Predictors of LV reverse remodeling after SAVR: insights from a prospective TTE and CMR study in patients with severe aortic stenosis

Abstract Background Left ventricular (LV) remodeling in patients with severe aortic stenosis (AS) is believed to be reversible after pressure overload relieve, as provided by both surgical (SAVR) or transcatheter aortic valve replacement. However, LV reverse remodeling (RR) is far from uniform in patients with AS and the same clinical indication for treatment. Aim To determine the prevalence and imaging predictors of LV RR after SAVR in patients with severe symptomatic AS. Methods single-center, prospective cohort study enrolling 119 patients with severe symptomatic AS (age 71±8y, 49% male; mean transaortic gradient 62±18mmHg, mean indexed aortic valve area (AVAi) 0.40±0.10 cm2/m2, mean LV ejection fraction by TTE 58±9%), with no previous history of ischemic cardiomyopathy, undergoing SAVR between 2019 and 2022. All patients underwent serial imaging assessment beyond clinical characterization (transthoracic echocardiogram – TTE, and 1.5T cardiac magnetic resonance - CMR) prior to surgery and at the 3rd to 6th month post-SAVR. Endomyocardial biopsy (EMB) for myocardial fibrosis quantification (Masson ́s Trichrome histochemistry) was performed during SAVR. LV RR was defined when in presence of at least one of the imaging criteria: > 15% reduction in LV end-diastolic indexed volume (LVEDV) by CMR; > 15% reduction in LV indexed mass (LVM) by CMR; > 15% reduction in LV geometric remodeling by CMR; >10% increase in LV ejection fraction (LVEF) either from TTE or CMR; > 50% increase in global longitudinal strain (GLS) at TTE. Clinical, imaging and histology derived data were compared in patients with post-SAVR LV RR and for each of the defined criteria. Results 107 patients (90%) met at least one criteria of LV RR. Morphological criteria were more prevalent than LV functional improvement after SAVR (Figure 1). In patients who met at least one criteria of LV RR, no differences were found between NT- proBNP, LV tissue characterization (by CMR) or myocardial fibrosis at EMB. Yet, those with LV RR had significantly smaller preoperative AVAi (0.4±0.9 vs 0.5±0.9 cm2/m2, p=0.007) and higher mean valvular gradients (62±18 vs 49±12 mmHg, p=0.017). Considering individual LV RR criteria, patients with higher NT-proBNP levels met LVEF and morphological (LVEDV and LVM) LV RR criteria (p=0.027, p=0.016 and p= 0.003, respectively). Patients with higher % of late gadolinium enhancement on pre-operative CMR [7.8 (IQR 2.5-11.3) vs. 1.6 (IQR 0-5.2)%, p=0.017] had significant GLS imp