MASH/NASH: Why Therapeutic Breakthroughs Demand a New Era of Patient Identification
By Léon Van Wouwe, Clinical Innovation Director, Volv Global
Metabolic dysfunction–associated steatohepatitis (MASH), historically known as non-alcoholic steatohepatitis (NASH), has rapidly shifted from an obscure hepatology concern to one of the fastest-growing causes of advanced liver disease worldwide. Around one-third of adults globally are estimated to have metabolic dysfunction–associated steatotic liver disease (MASLD/NAFLD), with a meaningful subset progressing to inflammatory, fibrotic MASH. [1,2]
Prevalence is even higher in at-risk groups such as people with obesity or type 2 diabetes, where NAFLD prevalence exceeds 50% in many cohorts. [3] MASH has become a leading contributor to liver transplantation and liver-related morbidity, particularly among women, and is projected to increase further over the next decade. [4]
Yet despite its clinical importance, MASH remains chronically underdiagnosed. Most patients are asymptomatic or present with non-specific features such as fatigue or mildly abnormal liver enzymes. [5–7] Diagnostic pathways frequently depend on opportunistic discovery (e.g., incidental imaging findings) and invasive liver biopsy in selected cases. Typical patterns include:
These delays are not trivial. Fibrosis stage is the strongest predictor of liver-related outcomes in MASH; once advanced fibrosis or cirrhosis is established, the opportunity for disease modification narrows. [1,3,5]
The typical MASH patient begins their journey unknowingly. The spectrum of metabolic dysfunction–associated steatotic liver disease runs from simple steatosis through MASH to progressive fibrosis, cirrhosis, and hepatocellular carcinoma (HCC). [3] Most patients traverse this trajectory without clear clinical signals until late stages.
Importantly, disease progression is heterogeneous: some individuals remain stable for years, while others progress rapidly to decompensated cirrhosis or HCC within a relatively short timeframe. [3,8] Real-world analyses show that a substantial minority of patients with MASH develop features of end-stage liver disease over just a few years of follow-up, underscoring the unpredictability and systemic burden of the condition. [8]
From a health-system perspective, the patient journey is misaligned with therapeutic opportunity:
Within this journey, outcome prediction becomes as important as diagnosis itself. Clinicians are being asked to decide:
Emerging work using data-driven clustering and artificial intelligence has shown distinct clinical phenotypes associated with rapid fibrosis progression in MASH. [9] This underscores a central message for this series: we must evolve from simple detection to trajectory-aware, prognosis-informed care.
The approval of resmetirom (Rezdiffra™) marked a watershed in the management of this disease. In March 2024, the US Food and Drug Administration (FDA) approved resmetirom as the first pharmacologic treatment for adults with noncirrhotic NASH/MASH with moderate to advanced liver scarring (F2–F3), to be used alongside diet and exercise. [10,11]
Resmetirom, a liver-directed thyroid hormone receptor-β agonist, has shown clinically meaningful improvements in steatohepatitis and fibrosis markers in pivotal trials, and regulators in Europe have since recommended conditional approval, signalling global momentum. [10,12,13]
At the same time, GLP-1–based and other investigational therapies are entering or expanding within the MASH space, including assets from Novo Nordisk, Eli Lilly, 89bio, Akero Therapeutics, Boehringer Ingelheim, and Boston Pharmaceuticals, among others. Wegovy (semaglutide) has recently received an additional indication in the US for MASH with fibrosis, becoming the second approved therapy in this setting. [14]
This is a striking moment: the therapeutic bottleneck is beginning to lift. But these innovations immediately expose the fragility of current identification practices:
To realise the promise of resmetirom and the broader pipeline, diagnostic prediction and outcome prediction must evolve in lockstep with therapeutics.
Machine-learning approaches have demonstrated that routinely collected data – demographics, comorbidities, laboratory values, medication patterns – can be used to identify patients with likely undiagnosed MASH/NASH from large at-risk populations. [15–17] Models trained on claims and electronic health records have successfully flagged patients for further diagnostic work-up, reducing reliance on chance discovery.
Advanced imaging and serum biomarkers are now being combined with AI-based tools to refine fibrosis staging. Regulatory bodies have even endorsed AI-supported histologic assessment (e.g., AIM-NASH) to reduce variability in biopsy interpretation and improve the reliability of treatment effect assessment in trials, a paradigm that can ultimately inform clinical practice. [18]
Beyond “does this patient have MASH?”, clinicians increasingly need tools that answer “how will this patient’s disease behave?” AI-driven models that classify patients into phenotypes with different risks of rapid fibrosis progression and adverse outcomes are beginning to emerge. [8,9,13] These models can underpin decisions such as:
For health systems, the long-term value of resmetirom and pipeline agents will depend on a precision hepatology infrastructure capable of:
In other words, therapeutic innovation and predictive innovation must be co-designed. Without earlier detection and robust risk stratification, even the most effective MASH therapies will be deployed too late, in too few patients, to meaningfully alter the population-level trajectory of liver disease.
MASH is poised to become one of the defining metabolic liver diseases of this century, with substantial clinical and economic burden. [1,2,4] The arrival of resmetirom and subsequent therapies has shifted the narrative from “we have nothing to offer” to “we must decide whom to treat, when, and how.”
If diagnosis continues to occur late and opportunistically, these therapies will be reserved for advanced disease that might otherwise have been prevented or attenuated. But if we build a prediction-enabled ecosystem – integrating AI-based early detection, non-invasive fibrosis staging, and outcome prediction into routine care – then resmetirom and its peers can fundamentally change the natural history of MASH.
The central theme of this series [LC1] is clear in MASH more than almost any other disease area:
Reflective Questions
Where in the MASH patient journey could predictive modelling create the largest near-term impact: early detection in at-risk populations, fibrosis staging, or treatment response prediction?
What policy, reimbursement, and workflow barriers currently prevent integration of non-invasive diagnostics and AI-enabled outcome prediction into routine metabolic and primary care pathways?
Léon van Wouwe has 20+ years’ global experience in clinical development and operations, uniting data science with pharma and research. He drives cross-functional collaboration to advance innovative treatments.
Photo by LPETTET on iStock.
Tech Tour Growth Health is one of Europe's most established gatherings at the intersection of…
Volv Global is pleased to be participating in the Life Sciences Innovation & Investment Corridor…
The idea that an AI system could detect a disease that appears nowhere in a…
The ISPOR International Annual Conference 2026 brings together global leaders in health economics and outcomes…
By Léon Van Wouwe, Clinical Innovation Director, Volv Global Inflammatory Bowel Disease (IBD) management has…
Introduction Advocacy plays a critical role in accelerating progress in rare diseases. As technology advances,…