GIPS-III Trial: No Benefit from Post-Myocardial Infarction Metformin in Non-Diabetic Patients

Trial results presented at ACC.14 demonstrated that treatment with metformin immediately after PCI in non-diabetic patients presenting with acute STEMI does not lead to improvements in left ventricular ejection fraction or other measure of cardiac function.

Despite hopes that metformin might provide an inexpensive, safe, and effective means to limit heart muscle damage and prevent heart failure after myocardial infarction (MI) in the general population, researchers reporting at the American College of Cardiology’s 2014 Scientific Session had negative findings from the GIPS-III trial. Chris Lexis, MD, of the University of Groeningen (The Netherlands) reported the results at the Late-Breaking Clinical Trials session on March 31, 2014, in Washington, DC.

For individuals who experience ST-segment elevation MI, prompt reperfusion of ischemic heart tissue by means of percutaneous coronary intervention (PCI) and antithrombotic treatment can help limit damage to the myocardium and preserve left ventricular ejection fraction (LVEF). Between 20% and 40% of patients will still develop heart failure post-MI.

Since metformin, a biguanide used in the treatment of type 2 diabetes mellitus (DM), improves outcomes for patients with diabetes who sustain MI, and since LVEF preservation had been seen in previous studies using a rat model, researchers hypothesized that metformin might have benefit for individuals without diabetes who sustain MI.

For the study, 380 patients receiving primary PCI after STEMI were enrolled in a double-blind, placebo controlled study. Patients had to be older than 18 years of age and have sustained STEMI, and received primary PCI. The stent implanted had to be at least 3 mm and result in a Thrombolysis in Myocardial Infarction (TIMI) flow grade of two or three. Patients with prior MI, known diabetes, a need for coronary artery bypass graft surgery, or severe renal dysfunction were all excluded. Patients had to be able to undergo magnetic resonance imaging (MRI).

Average age of participants was about 59 years old, with a mean BMI of 27 and mean HbA1c of 5.8%. The patient cohort was three-quarters male and overwhelmingly white (96.3%). Most (68.1%) had single-vessel disease, with the right coronary artery most commonly involved (44.6%). Median reported ischemia time was 161 minutes.

Primary efficacy was assessed by left ventricular ejection fraction after four months as measured by MRI imaging. N-terminal pro-brain natriuretic peptide (NT-proBNP) levels, along with incidence of major cardiovascular events, constituted secondary efficacy measures.

Subjects were randomized 1:1 to receive metformin 500 mg or placebo twice daily for four months. The first dose was started as soon as possible after PCI. Although nearly 25% of patients withdrew from the study (many for reasons related to MRI scanning) and the sample size had to be increased twice to maintain power, none of the remaining patients died, nor were any lost to follow-up in four months.

The difference in LVEF between the metformin in the placebo group was not significant, with LVEF actually slightly lower in the metformin arm (53.1% compared with 54.8%, P = .10). Similarly, major adverse cardiovascular events were more common, though not significantly so, in the metformin group than in the placebo group (3.1% compared with 1.1%, P = .16).There was no significant difference in NT-proBNP values between the two groups, nor in serum creatinine or HbA1c at four months.

Subgroup analysis included examining those individuals who had sustained more significant left ventricular (LV) wall damage, and metformin also failed to show benefits for these individuals, who may have had prolonged ischemia or a larger infarct. This subgroup was important for analysis because animal models had shown benefit for metformin in subjects with relatively larger LV wall damage.

In comments after the presentation, Lexis noted that the relatively dense population and highly developed pre-hospital and emergency medical system in the Netherlands meant that the patients enrolled were already receiving very good care, as evidenced by the lack of mortality and relatively high LVEF.

On further questioning at the press conference following the morning’s session, the author reiterated that he had enough confidence in the subgroup analysis to assert that using metformin as a strategy to preserve LVEF in non-diabetic patients with more significant damage post-STEMI was not warranted. The study’s principal investigator, Iwan van der Horst, MD, PhD, commented, “We had hoped to see benefit for this very cheap drug available as an oral agent.” However, panelists commended the investigators for a tightly-designed and well-powered study.