Postoperative course of S-100B protein and neuron-specific enolase in patients after implantation of continuous and pulsatile flow lvads

Abstract

Background In the early post-operative period after implantation of a continuous flow left ventricular assist device (LVAD) a non-pulsatile flow occurs. We compared the post-operative time-courses of protein S-100B (S100B) and neuron-specific enolase (NSE) as biochemical markers of brain injury in patients after implantation of a continuous flow LVAD and patients receiving a pulsatile flow LVAD. Methods Since 1998 the continuous flow DeBakey VAD has been implanted in 8 patients at our institution. For comparison purposes, a group of 7 consecutive patients in whom a pulsatile Novacor N100 LVAD was implanted were investigated. In both groups cardiopulmonary bypass (CPB) with cardiotomy suction was used. S100B and NSE were measured in serum pre-operatively, 4 hours after CPB, and on days 1, 3, 7, and 14 after implantation of the LVAD. A neurologic examination was performed pre-operatively and post-operatively on days 3 and 14. Results No differences were found between groups in pre-operative characteristics. The analysis of variance with repeated measurements for S-100B and NSE showed significant time effects ( p = 0.004, p = 0.009, respectively) but no group effects ( p = 0.06, p = 0.26, respectively) and no interaction between groups and time ( p = 0.12, p = 0.48, respectively). The pre-operative serum level of S100B was significantly higher ( p = 0.03) in the DeBakey VAD group. The pre-operative serum level of NSE was similar in the 2 groups ( p = 0.7). In both groups there was a significant increase of S100B and NSE immediately after surgery (S100B: p = 0.006, p = 0.019; NSE: p = 0.01, p = 0.001). The values returned to pre-operative levels in the DeBakey VAD group on day 1 after implantation and in the Novacor group for S100B on day 3 and NSE on day 1. Post-operatively the mean values of S100B and NSE in the DeBakey VAD group compared with the Novacor group were significantly elevated only on day 3 ( p = 0.005, p = 0.023). No neurologic complications were noted in patients with a continuous flow LVAD, whereas in the pulsatile LVAD group 2 patients presented neurologic abnormalities during the study period. Conclusion The similar course of biochemical markers of brain damage in both groups may indicate that the non-pulsatile flow in the early post-operative period does not lead to increased brain injury or permeability of the brain blood barrier. Elevated levels of S100B and NSE in the post-operative period can be used as diagnostic markers of brain injury in patients after implantation of both types of LVAD.