{
  "abstract": "Introduction Cardiac involvement in systemic sclerosis (SSc) is common and often subclinical. Cardiovascular magnetic resonance (CMR) provides multiparametric volumetric and tissue indices that enable data-driven phenotyping beyond conventional tests. In this analysis we aimed to identify CMR-defined myocardial phenotypes in a SSc cohort selected to minimise non-SSc cardiovascular confounding and to describe their clinical profiles.Material and Methods Eighty-nine SSc patients without ischaemic heart disease, overt cardiomyopathy, pulmonary arterial hypertension, diabetes, and without 2 or more traditional cardiovascular risk factors underwent comprehensive CMR on a 3 T system (Philips Achieva). Assessment included volumetric analysis, and tissue characterisation (T1 mapping and late gadolinium enhancement, LGE). Variables were standardised and clustered using agglomerative hierarchical clustering. Kruskal–Wallis test and Fisher’s exact or Chi-Squared tests were applied and effect sizes calculated.Results Median age was 55 years (IQR 49–64); 79.8% were female; 57 (64.0%) had lcSSC, 25 (28.1%) Scl-70+, and 30 (33.7%) ACA+; disease duration 7.3 years (1.8–15). Baseline NT-proBNP, high-sensitivity troponin I, CRP and creatine kinase were within normal range.Clustering identified three phenotypes. A ‘no remodelling with scar’ group (n=49) showed near-normal chamber size and function (LV EDVi 73.4 mL/m2; RV EDVi 76.9 mL/m2; LVEF 59.8%; RVEF 55.3%), the lowest native T1 (1220 ms), and mixed LGE (present 51.0%; RV insertion-point [RVIP] 18.4%, non-RVIP 32.7%). A ‘hyperdynamic low scar’ group (n=29) had similar volumes (LV EDVi 70.9 mL/m2; RV EDVi 74.2 mL/m2) but higher systolic function (LVEF 66.3%; RVEF 62.1%), low LGE prevalence (20.7%), and native T1 of 1277 ms, suggesting diffuse interstitial signal without overt remodelling. A ‘dilated–fibrotic’ (RVIP-scar dominant) group (n=11) exhibited the largest volumes (LV EDVi 108.4 mL/m2; RV EDVi 106.3 mL/m2) with mid-range EF (LVEF 57.7%; RVEF 55.9%), the highest LGE prevalence (81.8%) with RVIP enrichment (45.5%), and the highest native T1 (1288 ms). Effect sizes were largest for ventricular remodelling/function and tissue signal and smallest for ECV (0.057) and aortic distensibility (0.012). No differences in clinical characteristics were observed except for vasculopathy features (digital ulcers, pits, calcinosis) and ever-smoking status, which were most prevalent in the dilated–fibrotic phenotype and least in the hyperdynamic phenotype.Conclusions Unsupervised CMR-only clustering delineated three myocardial phenotypes in SSc spanning a continuum from preserved-function/scar disease through an intermediate state to a dilated–fibrotic pattern with RVIP scar. CMR-defined pathophysiological stages were largely independent of standard assessments. These data support phenotype-guided surveillance and warrant prospective, multi-centre validation with outcomes.Conclusions Initial therapy was associated with improved survival and reduced PAH progression in SSc-PAH, with consistent effects across haemodynamic thresholds and risk strata. These findings support guideline-recommended early intervention, highlight the importance of high-quality observational data in rare diseases, and underscore the need for RCTs to clarify treatment effects in patients with milder haemodynamic impairment.",
  "authors": [
    {
      "affiliations": [
        "University of Manchester, Centre for Musculoskeletal Research, Manchester, UK"
      ],
      "name": "Amr Mohammed"
    },
    {
      "affiliations": [
        "University Hospital Coventry & Warwickshire NHS Trust, Department of Rheumatology, Coventry, UK"
      ],
      "name": "Raluca Bianca Dumitru"
    },
    {
      "affiliations": [
        "Leeds Teaching Hospital NHS Trust, Department of Rheumatology, Leeds, UK",
        "University of Leeds, Leeds Institute of Rheumatic and Musculoskeletal Medicine, Leeds, UK"
      ],
      "name": "Lesley-Anne Bissell"
    },
    {
      "affiliations": [
        "Nottingham University Hospital NHS Trust, Nottingham, UK"
      ],
      "name": "Bara Erhayiem"
    },
    {
      "affiliations": [
        "Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK"
      ],
      "name": "Graham Fent"
    },
    {
      "affiliations": [
        "Leeds Teaching Hospital NHS Trust, Department of Medical Physics and Engineering, Leeds, UK"
      ],
      "name": "John P Greenwood"
    },
    {
      "affiliations": [
        "Baker Heart and Diabetes Institute, Melbourne, Australia"
      ],
      "name": "John Biglands"
    },
    {
      "affiliations": [
        "University of Leeds, Leeds Institute of Cardiovascular and Metabolic Medicine, Leeds, UK"
      ],
      "name": "Sven Plein"
    },
    {
      "affiliations": [
        "Leeds Teaching Hospital NHS Trust, Department of Rheumatology, Leeds, UK"
      ],
      "name": "Francesco Del Galdo"
    },
    {
      "affiliations": [
        "University of Manchester, Centre for Musculoskeletal Research, Manchester, UK",
        "Leeds Teaching Hospital NHS Trust, Department of Rheumatology, Leeds, UK"
      ],
      "name": "Maya H Buch"
    }
  ],
  "title": "OC.07 Unsupervised clustering identifies distinct CMR derived phenotypes in a cohort of SSC patients with low cardiovascular risk and no prior cardiac history",
  "uid": "23e57f4e-6b92-5e4c-b61f-a9a9985f7fa7"
}