Glitazones increase heart failure but not mortality in new meta-analysis

September 27, 2007

Burlington, MA - A new meta-analysis of studies with the diabetes drugs rosiglitazone (Avandia, GlaxoSmithKline) and pioglitazone (Actos, Takeda Pharmaceuticals North America) has shown an increased risk of heart failure but no increase in cardiovascular death with these drugs[1]. The authors conclude that heart failure associated with the glitazones might not carry the same risk as heart failure caused by progressive systolic or diastolic dysfunction of the left ventricle. But they acknowledge that longer follow-up and better characterization of such patients is needed to determine the effect of these drugs on overall cardiovascular outcome.

The meta-analysis is accompanied by two comments[2,3] and an editorial[4]. Both comments contend that there is very little evidence of any real benefits with the glitazones and because of this they should have no major role regardless of any safety concerns.

Does the fluid retention affect mortality?

The current study, published in the September 29, 2007 issue of the Lancet, was conducted by Drs Rodrigo Lago, Premranjan Singh, and Richard Nesto (Lahey Clinic Medical Center, Burlington, MA). They note that the glitazones, also known as thiazolidinediones (TZDs), are well known to cause fluid retention and therefore could potentially lead to development of congestive heart failure (CHF). But they point out that it is not known whether the heart failure caused by TZD-related fluid retention might ultimately affect survival. They therefore conducted a meta-analysis of pooled data from seven randomized trials of TZDs in subjects with prediabetes or type 2 diabetes to assess the risk of development of heart failure and death from cardiovascular causes.

Of the 20 191 patients included in the meta-analysis, 360 had CHF events (214 with TZDs and 146 with comparators). Compared with controls, patients given TZDs had increased risk for development of CHF across a wide background of cardiac risk. Results showed no heterogeneity of effects across studies, which indicated a class effect for TZDs. In contrast, the risk of cardiovascular death was not increased with either of the two TZDs.

Risk ratio (95% CI) for CHF and CV death with rosiglitazone/pioglitazone vs control

Outcome Rosiglitazone Pioglitazone Both agents
CHF 2.18 (1.44-3.32) 1.32 (1.04-1.68) 1.72 (1.21-2.42)
CV death 0.91 (0.63-1.32) 1.01 (0.51-2.01) 0.93 (0.67-1.29)

The authors note that longer follow-up and better characterization of patients in whom CHF develops because of fluid retention is needed to determine the effect of TZDs on overall cardiovascular outcome and whether CHF should be regarded as an adverse event or a characteristic cardiovascular end point.

They conclude: "Despite the glucose-lowering effect of TZDs, our data indicate that these drugs should not be used in patients with heart failure and should be cautiously used for glycemic control in patients with cardiovascular disease who do not have heart failure. In patients with type 2 diabetes without cardiovascular disease in whom the absolute risk for congestive heart failure is much lower, the use of TZDs should be weighed against the risks and benefits of other antidiabetic medications."

Nissen not impressed

Commenting on this study for heart wire , Dr Steven Nissen (Cleveland Clinic, OH) who conducted the previous meta-analysis suggesting harm with rosiglitazone[5], said, "The authors did not examine the risk of cardiac death in patients who developed CHF while taking a TZD drug. Instead, they looked at overall cardiovascular mortality. This finding neither confirms nor refutes the hypothesis that TZDs lead to major complications of heart failure, including death."

Kaul: "Reassuring but not definitive"

Dr Sanjay Kaul (Cedars-Sinai Medical Center, Los Angeles, CA) said the new data were reassuring in the sense that despite the increase in CHF, there was no increase in cardiovascular death. But he added that only randomized trials that are specifically designed to prospectively assess cardiovascular efficacy and safety would provide definitive answers.

Kaul pointed out that the lack of an increase in cardiovascular death with rosiglitazone in the current analysis is consistent with the results of the Singh et al meta-analysis published recently in the Journal of the American Medical Association[6], but inconsistent with Nissen's meta-analysis suggesting increased risk. He says these discordant results are perhaps best explained by the fact that the Nissen and Wolski meta-analysis included several small studies where zero events in the control group outnumbered zero events in the treatment group, without applying appropriate corrections, which could overestimate risk.

On whether there are meaningful differences between pioglitazone and rosiglitazone, Kaul says that both drugs increase fluid retention, but they have some minor differences with respect to changes in lipid profile. However, lack of a difference in cardiovascular mortality among the TZDs brings into question the clinical relevance of these differences in cardiometabolic profile.

Confidence intervals wide

In the first comment article, Drs John Cleland (University of Hull, UK) and Stephen Atkin (Hull York Medical School, UK) note that while neither rosiglitazone nor pioglitazone was associated with increased cardiovascular mortality in the current meta-analysis, the confidence intervals cannot exclude a 25% increase. They add that the current meta-analysis included far fewer trials than Nissen's, and it is not certain which set of trial selection criteria was least prone to bias. They point out that TZDs can cause fluid retention that might result in peripheral and pulmonary edema but there is little evidence that such drugs affect cardiac function adversely and that patients with edema caused by TZDs do not seem to have an unfavorable prognosis.

To heart wire , Cleland commented: "This meta-analysis is reassuring in one way—it suggests that the heart failure these drugs produce does not appear to be life threatening. It should not really be called heart failure. It is actually fluid retention that produces symptoms similar to heart failure and is not the same as actually having damage to heart muscle."

Effectiveness should be the key issue

But in their article, Cleland and Atkin stress that "all the meta-analyses consistently fail to spot the elephant in the room"—that treatments should be effective rather than merely innocuous.

Cleland told heart wire : "The glitazones, like many other diabetes drugs, have been licensed only on the basis that they reduce blood sugar. They have no evidence of hard clinical benefits, and just because a drug reduces blood sugar does not mean to say it will have any real benefits." He drew an analogy to HDL and torcetrapib. "We know high HDL is linked to a beneficial protective effect. Torcetrapib raised HDL but actually resulted in a worse outcome. So raising HDL does not necessary reduce event rates. The same with blood sugar. We have evidence to suggest that high blood sugar is linked to worse outcomes, but we don't know that lowering blood sugar will necessarily improve event rates."

He continued: "We have to think why we treat patients with diabetes. It is not to lower a number for blood sugar in a test tube. It is to reduce end-organ complications. The authors of this meta-analysis are suggesting that a treatment that lowers blood sugar and does not increase cardiovascular event rates is useful. But I fail to see why. I believe we should have a moratorium on these drugs until they have some evidence of a positive impact on end-organ function."

Is longer patent protection the answer?

Cleland suggests that companies need better incentives to do the right studies to show real effectiveness of drugs, and that means longer patent protection. "Under the current system, companies are not going to do a five- or 10-year study to prove effectiveness, as by the time the drug would be approved, the patent would be up. If drugs had 50 years of patent protection, then these studies would get done. It would also mean that companies would not have to charge such high prices, as they would have much longer in which to recoup their investment," he commented to heartwire .

These drugs give you heart failure and osteoporosis, and we don't know of any real benefits that they have, but these authors are asking that we should be reassured because they don't kill you. I don't think that is a good attitude.

In the second comment piece, Drs Victor Montori, Gunjan Gandhi (Mayo Clinic, Rochester, MN) and Gordon Guyatt (McMaster University, Hamilton, ON) make similar points about the inadequacy of blood sugar levels as a trial end point. "Surrogate outcomes allow smaller, shorter, and cheaper trials; provide a faster offering of more choices to patients and clinicians; save research money; and allow new drugs more rapid access to market. The apparent benefits are, however, a mirage, and the apparent savings represent false economy. Any savings are quickly overwhelmed by costs associated with potentially ineffective or even harmful (yet heavily advertised) expensive therapies and the incremental costs of treating the harms these interventions might cause. Patients and society may end up paying dearly for drugs that cause more harm than good," they write.

On the specific findings of this meta-analysis, Montori commented to heart wire : "These drugs give you heart failure and osteoporosis, and we don't know of any real benefits that they have, but these authors are asking that we should be reassured because they don't kill you. I don't think that is a good attitude."

Montori added: "We are not avid users of TZDs at the Mayo Clinic. I would certainly not use them to prevent diabetes, and they have many unattractive features in the treatment of diabetes. We are always telling our patients of the importance of losing weight, but these drugs make them put on weight. Patients don't like that."

He also pointed out that the discussion about whether the heart failure they induce is benign or malignant is not the point. "We shouldn't be using drugs that cause CHF at all. Even if it is benign, which I do not believe has been proven by this study, why should we give a drug that causes benign CHF when we don't know if it has any benefits? If they have truly major benefits like a reduction in dialysis, stroke, MI, or blindness, benign CHF may be an acceptable cost. But if we don't know the benefits, any risk is not worthwhile."

The editorial, by the Lancet, stresses the importance of further studies to assess safety of these drugs and says that regulatory agencies "must hold manufacturers' feet to the fire to ensure that these are performed, performed properly, thoroughly evaluated, and made available to guide decisions about prescribing." It adds: "Unless limitations on the understanding, analysis, and communication of drug safety issues are addressed, the thiazolidinediones might simply become the latest in a series of preventable drug disasters."

Nesto is involved in research funded by GlaxoSmithKline and Takeda. Lago and Singh declare no conflicts of interest, as do Kaul, Montori, Gandhi, and Guyatt. Cleland has received funding for research into heart failure and carvedilol from GlaxoSmithKline; has received speaker's honoraria and funding from Takeda for research on candesartan; was part of a GlaxoSmithKline advisory board on rosiglitazone-induced fluid retention; and has worked on a safety committee reviewing adverse events with rosiglitazone. Atkin has received speaker's honoraria and sponsorship to attend meetings from GlaxoSmithKline and Takeda.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.