With or Without Residual C-Peptide,Patients with Type 2 Diabetes Realize Glycemic Benefits from Real-Time Continuous Glucose Monitoring

Abstract

Real-time continuous glucose monitoring (RT-CGM) is superior to blood glucose monitoring (BGM) for adults with insulin-treated type 2 diabetes (T2D); however, the utility of C-peptide levels for predicting the magnitude of the glycemic benefits is controversial. Data were from a subset of 147 participants in the MOBILE study who were treated with basal-only insulin and who had baseline C-peptide levels ≥0.5 ng/mL. Participants were randomized to treatment with either RT-CGM (n = 100) or BGM (n = 47). Between-group differences in hemoglobin A1c (HbA1c) and time in range (TIR) changes were assessed. The between-group difference in HbA1c favored the RT-CGM group (by 0.58 percentage points, P = 0.004 at 3 months and by 0.42 percentage points, P = 0.04 at 8 months). TIR was 16% higher, and time >180 mg/dL was 16% lower, in the RT-CGM group at 8 months (P = 0.002 for each). In T2D managed with basal insulin, RT-CGM benefits occur for those with residual insulin secretory capacity. Clinical Trial Identifier: NCT03566693

Introduction

Type 2 diabetes (T2D) is a progressive disease associated with insulin resistance and β-cell dysfunction. Accordingly, therapy evolves over time and people with T2D often require treatment intensification to maintain good glycemic control. The paradigm for T2D management commonly begins with lifestyle changes, progressing to oral medications, followed by insulin therapy. Some amount of blood glucose monitoring (BGM) through fingerstick testing is often utilized for therapy adjustments and for patient safety once insulin is introduced. Significant immediate and long-term complications can arise when T2D is left untreated. However, globally, the prognosis for T2D remains poor and glycemic control remains elusive for many patients.^1,2 Barriers to treatment and glycemic control include provider inertia,³ medical bias, limited medical coverage,⁴ and poor medication adherence.⁵

The DIAMOND randomized controlled trial (RCT) demonstrated that real-time continuous glucose monitoring (RT-CGM; Dexcom G4 with Software 505) among participants with T2D using multiple daily insulin injections (MDI) resulted in significant reductions in hemoglobin A1c (HbA1c) when compared with usual care with BGM,⁶ whether or not participants had residual insulin secretory capacity at baseline.⁷ The MOBILE RCT⁸ was built on this foundation and showed that HbA1c reduction and improvement in time in range (TIR; 70–180 mg/dL) was also observed in participants with T2D who were treated with basal (but not prandial) insulin and randomized to a factory-calibrated RT-CGM system (Dexcom G6) for 8 months.

Despite the emerging evidence, guidelines by professional societies and coverage policies by payers have been slow to include RT-CGM for patients with T2D; coverage is still lacking in some regions and health systems. Indeed, C-peptide (indicative of endogenous insulin production) is still sometimes used as a dispositive factor in determining coverage eligibility for advanced diabetes technologies, and the arbitrary guidelines may exclude coverage for individuals with residual C-peptide values.⁹ Accordingly, it was of interest to determine whether a subset of participants in the MOBILE trial with T2D and endogenous insulin production at baseline (evidenced by C-peptide levels ≥0.5 ng/mL) also realized glycemic benefits with RT-CGM use. The 0.5 ng/mL cutoff was selected to align with CMS guidelines restricting access to diabetes technology.⁹ We examined HbA1c and RT-CGM metrics after 3 and 8 months of RT-CGM use or normal care with BGM for those with baseline nonfasting C-peptide levels ≥0.5 ng/mL.

Methods

Methods for the MOBILE study have been described previously (ClinicalTrials.gov Identifier: NCT03566693).⁸ In brief, the 8-month study was conducted at 15 primary care practices across the United States to determine the effectiveness of RT-CGM in adults with T2D treated with basal insulin. Participants were randomized 2:1 to RT-CGM (n = 116) or traditional BGM (n = 59). Participants randomized to the BGM group wore a blinded RT-CGM sensor at the 3- and 8-month visits. The protocol and informed consent form were approved by a central institutional review board for fourteen centers and a local board for one center.

In this subanalysis, participant data were analyzed according to their randomization assignment. The primary analysis was a treatment group comparison of HbA1c at 3 and 8 months in a longitudinal mixed effects linear model with baseline HbA1c as a fixed effect and clinical site as a random effect. Binary HbA1c outcomes were evaluated in mixed effects logistic regression models with baseline HbA1c as a fixed effect and clinical site as a random effect. The adjusted differences for the binary outcomes were calculated as described by Kleinman and Norton¹⁰ and confidence intervals were calculated using a bootstrap. The models for all outcomes handled missing data using direct likelihood analysis, which maximizes the likelihood function integrated over all possible values of the missing data.

Analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC). All P-values were two-sided. Statistical significance was defined as P < 0.05 for the primary outcome tested at 3 and 8 months. For secondary outcomes, confidence intervals and P-values were adjusted to control the false discovery rate using the two-stage Benjamini–Hochberg procedure.

Results

Most participants with T2D in the MOBILE study still had residual insulin production at baseline—100/112 (89%) and 47/51 (92%) of participants randomized to RT-CGM or BGM, respectively, had C-peptide levels ≥0.5 ng/mL (four in the RT-CGM group and eight in the BGM group had missing C-peptide values at baseline). Demographics, baseline HbA1c, and baseline TIR were similar among participants with C-peptide levels ≥0.5 ng/mL who were randomized to RT-CGM or BGM with mean ± SD age of 56 ± 9 and 58 ± 10 years, mean ± SD baseline HbA1c of 9.2% ± 1.0% and 9.1% ± 0.9%, and mean ± SD baseline TIR of 39% ± 26% and 39% ± 26%, respectively.

Participants with C-peptide levels ≥0.5 ng/mL randomized to RT-CGM achieved significantly lower HbA1c levels at 3 months (adjusted difference −0.58%, 95% CI −0.96 to −0.19, P = 0.004) and 8 months (adjusted difference −0.42%, 95% CI −0.83 to −0.01, P = 0.04) compared with those randomized to BGM (Table 1). Between months 3 and 8, several RT-CGM outcomes (mean glucose, TIR, and time above range) improved for those in the RT-CGM group and worsened for those in the BGM group. In the BGM group, the A1C difference from month 3 to 8 (−0.06%, 95% CI −0.41 to 0.29) was not significantly different from zero, and was of unlikely clinical significance. At 8 months, participants randomized to RT-CGM had a 3.8 h per day increase in TIR (adjusted difference 16%, 95% CI 7 to 25, P = 0.002), a 3.8 h per day decrease in time above 180 mg/dL (adjusted difference −16%, 95% CI −25 to −6, P = 0.002), and a similar decrease in time above 250 mg/dL (adjusted difference −16%, 95% CI −22 to −10, P < 0.001) without an increase in time below 70 or 54 mg/dL compared with participants randomized to BGM.

Table 1.

Comparison of HbA1c and RT-CGM Outcomes at 3 and 8 Months in Real-Time Continuous Glucose Monitoring and Blood Glucose Monitoring Groups for Participants with C-Peptide ≥0.5 ng/mL

	Month 3			Month 8
	RT-CGM	BGM	Adjusted difference (95% CI) [P-value]	RT-CGM	BGM	Adjusted difference (95% CI) [P-value]
HbA1c outcomes	N = 94	N = 46		N = 90	N = 39
Primary outcome^a	Mean ± SD			Mean ± SD
HbA1c (%)	8.0 ± 1.2	8.5 ± 1.5		8.0 ± 1.4	8.3 ± 1.3
Change from baseline	−1.2 ± 1.2	−0.7 ± 1.2	−0.58% (−0.96 to −0.19) [0.004]	−1.1 ± 1.5	−0.7 ± 1.2	−0.42% (−0.83 to −0.01) [0.04]
Secondary outcomes^b	n (%)			n (%)
HbA1c <7.0%	17 (18%)	3 (7%)	12% (0 to 22) [0.04]	17 (19%)	4 (10%)	11% (−4 to 27) [0.12]
HbA1c <7.5%	37 (39%)	17 (37%)	5% (−11 to 18) [0.53]	34 (38%)	10 (26%)	15% (−10 to 37) [0.22]
HbA1c <8.0%	54 (57%)	21 (46%)	14% (3 to 25) [0.01]	57 (63%)	16 (41%)	24% (8 to 40) [0.003]
Relative reduction ≥10%	61 (66%)	19 (42%)	23% (11 to 34) [<0.001]	57 (64%)	18 (46%)	18% (5 to 31) [0.006]
Reduction ≥1%	59 (63%)	18 (40%)	23% (11 to 34) [<0.001]	48 (54%)	17 (44%)	12% (−8 to 31) [0.24]
Reduction ≥1% or HbA1c <7.0%	60 (65%)	19 (42%)	22% (9 to 34) [0.001]	49 (55%)	17 (44%)	12% (−7 to 31) [0.20]
Reduction ≥0.5%	72 (77%)	32 (71%)	7% (−4 to 21) [0.24]	65 (73%)	26 (67%)	8% (−3 to 19) [0.18]
RT-CGM outcomes^c	N = 94	N = 42		N = 79	N = 41
	Mean ± SD			Mean ± SD
Hours of RT-CGM data	228 ± 50(9.5 ± 2.1 days)	211 ± 40(8.8 ± 1.7 days)		211 ± 65(8.8 ± 2.7 days)	215 ± 47(8.9 ± 1.9 days)
Mean glucose (mg/dL)	184 ± 47	191 ± 45	−11 (−25 to 4) [0.15]	179 ± 42	206 ± 55	−27 (−44 to −10) [0.003]
Time in range 70–180 mg/dL	56% ± 27%	50% ± 24%	8% (0 to 16) [0.06]	59% ± 25%	42% ± 27%	16% (7 to 25) [0.002]
Time >180 mg/dL	43% ± 27%	49% ± 25%	−7% (−16 to 1) [0.09]	41% ± 25%	57% ± 28%	−16% (−25 to −6) [0.002]
Time >250 mg/dL^d	14% ± 16%	18% ± 17%	−6% (−12 to −1) [0.02]	10% ± 11%	27% ± 24%	−16% (−22 to −10) [<0.001]
Time <70 mg/dL^d	0.18% ± 0.27%	0.74% ± 1.21%	−0.50% (−0.75 to −0.25) [<0.001]	0.22% ± 0.39%	0.48% ± 0.86%	−0.20% (−0.42 to 0.01) [0.06]
Time <54 mg/dL^d	0.02% ± 0.05%	0.17% ± 0.31%	−0.11% (−0.18 to −0.05) [<0.001]	0.02% ± 0.04%	0.05% ± 0.10%	−0.02% (−0.04 to 0.01) [0.13]

For HbA1c, point estimates, confidence intervals, and P-values were obtained using a mixed-effects linear regression model adjusting for a random site effect. Local HbA1c was included as an auxiliary variable in the model.

For secondary HbA1c outcomes, point estimates, confidence intervals, and P-values were obtained using a mixed-effects logistic regression model adjusting for baseline glycated hemoglobin and a random site effect. Confidence intervals and nominal (uncorrected) P-values were adjusted for multiple comparisons using the adaptive TST GBH method. All outcomes were prespecified except for HbA1c <8.0%, which was performed post hoc.

For RT-CGM outcomes, point estimates, confidence intervals, and P-values were obtained using a mixed-effects linear regression model adjusting for a random site effect. Confidence intervals and nominal (uncorrected) P-values were adjusted for multiple comparisons using the adaptive TST GBH method.

Winsorized at the 10th and 90th percentiles before reporting summary statistics.

BGM, blood glucose monitoring; HbA1c, hemoglobin A1c; RT-CGM, real-time continuous glucose monitoring; TST GBH, two-stage group Benjamini–Hochberg.

Conclusions

These data demonstrate that RT-CGM use improves glycemic control for participants with T2D who still have residual β-cell production and who are treated with exogenous basal (but not prandial) insulin. The vast majority of participants had residual insulin production, which is in agreement with a previous study assessing C-peptide concentrations in patients with insulin-managed T2D.¹¹ Results were consistent with a similar subanalysis of the DIAMOND trial⁷ and adds to the compendium of evidence showing that RT-CGM is beneficial for patients with insulin-treated type 1 diabetes or T2D, independent of a wide variety of baseline characteristics such as diabetes type, age, baseline HbA1c, education, diabetes numeracy, hypoglycemia awareness status, and residual insulin secretory capacity. Recent commentaries have discussed the need to modify eligibility criteria in coverage policies to include all individuals who could benefit from advanced diabetes technologies,^12
–14 and these findings should further challenge the arbitrary barriers preventing access.

Footnotes

Authors' Contributions

All authors contributed to the preparation of this article.

Acknowledgments

The authors thank Drs. David Price, Sarah Puhr, and John Welsh of Dexcom for comments on early drafts of this article.

Author Disclosure Statement

R.B. and P.C. are employees at the Jaeb Center for Health Research and received grant funding from Dexcom to their institution. T.C.W. and C.C. are employees of Dexcom.

Funding Information

Study funding and study devices were provided by Dexcom, Inc.

References

Aschner

, Gagliardino

, Ilkova

, et al.: Persistent poor glycaemic control in individuals with type 2 diabetes in developing countries: 12 years of real-world evidence of the International Diabetes Management Practices Study (IDMPS). Diabetologia, 2020; 63:711–721.

Khunti

, Chen

, Cid-Ruzafa

, et al.: Glycaemic control in patients with type 2 diabetes initiating second-line therapy: Results from the global DISCOVER study programme. Diabetes Obes Metab, 2020; 22:66–78.

Reach

, Pechtner

, Gentilella

, et al.: Clinical inertia and its impact on treatment intensification in people with type 2 diabetes mellitus. Diabetes Metab, 2017; 43:501–511.

Kruger

, Anderson

: Continuous glucose monitoring (CGM) is a tool, not a reward: unjustified insurance coverage criteria limit access to CGM. Diabetes Technol Ther, 2021; 23(S3):S45–S55.

Edelman

, Polonsky

: Type 2 diabetes in the real world: the elusive nature of glycemic control. Diabetes Care, 2017; 40:1425–1432.

Beck

, Riddlesworth

, Ruedy

, et al.: continuous glucose monitoring versus usual care in patients with type 2 diabetes receiving multiple daily insulin injections: a randomized trial. Ann Intern Med, 2017; 167:365–374.

Puhr

, Welsh

, Bauza

, Walker

: Patients with type 2 diabetes and residual insulin secretory capacity realize glycemic benefits from real-time continuous glucose monitoring. J Diabetes Sci Technol, 2021; 15:965–967.

Martens

, Beck

, Bailey

, et al.: Effect of continuous glucose monitoring on glycemic control in patients with type 2 diabetes treated with basal insulin: a randomized clinical trial. JAMA, 2021; 325:2262–2272.

Centers for Medicare and Medicaid Services: Insulin pump: C-peptide levels as a criterion for use. https://www.cms.gov/medicare-coverage-database/view/ncacal-decision-memo.aspx?proposed=N&NCAId=109 (accessed October 22, 2021 ).

10.

Kleinman

, Norton

: What's the risk? A simple approach for estimating adjusted risk measures from nonlinear models including logistic regression. Health Serv Res, 2009; 44:288–302.

11.

Uzunlulu

, Oguz

, Arslan Bahadir

, et al.: C-peptide concentrations in patients with type 2 diabetes treated with insulin. Diabetes Metab Syndr, 2019; 13:3099–3104.

12.

Aleppo

, Parkin

, Carlson

, et al.: Lost in translation: a disconnect between the science and Medicare coverage criteria for continuous subcutaneous insulin infusion. Diabetes Technol Ther, 2021; 23:715–725.

13.

Galindo

, Parkin

, Aleppo

, et al.: What's wrong with this picture? A critical review of current centers for Medicare & Medicaid Services coverage criteria for continuous glucose monitoring. Diabetes Technol Ther, 2021; 23:652–660.

14.

Anderson

, Gavin

, Kruger

: Current eligibility requirements for CGM coverage are harmful, costly, and unjustified. Diabetes Technol Ther, 2020; 22:169–173.