Quantification of coronary artery calcification in systemic sclerosis using visual ordinal and deep learning scoring:...

To investigate the association between systemic sclerosis (SSc) clinical features and the extent and progression of coronary artery calcifications.

We conducted a single-center retrospective cohort study of patients with SSc. In our primary aim, we investigated the association between SSc clinical features and the annual progression of coronary artery calcium (CAC) scores quantified using the visual ordinal scoring method. In our secondary aim, we utilized DeepCAC, a deep learning-based method, to quantify coronary artery calcifications ("deep learning CAC score"), and explored its association with SSc clinical features.

Eighty-six SSc patients were included in the primary aim and 171 in the secondary aim. SSc disease duration was inversely associated with annual ordinal CAC score progression in the demographics-adjusted model (coefficient = -0.004, 95 % CI -0.006 to -0.001, p-value = 0.01) and the demographics- and cardiovascular (CV) risk factor-adjusted model (coefficient = -0.004, 95 % CI -0.008 to -0.0004, p-value = 0.03). The presence of "fingertip ischemic ulcers or digital pitting scars" (demographics-adjusted model: coefficient = 1.07, 95 % CI 0.29 to 1.85, p < 0.01; demographics- and CV risk factor-adjusted model: coefficient = 1.39, 95 % CI 0.43 to 2.34, p < 0.01) and Group 1 pulmonary hypertension (demographics-adjusted model: coefficient = 1.34, 95 % CI 0.34 to 2.35, p < 0.01; demographics- and CV risk factor-adjusted model: coefficient = 1.52, 95 % CI 0.38 to 2.65, p < 0.01) were both associated with the deep learning CAC score.

Our results suggest that the progression of coronary artery calcification accelerates early during the SSc disease course and that severe microvasculopathy may be a risk factor for atherosclerotic CVD.