Pioglitazone was linked to a 63% increase in bladder cancer risk, compared with other noninsulin antidiabetic drugs in a large population-based cohort study.

Furthermore, incidence rose the longer patients took pioglitazone and the higher their dose, according to Dr. Marco Tuccori at Jewish General Hospital in Montreal, and his associates. “Rosiglitazone was not associated with an increased risk of bladder cancer in any analysis, suggesting the risk is drug-specific and not a class effect,” they wrote online March 30 in BMJ.

Both pioglitazone and rosiglitazone are thiazolidinediones that help treat insulin resistance in patients with type 2 diabetes. In 2005, Proactive trial investigators reported that patients were more likely to develop bladder cancer on pioglitazone, compared with placebo. But later observational studies reported mixed associations between pioglitazone and bladder cancer, perhaps because of biases and other methodological shortcomings, the researchers said (BMJ. 2016. doi: 10.1136/bmj.i1541).

Dr. Tuccori and his colleagues studied 145,806 patients with newly diagnosed type 2 diabetes, defined as a first-ever prescription for thiazolidinediones, metformin, sulfonylureas, prandial glucose regulators, acarbose, dipeptidyl peptidase-4 inhibitors, glucagonlike peptide agonists, or sodium-glucose cotransporter-2 inhibitors. The investigators excluded patients with any history of bladder cancer or less than a year of follow-up, and set a lag period of 1 year to help control for bladder cancer cases that were undiagnosed at study entry. The study included data for 2000-2013 from a national primary care registry that correlated well with the national cancer registry, the investigators said.

In all, 622 patients were diagnosed with bladder cancer (90 cases per 100,000 person-years), with a median of 4.4 years between study entry and cancer diagnosis (interquartile range, 2.5-6.5 years). Pioglitazone was associated with 121 cases of bladder cancer for every 100,000 person-years, compared with only 89 cases per 100,000 person-years for other antidiabetic drugs (hazard ratio, 1.6; 95% confidence interval, 1.2-2.2). Patients were no more likely to develop bladder cancer on rosiglitazone than other antidiabetic drugs (86.2 and 88.9 cases per 100,000 person-years of exposure, respectively; HR, 1.1; 95% CI, 0.8-1.5). Furthermore, pioglitazone was linked to nearly a 50% greater risk of bladder cancer, compared with rosiglitazone in a head-to-head comparison.

Not only did the link between pioglitazone and bladder cancer hold up in nine sensitivity analyses, but the hazard ratio increased to 1.73 when the researchers increased the required lag period to 2 years. Pharmacologic differences between pioglitazone and rosiglitazone might explain their distinct risk profiles in terms of bladder cancer, the investigators noted. Rosiglitazone is selective for peroxisome proliferator-activated receptor (PPAR) gamma, while pioglitazone has dual PPAR alpha/gamma activity. Only the latter increased expression of carcinogenic bladder biomarkers in rat models, they noted. This mechanism needs more exploration, they added.

The Canadian Institutes of Health Research funded the study. The researchers had no disclosures.