Pre LVAD CPET as Predictor of INTERMACS Profile and Early Outcomes Post LVAD

• Published in: The Journal of heart and lung transplantation Vol. 39; no. 4; pp. S182 - S183
• Main Authors: Brailovsky, Y., Mondellini, G.M., Javaid, A., Braghieri, L., Kleet, A.C., Shargarodskaya, R., Raikhelkar, J.K., Pineda, M.F., Fried, J.A., Kim, D.S., Takeda, K., Naka, Y., Sayer, G., Uriel, N., Goldsmith, R.L., Colombo, P.C., Yuzefpolskaya, M.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01-04-2020
• ISSN: 1053-2498; 1557-3117
• DOI: 10.1016/j.healun.2020.01.764

Cardiopulmonary Exercise Testing (CPET) is the clinical standard to select HF pts for cardiac replacement therapy. INTERMACS profile is used to risk stratify pts who undergo LVAD. We aimed to evaluate the association of CPET performance parameters with INTERMACS profile at implant as well as with clinical outcomes early post LVAD.
A retrospective analysis of LVAD pts implanted from 3/10 to 5/19 with CPET data available within 2y of surgery: peak oxygen consumption (pVO2) and minute ventilation − carbon dioxide production relationship (VE/VCO2). The primary endpoints were INTERMACS profile ≤2 at implant and a composite of post LVAD in-hospital mortality, right ventricular failure (RVF) and renal replacement therapy (RRT). Length of stay (LOS) in relation to CPET performance parameters was also evaluated. Receiver Operating Characteristics (ROC) curve assessed the ability of pVO2 and VE/VCO2 to predict the composite endpoint.
We studied 81 pts, median age 58 [IQR 16.5] y, 86% M, 41% ischemic, 48% HMII, 40% HM3 and 12% on other pumps. Median time from CPET to LVAD implant was 233 [388]d. Median pVO2 was 11.9[IQR 4.4] ml/kg/min and VE/VCO2 38 [IQR 10.3]. 46(56.8%) pts were INTERMACS ≤2 at implant. 6 pts died, 7 had RRT and 23 developed RVF during index admission. PVO2 and VE/VCO2 were not significantly different between INTERMACS ≤2 vs >2 groups (p=0.8 and p=0.15, respectively). ROC curves showed that pVO2 <12.8ml/kg/min had a sensitivity of 53% and specificity of 85% for the post LVAD composite endpoint (AUC 0.66, p=0.02), while VE/VCO2 was not a significant predictor (Figure). LOS was longer in pts with pVO2 <median (33[39] vs 22[7]d, p<0.001) and VE/VCO2 ≥median (30[29] vs 22[8]d, p=0.01).
In a real-world setting, CPET parameters obtained within 2 y of LVAD do not reliably predict INTERMACS profile at implant. However, pVO2 and VE/VCO2 can be helpful to identify pts at risk for in-hospital mortality, major complications and prolonged LOS post LVAD.