Bernard Sebastien (1), Zhaoling Meng (2), Tao Sheng (2), Dimple Patel (2), Sergej Ramusovic (3), Bruno Leroy (1), Jeff Barrett (2)
(1)Sanofi, Chilly-Mazarin, FRANCE (2)Sanofi , Bridgewater, NJ, USA, (3)Sanofi, Frankfurt, Germany
Objectives: The Edition 3 (NCT01676220)was a multicenter, randomized, open-label, two-arm parallel-group Phase 3 study comparing the safety of efficacy of glargine 300U/mL (Gla-300) versus glargine 100U/mL (Gla-100) once daily in insulin-naïve type 2 diabetes patients inadequately controlled on oral glucose-lowering drugs (1). A significant lower risk of hypoglycemia was found over the 6-month treatment period with Gla-300 versus Gla-100 while HbA1c levels decreased similarly with the two treatments. In an exploratory analysis, a modeling approach was applied to understand and compare the hypoglycemia risk kinetics between the two insulins when patients were titrating their insulin doses to achieve a protocol defined pre-breakfast plasma glucose level (f-SMPG).
Methods: Treatment emergent documented symptomatic hypoglycemia confirmed by blood glucose reading below 3 mmol/L (54 mg/dl)) was used in the analysis, where the threshold of 3 mmol/l was selected considering the glucose meters accuracy and ensuring the hypoglycemia alert value below 5.9 mmol/L (70 mg/dl). Time to first hypoglycemic event occurred within the 6 months of the study were modeled versus pre-breakfast self-monitored plasma glucose (f-SMPG) values as the time-varying covariate. F-SMPG was used since it was most frequently measured and almost daily during the dose titration period. It is more informative to track glycemic profile over time compared to other glucose measures in the study. A parametric time to event survival model with f-smpg at baseline, as time-varying and treatment (gla-300 and gla-100) as covariates was established with events assumed to follow a Weibull distribution. All hypoglycemia free patients were considered as censored at 6 months.
Results: There were 94 events overall in first 6 months in Edition 3 with 7.7% patients on gla-300 and 14.2% patients on gla-100 having at least 1 event. There was a statistically significant (p=0.001) trend for higher risk of hypoglycemic event associated with lower f-SMPG (2 treatments combined). A statistically significant (p= 0.002) treatment-by-time-varying f-SMPG interaction was also identified in the model indicating differential treatment effect of gla-300 compared to gla-100 during the insulin titration and f-SMPG reduction from the baseline phase. For gla-300, the global trend for decrease of risk over time is approximately compensated the hazard increase induced by the decrease in f-SMPG, leading to an estimated hypoglycemic hazard almost constant across time. For gla-100, the estimated hypoglycemic hazard is higher during the titration period, reduces and approaches plateaued at the later part of 6 month. Overall, the hypoglycemic risk is higher for patients on gla-100 than for patients on gla-300 during the titration phase. The beneficial effect of gla-300 is larger when time-varying fasting SMPG is higher (hazard decreased by 80% for f-SMPG=140 mg/dl as compared to hazard reduction of 36% for f-SMPG=100 mg/dl), which is during the titration phase of the study. A similar model was also obtained in Edition 2 (NCT01499095, 2), a similar Phase 3 study for prior insulin patients, for the first 6-month treatment period with a reduced magnitude in hypoglycemic risk benefit. The reduced hypoglycemic risk kinetics difference between gla-300 and gla-100 in Edition 2 could be due to the relatively lower f-SMPG in prior insulin patients at the beginning of the study.
Conclusions: The modeling approach, in particular the consideration of whole f-SMPG dynamics in the model, enabled to better characterize the kinetics of hypoglycaemic risk in patients receiving either of the two insulin glargine Gla-300 and Gla-100.
References:
[1] New insulin glargine 300 U/ml compared with glargine 100 U/ml in insulin-naïve people with type 2 diabetes on oral glucose-lowering drugs: a randomized controlled trial (EDITION 3). Bolli GB, Riddle MC, Bergenstal RM, Ziemen M, Sestakauskas K, Goyeau H, Home PD; on behalf of the EDITION 3 study investigators. Diabetes Obes Metab. 2015 Apr;17(4):386-94.
[2] New insulin glargine 300 units/mL versus glargine 100 units/mL in people with type 2 diabetes using oral agents and basal insulin: glucose control and hypoglycemia in a 6-month randomized controlled trial (EDITION 2).Yki-Järvinen H, Bergenstal R, Ziemen M, Wardecki M, Muehlen-Bartmer I, Boelle E, Riddle MC; EDITION 2 Study Investigators. Diabetes Care. 2014 Dec;37(12):3235-43.
Reference: PAGE 27 (2018) Abstr 8484 [www.page-meeting.org/?abstract=8484]
Poster: Drug/Disease Modelling - Safety