{
  "abstract": "Introduction Clinically significant portal hypertension (CSPH) critically influences prognosis in advanced metabolic dysfunction-associated steatotic liver disease (MASLD) but is underdiagnosed due to reliance on invasive hepatic venous pressure gradient (HVPG) measurement. The ANTICIPATE-NASH model, incorporating liver stiffness measurement, platelet count, and BMI, was developed as a non-invasive tool to estimate CSPH risk in patients with NASH. However, further validation is required, particularly regarding its rule-in criteria in the broader MASLD population. In this study, we evaluate the performance of the ANTICIPATE-NASH model for predicting CSPH in histologically confirmed MASLD patients from an independent multicentre European cohort.Methods Well-characterised MASLD patients with HVPG measurements and ANTICIPATE-NASH scores obtained within six months were identified from the European SLD (MASLD) Registry. CSPH was defined as HVPG ≥10 mmHg. Model performance was assessed using receiver operating characteristic (ROC) analysis and calibration. Diagnostic thresholds included the Youden-derived optimal cutoff and a high-specificity rule-in cutoff, with sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) calculated.Results Among 72 histologically validated MASLD patients (median age 55.0 years; 51% female; median BMI 36.2 kg/m 2), 49% had type 2 diabetes mellitus. Fibrosis stages were F0 (19.4%), F1 (15.3%), F2 (26.4%), F3 (20.8%), and F4 (18.1%). Fourteen patients (19.4%) had CSPH. Fibrosis stage distribution differed significantly between patients with and without CSPH (p=0.002), with CSPH occurring exclusively in patients with ≥F2 fibrosis and showing marked enrichment of advanced disease (F4: 50%). In contrast, non-CSPH patients were distributed across all fibrosis stages, including early disease (F0-F1: 43%), and had a lower prevalence of cirrhosis (F4: 10%).The ANTICIPATE-NASH model showed excellent discrimination for CSPH (AUROC=0.897) with good calibration (Hosmer-Lemeshow, p=0.426). The Youden-derived optimal cutoff (0.376) yielded a sensitivity of 85.7%, specificity of 81.0%, PPV of 52.2%, and NPV of 95.9%, supporting reliable exclusion of CSPH. For rule-in purposes, a high-specificity cutoff of 0.47 achieved 89.7% specificity with a PPV of 62.5%, while a cutoff of 0.52 increased specificity to 94.8% with a PPV of 76.9%. PPV estimates should be interpreted with caution given the limited number of CSPH cases.Conclusions The ANTICIPATE-NASH model showed robust external validity, with excellent discrimination and clinically meaningful risk stratification for CSPH in MASLD patients. The Youden cutoff (0.376) enables accurate exclusion of CSPH with very high NPV, while a high-specificity cutoff (0.52) allows confident identification of patients at high risk.",
  "authors": [
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom"
      ],
      "name": "Xiaowen Ma"
    },
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom"
      ],
      "name": "Kristy Wonders"
    },
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom"
      ],
      "name": "Alasdair Blain"
    },
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom"
      ],
      "name": "Rachael Redmond"
    },
    {
      "affiliations": [
        "University of Turin, Turin, Italy"
      ],
      "name": "Elisabetta Bugianesi"
    },
    {
      "affiliations": [
        "Instituto de Salud Carlos III, Madrid, Spain",
        "Universitat Autònoma de Barcelona, Barcelona, Spain"
      ],
      "name": "Joan Genescà"
    },
    {
      "affiliations": [
        "Hospital Universitario Marqués de Valdecilla, Santander, Spain"
      ],
      "name": "Javier Crespo"
    },
    {
      "affiliations": [
        "Università degli Studi di Milano, Milan, Italy",
        "Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy"
      ],
      "name": "Luca Valenti"
    },
    {
      "affiliations": [
        "University of Oxford, Oxford, United Kingdom"
      ],
      "name": "Jeremy Cobbold"
    },
    {
      "affiliations": [
        "University Hospital Wurzburg, Würzburg, Germany"
      ],
      "name": "Andreas Geier"
    },
    {
      "affiliations": [
        "Saarland University Medical Center, Homburg, Germany",
        "Saarland University, Saarbrücken, Germany"
      ],
      "name": "Jörn Schattenberg"
    },
    {
      "affiliations": [
        "University of Bern, Bern, Switzerland"
      ],
      "name": "Annalisa Berzigotti"
    },
    {
      "affiliations": [
        "NIHR Nottingham Biomedical Research Center, Nottingham, United Kingdom"
      ],
      "name": "Guruprasad P Aithal"
    },
    {
      "affiliations": [
        "University of Antwerp, Antwerp, Belgium"
      ],
      "name": "Sven Francque"
    },
    {
      "affiliations": [
        "Hospital Pitié Salpêtrière, Paris, France"
      ],
      "name": "Vlad Ratziu"
    },
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom"
      ],
      "name": "Simon J Cockell"
    },
    {
      "affiliations": [
        "Universitat Autònoma de Barcelona, Barcelona, Spain"
      ],
      "name": "Juan M Pericàs"
    },
    {
      "affiliations": [
        "University of Edinburgh, Edinburgh, United Kingdom"
      ],
      "name": "Neil Henderson"
    },
    {
      "affiliations": [
        "Newcastle University, Newcastle Upon Tyne, United Kingdom",
        "Newcastle upon Tyne Hospitals NHS Trust, Newcastle upon Tyne, United Kingdom"
      ],
      "name": "Quentin M Anstee"
    }
  ],
  "title": "O11 External validation of ANTICIPATE-NASH model for non-Invasive CSPH detection in MASLD: a multicentre European study",
  "uid": "f5241ef1-1305-5db2-86d4-cc2349573730"
}