Early feasibility study with an implantable near-infrared spectroscopy sensor for glucose, ketones, lactate and ethanol.

OBJECTIVE: To evaluate the safety and performance of an implantable near-infrared (NIR) spectroscopy sensor for multi-metabolite monitoring of glucose, ketones, lactate, and ethanol. RESEARCH DESIGN AND METHODS: This is an early feasibility study (GLOW, NCT04782934) including 7 participants (4 with type 1 diabetes (T1D), 3 healthy volunteers) in whom the YANG NIR spectroscopy sensor (Indigo) was implanted for 28 days. Metabolic challenges were used to vary glucose levels (40-400 mg/dL, 2.2-22.2 mmol/L) and/or induce increases in ketones (ketone drink, up to 3.5 mM), lactate (exercise bike, up to 13 mM) and ethanol (4-8 alcoholic beverages, 40-80g). NIR spectra for glucose, ketones, lactate, and ethanol levels analyzed with partial least squares regression were compared with blood values for glucose (Biosen EKF), ketones and lactate (GlucoMen LX Plus), and breath ethanol levels (ACE II Breathalyzer). The effect of potential confounders on glucose measurements (paracetamol, aspartame, acetylsalicylic acid, ibuprofen, sorbitol, caffeine, fructose, vitamin C) was investigated in T1D participants. RESULTS: The implanted YANG sensor was safe and well tolerated and did not cause any infectious or wound healing complications. Six out 7 sensors remained fully operational over the entire study period. Glucose measurements were sufficiently accurate (overall mean absolute (relative) difference MARD of 7.4%, MAD 8.8 mg/dl) without significant impact of confounders. MAD values were 0.12 mM for ketones, 0.16 mM for lactate, and 0.18 mM for ethanol. CONCLUSIONS: The first implantable multi-biomarker sensor was shown to be well tolerated and produce accurate measurements of glucose, ketones, lactate, and ethanol. TRIAL REGISTRATION: Clinical trial identifier: NCT04782934.

The 13C glucose breath test accurately identifies insulin resistance in people with type 1 diabetes.

OBJECTIVE: This study investigated whether the delta-over-baseline of exhaled 13CO2 (Δ13CO2), generated from a 13C glucose breath test (13C-GBT), measured insulin resistance (IR) in people with type 1 diabetes, using the hyperinsulinemic-euglycemic clamp (HEC) as reference method. Secondary objective was to compare the 13C-GBT with the estimated glucose disposal rate (eGDR). METHODS: A 40 mU/m2/min HEC and two separate 13C-GBTs (euglycemic with insulin bolus and hyperglycemic without bolus) were consecutively performed in 44 adults with type 1 diabetes with varying body compositions. eGDR was calculated based on HbA1c, presence of hypertension and waist circumference. RESULTS: Mean M-value was 5.9 ± 3.1 mg/kg/min, mean euglycemic Δ13CO2 was 6.4 ± 2.1 δ‰, while median eGDR was 5.9 [4.3 - 9.8] mg/kg/min. The hyperglycemic Δ13CO2 did not correlate with the M-value, while the euglycemic Δ13CO2 and the M-value correlated strongly (r = 0.74, p < 0.001). Correlation between M-value and eGDR was more moderate (Spearman's rho = 0.63, p < 0.001). Linear regression showed an association between Δ13CO2 and M-value, adjusted for age, sex and HbA1c (adjusted R² = 0.52, B = 1.16, 95% CI 0.80 - 1.52, p < 0.001). The AUROC for Δ13CO2 to identify subjects with IR (M-value < 4.9 mg/kg/min) was 0.81 (95% CI 0.68 - 0.94, p < 0.001). The optimal cut-off for Δ13CO2 to identify subjects with IR was ≤ 5.8 δ‰. CONCLUSIONS: Under euglycemic conditions, the 13C-GBT accurately identified individuals with type 1 diabetes and concurrent IR, suggesting its potential as a valuable non-invasive index.

Trends of glucose, lactate and ketones during anaerobic and aerobic exercise in subjects with type 1 diabetes: The ACTION-1 study.

BACKGROUND: Exercise is part of type 1 diabetes (T1D) management due to its cardiovascular and metabolic benefits. However, despite using continuous glucose monitoring, many patients are reluctant to exercise because of fear for hypoglycaemia. AIMS: We assessed trends in glucose, lactate and ketones during anaerobic and aerobic exercise in people with T1D and compared incremental area under the curve (AUC) between both exercises. METHODS: Twenty-one men with T1D (median [IQR]: age 29 years [28-38], body mass index (BMI) 24.4 kg/m2 [22.3-24.9], HbA1c 7.2% [6.7-7.8]), completed a cardiopulmonary exercise test (CPET) and a 60-min aerobic exercise (AEX) at 60% VO2 peak on an ergometer bicycle within a 6-week period. Subjects consumed a standardised breakfast (6 kcal/kg, 20.2 g CHO/100 ml) before exercise without pre-meal insulin and basal insulin for pump users. RESULTS: During CPET, glucose levels increased, peaking at 331 mg/dl [257-392] 1-3 h after exercise and reaching a nadir 6 h after exercise at 176 mg/dl [118-217]. Lactate levels peaked at 6.0 mmol/L [5.0-6.6] (max 13.5 mmol/L). During AEX, glucose levels also increased, peaking at 305 mg/dl [245-336] 80 min after exercise and reaching a nadir 6 h after exercise at 211 mg/dl [116-222]. Lactate levels rose quickly to a median of 4.3 mmol/L [2.7-6.7] after 10 min. Ketone levels were low during both tests (median ≤ 0.2 mmol/L). Lactate, but not glucose or ketone AUC, was significantly higher in CPET compared to AEX (p = 0.04). CONCLUSIONS: Omitting pre-meal insulin and also basal insulin in pump users, did prevent hypoglycaemia but induced hyperglycaemia due to a too high carbohydrate ingestion. No ketosis was recorded during or after the exercises. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov: NCT05097339.