GDF-15 improves the predictive capacity of Steatotic liver disease non-invasive tests for incident morbidity and mortality risk for cardio-renal-metabolic diseases and malignancies

Michail Kokkorakis; Pytrik Folkertsma; José Castela Forte; Bruce H R Wolffenbuttel; Sipko van Dam; Christos S Mantzoros

doi:10.1016/j.metabol.2024.156047

GDF-15 improves the predictive capacity of Steatotic liver disease non-invasive tests for incident morbidity and mortality risk for cardio-renal-metabolic diseases and malignancies

Metabolism. 2024 Oct 11:156047. doi: 10.1016/j.metabol.2024.156047. Online ahead of print.

Authors

Michail Kokkorakis¹, Pytrik Folkertsma², José Castela Forte³, Bruce H R Wolffenbuttel², Sipko van Dam², Christos S Mantzoros⁴

Affiliations

¹ Department of Clinical Pharmacy and Pharmacology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands; Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA. Electronic address: m.kokkorakis@umcg.nl.
² Department of Endocrinology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands.
³ Department of Clinical Pharmacy and Pharmacology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands.
⁴ Department of Clinical Pharmacy and Pharmacology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands; Department of Medicine, Boston VA Healthcare System, Boston, MA, USA.

PMID: 39396641
DOI: 10.1016/j.metabol.2024.156047

Abstract

Background & aims: Noninvasive tools (NITs) are currently used to stratify the risk of having or developing hepatic steatosis or fibrosis. Their performance and a proteomic-enabled improvement in forecasting long-term cardio-renal-metabolic morbidity, malignancies, as well as cause-specific and all-cause mortality, are lacking. Therefore, the performance of established NITs needs to be investigated in identifying cardio-renal-metabolic morbidity, malignancies, cause-specific and overall mortality and improve their performance with novel, proteomic-enabled NITs, including growth differentiation factor 15 (GDF-15), allowing multipurpose utilization.

Methods: 502,359 UK Biobank participants free of the study outcomes at baseline with a 14-year median follow-up were grouped into three categories: a) general population, b) potentially metabolic dysfunction-associated steatotic liver disease (MASLD) population, c) individuals with type 2 diabetes mellitus. The investigated NITs include Aspartate aminotransferase to Platelet Ratio Index (APRI), Fibrosis 4 Index (FIB-4), Fatty Liver Index (FLI), Hepatic Steatosis Index (HSI), Lipid Accumulation Product (LAP), and metabolic dysfunction-associated fibrosis (MAF-5) score.

Results: Adding GDF-15 to the existing NITs led to significantly increased prognostic performance compared to the traditional NITs in almost all instances, reaching substantially high C-indices, ranging between 0.601 and 0.808, with an overall >0.2 improvement in C-index. Overall, with the GDF-15 enhanced NITs, up to more than seven times fewer individuals need to be screened to identify more incident cases of adverse outcomes compared to the traditional NITs. The cumulative incidence of all outcomes, based on the continuous value percentiles of NITs, is increasing exponentially in the upper quintile of the GDF-15 enhanced NITs.

Conclusions: The herein-developed GDF-15 enhanced indices demonstrate higher screening effectiveness and significantly improved prognostic abilities, which are reduced to practice through an easy-to-use web-based calculator tool (https://clinicalpredictor.shinyapps.io/multimorbidity-mortality-risk/).

Keywords: AI; MASLD; ML; NITs; Omics.