NEW YORK (Reuters Health) – For patients with type 2 diabetes not controlled with metformin and a sulfonylurea, the choice of an additional agent should be based on individual clinical characteristics since improvements in glucose control are similar.

That conclusion comes from a meta-analysis conducted by Dr. Jorge L. Gross, at Hospital de Clinicas de Porto Alegre in Brazil, and colleagues, reported in the May 17 issue of the Annals of Internal Medicine.

The authors explain that up to 70% of patients who initially achieve glycemic control with metformin have progressive deterioration in 2-3 years, and adding a sulfonylurea does not usually maintain long-term control. “Options for third agents include insulin, alpha-glucosidase inhibitors (acarbose), thiazolidinediones, glucagon-like peptide-1 (GLP-1) agonists, and dipeptidyl peptidase-4 inhibitors.”

To compare the efficacy of these drug classes, the team identified 18 trials involving 4535 participants. The pooled results of direct and indirect comparisons showed that the additive effect of the different tertiary agents on HbA1c were similar.

Specifically, the change in A1C was -0.71% with the addition of insulin, -1.15% with thiazolidinediones, -0.60% with acarbose, -1.04% with GLP-1 agonists, and -0.89% with DPP-4 inhibitors, according to the report.

Side effects, however, differed considerably between the different agents, the investigators found. Body weight increased with insulin and thiazolidinediones, but decreased with acarbose and GLP-1 agonists, for example, while hypoglycemia was more likely with add-on insulin, followed by GLP-1 agonists.

“The available limited evidence does not clearly identify a preferred antihyperglycemic drug class among drugs represented in clinical trials,” Dr. Gross and colleagues conclude.

“When choosing a third drug to be added to metformin and sulfonylurea therapy in patients requiring additional glycemic control,” they advise, “the patient’s clinical features, such as importance of weight changes and incidence of hypoglycemia, should be taken into account.”

Ann Intern Med 2011; 154:672-679.