Baseline cardiac magnetic resonance–derived left atrial strain predicts 6-month improvement in exercise capacity after disease-modifying therapy in transthyretin amyloid cardiomyopathy

Background Disease-modifying therapy has improved clinical outcomes in patients with transthyretin amyloid cardiomyopathy (ATTR-CM), however, predictors of functional response remain unclear. Left atrial (LA) dysfunction is common in ATTR-CM and may reflect impaired cardiac reserve. LA reservoir strain assessed by cardiac magnetic resonance (CMR) provides a quantitative marker of LA function, yet its relationship with subsequent improvement in exercise capacity after therapy has not been established. Purpose To determine whether baseline CMR-derived LA reservoir strain is associated with 6-month improvement in cardiopulmonary exercise testing (CPX) parameters after initiation disease-modifying therapy in patients with ATTR-CM. Methods We screened consecutive patients admitted to a single center between August 2021 and May 2025 who underwent endomyocardial biopsy for evaluation of suspected cardiac amyloidosis. Patients with AL amyloidosis, CMR not performed, missing baseline or 6-month CPX data, atrial fibrillation at baseline, or inadequate image quality were excluded. Among 62 biopsy-proven cases, 30 patients met eligibility criteria and constituted the analytic cohort. All patients continued disease-modifying therapy throughout the 6-month follow-up. Demographic, laboratory and echocardiographic data were collected from medical records. LA reservoir strain was measured from 4-chamber cine CMR images. Improvement was defined as Δ (6 months − baseline) > 0 for anaerobic threshold (AT) and peak VO2. Group comparisons were performed using the Mann–Whitney U test. Multivariable logistic regression models adjusted for either log-transformed BNP or eGFR were used to evaluate independent associations. BNP was log-transformed due to a skewed distribution. Results In the analytic cohort (n=30, 29 wild type ATTR and 1 variant ATTR), mean age was 79.1 ± 5.6 years and mean echocardiographic left ventricular ejection fraction was 52.8 ± 11.4%. AT improved in 14 patients (47%) and peak VO2 improved in 13 patients (43%). Baseline LA reservoir strain was significantly higher in patients with AT improvement compared with those without (median [IQR] 12.8 [10.4–16.1] vs 7.7 [5.8–10.2]; p=0.026). Similarly, LA reservoir strain was higher in patients with peak VO2 improvement (13.3 [9.9–16.1] vs 7.7 [5.0–11.8]; p=0.013). After adjustment for log BNP, baseline LA reservoir strain independently predicted AT improvement (OR 1.27 per 1% increase, 95% CI: 1.04–1.55; p=0.020) and peak VO2 improvement (OR 1.27, 95% CI: 1.04–1.56; p=0.019). These associations remained significant after adjustment for eGFR. Conclusions Baseline CMR-derived LA reservoir strain independently predicts improvement in exercise capacity after disease-modifying therapy in patients with ATTR-CM, suggesting that preserved atrial function may reflect residual cardiac reserve and identify patients more likely to experience functional benefit.