving more than 10 000 IEQ/kg with BAS, FK506, and Rapa therapy had a median graft survival of 9·7 years (3·1–16·0) and 73% (16 of 22) insulin independence. Kaplan–Meier estimates indicated graft survival rates of 86% at 1 year, 65% at 5 years, 47% at 10 years, 47% at 15 years, and 40% at 20 years. Overall survival was 92% (73 of 79) over a median follow-up of 13·1 years, with a 20-year survival probability of 84%. Adverse events related to immunosuppressive therapy were reported in 44% (35 of 79) of patients, with allosensitisation rates increasing from 6% at baseline to 42% after therapy discontinuation. Interpretation: This analysis of a large islet transplantation alone cohort provides valuable insights into factors influencing outcomes and highlights potential risks, supporting informed clinical decision making and the optimisation of future β-cell replacement strategies. Funding: None.