December 12, 2017
2 min read
Save

HbA1c reductions differ with type of second-line drug class

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Assaf Gottlieb
Assaf Gottlieb

 

HbA1c is significantly reduced in adults with type 2 diabetes when treated with a second-line drug class, but reductions are greater with sulfonylureas compared with thiazolidinediones, DPP-IV inhibitors and GLP-1 receptor agonists, according to findings published in BMJ Open Diabetes Research & Care.

“While metformin is the preferred choice for initial drug treatment in type 2 diabetes, there is no clearly preferred choice for an additional drug when indicated,” Assaf Gottlieb, PhD, assistant professor in the School of Biomedical Informatics at the University of Texas Health Science Center at Houston, told Endocrine Today. “Using anonymized and de-identified electronic health records, we performed a retrospective analysis of more than 40,000 patients to compare the counterfactual drug effectiveness in lowering HbA1c levels and effect on BMI of four diabetes second-line drug classes.”

Gottlieb, Chen Yanover, PhD, research staff member in the machine learning for healthcare and life sciences group at IBM Research — Haifa in Israel, and colleagues evaluated data on 40,871 adults with type 2 diabetes based on the Northwestern University diabetes phenotyping algorithm to determine drug effectiveness in lowering HbA1c and effect on BMI of four diabetes second-line drug classes using EHRs. Participants were included if they were first prescribed metformin and subsequently prescribed a second-line drug belonging to one of four classes, including sulfonylureas (n = 26,684; mean age, 61.2 years; 47.7% women), TZDs (n = 4,794; mean age, 59.6 years; 48.2% women), GLP-1 receptor agonists (n = 1,532; mean age, 52.8 years; 66.6% women) and DPP-IV inhibitors (n = 7,861; mean age, 58.9 years; 51.1% women) between 2000 and 2015. Changes in HbA1c and BMI were measured at 6 and 12 months after the index date.

After 12 months of treatment, all four second-line drug classes were predicted to reduce HbA1c to less than 7% with a reduction of 0.6% to 0.61% in HbA1c from baseline. HbA1c levels inferred for sulfonylureas were significantly higher than for TZDs, DPP-4 and GLP-1 by 0.09% to 0.24%. HbA1c levels were higher with DPP-IV inhibitors than TZDs after 12 months and higher than GLP-1 receptor agonists after 6 months, but the latter difference was no longer significant after 12 months.

Predicted BMI was significantly lower with DPP-IV inhibitors after 12 months compared with sulfonylureas or TZDs (0.47 to 0.81 kg/m2). No significant difference in BMI was found between DPP-IV inhibitors and GLP-1 receptor agonists.

“Our study demonstrates that while sulfonylureas are the preferred choice for second-line drugs, they may not be the most effective, and newer drugs might be preferable,” Gottlieb said. “Additionally, our results show that secondary use of electronic health records can allow replication and expansion of clinical trial results,” he said. “The advantages of this approach in terms of the labor and costs required to expand evidence-based medicine are evident. We expect that causal inference based on observational data will become more widely used in the future for such purposes.”

Gottlieb added that a more personalized risk-benefit approach is needed that can address the effectiveness and potential adverse effects of each drug.

“Such approach will allow physicians to make more informed decisions and tailor the best drug for each patient,” he said. – by Amber Cox

For more information:

Assaf Gottlieb, PhD, can be reached at assaf.gottlieb@uth.tmc.edu. Chen Yanover, PhD, can be reached at cheny@il.ibm.com.

Disclosures: The authors report no relevant financial disclosures.

Editor's Note: On Dec. 13, we updated the 4th paragraph to reflect correct HbA1c level information. The Editors regret this error.