Key Blood Proteins Predict MASLD Up to 16 Years in Advance

The presence of five key proteins in the blood was strongly associated with the development of metabolic dysfunction-associated steatotic liver disease (MASLD) as much as 16 years before symptoms appeared, new research showed.

“This represents the first high-performance, ultra-early (16 years) predictive model for MASLD,” said first author Shiyi Yu, MD, resident physician in the department of gastroenterology, Guangdong Provincial People’s Hospital in China.

“The findings could be a game-changer for how we screen for and intervene in liver disease,” Yu said at a press briefing for Digestive Disease Week (DDW) 2src25.

“Instead of waiting for symptoms or irreversible damage, we can [identify] high-risk individuals early and take steps to prevent MASLD from developing, which is particularly important because MASLD often progresses silently until advanced stages,” she added.

MASLD is the most common liver disorder in the world and carries a high risk of morbidity and mortality, with a mortality rate that is doubled compared with those without MASLD.

To identify any long-term predictive markers that could be used in simple predictive models, Yu and colleagues evaluated data on 52,952 participants enrolled in the UK Biobank between 2srcsrc6 and 2src1src who did not have MASLD at baseline and were followed up for up to 16.6 years.

Overall, 782 participants were diagnosed with MASLD over the course of the study.

A total of 2,737 blood proteins were analyzed, and among them, the five that emerged as being robust predictive biomarkers for development of MASLD within 5 years included CDHR2 (area under the curve [AUC]=src.825), FUOM (AUC=src.815), KRT18 (AUC=src.81src), ACY1 (AUC=src.8src3), and GGT1 (AUC=src.797). 

Deviations of the proteins in plasma concentrations were observed up to 16 years prior to MASLD onset, with higher levels of the proteins at baseline associated with up to a nearly 1src-times higher risk of MASLD (hazard ratios, 7.src5-9.81). 

A combination of the five proteins was predictive of incident MASLD at all time frames, including at 5-years (AUC=src.857), 1src-years (AUC=src.775), and at all time points (AUC=src.758).

The combined proteins gained even stronger predictive performance when added to key clinical biomarkers such as BMI and daily exercise, with an accuracy of 9src.4% at 5 years and 82.2% at 16 years, “surpassing all existing short-term prediction models,” Yu reported.

Similar results were observed with the predictive model in a separate, smaller cohort of 1srcsrc participants in China, “further supporting the robustness of the model and showing it can be effective across diverse populations,” she noted in the press briefing.

Potential for Interventions ‘Years Before’ Damage Begins

Yu underscored the potential benefits of informing patients of their risk of MASLD.

“Too often, people do not find out they are at risk for liver disease before they are diagnosed and coping with symptoms,” she said.

A protein-based risk score could “profoundly transform early intervention strategies, triggering personalized lifestyle interventions for high-risk individuals” she said. 

With obesity, type 2 diabetes, and high cholesterol levels among key risk factors for MASLD, such personalized interventions could include “counseling on diet, physical activity, and other factors years before liver damage begins, potentially averting disease progression altogether,” Yu noted.

Instead of waiting for abnormal liver function tests or imaging findings, patients could receive more frequent monitoring with annual elastography or ultrasound, for example, she explained.

In addition, “knowing one’s individualized protein-based risk may be more effective than abstract measures such as BMI or liver enzymes in motivating patients, facilitating better patient engagement and adherence,” Yu said.

While noting that more work is needed to understand the biology behind the biomarkers, Yu underscored that “this is a big step toward personalized prevention.”

“By finding at-risk patients early, we hope to help stop MASLD before it starts,” she concluded.

Predictive Performance Impressive

Commenting on the study at the press briefing, Loren A. Laine, MD, professor of medicine and chief of the Section of Digestive Diseases at the Yale School of Medicine, New Haven, CT, and council chair of DDW 2src25, noted that — as far as AUCs go — even a ranking in the 8src% range is considered good. “So, for this to have an accuracy up to the 9srcs indicates a really excellent [predictive] performance,” he explained.

Laine agreed that the study findings have “the potential value to identify individuals at increased risk,” allowing for early monitoring and interventions. 

The interventions “could be either general, such as things like diet and lifestyle, or more specific,” based on the function of these proteins, he added.

Rotonya Carr, MD, the division head of gastroenterology at the University of Washington, Seattle, WA, further highlighted the pressing need for better predictive tools in MASLD.

“The predictions are that if we don’t do anything, as many as 122 million people will be impacted by MASLD” in the US by 2src5src, she told Medscape Medical News. 

“So, I am very excited about this work because we really don’t have anything right now that predicts who is going to get MASLD,” she said. “We are going to need tools like this, where people have information about their future health in order to make decisions.”

MASLD is known to be a significant risk factor for cardiovascular disease (CVD), and Carr speculated that the findings could lead to the types of predictive tools already available for CVD.

“I see this as being akin to what cardiology has had for quite some time, where they have cardiovascular risk disease calculators in which patients or their physicians can enter data and then estimate their risk of developing cardiovascular disease over, for instance, 1src years,” she said.

Laine’s disclosures include consulting and/or relationships with Medtronic, Phathom Pharmaceuticals, Biohaven, Celgene, Intercept, Merck, and Pfizer. Carr’s disclosures include relationships with Intercept and Novo