The proposed project aims at better differentiating clusters of pre-capillary pulmonary hypertension (PH) in systemic sclerosis (SSc) patients, using radiomics automated computer technology for the quantification of the extent and the severity of lung fibrosis. The aim is to create clusters of SSc-PH patients, in the context of possible coexisting lung fibrosis, to really define prognosis and treatment impact.
We showed that an automatic quantification of pulmonary fibrosis and pulmonary vessels is as good as the visual assessment combined with functional decline in identifying group 3 SSc-PH. In addition, the combination of functional impairment and automated radiomic estimation of pulmonary fibrosis and lung vessels was statistically superior to the current practice in achieving the same aim. This may help to homogenize the repeatability of patients’ assessments and perform specific studies, such as testing medications, which are a big unmet need in particular in patients with pulmonary hypertension and extensive pulmonary fibrosis. These studies should always take into consideration the quantification of radiological and functional involvementrelated to ILD and its patterns. Our cluster analysis showed that the different impact on survival might be possibly related to differences in these of variables, more that on hemodynamic features.
The current practice to identify if pre-capillary pulmonary hypertension in systemic sclerosis mostly relates to lung tissue disease or to “pure” vascular disease, relies on the functional respiratory assessments and the visual estimation of pulmonary fibrosis extent on high-resolution chest CT. Reproducibility issues affect both assessments, in particular the latter that also relies on local expertise. We showed that an automatic quantification of pulmonary fibrosis and pulmonary vessels is as good as the visual assessment combined with functional decline. In addition, the combination of functional impairment and automated radiomic estimation of pulmonary fibrosis and lung vessels was superior to the current practice. This may homogenize the repeatability of patients’ assessments. This is also in line with the results of our cluster analysis, in which the groups of pre-capillary pulmonary hypertension associated with systemic sclerosis represent different extents of radiological and functional involvements due to pulmonary fibrosis. These parameters, together with the pulmonary hypertension features, should therefore be taken into account when performing studies and testing medications, which are a high unmet need in this cohort. Certain groups may, probably, benefit from a more vascular-oriented treatment regimen while other from an anti-fibrosis targeting schema, or very likely the combination of both.