Department of Thoracic and Cardiovascular Surgery, West German Heart and Vascular Center Essen, University Hospital Essen, Essen, Germany.
Department of Critical Care Medicine, imland Klinik Rendsburg, Rendsburg, Germany.
Eur J Cardiothorac Surg. 2017 Oct 1;52(4):781-788. doi: 10.1093/ejcts/ezx247.
Preoperative liver dysfunction is a well-known risk factor for adverse events after major surgery. However, there is only little data regarding the precise role of the Model of End-Stage Liver Disease (MELD) score and the De Ritis ratio (DRR, alanine transaminase/aspartate aminotransferase) as a predictor for outcome after left ventricular assist device (LVAD) implantation.
A retrospective analysis of all patients undergoing LVAD implantation at our institution between January 2012 and August 2014 was performed. The primary outcome was survival at 180 days after surgery.
During the observation period, 63 patients underwent LVAD implantation (mean age 59.9 ± 8.3 years, 50% male). Mean preoperative ejection fraction was 16.3 ± 7.7, 13 patients required preoperative renal replacement therapy and 9 patients were on extracorporeal life support. Mean Interagency Registry for Mechanically Assisted Circulatory Support level was 2.8 ± 1.3, mean preoperative MELD was 12.7 ± 7.2, mean preoperative DRR was 2.01 ± 4.4. Aspartate aminotransferase (102 ± 220.8 vs 57.8 ± 123.4 U/l, P = 0.041), MELD score (16.1 ± 8.8 vs 11.4 ± 6.1, P = 0.017) and DRR (4.2 ± 7.8 vs 1.1 ± 1.1, P = 0.001) were significantly higher in non-survivors than in survivors after 180 days. Using logistic regression analyses, a DRR >1.37 was an independent predictor for 30-day mortality [odds ratio (OR) 4.5] and 180-day mortality (OR 4.1). In addition, the DRR was associated with postoperative acute kidney injury with need for renal replacement therapy (OR 4.2) and prolonged postoperative ventilation time >72 h (OR 3.8). Using receiver operator characteristics analyses, DRR showed a sensitivity of 0.80 and a specificity of 0.81 (area under the curve 0.834, cut-off 1.37) for 180-day mortality.
The DRR is predictive of early and mid-term mortality as well as relevant morbidities in patients undergoing LVAD implantation. Therefore, the DRR should be considered within the preoperative risk stratification and patient selection for LVAD implantation.
术前肝功能障碍是大手术后发生不良事件的一个众所周知的危险因素。然而,关于终末期肝病模型(MELD)评分和 De Ritis 比值(DRR,丙氨酸氨基转移酶/天冬氨酸氨基转移酶)作为左心室辅助装置(LVAD)植入术后结果预测因子的确切作用的数据很少。
对 2012 年 1 月至 2014 年 8 月期间在我院接受 LVAD 植入术的所有患者进行回顾性分析。主要观察终点为术后 180 天的生存率。
在观察期间,63 名患者接受了 LVAD 植入术(平均年龄 59.9±8.3 岁,50%为男性)。术前平均射血分数为 16.3±7.7,13 名患者需要术前肾脏替代治疗,9 名患者接受体外生命支持。平均 INTERMACS 机械循环支持级别为 2.8±1.3,术前 MELD 平均为 12.7±7.2,术前 DRR 平均为 2.01±4.4。术后 180 天,非幸存者的天冬氨酸氨基转移酶(102±220.8 vs 57.8±123.4 U/L,P=0.041)、MELD 评分(16.1±8.8 vs 11.4±6.1,P=0.017)和 DRR(4.2±7.8 vs 1.1±1.1,P=0.001)明显高于幸存者。使用逻辑回归分析,DRR>1.37 是 30 天死亡率(优势比[OR]4.5)和 180 天死亡率(OR4.1)的独立预测因子。此外,DRR 与术后需要肾脏替代治疗的急性肾损伤(OR4.2)和术后通气时间延长>72 小时(OR3.8)有关。使用受试者工作特征曲线分析,DRR 对 180 天死亡率的敏感性为 0.80,特异性为 0.81(曲线下面积 0.834,临界值 1.37)。
DRR 可预测接受 LVAD 植入术患者的早期和中期死亡率以及相关发病率。因此,DRR 应在 LVAD 植入术的术前风险分层和患者选择中考虑。