OR03-5 Identification of Four New Glucotypes during the First Year after Type 1 Diabetes Onset Using Continuous Glucose Monitoring Metrics

Current markers of β-cell function poorly reflect glucose homeostasis after diabetes onset, mainly because of the lack of integration of parameters of insulin sensitivity and β-cell responsiveness to glucose. Recently, the widespread use of continuous glucose monitoring (CGM) helped stratify glucose control in patients with long-term diabetes in new clinically relevant glucotypes. In our DIATAG study, we investigated how CGM metrics may help to segregate patients with new-onset type 1 diabetes (T1D) entering or not in partial remission (PR). We collected data from 66 pediatric patients with T1D (10.4±4.7 yo) during the first year after diabetes onset (Δ). Clinical parameters (i.e. A1C, insulin daily dose, insulin-dose adjusted A1c [IDAA1C]) and CGM data were collected at Δ+3, +6, +9, +12 months (n=168, 57% remitters [IDAA1C <9]). A panel of 46 CGM metrics was calculated on an hourly and daily basis using iglu package (1). Using unsupervised hierarchical clustering based on daily CGM metrics and clinical parameters, we identified four clusters of glucose metrics that differed from each other (p<0.05). Cluster 1 was characterized by an increased glucose stability (coefficient of variation [CV] 32±5%) within time in target (TIT [63-140 mg/dL], 83±6%) with 9% of sensor values in 63-70 mg/dL range in the early morning period. Cluster 2 exhibited a progressive decrease of TIT (60±14%) and an increase of target above range (TAR [>180 mg/dL], 20±9%) especially during the day while target in range (TIR [70-180 mg/dL]) was nearly equivalent to Cluster 1 (80±8%). Notably, CV remained low (32±5%) with little time spent below range (TBR [<70 mg/dL], 5±2%). Cluster 3 demonstrated a net increase of CV during all the nycthemere (46±7%) concomitantly to an increase of TBR (14±7%). Interestingly, TIT and TAR remained close to values from Cluster 2 (55±11%; 23±10%). Finally, Cluster 4 demonstrated major hyperglycemias (TAR, 55±13%) with increased CV (47±9%) and decreased TBR (8±6%). Interestingly, clustering allowed clinically relevant segregation of intermediate values of IDAA1C (i.e. 7.5-10) between Cluster 2 and Cluster 3, independently of the remission status. Moreover, participants at Δ+3 months distributed across all clusters (with respectively 38% in Cluster 1, 26% in Cluster 2, 29% in Cluster 3 and 7% in Cluster 4) highlighting the usefulness of CGM in characterizing disease heterogeneity from diabetes onset. In our study, a combination of CGM metrics and clinical parameters unraveled key clinical milestones of glucose homeostasis and remission status during the first year of T1D. 1. Broll et al, 2021;16(4): e0248560. Presentation: Saturday, June 11, 2022 12:30 p.m. - 12:45 p.m.