Early-phase insulin secretion during mixed-meal tolerance testing predicts β-cell function and secretory capacity in cystic fibrosis

Abstract

Insulin secretion within 30 minutes of nutrient ingestion is reduced in people with cystic fibrosis (PwCF) and pancreatic insufficiency and declines with worsening glucose tolerance. The glucose potentiated arginine (GPA) test is validated for quantifying β-cell secretory capacity as an estimate of functional β-cell mass but requires technical expertise and is burdensome. This study sought to compare insulin secretion during mixed-meal tolerance testing (MMTT) to GPA-derived parameters in PwCF.MethodsSecondary data analysis of CF-focused prospective studies was performed in PwCF categorized as 1) pancreatic insufficient [PI-CF] or 2) pancreatic sufficient [PS-CF] and in 3) non-CF controls. MMTT: insulin secretory rates (ISR) were derived by parametric deconvolution using 2-compartment model of C-peptide kinetics, and incremental area under the curve (AUC) was calculated for 30, 60 and 180-minutes. GPA: acute insulin (AIR) and C-peptide responses (ACR) were calculated as average post-arginine insulin or C-peptide response minus pre-arginine insulin or C-peptide under fasting (AIRarg and ACRarg), ~230 mg/dL (AIRpot and ACRpot), and ~340 mg/dL (AIRmax and ACRmax) hyperglycemic clamp conditions. Relationships of MMTT to GPA parameters were derived using Pearson’s correlation coefficient. Predicted values were generated for MMTT ISR and compared to GPA parameters using Bland Altman analysis to assess degree of concordance.Results85 PwCF (45 female; 75 PI-CF and 10 PS-CF) median (range) age 23 (6-56) years with BMI 23 (13-34) kg/m2, HbA1c 5.5 (3.8-10.2)%, and FEV1%-predicted 88 (26-125) and 4 non-CF controls of similar age and BMI were included. ISR AUC30min positively correlated with AIRarg (r=0.55), AIRpot (r=0.62), and AIRmax (r=0.46) and with ACRarg (r=0.59), ACRpot (r=0.60), and ACRmax (r=0.51) (all P<0.001). ISR AUC30min strongly predicted AIRarg (concordance=0.86), AIRpot (concordance=0.89), and AIRmax (concordance=0.76) at lower mean GPA values, but underestimated AIRarg, AIRpot, and AIRmax at higher GPA-defined β-cell secretory capacity. Between test agreement was unaltered by adjustment for study group, OGTT glucose category, and BMI.ConclusionEarly-phase insulin secretion during MMTT can accurately predict GPA-derived measures of β-cell function and secretory capacity when functional β-cell mass is reduced. These data can inform future multicenter studies requiring reliable, standardized, and technically feasible testing mechanisms to quantify β-cell function and secretory capacity.