Loading ...

TOPLINE:

In a network meta-analysis of 46 observational studies involving more than 7 million patients with type 2 diabetes (T2D), the use of thiazolidinediones was most strongly associated with a reduced risk for hepatocellular carcinoma (HCC). The use of GLP-1 receptor agonists (GLP-1 RAs) showed the strongest association with the reduced risk for decompensation of cirrhosis, whereas the use of SGLT2 inhibitors was most strongly associated with protection against progression to cirrhosis.

METHODOLOGY:

• T2D has been recognized as a key cardiometabolic risk factor contributing to liver disease; however, the comparative effects of different antidiabetic drug classes on major adverse liver outcomes remain unclear.
• Researchers conducted a systematic review and network meta-analysis to compare the associations between the use of different antidiabetic drug classes and major adverse liver outcomes in patients with T2D.
• The analysis included 46 observational studies involving 7,124,845 adults with T2D (mean age, 55.7 years; 33.7% male), comprising 77 comparisons across seven drug classes: metformin, GLP-1 RAs, SGLT2 inhibitors, DPP-4 inhibitors, thiazolidinediones, sulfonylureas, and insulin.
• The primary outcome was the occurrence of three major adverse liver events: incident HCC, hepatic decompensation (a composite of spontaneous bacterial peritonitis, ascites, variceal bleeding, or hepatic encephalopathy), and incident cirrhosis, each assessed as time-to-event endpoints.
• Secondary outcomes included variceal bleeding, hepatic encephalopathy, development of ascites, hepatic failure (including coagulopathy and multiorgan failure), the need for liver transplantation, and liver-related death.

TAKEAWAY:

• In 38 studies including 61 comparisons, thiazolidinediones ranked highest for reducing the risk for HCC (surface under the cumulative ranking curve [SUCRA], 92%), with their use associated with a lower incidence of HCC than the use of DPP-4 inhibitors (hazard ratio [HR], 0.50), GLP-1 RAs (HR, 0.71), insulin (HR, 0.19), and sulfonylureas (HR, 0.69).
• In 25 studies including 44 comparisons, GLP-1 RAs ranked highest for reducing the risk for decompensation of cirrhosis (SUCRA, 100%), and their use was associated with fewer hepatic decompensation events than the use of DPP-4 inhibitors (HR, 0.73), insulin (HR, 0.16), metformin (HR, 0.86), SGLT2 inhibitors (HR, 0.90), sulfonylureas (HR, 0.66), and thiazolidinediones (HR, 0.75). GLP-1 RAs also ranked lowest for the risk for variceal bleeding (SUCRA, 79%; 11 studies) and hepatic encephalopathy (SUCRA, 80%; nine studies).
• SGLT2 inhibitors ranked highest for preventing progression to cirrhosis (SUCRA, 88%; 12 studies comprising 22 comparisons), and their use was associated with a lower risk of developing cirrhosis than the use of insulin (HR, 0.45), DPP-4 inhibitors (HR, 0.66), and GLP-1 RAs (HR, 0.66). They also ranked highest for preventing liver-related death (SUCRA, 73%; five studies).
• The use of insulin was consistently associated with the worst outcomes for both the development of HCC and the decompensation of cirrhosis (SUCRA < 1% for both).

IN PRACTICE:

“Antihyperglycemic agents are not equivalent with respect to [major adverse liver outcomes], and that consideration of hepatic endpoints may meaningfully complement established cardiorenal criteria when selecting therapy in patients at risk for progressive liver disease,” the authors of the study wrote.

SOURCE:

The study was led by Pedro Robson Costa Passos, Federal University of Ceará, Fortaleza, Brazil. It was published online in Diabetes Care.

LIMITATIONS:

All included studies were observational, limiting causal inference. The use of similar databases likely resulted in some duplication of patient populations. Studies varied widely in baseline liver disease status, comorbidities, outcome definitions, and covariate adjustments. Additionally, most studies lacked data on drug dose, adherence, or treatment duration.

DISCLOSURES:

The study was supported by the Medical Research Council and the National Institute for Health and Care Research Oxford Biomedical Research Centre. The authors declared having no relevant conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.