OUP accepted manuscript

This study aimed to determine whether lower values of feature-tracking cardiovascular magnetic resonance (CMR)-derived left atrial (LA) reservoir strain (LARS) and impaired left ventricular (LV) global longitudinal strain (GLS) were associated with the presence of symptoms and long-term prognosis in patients with systemic sclerosis (SSc).A total of 100 patients (54[IQR 46-64] years, 42% male) with SSc who underwent CMR imaging at two tertiary referral centres were included. All patients underwent analysis of LARS and LV GLS using feature-tracking on CMR and were followed-up for the occurrence of all-cause mortality.The median LV GLS was -21.8% and the median LARS was 36%. On multivariable logistic regression, LARS (OR 0.964 per %, 95%CI 0.929-0.998, p = 0.049) was independently associated with New York Heart Association (NYHA) class II-IV heart failure symptoms. Over a median follow-up of 37 (21-62) months, a total of 24 (24%) patients died. Univariable Cox regression analysis demonstrated that LARS (HR 0.94 per %, 95%CI 0.91-0.97, P < 0.0001) and LV GLS (HR 1.10 per %, 95%CI 1.03- 1.17, P = 0.005) were associated with all-cause mortality, while LV ejection fraction was not. Likelihood ratio tests demonstrated that LARS provided incremental value over prognostically important clinical and imaging parameters, including late gadolinium enhancement.In patients with SSc, LARS was independently associated with the presence of NYHA class II-IV heart failure symptoms. Although both LARS and LV GLS were associated with all-cause mortality, only LARS provided incremental value over all evaluated variables known to be prognostically important in patients with SSc.