Pancreatic islet transplantation in type 1 diabetes: 20-year experience from a single-centre cohort in Canada.

BACKGROUND: Islet transplantation offers an effective treatment for selected people with type 1 diabetes and intractable hypoglycaemia. Long-term experience, however, remains limited. We report outcomes from a single-centre cohort up to 20 years after islet transplantation. METHODS: This cohort study included patients older than 18 years with type 1 diabetes undergoing allogeneic islet transplantation between March 11, 1999, and Oct 1, 2019, at the University of Alberta Hospital (Edmonton, AB, Canada). Patients who underwent islet-after-kidney transplantation and islet transplantation alone or islet transplantation before whole-pancreas transplantation (follow-up was censored at the time of whole-pancreas transplantation) were included. Patient survival, graft survival (fasting plasma C-peptide >0·1 nmol/L), insulin independence, glycaemic control, and adverse events are reported. To identify factors associated with prolonged graft survival, recipients with sustained graft survival (≥90% of patient follow-up duration) were compared with those who had non-sustained graft survival (<90% of follow-up duration). Multivariate binary logistic regression analyses were done to determine predictors of sustained graft survival. FINDINGS: Between March 11, 1999, and Oct 1, 2019, 255 patients underwent islet transplantation and were included in the analyses (149 [58%] were female and 218 [85%] were White). Over a median follow-up of 7·4 years (IQR 4·4-12·2), 230 (90%) patients survived. Median graft survival was 5·9 years (IQR 3·0-9·5), and graft failure occurred in 91 (36%) patients. 178 (70%) recipients had sustained graft survival, and 77 (30%) had non-sustained graft survival. At baseline, compared with patients with non-sustained graft survival, those with sustained graft survival had longer median type 1 diabetes duration (33·5 years [IQR 24·3-41·7] vs 26·2 years [17·0-35·5]; p=0·0003), median older age (49·4 years [43·5-56·1] vs 44·2 years [35·4-54·2]; p=0·0011), and lower median insulin requirements (0·53 units/kg per day [0·45-0·67] vs 0·59 units/kg per day [0·48-0·70]; p=0·032), but median HbA1c concentrations were similar (8·2% [7·5-9·0] vs 8·5% [7·8-9·2]; p=0·23). 201 (79%) recipients had insulin independence, with a Kaplan-Meier estimate of 61% (95% CI 54-67) at 1 year, 32% (25-39) at 5 years, 20% (14-27) at 10 years, 11% (6-18) at 15 years, and 8% (2-17) at 20 years. Patients with sustained graft survival had significantly higher rates of insulin independence (160 [90%] of 178 vs 41 [53%] of 77; p<0·0001) and sustained improvements in glycaemic control mixed-main-effects model group effect, p<0·0001) compared with those with non-sustained graft survival. Multivariate analyses identified the combined use of anakinra plus etanercept (adjusted odds ratio 7·5 [95% CI 2·7-21·0], p<0·0001) and the BETA-2 score of 15 or higher (4·1 [1·5-11·4], p=0·0066) as factors associated with sustained graft survival. In recipients with sustained graft survival, the incidence of procedural complications was lower (23 [5%] of 443 infusions vs 17 [10%] of 167 infusions; p=0·027), whereas the incidence of cancer was higher (29 of [16%] of 178 vs four [5%] of 77; p=0·015) than in those with non-sustained graft survival; most were skin cancers (22 [67%] of 33). End-stage renal disease and severe infections were similar between groups. INTERPRETATION: We present the largest single-centre cohort study of long-term outcomes following islet transplantation. Although some limitations with our study remain, such as the retrospective component, a relatively small sample size, and the absence of non-transplant controls, we found that the combined use of anakinra plus etanercept and the BETA-2 score were associated with improved outcomes, and therefore these factors could inform clinical practice. FUNDING: None.