Log-linear relationship between endogenous insulin secretion and glycemic variability in patients with type 2 diabetes on continuous glucose monitoring

The contribution of endogenous insulin secretion to glycemic variability (GV) may differ between patients with impaired insulin secretion and those with preserved secretion. Our objective was to determine the linearity of the relationship between fasting C-peptide (CPR) as a marker of endogenous insulin secretion and GV in type 2 diabetes (T2DM), regardless of the type of antidiabetic treatment. We conducted a prospective observational study using continuous glucose monitoring obtained from 284 Japanese outpatients with T2DM with various HbA1c values and antidiabetic treatment. We constructed a prediction curve of base-line CPR versus coefficient of variation (CV) and identified the clinical factors associated with CV using multiple regression analysis. Fasting CPR showed a significant negative log-linear relationship with CV (P < 0.0001), and the latter being strikingly high in the low-CPR group. The multiple regression analysis showed that low CPR was an independent predictor of high CV (P < 0.0001). The significant correlations were sustained in both patients with/without insulin treatment. The contribution of endogenous insulin secretion to GV depends on the extent of insulin secretion impairment. Fasting CPR may represent a useful indicator of GV instability in T2DM.

MON-640 Endogenous Insulin and C-Peptide Suppression Test Using a Rapid-Acting Insulin Analog in the Diagnosis of Insulinoma

C-peptide suppression test (CPS) was shown to diagnose the cause of hyperinsulinemic hypoglycemia, i.e. insulinoma, as effectively as supervised 72-hour fast test with less time consuming and cost. In the conventional CPS, regular insulin (RI) is used to induce hypoglycemia that subsequently suppresses endogenous insulin secretion. As RI is measurable in plasma insulin (PI) assay, plasma C-peptide (PCP) but not PI response is therefore used for assessment of endogenous insulin secretion in CPS using RI. As rapid acting insulin analogs (RA) are not measurable in a selected PI assay, both PCP and PI levels can be used to assess endogenous insulin secretion if an RA is used instead of RI in CPS. There is no study on PI and PCP responses to RA in insulinoma. This study aimed to examine efficacy of modified CPS, so-called insulin and C-peptide suppression test (ICPS) by using an RA (insulin aspart) in the diagnosis of insulinoma. Ten patients, 7 with histopathological proven insulinoma (IN) and 3 with non-insulinoma (non-IN), underwent ICPS. Blood samples were collected for measurement of plasma glucose (PG), PCP, and PI using a Roche Cobas 8000 module e602 which does not measure insulin aspart at before and every 10-20 minutes during intravenous infusion of insulin aspart at the rate of 0.05-0.075 unit/kg/hr. The test was terminated if the patients had a PG level of &lt;40 mg/dL. There was no difference in baseline biochemical data between two groups. At the end of the test, IN had significantly higher PCP level (range: 0.60-10.20 vs. 0.31-0.48 ng/mL and median: 4.67 vs. 0.47 ng/mL; p = 0.017), lower percentage of PCP suppression (range: 2.10-53.10% vs. 63.08-91.24% and mean: 24.78 ± 19.4% vs. 73.17 ± 15.7%; p = 0.005), higher PI level (range: 1.13-113.70 vs. 0.22-0.73 µIU/mL and median: 21.29 vs. 0.49 µIU/mL; p = 0.017) and lower percentage of PI suppression (range: -14.63-81.32% vs. 88.14-96.10% and mean: 28.45 ± 36.2% vs. 93.31 ± 4.5%; p = 0.003) than did the non-IN. No overlapping of these parameters was observed between IN and non-IN. Using the same cut-off levels as the supervised 72-hour fast test, the insulin criterion (≥3 µIU/mL) in ICPS had a sensitivity of 86% and a specificity of 100% and the C-peptide criterion (≥0.6 ng/mL) in ICPS had a sensitivity of 100% and a specificity of 100%. There was a high correlation of 89% between using PCP and PI responses. In conclusion, ICPS using an RA is effective in the diagnosis of insulinoma.