C-peptide level concomitant with hypoglycemia gives better performances than insulin for the diagnosis of endogenous hyperinsulinism: a single-center study of 159 fasting trials

INTRODUCTION: Diagnosis of endogenous hyperinsulinism relies on the occurrence of a hypoglycemia, concomitant with inadequate high insulin and C-peptide levels. However, diagnostic cutoffs are not consensual among the different learned societies. The objective of this work was to propose optimized cutoffs for these three parameters for the diagnosis of endogenous hyperinsulinism. METHODS: All the patients having performed a fasting trial in Cochin Hospital Endocrinology Department between February 2012 and August 2022 were included. The results of glycemia, insulin and C-peptide levels during fasting trial were collected and analyzed. RESULTS: One hundred and fifty-nine patients were included: 26 with endogenous hyperinsulinism and 133 without endogenous hyperinsulinism. ROC analysis of glycemia nadir during fasting trial identified the value of 2.3 mmol/L as the optimal cutoff, ensuring a sensitivity of 100% associated with a specificity of 81%. ROC analysis of insulin and C-peptide levels concomitant with hypoglycemia <2.3 mmol/L showed very good diagnostic performances of both parameters with respective cutoffs of 3.1 mUI/L (=21.5 pmol/L; sensitivity = 96%; specificity = 92%) and 0.30 nmol/L (sensitivity = 96%; specificity = 100%). Insulin to glycemia ratio as well as C-peptide to glycemia ratio (in pmol/mmol) at the time of glycemia nadir did not show better diagnostic performances than C-peptide alone. CONCLUSION: A C-peptide level 0.3 nmol/L concomitant with a hypoglycemia <2.3 mmol/L appears as the best criterion to make the diagnosis of endogenous hyperinsulinism. Insulin level can be underestimated on hemolyzed blood samples, frequently observed in fasting trial, and thus shows lower diagnostic performances.