Thiazolidinediones for people with chronic kidney disease and diabetes

Type 2 diabetes is a risk factor for cardiovascular disease and chronic kidney disease (CKD), and it affects quality of life and contributes to substantial costs for healthcare systems. Approximately a third of people with type 2 diabetes develop CKD. Thiazolidinediones are associated with improved glucose management and a lower risk of progression to kidney failure, requiring long-term dialysis. However, evidence of safety in people with CKD and type 2 diabetes is still lacking and is based on low-certainty studies.

This review aims to assess the benefits and harms of thiazolidinediones in people with CKD and type 2 diabetes.

The Cochrane Kidney and Transplant Register of Studies was searched up to October 2024 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and Embase, conference proceedings, the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Randomised controlled studies were eligible if they compared treatment with thiazolidinediones against placebo, usual medical care or another glucose-lowering agent in people with CKD and type 2 diabetes. We included all CKD stages (from 1 to 5), including people treated with dialysis.

The critical outcomes included cardiovascular death and severe hypoglycaemia. The important outcomes included death (any cause), 3-point and 4-point major cardiovascular events (MACE), non-fatal myocardial infarction (MI), nonfatal stroke, kidney failure requiring kidney replacement therapy (KRT), and adverse events.

Three authors independently extracted data and assessed the risk of bias using the Risk of Bias 2 tool.

Pooled analyses using summary estimates of effects were obtained using a random-effects model, and results were expressed as risk ratios (RR) and their 95% confidence intervals (CI) for dichotomous outcomes and mean difference (MD) and 95% CI for continuous outcomes. Confidence in the evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.

Eighty-five studies involving 3044 participants were included. The median age was 62 years, and the median follow-up was 24 weeks; one study included children. Five studies were conducted in people with CKD stages 1 and 2, six studies in people with CKD stages 3 to 5, seven studies in people on dialysis, and the remaining studies included people with both CKD stages 1 and 2 and CKD stages 3 to 5. Overall, the risks of bias in the included studies for all critical outcomes in studies that compared thiazolidinediones to placebo or standard care alone were high due to concerns about blinding. The overall risk of bias for cardiovascular death reported in studies that compared thiazolidinediones to placebo or standard care alone was assessed as low or uncertain, except for two studies that were considered at high risk of bias due to concerns about lack of blinding, whilst other domains seemed to be free from other sources of bias. Two studies reported severe hypoglycaemia: one study was assessed as unclear risk of bias for all the risk of bias domains, while the other study was considered at high risk of bias due to concerns about lack of blinding. Other critical outcomes, including 3-and 4-point MACE, non-fatal MI, non-fatal stroke, or kidney failure requiring KRT were not reported in the included studies.

Compared to placebo or standard care alone, thiazolidinediones may have little or no effect on the risk of cardiovascular death (RR 0.20, 95% CI 0.01 to 3.97; 4 studies, 226 participants; low-certainty evidence), whilst the effect of severe hypoglycaemia was not estimable (2 studies, 107 participants) in people with all CKD stages and type 2 diabetes. No study evaluated the effects of 3- and 4-point MACE, non-fatal MI, non-fatal stroke, or kidney failure requiring KRT. Compared to sulphonylureas, thiazolidinediones may have little or no effect on cardiovascular death (RR 2.78, 95% CI 0.11 to 67.92; 4 studies, 483 participants; low-certainty evidence) in people with CKD stages 1 and 2 and type 2 diabetes, but no study evaluated the effects of severe hypoglycaemia, 3- and 4-point MACE, non-fatal MI, non-fatal stroke, or kidney failure requiring KRT. The effects of thiazolidinediones compared to other hypoglycaemic agents or different doses of thiazolidinediones were uncertain. The risk of bias in the included studies was high or unclear, leading to low to very low-certainty evidence.

Thiazolidinediones may have little or no effect on the risk of cardiovascular death, whilst the effects of severe hypoglycaemia or other cardiovascular and kidney outcomes were uncertain in people with CKD and type 2 diabetes. The effects of thiazolidinediones in people with CKD and type 2 diabetes are insufficient to provide firm conclusions. Future studies will address the benefits and harms of thiazolidinediones in this setting, reporting the core outcomes as prioritised by stakeholders to better inform decision-making.

No funding was received for the conduct of this review.

Protocol (2023): https://doi.org/10.1002/14651858.CD015907.