Imaging Predictors of Pancreatic Health and Islet Yield in Pediatric Total Pancreatectomy with Islet Autotransplantation

INTRODUCTION: Total pancreatectomy and islet autotransplantation (TPIAT) can alleviate symptoms in chronic pancreatitis. We aimed to identify pre-TPIAT imaging markers predicting explanted pancreas health and islet yield. METHODS: We retrospectively analyzed 104 pediatric TPIAT patients, excluding those with presurgical diabetes or pancreatic surgeries. Pancreas parenchymal volume was manually segmented, and T1 signal intensity ratio pancreas to spleen (T1 SIRp/s) was calculated. An islet biologist assessed fat infiltration, fibrosis, and islet yield. Logarithmic transformation and regression analyses were used for variance stabilization and predictive modeling. RESULTS: Ninety-four patients (60% female, median age 12.5 years) were included. Univariate analyses revealed that an increase in pancreas volume was associated with less fibrosis (odds ratio [OR] = 0.88 per 5 mL, 95% CI: 0.77-0.99, P < 0.05), higher pellet volume, total islet equivalent and total islet count. For advanced fibrosis, an increase in T1 SIRp/s was linked to decreased odds (OR = 0.74 per 0.1 unit, 95% CI: 0.59-0.92, P < 0.05), whereas a higher Cambridge score was associated with increased odds (OR = 1.34 per 1 unit of Cambridge, 95% CI: 1.01-1.77, P < 0.05). A model incorporating segmented pancreas volume and T1 SIRp/s predicted advanced fibrosis with an area under receiver operating curve (AUC) of 0.75 (95% CI: 0.64-0.87). In addition, models that included larger pancreas volume and the absence of acute pancreatitis predicted total islet count and total islet equivalent. DISCUSSION: In children with chronic pancreatitis, noninvasive cross-sectional imaging focused on the parenchyma can guide the management, as a smaller parenchymal bulk and lower T1 SIRp/s predict advanced fibrosis, whereas larger pancreas volume and T1 SIRp/s predict larger pellet volumes.