Guy A. MacGowan, Robert L. Kormos, Dennis M. McNamara, Rene J. Alvarez, Warren D. Rosenblum, Si Pham, Arthur M. Feldman, Srinivas Murali

The purpose of this study was to determine which patients on a cardiac transplantation list required a ventricular assist device. In a preliminary study, 26 patients with decompensated severe New York Heart Association class IV chronic heart failure were studied. Blood levels for sodium, hemoglobin, cytokines, neurohormones, and hemodynamics were obtained. During short-term follow-up of 40 days, 12 patients had undergone emergent implantation of a ventricular assist device (range 1-27 days, mean 5 days), 4 died (range 14-38 days, mean 26 days), and 5 were alive and receiving only medical therapy while waiting for a transplantation. In addition, five patients had undergone transplantation (range 5-29 days, mean 18 days, excluded from further analysis). Survival curves were constructed by comparing the incidence of death and the implantation of an emergent ventricular assist device in patients with values of a variable above or below the mean value (or median for nonnormally distributed data). There was a significantly greater incidence of death or need for a ventricular assist device in patients with higher levels of tumor necrosis factor-alpha (P = .008), lower levels of serum sodium and hemoglobin (P = .02 and P = .03, respectively), higher heart rates (P = .03), and higher plasma norepinephrine levels (P = .01). The Cox proportional hazards model demonstrated that only serum sodium (P = .03) independently predicted those patients who died or who required emergent left ventricular assist device. Numerous variables, particularly serum sodium, need to be considered when evaluating which patients on the transplant list require early assist device implantation or urgent transplantation. These preliminary observations merit confirmation in a larger patient population.