A UK population-based cohort study, published in the BMJ, concludes that pioglitazone is associated with an increased risk of bladder cancer in patients with type 2 diabetes.

Using data from the UK Clinical Practice Research Datalink, 145,806 patients newly treated with diabetes drugs between 1 January 2000 and 31 July 2013 were followed until the end of July 2014. The cohort was followed for a mean of 4.7 years and resulted in 689,616 person-years of follow-up. The main outcome measure was incident bladder cancer.

A total of 622 patients were diagnosed with bladder cancer during the study. Statistical modelling was used to estimate hazard ratios (HR) for bladder cancer with adjustment for a large number of confounding factors.

Compared with no thiazolidinedione use, pioglitazone was associated with an increased risk of bladder cancer: 121 vs 88.9 per 100,000 person-years; HR 1.63 (95% CI: 1.22-2.19).

A duration response relation was also seen (P < 0.01 for trend) with use of pioglitazone for more than two years associated with a raised risk of bladder cancer: HR 1.78 (95% CI: 1.21-2.64). Rosiglitazone was not associated with a raised risk of bladder cancer: 86.2 vs 88.9 per 100,000 person-years; HR 1.1 (95% CI: 0.83-1.47). Similarly no duration response relation was seen.

The authors acknowledge that there may be sources of error from retrospective data analysis. Confounding factors such as family history, ethnicity, occupational exposure, diet and exercise may not be controlled for, and the quality of data input may vary. However, the clear difference between the two drugs of the same class suggests a true association between pioglitazone and bladder cancer.

‘Additional studies are needed to better understand the biological mechanism behind the possible pioglitazone specific effect on the bladder,’ the study authors conclude.

Although most GPs have been aware of the possible link, and have acted to counsel patients on the drug, the research evidence linking pioglitazone with bladder cancer has been contradictory.

It is not easy for frontline GPs to review the evidence for themselves when the debate involves statistical analyses of huge retrospective data sets. In practice, we should ensure that any patients who still take pioglitazone are aware of the continuing debate about the risks of bladder cancer, and leave the drug for specialist initiation only. We should continue to be vigilant for bladder cancer in patients who have ever taken pioglitazone.

Dr Matthew Lockyer