Two of the most commonly used drugs in the management of type 2 diabetes, rosiglitazone and pioglitazone (brand names Avandia and Actos), have been linked to an increased risk of heart failure, The Independent reported. Taking one of the two drugs, prescribed for type 2 diabetes sufferers, could “double the risk of a heart failure”, it said on July 27 2007.
The Daily Telegraph quoted the researchers as saying that one in 50 patients taking the drug over a period of 26 months would “suffer heart failure and need admission to hospital”. Use of these drugs has apparently doubled in the past three years, and they say that this research calls for a rethink of the regulation of these drugs by authorities.
The original research was carried out to investigate the risk of heart failure with use of the thiazolidinedione class of drugs – rosiglitazone and pioglitazone – and to see how this changes with dose and duration of use. Questions are raised about the safety of these drugs, but the results of this research need careful interpretation, and balancing against the benefits that these drugs bring to people with diabetes.
Where did the story come from?
Sonal Singh and Curt Furberg of Wake Forest University Medical School, North Carolina, USA and Yoon Loke of University of East Anglia, England carried out this research and it was published in the peer reviewed journal, Diabetes Care.
What kind of scientific study was this?
The researchers collected information from randomised controlled trials (RCTs), observational studies, case series', case reports, and reports from the Canadian Drug Reaction Monitoring Program.
These reports were then combined using teleo-analysis. This is described by the authors as a method that “attempts to determine the adverse effect of a drug by complementing information from different study designs across all grades of evidence”.
To identify the RCTs, the authors looked for potential studies that were referenced in previous research papers, and searched a single database (PubMed) between January 2003 to September 2006 to find recent research that had looked at patients taking either of the two thiazolidinedione drugs for longer than 6 months, where they had been compared with an inactive placebo drug and there was information available on the number of people who experienced heart failure.
The researchers also searched PubMed for observational studies and case reports containing information about diabetic patients taking these drugs who developed heart failure compared with the use of other drugs to treat type 2 diabetes.
This search obtained three RCTs, four observational studies and 162 single patient cases. Computer methods were used to calculate a combined odds ratio, a statistical measure of the risk of developing heart failure while taking the drugs, for the three RCTs and four observational studies. From the individual patient case reports, the researchers combined results to look at the average time that it took for the patient to develop heart failure after starting the drugs.
What were the results of the study?
The researchers calculated from the combined RCTs that a person was just over twice as likely (110% increased risk) to develop heart failure if taking one of the thiazolidinedione drugs compared with an inactive placebo drug. From the observational studies, they calculated that a person was at a 50% greater risk of heart failure. To further clarify the risk, they calculated that over a period of 2.2 years, if 50 people were treated with the drugs, one would be affected by heart failure.
The individual case studies and reports from the Canadian Drug Reaction Monitoring Program found that the average time to development of heart failure from time of starting a thiazolidinedione drug was 24 weeks, regardless of the dose taken.
What interpretations did the researchers draw from these results?
The researchers conclude from their teleo-analysis that there is an increased risk of heart failure when either of the thiazolidinedione drugs, rosiglitazone or pioglitazone, are taken. They suggest that practice guidelines and drug packaging information acknowledge this risk.
What does the NHS Knowledge Service make of this study?
This research raises several questions over the risks of taking the thiazolidinedione drugs. However, when considered alone, this study has a few drawbacks which the authors themselves acknowledge:
- These study results cannot prove the thiazolidinediones as the cause of heart failure. Numerous other medical conditions could have caused the condition in a patient who happened to be taking these diabetic drugs, e.g. recent heart attack. Also other drugs that have not yet been considered could contribute a risk (the studies in this report compared the drugs to an inactive placebo drug only, not to another active drug).
- This research has combined data from several studies of different size and design. Different classifications of heart failure were used, for example, sometimes patients considered as new cases of heart failure by one study may not have been considered the same by another study.
- The benefit of not taking the drug in reducing the risk of heart failure would need to be carefully weighed against the risk to health of not taking the drugs for type 2 diabetes.
- The reliability of this method of combining several study types, and the methods used for retrieving the studies (e.g. using only one computer database) may be questioned. For example, the case reports that were identified may have been subject to publication bias, for example, only those showing a harmful effect were published.
However, the fact that these studies have shown a possible link between certain diabetic drugs and the risk of heart failure cannot be excluded, and suggests that further research is needed into this area of medical care.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Independent, 27 July 2007
Daily Mail, 27 July 2007
The Times, 27 July 2007
BBC News, 27 July 2007
The Daily Telegraph, 27 July 2007
The Guardian, 27 July 2007
Links to the science
Diabetes Care 2007; 30:2148-2153.