Editorial: Beyond Clinical Trials—Real-World Data Suggest Usefulness of GLP-1 RAs in MASLD Treatment. Authors' Reply

Abstract

We sincerely appreciate the opportunity to respond to the invited editorial by Candels and Strnad discussing our study on the effectiveness of GLP-1 receptor agonists (GLP-1 RAs) in metabolic dysfunction-associated steatotic liver disease (MASLD) [1, 2]. We are grateful for their insightful commentary and for highlighting the strengths of real-world data in capturing long-term outcomes beyond the constraints of clinical trials. We acknowledge their concerns regarding the short follow-up period and potential bias from including patients with advanced hepatic fibrosis at baseline. To address these concerns, we conducted this retrospective study to evaluate the effect of GLP-1 RAs on MASLD patients with a lower fibrosis burden at baseline.

As in our previous studies [2-4], we utilised the TriNetX Research Network, a multinational and multi-institutional database. We identified patients with MASLD and type 2 diabetes (T2D) from this database. We included only those who had undergone abdominal ultrasound imaging (CPT 1010775) and had no prior diagnostic codes for hepatic fibrosis or cirrhosis (ICD-10 K74), representing a population with a lower likelihood of pre-existing significant fibrosis. We then identified patients with their first prescription for either a GLP-1 RA or a sodium-glucose cotransporter-2 inhibitor (SGLT2i) and conducted a propensity score-matched (PSM) cohort study to balance baseline characteristics between the two groups. We measured the risk of major adverse liver outcomes (MALOs), a composite endpoint consisting of decompensated cirrhosis events, hepatocellular carcinoma, and liver transplantation during the follow-up period.

After PSM, 5993 individuals remained in each of the GLP-1 RA and SGLT2i groups for outcome analyses, with a mean follow-up time of 34.9 months. Compared to SGLT2i new users, GLP-1 RA new users showed a significantly lower risk of MALOs (hazard ratio [HR], 0.77; 95% CI: 0.59–0.99, Table 1). The GLP-1 RA group also demonstrated significantly lower risks of all-cause mortality (HR, 0.72; 95% CI: 0.72–0.85) and decompensating events (HR, 0.78; 95% CI: 0.59–1.00). Additionally, the GLP-1 RA group showed non-significantly lower risks of variceal bleeding, hepatic encephalopathy, ascites-related complications, hepatocellular carcinoma, and liver transplant (Table 1).

TABLE 1. Adjusted hazard ratios of primary and secondary outcomes for the comparison of GLP-1 receptor agonists versus SGLT2 inhibitors.

Outcomes	No. of patients with outcome		Adjusted hazard ratio (95% CI)	p
	GLP-1 RA (n = 5993)	SGLT2i (n = 5993)
Primary outcome
Major adverse liver events	104	132	0.77 (0.59–0.99)	0.042
Secondary outcomes
Decompensated events	97	123	0.78 (0.59–1.00)	0.046
Variceal bleeding	< 10	< 10	0.84 (0.28–2.50)	0.75
Hepatic encephalopathy	34	47	0.70 (0.45–1.09)	0.116
Ascites-related complications	77	86	0.87 (0.64–1.19)	0.39
Hepatocellular carcinoma	< 10	< 10	0.97 (0.40–2.32)	0.94
Liver transplant	< 10	< 10	0.49 (0.09–2.66)	0.40
All-cause mortality	231	313	0.72 (0.61–0.85)	0.0001

These findings indicate the potential protective effects of GLP-1 RAs against MALOs in MASLD patients with a lower fibrosis burden. These additional analyses strengthen our initial findings and demonstrate that the beneficial impact of GLP-1 RAs extends beyond patients with advanced fibrosis to a broader MASLD population. This reinforces the clinical utility of GLP-1 RAs in MASLD management, complementing evidence from a randomised controlled trial [5] and supporting their consideration in hepatology practice [6].

In conclusion, we appreciate the opportunity to engage in this discussion and value the editorial's role in fostering discourse on this important topic. We believe our additional analyses provide clarity and reassurance regarding the applicability of our findings.