NPC Archive Item: Effects of pioglitazone▼ on major cardiovascular adverse events in high risk patients with type 2 diabetes

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

Wilcox R, Kupfer S, Erdmann E. Effects of pioglitazone on major adverse cardiovascular events in high-risk patients with type 2 diabetes: Results from prospective pioglitazone clinical trial in macrovascular events (PROactive 10). Am Heart J 2008;155:712–7.

The PROactive study failed to show a significant benefit for pioglitazone compared with placebo in people who had type 2 diabetes as well as other evidence of cardiovascular disease. Despite a re-analysis of secondary endpoint data from this study, claims of this showing a significant benefit for pioglitazone appear unjustified.

We should continue to focus our attention on reducing overall cardiovascular risk in patients with type 2 diabetes, rather than pursuing intensive strategies to achieve tight blood glucose control. Priority should be given to treating patients with evidence-based interventions such as smoking cessation, blood pressure control, metformin, aspirin and simvastatin while seeking to control the symptoms associated with high blood glucose levels.

Glitazones have been the subject of an MHRA warning about their use in people with, or a history of heart failure, particularly in those using insulin. The NICE guideline on the management of type 2 diabetes defines a limited role only for these agents as third-line if metformin and suphonylureas have failed to adequately control HbA1c (to <7.5% or other higher level agreed with the individual).

What is the background to this?
Metformin is the only oral hypoglyacemic agent with robust evidence from randomised controlled trials (RCTs) showing it reduces the risk of macrovascular end points. In the NICE guideline on the management of type 2 diabetes, it is the first-choice oral hypoglyacemic in overweight patients and may be considered in the non-overweight. Sulphonylureas may be considered in the non-overweight, or if metformin is contraindicated or not tolerated. A sulphonylurea can be added to metformin as second-line therapy, if blood glucose control remains or becomes inadequate with metformin alone. Glitazones (pioglitazone [Actos®] and rosiglitazone [Avandia®]) are third-line agents, as dual or triple therapy with metformin and/or a sulphonylurea if HbA1c targets are not reached.

Glitazones are associated with an increased risk of heart failure (HF), and they should not be used in patients with HF, or a history of HF. Particular caution is advised in those using insulin. The Medicines and Healthcare Products Regulatory Agency (MHRA) has advised on their use in people at risk from HF, and for the close monitoring of patients during treatment for signs and symptoms of fluid retention, including weight gain and oedema. Rosiglitazone may be also be associated with a small increased risk of cardiac ischaemia (see Drug Safety Update 2007; 1(5) and MeReC Rapid Review blog) and is contraindicated in patients with acute coronary syndrome (ACS), and not recommended for use in patients with ischaemic heart disease or peripheral arterial disease.

For further information on type 2 diabetes and its treatment see the type 2 diabetes section of NPC.

What does this study claim?
This study was a re-analysis of data from the PROactive study (PROspective pioglitAzone Clinical Trial In macroVascular Events), a study that failed to show a convincing benefit for pioglitazone. The authors of this re-analysis claim that in patients with type 2 diabetes at high risk of cardiovascular (CV) events, pioglitazone reduced the risk of major CV events in various composite end points at 3 years. At final visit, 257 (9.9%) pioglitazone-treated patients and 313 (11.9%) placebo-treated patients had a first event of CV death, nonfatal myocardial infarction (MI) (excluding silent MI), or nonfatal stroke (hazard ratio [HR] 0.82, 95% confidence interval [CI] 0.70 to 0.97; P=0.02; number needed to treat [NNT] 48 over 34.5 months). Note: this was a secondary endpoint of the PROactive study (see below).

How does this relate to other studies?
In the original published analysis of the PROactive study, for the primary endpoint, a composite of all-cause mortality, non-fatal MI (including silent MI), stroke, ACS, endovascular or surgical intervention in the coronary or leg arteries, and amputation above the ankle, there was no statistically significant difference between the pioglitazone and placebo groups (HR 0.90, 95%CI 0.80 to 1.02; P=0.095). Heart failure requiring hospital admission was significantly higher with pioglitazone, compared with placebo (5.7% vs. 4.1%; P=0.007; number needed to harm [NNH] 62 over 34.5 months). Note that this P value is highly statistically significant meaning that this difference is unlikely to have arisen by chance.

So what?
The validity of this re-analysis of PROactive data is highly questionable. The practice of evaluating secondary outcomes where the primary endpoint has failed to reach statistical significance is controversial and the results should be interpreted very cautiously. It is especially important in this study as the P value for the “main prespecified secondary endpoint” (a composite of all-cause death, non-fatal MI (excluding silent MI), and non-fatal stroke) was not particularly low (P=0.027), and the observed difference between groups could have occurred by chance (the chance of throwing a double six with two dice is 1 in 36 or P=0.028).

Eight composite endpoints were evaluated in this re-analysis. When each component was assessed individually, there were no statistically significant differences between pioglitazone and placebo for any of the components, which further questions the validity of these composite endpoints. Furthermore, none of the eight outcomes were actually pre-defined in the original study design [1], which raises the possibility that the data from the study has been analysed in many ways to find any kind of positive result in favour of the intervention.

For more information on primary and secondary end points and post hoc analyses see MeReC Briefing 2005;30.

Given all of these design flaws, claims that pioglitazone reduces the risk of CV events, compared with placebo in patients with type 2 diabetes and evidence of macrovascular complications appear to be unfounded. We have previously blogged about rosiglitazone’s associated increased risk of CV events compared with placebo. This excess risk was not seen in the PROactive study with pioglitazone. However, we must also consider the possible harms of all glitazones, particularly with respect to the increased risk of heart failure as highlighted by the MHRA.

Study details
Patients – this was a prospective RCT of adults (35–75 years) with type 2 diabetes and HbA1c >6.5%, with evidence of macrovascular disease before recruitment.

Interventions/comparators – patients were randomised to oral pioglitazone (n=2605), titrated from 15 to 45mg/day based upon tolerability, or matching placebo (n=2633), in addition to their existing glucose-lowering and cardiovascular medication(s). Mean follow up was 34.5 months. Concealed allocation was not described. Evidence-based interventions to reduce CV risk had not been optimised at baseline, e.g. 14% were smokers, only 43% were on a statin, and over 70% had blood pressure >140/85mmHg.1

Results – This re-analysis of the PROactive study reported the effects of pioglitazone on the “main secondary endpoint” (not defined in the original study design1), a composite of all-cause mortality, nonfatal MI, and nonfatal stroke (not reported in abstract; HR 0.84, 95%CI 0.72 to 0.98; P=0.027), a “pre-specified” secondary end point (not defined in the original study design [1]) of CV death, nonfatal MI, or nonfatal stroke (HR 0.82, 95%CI 0.70 to 0.97; P=0.02), and six other post hoc composite endpoints (various combinations of all-cause, CV, or cardiac mortality; plus nonfatal MI; plus nonfatal stroke; and/or ACS). There were statistically significant differences in favour of pioglitazone for five of these outcomes (P<0.05), but no significant benefit for pioglitazone when any of the individual components were analysed separately. Secondary end points from the original study design included the individual components of the primary end point and cardiovascular mortality.

The PROactive study was sponsored by Takeda and Eli Lilly.

Please comment on this blog in the NPC discussion rooms, or using our feedback form