Titration of Basal and Prandial Insulin Doses With the Initiation of Glucagon-Like Peptide-1 Receptor Agonist Therapy

Objective Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) have demonstrated significant efficacy in improving glycemic control in type 2 diabetes mellitus, which often results in decreased insulin dose requirements. The purpose of this study was to examine the changes in basal and prandial insulin dose requirements from baseline to three months following initiation of a GLP-1 RA. Methodology A retrospective chart review was conducted of adult insulin-treated patients at the Chertow Diabetes Center, Huntington, WV, who were started on GLP-1 RAs for 24 months. Results Mean daily basal insulin doses decreased by 8.7 units (P = 0.29; mean 8.3% change) and mean daily prandial insulin doses decreased by 9.4 units (P = 0.10; mean 18.4% change) from baseline to three months after starting a GLP-1 RA. Average hemoglobin A1c significantly decreased from 8.8% (73 mmol/mol) at baseline to 8.0% (64 mmol/mol) at three months (P < 0.001). Significant decreases from baseline to three months were also observed in mean body weight, mean low-density lipoprotein (LDL) cholesterol, and mean total cholesterol. Conclusions GLP-1 RA therapy was associated with a significant decrease in hemoglobin A1c, body weight, and LDL-cholesterol from baseline to three months after initiation. Therapy with GLP-1 RAs was also associated with an overall decrease in daily basal and prandial insulin dose requirements, although this finding did not reach statistical significance.


Introduction
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) considerably improve glycemic control by improving insulin secretion in a glucose-dependent way and decreasing glucagon secretion, causing an average hemoglobin A1c reduction of approximately 1.2% [1].These agents have also demonstrated several other benefits in addition to decreasing blood glucose levels.Through delaying gastric emptying and consequently decreasing appetite, GLP-1 RAs can promote weight loss [1][2].Studies have also demonstrated positive effects on blood pressure and cholesterol levels.A meta-analysis of 60 trials showed that, on average, GLP-1 RAs were associated with a positive effect on blood pressure [3].Another meta-analysis of 35 trials showed that GLP-1 RAs were associated with significant reductions in low-density lipoprotein (LDL, about 3.1 mg/dL) and total cholesterol (about 5 mg/dL) [4].Another benefit of GLP-1 RAs is a lower risk of hypoglycemia when used as monotherapy compared with many other antidiabetic agents, such as sulfonylureas [5].
Combining a GLP-1 RA and insulin is an appealing treatment strategy for type 2 diabetes because of its potential for achieving substantial improvements in glycemic control with a relatively low risk of hypoglycemia or weight gain.A meta-analysis that examined randomized controlled trials found that combination treatment with GLP-1 RAs and basal insulin led to a mean greater reduction in A1c compared with any other anti-diabetic treatment strategy (P = 0.0047) [3].Additionally, there was no difference found in the relative risk of hypoglycemia with GLP-1 RA and basal insulin combination treatment compared with other treatments.
Because of their potent blood glucose-lowering efficacy, the effect of GLP-1 RAs on insulin requirements in insulin-treated patients can be extremely variable and represent a challenge, especially in patients who are receiving high daily doses of insulin.A randomized controlled trial compared exenatide 10 mcg twice daily to placebo in type 2 diabetic patients for 30 weeks, with both interventions given concomitantly with basal insulin [6].At randomization, patients with a hemoglobin A1c greater than 8% (64 mmol/mol) continued to receive their current insulin doses, while patients with an A1c less than 8% (64 mmol/mol) decreased their insulin doses by 20%.At 30 weeks, the reduction of A1c was greater in the exenatide group than placebo (P < 0.001).
The prescribing information for all currently commercially available GLP-1 RAs states that patients may need a decrease in insulin doses upon initiation of a GLP-1 RA to avoid hypoglycemia [7][8][9][10][11][12][13].However, there are insufficient data on appropriate methods for titrating and decreasing insulin doses upon GLP-1 RA initiation [1].The purpose of this study was to examine the changes in basal and prandial insulin requirements within the first three months following initiation of GLP-1 RAs.

Study design
A retrospective chart review using patients' electronic medical records was conducted on adult insulintreated patients who were started on GLP-1 RAs over 24 months at the Chertow Diabetes Center, Huntington, WV.Patient weight, hemoglobin A1c, blood pressure, serum lipids, and basal and prandial insulin daily doses were all collected at baseline before initiation of GLP-1 RA and after three months of therapy.Other baseline characteristics collected before initiation of GLP-1 RA included glomerular filtration rate (GFR) and C-peptide level.Changes in basal and prandial insulin dose requirements, hemoglobin A1c, weight, blood pressure, and serum lipids were calculated by comparing the data from the visit before starting the GLP-1 RA to the following visit (3-12 months from baseline data).
Patients were included if they had a diagnosis of type 1 or type 2 diabetes and were over the age of 18 years.Patients were excluded if they stopped the GLP-1 RA after less than four weeks from initiation.This study was approved by the institutional review board of Marshall University.
When patients were started on GLP-1 RAs, the clinical pharmacist was available for consultation, which could include counseling on the new medication and ensuring insurance coverage and affordability of the GLP-1 RA.Consultation services could also include diabetes therapy management following GLP-1 RA initiation, which involved the clinical pharmacist following up with patients regularly to monitor their blood glucose readings and then consult with the patient's provider to make any appropriate therapy changes.Referral to clinical pharmacist consultation was based on provider discretion; for example, providers would often refer patients if they felt they needed closer monitoring of response and need for insulin dose adjustments after starting GLP-1 RA therapy.

Outcome measures
The primary outcome of this study was the change in insulin dose requirements from baseline to three months after GLP-1 RA initiation.Secondary outcomes were the changes in hemoglobin A1c, weight, blood pressure, and serum lipids from baseline to three months after GLP-1 RA initiation.

Statistical analyses
Statistical analysis was performed using SPSS version 18 (SPSS Inc., Chicago), a statistical software, frequently employed in biomedical sciences [14].We also used the R base package for data cleaning, cachenation, and tabulation [15].Mediation analysis is a tool for statistical inferences, widely applied recently in psychological, social sciences [16].SPSS has a macro named PROCESS that we employed for the mediation analysis [17].It is used for estimation of indirect, direct, and complete effects with categorical independent variables [18].We used mediation analysis in this study to examine these effects on shortacting insulin doses following initiation of GLP-1 RA.A paired t-test was used to compare daily basal and prandial insulin doses, hemoglobin A1c, weight, blood pressure, and serum lipids before and three months following GLP-1 RA initiation.Statistical significance was set at a P-value less than or equal to 0.05.

Study population
Initially, 135 patients were screened for eligibility, with 83 included in the final analysis.A total of 52 patients were not included, with most not included because they were not receiving insulin therapy.Population baseline characteristics are listed in Table

Outcome measures
Insulin doses before and three months after starting GLP-1 RA therapy are detailed in Table 3. Mean daily basal and prandial insulin doses both decreased from baseline to three months following GLP-1 RA initiation, by 8.7 units (P = 0.29; mean 8.3% change) and by 9.4 units (P = 0.10; mean 18.4% change), respectively.Results of the mediation analysis to examine the total, direct, and indirect effects on short-acting insulin doses following initiation of GLP-1 RA are detailed in Table 5.

TABLE 1 : Population baseline characteristics.
HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterolInsulins were categorized as basal or prandial.Insulin lispro and aspart were categorized as prandial insulins, while insulin regular U-500, glargine, Neutral Protamine Hagedorn (NPH), detemir, and degludec were classified as basal insulins.The distribution of GLP-1 RAs received by patients is detailed in

TABLE 3 : Primary outcome results.
Results of the secondary outcomes are detailed in Table4.Average hemoglobin A1c, body weight, LDL cholesterol, and total cholesterol all significantly decreased from baseline to three months following GLP-1 RA initiation.Systolic and diastolic blood pressure, HDL cholesterol, and triglyceride levels all decreased as well, but these results did not meet statistical significance.