Long Term Prediction of MASH-related HCC Risk by Proteomics

Introduction MASH is a progressive liver disease, recognized as a severe form of MASLD, characterized histologically by hepatic steatosis accompanied by hepatocellular inflammation and ballooning degeneration. It represents a growing public health concern due to its strong association with metabolic syndrome components such as obesity, insulin resistance, type 2 diabetes mellitus, hypertension, and dyslipidemia. Globally, the prevalence of MASLD is estimated to be approximately 25%, with MASH affecting a substantial subset of these patients, particularly those with comorbid metabolic disorders.

MASH progression is heterogeneous and can lead to advanced fibrosis, cirrhosis, and hepatocellular carcinoma (HCC). Importantly, while cirrhosis is a well-known risk factor for HCC, recent epidemiological and clinical studies have demonstrated that HCC can also develop in non-cirrhotic livers affected by MASH. This phenomenon complicates clinical decision-making regarding surveillance and management, as current guidelines primarily recommend HCC screening in patients with cirrhosis regardless of etiology.

The pathogenesis of MASH and its progression to HCC involves complex interactions between metabolic derangements, chronic hepatic inflammation, oxidative stress, mitochondrial dysfunction, lipotoxicity, and activation of fibrogenic pathways. Excessive accumulation of free fatty acids and toxic lipid metabolites in hepatocytes induces cellular stress, triggering inflammatory cascades and apoptosis. Kupffer cells and recruited immune cells perpetuate the inflammatory milieu, contributing to progressive fibrosis.

The increasing incidence of MASH-related HCC underscores the urgent need for novel biomarkers capable of accurately stratifying patients according to their risk of HCC development. Early identification of high-risk individuals would allow for tailored surveillance strategies, early diagnosis, and timely therapeutic interventions, potentially improving clinical outcomes. Unfortunately, serological biomarkers, including alpha-fetoprotein (AFP), have limited sensitivity and specificity for early HCC detection in the MASH population. Thus, novel molecular markers reflecting the underlying pathophysiology of MASH and hepatocarcinogenesis are being actively investigated.

Proteomics, the large-scale study of proteins expressed by cells or tissues, offers unique advantages for biomarker discovery in MASH and HCC. Proteins are the functional effectors of cellular processes and may better reflect dynamic pathological changes compared to genomic or transcriptomic data alone.

Tissue proteomic profiling by mass spectrometry is an innovative and promising tool well-suited for the identification of prognostic proteomic signatures. This methodology was developed at the U1312 BRIC laboratory (team 3, Liver Cancers and Tumor Invasion) in collaboration with the Oncoprot platform (UMS 005-TBMcore). Starting from a formalin-fixed, paraffin-embedded (FFPE) tissue sample, a region of interest is selected and laser microdissected. Proteins are then extracted, fixation is reversed, and proteins are digested using trypsin. The resulting peptides are subsequently analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) on a high-resolution mass spectrometer. Data processing is performed by a bioinformatician: each sample is compared to a reference sample (e.g., healthy liver tissue), allowing calculation of relative protein expression intensities (protein abundance ratios). Two samples can then be compared using a triangulation quantification method, which involves comparing protein abundance ratios between the samples. All proteins exhibiting significantly different abundance ratios between the two samples constitute the proteomic profile.

Advances in mass spectrometry-based proteomic technologies now enable high-throughput, sensitive, and quantitative analysis of protein expression profiles from small amounts of tissue, including formalin-fixed paraffin-embedded (FFPE) samples routinely collected during diagnostic liver biopsies. This allows retrospective analyses of well-characterized patient cohorts and facilitates integration of proteomic data with clinical outcomes.

Study Design and Methodology This study is based on the retrospective analysis of a cohort of n=30 patients from Bordeaux University Hospital (CHU), with biological samples and clinical data already collected as part of routine care. Since the malignant transformation of MASLD (Metabolic Associated Steatohepatitis) is a long process (several years), this retrospective approach ensures the feasibility of the study and the rapid acquisition of results.

To address the stated hypotheses, this study will rely on the retrospective analysis of clinical data from patients in the hepatology department cohort at CHU Bordeaux (registered under ClinicalTrials.gov identifiers NCT01241227 and NCT02060565), as well as on proteomic analysis of liver biopsies from the same patients, available at the CHU Bordeaux Cancer Biobank (CRB). The Bordeaux CHU cohort currently includes over 7000 MASLD patients enrolled at diagnosis and followed prospectively to date. The first patient was included in 2003, providing up to 20 years of follow-up for early enrollees, making this cohort a valuable resource for studying the risk of malignant transformation.

Clinical data have been collected using a standardized procedure within an existing database, ensuring data quality and completeness for this project. Proteomic data from liver biopsies of patients included in the ProteoMASH study will be retrieved from a dedicated database established at Inserm U1312 and cross-referenced with clinical data from the CHU database (such as development of hepatocellular carcinoma [HCC] during follow-up, duration of follow-up since diagnosis, presence of diabetes, etc.), aiming to validate the diagnostic and prognostic value of these novel biological markers.

For this study, we will use biopsy fragments obtained during routine clinical care. Clinical data will be extracted from patient medical records within the service database; no additional procedures or follow-up visits will be imposed on patients, as follow-up data are collected as part of standard care.

Patients from the Bordeaux CHU cohort will be classified into two subgroups:

• Group 1 (n=15): patients with a liver biopsy diagnostic of MASLD prior to HCC development, with a time interval between biopsy and HCC diagnosis ranging from 1 to 15 years.
• Group 2 (n=15): control patients with a liver biopsy diagnostic of MASLD who did not develop HCC during follow-up.

Potential biases inherent to retrospective cohorts will be controlled as effectively as possible:

Control patients (Group 2) will be selected accounting for known HCC risk factors such as age, sex, presence and control of diabetes (defined by HbA1c <8% according to French guidelines), and fibrosis stage (histologically defined from F0: no fibrosis to F4: cirrhosis). The objective is to ensure these risk factors are similarly distributed between groups.

Additionally, we will ensure Group 2 patients have been followed for at least 5 years with imaging support (ultrasound, CT scan, or MRI) to confirm that no HCC could have been missed.

For Group 1, objective diagnostic criteria for HCC will be used, based on imaging (CT or MRI following LI-RADS criteria from the American College of Radiology) or histopathology (new biopsy or surgical resection), to minimize classification bias. Furthermore, to exclude the possibility of occult HCC at the time of liver biopsy, only patients with HCC diagnosed more than one year after biopsy will be included, with imaging (ultrasound, CT, or MRI) confirming absence of HCC at least one year post-biopsy and prior to HCC diagnosis.

Another potential bias relates to patients lost to follow-up (e.g., followed at other centers, deceased before 15 years without HCC development), who will be excluded from analysis and might have different HCC risk. Therefore, we cannot be certain that Group 2 patients fully represent all MASLD patients who do not develop HCC. However, this limitation is justified by the exploratory nature of the ProteoMASH study and will be considered in future projects.

Moreover, patients will be included over homogeneous time periods to avoid confounding due to changes in medical management over time.

Finally, routinely collected clinical data already available in the existing database (e.g., HbA1c values for diabetes control, histologically defined fibrosis stage) will be used to minimize missing data bias.

Liver FFPE biopsies will undergo optimized proteomic workflows and the resulting proteomic datasets will be analyzed using bioinformatics pipelines to identify proteins or panels differentially expressed between the HCC and non-HCC groups. Identified proteins will be annotated for their biological functions, pathways, and involvement in hepatocarcinogenesis.

Clinical and Research Implications The identification of robust proteomic biomarkers for HCC risk in MASH patients may have profound clinical implications. Incorporating such biomarkers into clinical workflows could refine risk stratification algorithms beyond fibrosis staging alone, enabling personalized surveillance schedules tailored to individual molecular risk profiles.

Moreover, proteomic biomarkers may serve as therapeutic targets or indicators of treatment response in emerging pharmacological interventions for MASH and early HCC, facilitating precision medicine approaches.

Future studies should focus on validating these biomarkers in larger, multicentric cohorts and integrating proteomic data with other omics modalities (genomics, metabolomics) and clinical parameters. Longitudinal sampling and serial proteomic analyses could elucidate temporal dynamics of disease progression and biomarker fluctuations.

Conclusion In conclusion, the application of tissue proteomics to FFPE liver biopsy specimens from patients with metabolic-associated steatohepatitis represents a promising strategy for identifying novel biomarkers predictive of hepatocellular carcinoma risk. This approach addresses critical limitations of current diagnostic methods and offers a pathway toward improved early detection, personalized risk assessment, and ultimately better patient outcomes in the growing population affected by MASH.