Abstract
Patients with type 2 diabetes (T2D) have deficits in insulin production, insulin sensitivity, or both, resulting in persistently high blood glucose and long-term complications when left untreated. Treatment of T2D typically involves lifestyle changes, oral medications, and/or insulin therapy and glucose monitoring with self-monitoring of blood glucose (SMBG) via fingerstick testing. However, glycemic control remains suboptimal for many patients with T2D, even those who utilize insulin, and adherence to frequent SMBG testing and treatment regimens can be challenging. The DIAMOND randomized controlled trial demonstrated that real-time continuous glucose monitoring (rtCGM; Dexcom G4 with Software 505) was superior to SMBG for patients with T2D using multiple daily insulin injections (MDI)—resulting in significant reductions in hemoglobin A1c (HbA1c) compared to usual care with SMBG. 1 It was of interest to determine whether a subset of participants with T2D in the DIAMOND trial with endogenous insulin production at baseline (evidenced by C-peptide levels ≥0.5 ng/mL) also realized glycemic benefit with rtCGM use. We examined HbA1c and CGM metrics after 12 and 24 weeks of CGM use or normal care with SMBG for those with baseline, non-fasting C-peptide levels ≥0.5 ng/mL.
Methods for the T2D arm of the DIAMOND trial have been described previously. 1 In the current analysis, p values were calculated for changes in HbA1c (primary outcome) but not CGM metrics.
The majority of participants in the T2D arm of the DIAMOND trial still had residual insulin production at baseline—67/79 (84.8%) and 72/79 (91.1%) of participants randomized to rtCGM and SMBG, respectively, had C-peptide levels ≥0.5 ng/mL—confirming previous studies showing that most patients with T2D retain functional pancreatic beta cell reserves. 2 Demographics and baseline HbA1c levels were similar among participants with C-peptide levels ≥0.5 ng/mL who were randomized to rtCGM and SMBG with mean ± SD age of 59 ± 11 years and 60 ± 9 years and mean ± SD baseline HbA1c of 8.5 ± 0.6% and 8.5 ± 0.7%, respectively. Participants with C-peptide levels ≥0.5 ng/mL randomized to rtCGM demonstrated significantly better HbA1C at 12 and 24 weeks (P < 0.01 and P = .01, respectively) and significantly more of them achieved an HbA1c reduction of ≥0.5% (P < .01 at both 12 and 24 weeks; Table 1) than those randomized to SMBG. Glycemic improvement was evidenced by fewer sensor glucose values >180 mg/dL and more in the 70 to 180 mg/dL target range (Table 1). These data demonstrate that rtCGM use improves glycemic control for patients with T2D who still have residual beta cell production and are treated with exogenous insulin. This finding is not surprising, given the primary results of the DIAMOND trial 1 and the fact that the majority of the trial participants had C-peptide levels ≥0.5 ng/mL (shown here). Patients with T2D and C-peptide <0.5 ng/mL more closely resemble patients with type 1 diabetes (T1D); numerous studies, including the T1D arm of the DIAMOND trial, 3 have demonstrated the efficacy of rtCGM for patients with T1D.
HbA1c values imputed using site-measured HbA1c for 1 in CGM Group and 0 in Control Group at 12 weeks and 1 and 0, respectively at 24 weeks; and multiple imputation used for 2 and 4 additional missing data points, respectively at 24 weeks.
For secondary HbA1c outcomes at 24 weeks, N = 65 for the CGM Group and N = 68 for the Control Group.
All analyses were pre-specified except for the outcome for the reduction in HbA1c ≥0.5%, which was performed post-hoc.
P value for change in HbA1c is from a mixed effects linear model with baseline HbA1c as a fixed effect and clinical site as a random effect (results similar when also adjusting for race). For the binary outcomes, P values are from mixed effects logistic regression models with baseline HbA1c as a fixed effect and clinical site as a random effect using an adaptive quadrature estimation routine. The adjusted differences for the binary outcomes are calculated with a bootstrap.
SI conversion: to convert glucose from mg/dL to mmol/L, multiply the values by 0.0555.
CGM outcomes were not calculated for participants with <72 hours of data: 0 CGM/1 Control at 24 weeks.
For CGM outcomes at 12 weeks, N = 65 for the CGM Group and N = 67 for the Control Group. At 24 weeks, N = 62 for the CGM Group and N = 66 for the Control Group.
Footnotes
Author Contributions
All authors contributed significantly and have reviewed this submission: Sarah Puhr, John B. Welsh, and Tomas C. Walker prepared the letter and Colleen E. Bauza performed the analysis.
Abbreviations
CGM, continuous glucose monitoring; HbA1c, hemoglobin A1C; T1D, type 1 diabetes; T2D, type 2 diabetes; TIR, time in range.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Sarah Puhr, John B. Welsh, and Tomas C. Walker are employees of Dexcom, Inc.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The DIAMOND randomized controlled clinical trial, from which the data derive, is registered at clinicaltrials.gov (NCT02282397) and was supported by Dexcom, Inc. The protocol and Health Insurance Portability and Accountability Act–compliant informed consent forms were approved by one central and multiple local institutional review boards. Dexcom and Dexcom G4 are registered trademarks of Dexcom, Inc. in the United States and/or other countries.
