Left ventricular reverse remodeling in post operative aortic stenosis patients: prevalence and predictor(s)

Abstract

Abstract Background In patients with severe aortic stenosis (AS), left ventricular (LV) remodeling is believed to be a compensatory adaptive process which should reverse after aortic valve intervention. However, this is not always the rule and remodeling persistence may negatively impact post-procedural outcomes and survival. Aim To assess the prevalence and predictors of morphological LV reverse remodeling in severe symptomatic AS patients after surgical aortic valve replacement (AVR). Methods We prospectively studied 75 patients (72y [68–77y], 45% male) with severe symptomatic AS - mean gradient (AVM): 61±17mmHg; mean indexed aortic valve area (AVAi) 0.41±0.10 cm2/m2 with no previous history of ischemic cardiomyopathy, all with high gradient, 4 with low-flow, 81% with hypertension, 27% with type 2 diabetes mellitus and 35% patients with stage 3 chronic kidney disease: median MDR creat clearance: 70.4mL/min [40–102]. All patients performed pre-operative cardiac magnetic resonance (CMR) at a mean period of 3.4 months (0–17 months) before AVR and at the 3–6th months after AVR, for LV reverse remodeling assessment. It was defined as at least the occurrence of one of the following: >15% reduction in LVEDVi; >15% reduction in LVMi by CMR; >10% reduction in geometric remodeling ratio. Clinical, AV severity data, preoperative functional LV and tissue characterization data were analyzed at multivariate regression to predict the occurrence of LV reverse remodeling. Results Overall, at pre-operative CMR: mean LV indexed mass (LVMi): 82±28.9 g/m2; mean end-diastolic LV indexed volume (LVEDVi): 87.4±26.6 mL/m2; mean geometric remodeling (LV mass/end-diastolic volume): 0.92±0.2 g/mL. After AVR, at echocardiographic evaluation, no patient had prosthetic obstruction or prosthetic patient mismatch: median LV-Ao gradient 12mmHg [9.1–14 mmHg]; 5 of them had mild paravalvular regurgitation. LV reverse remodeling occurred in 65 patients (88%) (Figure 1A) and these were younger, had significantly smaller preoperative AVAi and higher valvular gradients (Figure 1B). At multivariate analysis, only preoperative AVAi remained an independent predictor (odds ratio 0.85, 95% CI 0.735–0.984, p=0.029). Conclusions In this prospective cohort of patients LV reverse remodeling after surgical AVR was highly frequent, occurring in almost nine out of every ten patients. Funding Acknowledgement Type of funding sources: None.