Glitazones may be beneficial post-PCI

Clinical Question

Is treatment with a glitazone after angioplasty effective in decreasing the need for revascularization?

Bottom Line

The use of a thiazolidinedione (glitazone) following percutaneous coronary intervention (PCI) will decrease the need for revascularization over the following 6 months. The effect occurs in patients with and without diabetes and is likely due to an effect other than on blood glucose control. The studies in this meta-analysis were small, and further research is necessary before putting these results into practice. (LOE = 1a)


Riche DM, Valderrama R, Henyan NN. Thiazolidinediones and risk of repeat target vessel revascularization following percutaneous coronary intervention. Diabetes Care 2007;30:384-88.  [PMID:17259517]

Study Design

Meta-analysis (randomized controlled trials)




Various (meta-analysis)


In addition to improving insulin sensitivity, the glitazones may have an anti-inflammatory effect and may inhibit vascular smooth muscle cells. On the basis of these findings, several studies have been performed to evaluate their effectiveness following PCI in patients with and without diabetes. After a complete search of several databases, citation lists, and abstracts from major meetings, the researchers identified 7 studies for inclusion, evaluating 608 patients, which is small by cardiology standards. The researchers mention that they evaluated the quality of the research, but they did not include this analysis in their report. The use of stents or type of stents with PCI is not standardized in the studies. All studies compared the use of a glitazone with placebo, starting before or shortly after surgery, and used for 6 months. Five studies enrolled patients with diabetes, and most of the studies were conducted in an Asian population. Treatment decreased the need for revascularization from 21% to 6.6%. One reavascularization would be avoided for every 7 patients treated with a glitazone (number needed to treat = 6.9; 95% CI 5.3 - 12.5). Results were similar in the studies enrolling diabetic and nondiabetic patients. There was no heterogeneity among the studies and no evidence of publication bias.