Have we over reacted about Pio?

I'm going to apologise in advance because I'm about to start banging on about one of my old favourites: TZDs. They're a highly attractive group of drugs for treating T2DM.

The PPARgamma receptor is stimulated, adipogenesis ensues, OK the patient puts on a bit of weight, but much more importantly, they are able to mop up all those extra calories in nice safe little adipocytes rather than depositing ectopic triglyceride in the liver, pancreas and muscles where it will drive insulin resistance.

What’s not to like?

Alright back in 1997, Troglitazone had to be withdrawn because of several cases of fatal hepatic necrosis. Admittedly, this wasn’t a brilliant start to a new drug class’s life but lactic acidosis with Fenformin hasn’t been the death nell of Metformin, on the contrary, Metformin is most diabetologist’s favourite drug.

It was also a shame that Rosiglitazone didn’t come out of the Nissen 2007 NEJM meta analysis smelling of roses, but perhaps the real mistake there was GSK not sharing everything they had on file with the FDA. Of course, increasing the risk of post menopausal osteoporotic fracture with the glitazones is also a problem, but, does PIO really increase the risk of bladder cancer?

At the EASD in Berlin a fortnight ago Dr Peter Boyle from the international Prevention Research Institute in France gave a fantastic overview of the epidemiology of diabetes and cancer. Perhaps most reassuring for many in the audience was the firm conclusion that Insulin Glargine appears not, after all, to be associated with increased risk of breast, or indeed a whole range of other, cancers.

Indeed, after much careful dissection of the epidemiology of cancer and T2DM, one of Dr Boyle’s firm conclusions was that incidence of a whole range of cancers rises in the first year after diagnosis with T2DM and then this slowly reverts to baseline over the next few years and it appears that the increased incidence in the first year of diabetes is probably a form of ascertainment bias brought about by increased medical attention.

So, although some reports have linked PIO with a 1.39-fold increase in risk of bladder cancer, not all studies have consistently shown this affect and there is a risk that the epidemiologic signal in the data might just be a manifestation of the “first year of diabetes increases the risk of many cancer diagnoses” phenomena.

Should we continue to use PIO? The MHRA obviously thinks that where we judge the potential benefits to outweigh the risks the answer may still be yes. Do I agree with this approach? Absolutely. But there’s a bigger lesson to be learnt here; association is not causation, when Insulin Glargine initially looked to be associated with breast cancer we were worried, but fortunately we didn’t panic. A year or so later and the facts seem to be much clearer. Perhaps, the sole remaining TZD might be a victim of the same phenomena?

Tags for this article: Pioglitazone

The author - Dr Jeremy Turner

Jeremy Turner A consultant diabetologist and endocrinologist in Norfolk, and author of Diabetes Bible

A bit more about the blog...

My blog expresses my personal views on the rapidly advancing field of diabetes. It is aimed at fellow physicians and is not offering medical advice to readers. I will not respond to requests for clinical advice. If you have health concerns please contact your GP or specialist.

Recent Articles


Diabetesbible is for health professionals