Abstract

Islet yield is an important predictor of acceptable glucose control after total pancreatectomy with islet autotransplantation (TP-IAT). We assessed if pancreas volume calculated with preoperative MRI could assess islet yield and postoperative outcomes. We reviewed dynamic MRI studies from 154 adult TP-IAT patients (2009-2016), and associations between calculated volumes and digest islet equivalents (IEQs) were tested. In multivariate regression analysis, pancreas volume (P < .001) and preoperative HbA1c levels (P = .009) were independently associated with digest IEQs. The IEQ prediction formula was calculated according to each preoperative HbA1c level, (a) pancreas volume × 5800 for HbA1c = 6.5, (b) pancreas volume × 10 000 for HbA1c =5.7/<6.5 and (iii) pancreas volume × 11 400 for HbA1c < 5.7. The formula was internally validated with 28 TP-IAT patients between 2017 and 2018 (r2  = .657 and r2  = .710 when restricted to 24 patients without prior pancreatectomy). An estimated IEQs/Body Weight (kg) =3700 predicted HbA1c =6.5 and insulin independence at 1 year after TP-IAT with 77% and 88% sensitivity and 55% and 43% specificity, respectively. The combination of pancreas volume and preoperative HbA1c levels may be useful to estimate islet yield. Estimated IEQs were reasonably sensitive to predict acceptable glucose control at 1 year.

View details for DOI 10.1111/ctr.14008

View details for Web of Science ID 000545125100001

View details for PubMedID 32530540