Repeated hypoglycemia exposure leads to impaired awareness of hypoglycemia (IAH) and the development of defective counterregulatory responses. To date, only islet transplantation has demonstrated normalization of endogenous glucose production (EGP) during insulin-induced hypoglycemia, in addition to improving hypoglycemia awareness. This study aims to validate clinical metrics of IAH (Clarke score), hypoglycemia severity (HYPO score), glycemic lability (lability index, LI), and CGM measures of hypoglycemia exposure and glucose variability as predictors of defective glucose counterregulation, defined by absent EGP, and hypoglycemia unawareness, defined by absent autonomic symptom (AS) recognition, measured in response to insulin-induced hypoglycemia. 

42 subjects with mean±SD age 43±13 years and T1D duration 29±12 years including 32 with IAH (Clarke unaware;19 pre-islet [Cohort 0], 13 pre-CGM [Cohort 1]) and 10 with intact hypoglycemia awareness (aware), and 12 non-diabetic control subjects, matched for age, sex and BMI (Cohort 3) underwent single-blinded randomized-paired hyperinsulinemic euglycemic and hypoglycemic clamp experiments using a stable glucose isotope tracer at the University of Pennsylvania between 2008-2021. Glycemic thresholds for EGP and AS responses were determined by the plasma glucose at which the response >95%CI of the euglycemic clamp control experiment. ROC and sensitivity analyses were performed to assess metric prediction of defective counterregulation.

HYPO and LI were greater in Clarke unaware (Cohorts 0 and 1) vs aware cohort (Cohort 3; p<0.0001 and p=0.03, respectively). During hypoglycemic clamp experiments, final step EGP response was greatest in controls (1.4±0.4 mg/kg/min) and greater in aware (0.9±0.4 mg/kg/min) than unaware cohorts (0.5±0.4 mg/kg/min, p<0.01) which remained greater than in euglycemia (0.2±0.6 mg/kg/min, p=0.004). AS responses were comparable in aware and control groups (7.9±2.9 vs 8.0±6.7) however significantly less in the unaware group (2.4±3.0, p<0.001) which were no different than during euglycemia (1.0±2.2).

Dichotomized Clarke score (</≥4) and Composite HYPO score/LI (</≥75th%) demonstrated good predictive value for determining glycemic threshold for EGP (Clarke: AUC=0.73, p=0.03; HYPO/LI: AUC=0.71, p=0.03) and AS (Clarke: AUC=0.71, p=0.04; HYPO/LI: AUC=0.71, p=0.03) responses.

CGM time spent in hypoglycemia showed good ability to predict AS by 240 minutes but not EGP response. Parallel assessment of Clarke and hypoglycemia metrics increased the sensitivity and specificity to predict absent AS response at 240 minutes. Defective EGP response at 240 minutes was best predicted by a composite of Clarke and variability / lability metrics however at a lower specificity.

These data suggest presence of IAH together with hypoglycemia exposure or increased glucose variability may identify individuals with defective counterregulation appropriate for early consideration of beta-cell replacement therapy.