to the editor: I applaud the authors for focusing their review on clinically relevant and useful guidance for managing blood glucose levels with noninsulin therapies. The streamlined algorithm in Figure 1, however, is misleading in two regards.

First, the algorithm suggests that a dipeptidyl-peptidase-4 (DPP-4) inhibitor and a glucagon-like peptide-1 (GLP-1) receptor agonist are a recommended combination. Combining medications from these two classes has been investigated in one animal study and one randomized controlled trial in humans.¹ In the animal study, adding sitagliptin (Januvia) to liraglutide (Victoza) did not increase the blood concentration of liraglutide, change its pharmacokinetics, or reduce its breakdown. In the human trial, patients taking metformin plus sitagliptin were randomized to either add exenatide (Byetta, Bydureon) or substitute sitagliptin with exenatide. Although the triple-therapy group had an additional 0.3% decrease in A1C level, this change is smaller than what would be expected after adding other recommended agents. Also, it is much lower than the average A1C decrease with GLP-1 receptor agonists (0.8% to 2%) that would be expected in the absence of coadministered DPP-4 inhibitors.

The best available evidence suggests that coadministration of DPP-4 inhibitors and GLP-1 receptor agonists is clinically inferior to other options for enhancing glycemic control. As this review points out, this combination is not approved by the U.S. Food and Drug Administration, nor recommended in the algorithm from the American Diabetes Association (ADA), which is a source for the simplified algorithm in Figure 1.²

Second, Figure 1 suggests that adding a sulfonylurea to a basal insulin is a recommended combination. However, the ADA algorithm does not recommend this combination. In the algorithm's initial publication, the authors noted, “Insulin secretagogues do not seem to provide for additional HbA_1c reduction or prevention of hypoglycemia or weight gain after insulin is started, especially after the dose is titrated and stabilized.”³ The American Association of Clinical Endocrinologists/American College of Endocrinology comprehensive diabetes management algorithm from 2015 similarly states, “Consider discontinuing or reducing sulfonylurea after basal insulin started.” ⁴

Given the variety of recommended combinations to help patients with type 2 diabetes mellitus achieve their personalized goals, we should not be using nonrecommended combinations that add little or no clinical benefit.

in reply: We thank Dr. Geurin for bringing up two issues that may have caused confusion. The intent of the algorithm was to simplify the available medication choices, rather than recommend specific medications or combinations as preferred add-on therapy. We recommended that a comprehensive, patient-centered approach be used to select drug therapy, while considering A1C levels, fasting or postprandial glucose control, weight loss, and adverse effects. We thought the similar mechanisms of action for drug classes listed in eTable B would lead readers to conclude that concomitant therapy would not be acceptable, but we should have been more explicit.

Dr. Geurin's second concern was that Figure 1 also suggested that the addition of a sulfonylurea to established basal insulin therapy was a recommended combination. Although we concur that this drug combination is not ideal, we also recognize that some patients may have limited medication choices because of cost, adverse effects, contraindications, or personal preference. An exhaustive review of the evidence for and against various combination therapies was beyond the scope of our article.