Sci Rep. 2020 Sep 25;10(1):15781. doi: 10.1038/s41598-020-72967-8.
The safety and usefulness of pioglitazone (Pio) is repeatedly called into question due to the contradictory information available about it. A meta-analysis and risk benefit assessment was conducted to address the various points of debate regarding Pio. Electronic database search (Cochrane library, Embase & PubMed) resulted in 10 citations eligible for this meta-analysis (prospective, randomised studies), which was conducted using CMA software version 3 (Biostat Inc., Englewood, NJ, USA). The meta-analysis was registered with PROSPERO (ID: CRD42019122403) and compared pioglitazone with a control (anti-hyperglycemic agents without pioglitazone) in patients with either established cardiovascular disease or having high cardiovascular risk. Sensitivity and subgroup analysis were conducted to differentiate the effect of Pio against active controls and placebo. The use of Pio compared to the control group that did not use Pio resulted in a 14% and 23% significant reduction in odds of major adverse cardiac events (MACE: MH-OR, 0.86; 95% CI 0.75-0.98), and stroke (MH-OR, 0.77; 95% CI 0.60-0.99), respectively. This reduction in stroke was not significant in comparison to placebo on subgroup analysis. However, Pio significantly increased odds of heart failure (HF) (MH-OR, 1.47; 95% CI 1.26-1.71) as well as hospitalization for heart failure (hHF) (MH-OR, 1.48; 95% CI 1.21-1.81). In addition, the use of Pio was associated with a significant increase in odds of fractures in women (MH-OR, 2.05; 95% CI 1.28-3.27) and anaemia (MH-OR, 2.56; 95% CI 1.55-4.21). Pio has no significant effect on bladder cancer nor macular oedema. Pio has salutary effects on MACE. The positive effects are completely offset by the harm they seem to cause by way of heart failure, fractures, and anaemia. Pio should therefore be reserved for treatment of T2D with high CV risk or established cardiovascular (CV) disease only in selected patients where other antidiabetics are precluded and not routinely.