Right atrial volume indexed by cardiovascular magnetic resonance as a predictor of mortality in patients with heart failure with reduced ejection fraction

Background

Right Atrial Volume Indexed (RAVI) measured by echocardiogram was identified to be an independent predictor of morbidity in patients with heart failure (HF) with reduced ejection fraction (HFrEF). Meta-Analysis Global Group in Chronic heart failure (MAGGIC) risk score is a robust tool in predicting mortality in HF patients. The aim of this study is to evaluate RAVI by cardiac magnetic resonance (CMR) imaging as an independent predictor of all-cause mortality in patients with HFrEF and to compare it with the validated risk score

Methods

We identify 488 patients with left ventricular ejection fraction (LVEF) < 35% assessed by CMR. We excluded patients undergoing open-heart surgery, severe valvular disease and patients with inadequate imaging. Right atrial volume was calculated based on a measured area in two and four chamber views utilizing a validated equation and indexed to body surface area. MAGGIC risk score was calculated using an online calculator. Follow-up information was acquired via phone call questionnaire and social security death index in patients not able to be reached directly. Our primary outcome was all-cause mortality or an appropriate implantable cardioverter defibrillator therapy.

Results

Two hundred forty four patients (mean age 60 ± 15; 33% women) were followed for a mean period of 2.1 years. Thirty-two patients (13%) had a primary outcome. The mean RAVI was 53 ± 26 ml/m2. The RAVI was significantly larger in patients with an event than without (75.9 ± 31 ml/m2 vs. 49 ± 23 ml/m2, p < 0.001). The mean MAGGIC score was 19.7 ± 7 (mean one-year mortality 10.2%). RAVI assessed continuously was an independent predictor of mortality controlled for MAGGIC risk, right ventricular ejection fraction (RVEF[JH1]) with HR 1.02(1.01-1.03), p = 0.002. RAVI (0.76 ± 0.08) has a greater C statistic than LVEF (0.55 ± 0.1, p < 0.004), left atrial volume indexed (0.64 ± 0.1, p < 0.025) and has a trend towards greater C statistics than RVEF (0.65 ± 0.11, p < 0.06). The addition of RAVI to the MAGGIC risk score significantly reclassify risk (integrated discrimination improvement by 9%, p < 0.0015 and category-free net reclassification improvement by 58%, p < 0.0029).

Conclusions

RAVI measured by CMR is an independent predictor of mortality in patients with HFrEF. The addition of RAVI to MAGGIC score significantly reclassify risk.