Adjuvant Metformin in Type 1 Diabetes

April 24, 2017

This study is the first retrospective, long-term analysis of metformin as adjuvant therapy to insulin in patients with T1DM.

Adjunctive metformin may have small short-term benefits, but it provides no added benefit during 10-year follow-up, according to a study published online in Diabetes, Obesity and Metabolism.1

The study is the first retrospective, long-term analysis of metformin as adjuvant therapy to insulin in type 1 diabetes mellitus (T1DM).

“We conclude that metformin is not associated with long-term beneficial effects on BMI, HbA1c or daily insulin dose when added as adjunct therapy to intensive insulin therapy in T1DM patients,” wrote lead author Chantal Mathieu, MD, PhD, of the Laboratory for Clinical and Experimental Endocrinology, KU Leuven, Belgium, and colleagues.

Adjuvant metformin is often used off-label in T1DM, owing to its mechanism of action, which may improve glycemic control and decrease insulin dose, weight, and risk of hypoglycemia. While several short-term studies have supported the beneficial effect of adjuvant metformin in T1DM, whether these effects are sustained long term has remained unknown.

To evaluate the long-term effects of adjuvant metformin in a real-world setting, researchers used electronic medical records of patients with T1DM seen at University Hospitals Gasthuisberg, Leuven between September 2000 and December 2014.

They separated patients into two groups: those who received adjuvant metformin for 6 months or longer (metformin group, n=186) and those who refused metformin or took it for less than 6 months (control group, n=62). They also compared these groups to a baseline reference group of patients with T1DM who were never offered adjuvant metformin (n=961). Hemoglobin A1c (HbA1c), body mass index (BMI), and daily insulin dose were recorded yearly.

At baseline, the metformin group had higher BMI and lower insulin doses than the control group. Both groups had comparable baseline HbA1c levels (>7%).

Key Results:

• First year:

♦ Small, nonsignificant decreases in BMI (P=0.250) and HbA1c (P=0.52) for metformin vs control

♦ Significant decrease in total insulin dose for metformin vs control (P=0.469)

• 10th year: Graphs for BMI, HbA1c, and insulin dose showed a converging pattern, which suggests no persistent effect of metformin

The authors noted several limitations. The study could not measure confounders such as physical activity, comorbid conditions, use of a dietician, or use of other medications that affect weight (eg, antipsychotics and antidepressants). Also, the study could not evaluate macrovascular or microvascular complications or overall survival because of the existence of confounders such as statin, angiotensin-converting enzyme inhibitor, antihypertensive, and tobacco use.

However, the authors noted that some studies have suggested that metformin may decrease microvascular and macrovascular complications independently of glycemic control. REMOVAL (REducing with MetfOrmin Vascular Adverse Lesions in type 1 diabetes), a double-blind, randomized, placebo-controlled trial, is currently evaluating the effects of 3 years of adjuvant metformin on atherosclerosis in T1DM.2

Take-home Points

• The first long-term analysis shows that the short-term benefits of adjuvant metformin on BMI, insulin dose, and HbA1c are not maintained over 10 years in T1DM.

• The study could not evaluate long-term microvascular and macrovascular benefits of adjuvant metformin in T1DM.

• The REMOVAL trial is currently evaluating the effects of 3 years of adjuvant metformin on atherosclerosis in T1DM.

References:

1. Staels F, Moyson C, Mathieu C. Metformin as add-on to intensive insulin therapy in type 1 diabetes mellitus. Diabetes Obes Metab. doi: 10.1111/dom.12948. Published online March 20, 2017.

2. Petrie JR, Chaturvedi N, Ford I, et al. Metformin in adults with type 1 diabetes: design and methods of REducing with MetfOrmin Vascular Adverse Lesions (REMOVAL): an international multicentre trial. Diabetes Obes Metab. 2017;19:509-516. doi:10.1111/dom.12840.