16 December 2010
Rosiglitazone was withdrawn from clinical use across Europe because of concerns over excess cardiovascular (CV) risk. This cohort study found that pioglitazone▼ was associated with the same level of risk of acute myocardial infarction, acute heart failure and death as rosiglitazone. The study has some limitations and its findings conflict with a previous observational study which found that pioglitazone had a lower CV risk than rosiglitazone. However, as prospective head-to-head randomised controlled trials are not available on the comparative CV safety of these drugs, we should still be cautious with pioglitazone use.
Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.
Healthcare professionals should continue to follow NICE guidance on type 2 diabetes and agree individual HbA1c targets taking into account patient preference, and the balance of likely benefits and burden of treatment. Reducing HbA1c to levels of about 7.5% (59mmol/L) by diet, other lifestyle measures or treatment with metformin and/or a sulphonylurea would seem optimal based on current evidence. Uncertainty remains about the benefits of reducing HbA1c from around 9% (75mmol/L) to 7.5% using other glucose lowering interventions when other major CV risk factors (such as smoking, blood pressure, cholesterol) are also being managed actively. Pursuing HbA1c below a level of about 7.5% by adding insulin or any other third-line drug may not confer any benefit, adds to the risks, and requires caution.
People with diabetes are already at an increased risk of CV disease, and the aim of diabetic treatment is to reduce these risks, not increase them or simply affect biochemical markers, such as blood glucose levels. The withdrawal of rosiglitazone is an opportunity to review patients’ diabetes medication on an individual basis. Pioglitazone may be a suitable alternative for some patients, but it is not a straightforward decision to swap to this agent. Prescribers should note that pioglitazone is contraindicated in patients with heart failure or a history of heart failure. Furthermore, NICE advises that pioglitazone should not be commenced or continued in people with a higher risk of fracture, because it may increase this risk further. In addition, the Summary of Product Characteristics states that pioglitazone should be avoided in hepatic impairment, and liver function should be monitored before treatment and periodically thereafter, or if the patient develops symptoms suggesting hepatic dysfunction. Pioglitazone remains a black triangle drug and any suspected adverse reactions should be reported through the Yellow Card system.
What is the background to this?
The potential CV harms of glitazones fall into two categories, those associated with heart failure and those associated with ischaemic heart disease, such as myocardial infarction (MI).
Following mounting evidence that rosiglitazone is associated with an increased risk of both heart failure and MI, this drug was withdrawn from clinical use in Europe in September 2010. At that time, the MHRA stated that studies in patients taking pioglitazone suggest it can cause fluid retention, which may lead to heart failure in some people, but do not suggest that pioglitazone is associated with an increased risk of MI.
However, as an American Science Advisory pointed out earlier this year, there are currently no prospective randomised controlled trials (RCTs) that have directly examined the risk of ischaemic heart disease events associated with pioglitazone compared with rosiglitazone. There are observational studies, such as the one detailed here, but these have reached differing conclusions about the relative CV safety of these two drugs.
What does this study claim?
In this US cohort study of 36,628 (28,938 matched) adult patients newly initiated on rosiglitazone or pioglitazone, there was no significant difference between groups in the composite risk of acute MI, acute heart failure, or all-cause death.
Of the rosiglitazone-treated patients, 602 (4.16%) had an acute MI, acute HF, or death compared with 599 (4.14%) propensity score-matched pioglitazone-treated patients (hazard ratio [HR] 1.03; 95% confidence interval [CI] 0.91 to 1.15). There were also no significant differences between groups in the individual risks of acute MI, acute heart failure, or death; or in the risks in a sub-population of patients over 65 years of age (see below for details).
Following rosiglitazone’s withdrawal it is important that health professionals consider carefully the evidence around the possible CV effects of pioglitazone. People with diabetes are already at an increased risk of CV disease, and the aim of diabetic treatment is to reduce these risks, not increase them or simply affect biochemical markers, such as blood glucose levels.
Pioglitazone▼ and heart failure
In MeReC Extra No. 30 (2007) we reminded prescribers that there is consistent evidence that pioglitazone (and rosiglitazone) can cause weight gain and fluid retention, and lead to new or worsening heart failure. This is not a rare occurrence, and it can be serious and sometimes fatal. A meta-analysis of seven RCTs found that pioglitazone or rosiglitazone increased the risk of heart failure compared with controls across a wide range of cardiac risk (relative risk [RR] 1.72; 95%CI 1.21 to 2.42), but there was no increased risk of CV death.
The clinical outcome study with pioglitazone▼, PROactive, also found an increased risk of heart failure. Heart failure requiring admission to hospital occurred in 5.7% of patients in the pioglitazone arm and 4.1% of the placebo arm (P=0.007; number needed to harm [NNH] 62 over three years).
Pioglitazone▼ and ischaemic heart disease
The majority of published studies do not suggest that there is an increased risk of ischaemic heart disease with pioglitazone.
In PROactive, pioglitazone did not statistically significantly reduce the primary vascular composite outcome (all-cause mortality, non-fatal MI, stroke, acute coronary syndrome, vascular interventions or amputation) compared with placebo. However, neither did it increase this. The narrower secondary endpoint of all-cause mortality, non-fatal MI or stroke was reduced (HR 0.84; 95% CI 0.72–0.98; number needed to treat [NNT] 49. However, this result must be viewed with caution, and it is only hypothesis generating.
In a meta-analysis of 19 RCTs of pioglitazone versus an active comparator (in which nearly 80% of pooled events were contributed by PROactive), pioglitazone significantly reduced the composite end point of MI, stroke, or all-cause death (HR 0.82, 95%CI 0.72 to 0.94), but again heart failure was increased.
Pioglitazone, rosiglitazone and CV risk
This cohort study by Wertz et al found that pioglitazone was associated with the same level of CV risk as rosiglitazone. However, three other comparative cohort studies of elderly patients taking glitazones found pioglitazone was associated with a lower CV risk than rosiglitazone (Graham et al 2010, Winkelmayer et al, 2008 and Juurlink et al, 2009). In the study by Graham et al, compared with pioglitazone, rosiglitazone was associated with an increased risk of acute MI, stroke, heart failure or all-cause mortality (HR 1.18, 95%CI 1.12 to 1.23; number needed to harm [NNH] 60 over 1 year).
So what do the results of these studies, including this new study by Wertz tell us? Drawing definitive conclusions is not possible, but several explanations need to be considered:
- Neither rosiglitazone nor pioglitazone is associated with an increased CV risk compared to placebo.
- Rosiglitazone is associated with an increased CV risk compared to placebo but pioglitazone is not.
- Both rosiglitazone and pioglitazone are associated with a similar increased CV risk compared to placebo.
- Both rosiglitazone and pioglitazone are associated with an increased CV risk compared to placebo, but the risk with pioglitazone is less than that with rosiglitazone.
The first explanation can be rejected in view of other evidence of harm from rosiglitazone, and rosiglitazone’s withdrawal from the market because of these concerns. The second explanation is a possibility, but the evidence available currently is not robust enough to be confident that this is the case. All observational studies have limitations because treatment assignment is not randomised and results may reflect unrecognised biases or confounding. The limitations of the Wertz study are discussed in more detail in the related editorial.
The third and fourth explanations cannot be ruled out entirely. Pioglitazone did not increase CV risk in PROactive. However, as head-to-head comparisons from prospective RCTs regarding the comparative CV safety of rosiglitazone and pioglitazone are not available, we should still be cautious with pioglitazone use.
Pioglitazone▼ and other safety issues
The other main safety issue with pioglitazone (and rosiglitazone) is an increased risk of fractures. In October 2007, the MHRA stated that the risk of fracture should be considered in the care of patients, especially women, treated with either glitazone. More recently, a Canadian cohort study suggested that both men and women are at increased risk of fracture from glitazones, and that pioglitazone may actually be more strongly associated with fractures than rosiglitazone.
There are also rare reports of liver dysfunction with pioglitazone. In addition, the FDA is reviewing data from an ongoing, ten-year epidemiological study designed to evaluate whether or not pioglitazone is associated with an increased risk of bladder cancer.
Study details –
Wertz DA, et al. Risk of cardiovascular events and all-cause mortality in patients treated with thiazolidinediones in a managed-care population. Circ Cardiovasc Qual Outcomes 2010;3:538-45
American retrospective cohort study of patients using the HealthCore Integrated Research Database, which used propensity score-matching to to control for potential confounders.
Patients, intervention and comparison
36,628 patients over 18 years of age, newly initiated on rosiglitazone or pioglitazone between January 2001 and December 2005. 58% male, mean age 54. After matching, 14,469 patients included in each cohort (total 28,938). Compared risk of acute myocardial infarction (AMI), acute heart failure (AHF), or all-cause death among pioglitazone- and rosiglitazone-treated patients in a managed-care population.
Outcomes and results
Primary outcome measure was time to composite event of AMI, AHF or death among pioglitazone- and rosiglitazone-treated patients. Cox proportional hazards model was used to evaluate effects of exposure to rosiglitazone and pioglitazone on time to event.
Of the rosiglitazone-treated patients, 602 (4.16%) had an AMI, AHF, or death compared with 599 (4.14%) propensity score–matched pioglitazone-treated patients (hazard ratio 1.03; 95%CI 0.91 to 1.15).
There were no significant differences in individual risk for AMI (0.66% with rosiglitazone vs. 0.84% with pioglitazone), AHF (1.83% with rosiglitazone vs. 1.68% with pioglitazone) or death (0.17% with rosiglitazone vs. 0.12% with pioglitazone) between groups.
In a sub-population of patients over 65 years of age, crude event rates were much higher, but there was still no significant differences between groups. Of the rosiglitazone treated patients over 65 years of age, 355 (13.88%) had an had an AMI, AHF, or death compared with 393 (13.94%) of pioglitazone-treated patients (hazard ratio 0.97; 95%CI 0.83 to 1.12).
This was an internally funded study (data from HealthCore/WellPoint) with no external support.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you