Risk stratification of ambulatory patients with advanced heart failure undergoing evaluation for heart transplantation

Abstract

Risk stratification of ambulatory heart failure (HF) patients has relied on peak VO(2)<14 ml/kg/min. We investigated whether additional clinical variables might further specify risk of death, ventricular assist device (VAD) implantation (INTERMACS <4) or heart transplantation (HTx, Status 1A or 1B) within 1 year after HTx evaluation. We hypothesized that right ventricular stroke work index (RVSWI), pulmonary capillary wedge pressure (PCWP) and the model for end-stage liver disease-albumin score (MELD-A) would be additive prognostic predictors. We retrospectively collected data on 151 ambulatory patients undergoing HTx evaluation. Primary outcomes were defined as HTx, LVAD or death within 1 year after evaluation. Average age in our cohort was 55 ± 11.1 years, 79.1% were male and 39% had an ischemic etiology (LVEF 21 ± 10.5% and peak VO(2) 12.6 ± 3.5 ml/kg/min). Fifty outcomes (33.1%) were observed (27 HTxs, 15 VADs and 8 deaths). Univariate logistic regression showed a significant association of RVSWI (OR 0.47, p = 0.036), PCWP (OR 2.65, p = 0.007) and MELD-A (OR 2.73, p = 0.006) with 1-year events. Stepwise regression showed an independent correlation of RVSWI<5gm-m(2)/beat (OR 6.70, p < 0.01), PCWP>20 mm Hg (OR 5.48, p < 0.01), MELD-A>14 (OR 3.72, p< 0.01) and peak VO(2)<14 ml/kg/min (OR 3.36, p = 0.024) with 1-year events. A scoring system was developed: MELD-A>14 and peak VO(2)<14-1 point each; and PCWP>20 and RVSWI<5-2 points each. A cut-off at≥4 demonstrated a 54% sensitivity and 88% specificity for 1-year events. Ambulatory HF patients have significant 1-year event rates. Risk stratification based on exercise performance, left-sided congestion, right ventricular dysfunction and liver congestion allows prediction of 1-year prognosis. Our findings support early and timely referral for VAD and/or transplant.