Prof Erland Erdmann

Although it was a subgroup analysis, Prof Erland Erdmann (University of Cologne, Germany) concluded, "Pioglitazone was effective in the secondary prevention of atherosclerotic disease in high-risk patients with type 2 diabetes and a previous myocardial infarction."

Adding pioglitazone to the medication of 1000 of these patients would prevent 22 recurrent MIs and 23 acute coronary syndrome (ACS) events over three years, Erdmann noted. "We think that it's a very positive effect of pioglitazone, and it is one of the few evidence-based trials for medication in diabetes and previous myocardial infarction."

There was an increase in hospitalizations for heart failure with pioglitazone treatment—as was seen in the overall trial—but all-cause deaths were not different between groups, a finding that some see as reassuring.

The results were presented here at the American Heart Association Scientific Sessions 2005. The PROactive study was funded by Takeda Pharmaceutical Company Ltd and Eli Lilly and Co.

Pioglitazone is a peroxisome proliferator-activated receptor- (PPAR-) agonist used to treat type 2 diabetes. It has a variety of actions that could potentially positively affect cardiovascular risk, including reduction of glucose, triglycerides, and C-reactive protein and raising of HDL.

The main PROactive study was a randomized, double-blind, placebo-controlled trial in 5238 patients with type 2 diabetes considered at high risk for CV events because of the presence of macrovascular disease, including previous stroke, MI, revascularization, or peripheral artery disease, and objective evidence of coronary artery disease.

Patients were randomized to 45 mg/day of pioglitazone or placebo given on top of standard medications recommended in this patient population, including antihypertensive drugs; glucose-lowering medications such as metformin, sulfonylurea, and insulin; antiplatelet drugs including aspirin; and lipid-lowering agents such as statins and fibrates.

The main study results, first presented at the European Association for the Study of Diabetes (EASD) in September 2005 and subsequently published in the Lancet [1], showed a nonsignificant 10% reduction in the study's primary end point, a composite of all-cause mortality, nonfatal MI (including silent MI), stroke, major leg amputation (above the ankle), acute coronary syndrome, cardiac intervention including CABG or PCI, or leg revascularization. However, there was a significant 16% reduction in the secondary composite end point of death, MI (excluding silent MI), and stroke seen with pioglitazone.

At the time, the investigators speculated that the primary end point might not have been ideal, and many have criticized the decision since the results were presented. Erdmann told heartwirethat PROactive steering committee chair Prof John Dormandy (St George's Hospital, London, UK) is a peripheral arterial surgeon, and it was Dormandy's opinion that peripheral and coronary arterial disease are the same process. "That is textbook knowledge also, but there are many colleagues who say there is a difference in diabetes between leg and heart and brain perfusion, and that may be true," Erdmann said. Procedure-related end points like revascularization or amputation are also to some degree at the discretion of the treating physician, unlike "hard" end points such as death or MI, he pointed out.

In the PROactive subgroup analysis presented here, the investigators looked further at the effects of treatment in a subgroup of 2445 of these patients, 46.7% of the overall population, with type 2 diabetes and a previous MI at least six months before enrollment in the study.

Prespecified analyses for the overall study group included time to fatal or nonfatal MI; time to cardiovascular death or nonfatal MI; and time to CV death, nonfatal MI, or stroke. This analysis then was "a post hoc, exploratory analysis for the diabetes and MI subgroup, in which some of the factors were prespecified and some we added," Erdmann noted. About 63% of this high-risk study group were on statin therapy, he said.

They found a statistically significant 28% reduction in recurrent fatal or nonfatal MI with treatment. They constructed a composite cardiac end point in this group, including cardiac death, nonfatal MI, coronary revascularization, and acute coronary syndrome, and found there was a statistically significant difference of 19% after two years favoring pioglitazone, he said. Acute coronary syndrome was also significantly reduced by 37%.

Pioglitazone also reduced HbA1_c, although there was also a reduction of 0.8% in the control group due to the good treatment of these patients, he said. Triglycerides were reduced by 11%, and HDL cholesterol was increased by 18%, producing improvement in the HDL:LDL ratio.

There was a higher incidence of heart failure reported with pioglitazone in the main study, leading to some debate about the net effect of this treatment, Erdmann noted. In this analysis, there were fewer adverse events in the pioglitazone group because of the lower incidence of end-point events. Heart-failure hospitalizations were higher with pioglitazone, but all-cause death among heart-failure patients was not different between pioglitazone and control patients.

"We think, and we suggest, that some of these patients did not have heart failure but rather edema, and it's quite natural for a doctor to see a patient with type 2 diabetes after myocardial infarction and edema to maybe be a bit quick [to] say it's heart failure," Erdmann said. "At least we can say that the death rate is not different between the two groups, and I personally think that heart failure has nothing to do with this."

It was also pointed out during wider discussion of the trial since it was first presented that the effect of pioglitazone was mitigated in those patients who were already on a statin. About half the patients were receiving a statin in the main study, and in these patients, the beneficial effect of pioglitazone was reduced compared with those not on statins, Erdmann acknowledged.

However, 63% of patients in the current analysis were on statins, he noted, "and nevertheless we have this high effect." Multivariate analysis of these data showed the benefit on these end points seen even in the presence of statins, he told heartwire.

During his talk here, Erdmann presented two Kaplan-Meier curves from the CARE and HPS statin trials in similar patient populations, superimposing curves from the present subgroup analysis, suggesting the risk reduction seen with pioglitazone in this PROactive subgroup analysis was additive to statin treatment. "Today's practice is to give both [agents] in these patients," he said.

Invited discussant for the PROactive study was Dr Jorge Plutzky (Brigham and Women's Hospital, Boston, MA), whose group has previously done extensive research into the activity of PPAR receptors and agonists. He pointed to several points of caution in the PROactive subgroup analysis: for example, there were no data given for the original primary end point in this subgroup, and some of the end points presented were not prespecified, although, he added, "the investigators have appropriately labeled this as an exploratory analysis."

Plutzky also noted that the primary end point for PROactive included both coronary- and peripheral-disease end points. "Many who pursue mechanisms of atherosclerosis would point out that peripheral vascular disease and coronary disease, including the response to therapy, are not the same."

Of the heart-failure finding, Plutzky said the glitazones have not been known to cause myocardial dysfunction, and the equivalent mortality seen among those patients would argue for the interpretation of an increase in edema rather than heart failure. "We need a distinction between an untoward myocardial effect and what is more likely here, a side effect of the drug involving fluid retention."

"In the past three months since PROactive has been released, I have been surprised at the animated, if not at times heated, discussion on both sides," Plutzky said. From the perspective of patients, the most important issue surrounding the use of these drugs is that ultimately the correct answer is found, and that will not likely be from a single study, he said. "Analyses such as we just heard may help guide us as we go forward," he said, but fortunately, other studies with these agents, among them IRIS, RECORD, DREAM, and BARI-2D, are now under way.

Dr Nikolaus Marx (Ulm University Medical Center, Germany) did some of the early work on the effects of pioglitazone, particularly in restenosis. He told heartwire that after the "disappointing" results of the primary PROactive trial, the present analysis is more in line with their own work with this agent. However, he said, "it's interesting, and it's intellectually very satisfying, but it doesn't change clinical practice because it's a post hoc analysis of a secondary end point, later defined," he said. Their previous work has suggested that statins and glitazones are additive, but "from this study we can't conclude that."

He said he would like to see a "real" cardiac-end-points trial with pioglitazone: "type 2 diabetics, post-MI, just hard end points—death, myocardial infarction, stroke." As for clinical practice now, Marx said his interpretation is that these patients must be treated to target on parameters such as blood pressure and LDL cholesterol, "and if you have to take an antidiabetic drug, the glitazones may have benefits beyond glucose lowering," evidence based not only on surrogate end-point findings, but now on secondary end points of a large outcome trial. "But I think you can't go further than that."

Commenting for heartwireon these findings, Dr Roger Blumenthal (Johns Hopkins University Medical Center, Baltimore, MD) found some comfort in the heart-failure results in this analysis. "The overall findings weren't so impressive, partly because of this increase in heart failure and the fact that the absolute risk reductions were a lot more modest than a lot of us anticipated," he said.

"I think cardiologists will find these data reassuring, that we're not trading one source of mortality for another," Blumenthal said.