Cho Jai Young, Suh Kyung-Suk, Lee Hae Won, Cho Eung-Ho, Yang Sung Hoon, Cho Yong Beom, Yi Nam-Joon, Kim Min A, Jang Ja-June, Lee Kuhn Uk
Department of Surgery, Seoul National University College of Medicine, Seoul, Korea.
Liver Transpl. 2006 Sep;12(9):1403-11. doi: 10.1002/lt.20772.
Early postoperative graft function assessments are essential after living donor liver transplantation (LDLT) to predict patient and graft outcome. Computed tomography (CT) is usually used to evaluate various complications and parenchymal abnormalities after LDLT. Here, we attempted to determine the prognostic values of CT attenuation changes of grafts for predicting 1-year patient survival. Liver attenuation indices (LAIs), derived from differences between hepatic and splenic attenuations, were calculated on unenhanced CT images obtained 10 days after LDLT in 62 adult LDLT recipients between September 2002 and August 2004. Patients were assigned to 1 of 2 groups according to LAI value on the 10th postoperative day, as follows: group L (LAI < or = 5, n = 14) or group H (LAI > 5, n = 48). Parenchymal dysfunction scores, summed parameters for histological dysfunction including both portal tract and centrilobular features, were also assessed on the 10th postoperative day using liver biopsy specimens. Histological parenchymal dysfunction, especially in the centrilobular area, in terms of cholestasis, centrilobular necroinflammation, central vein fibrosis, steatosis, mononuclear infiltrates, and hepatocyte ballooning, was more prominent in group L than in group H, while that in the portal area was similar between the 2 study groups. Significant negative linear correlations were observed between LAI and parenchymal dysfunction scores (r = 0.486, P < 0.001). Group L patients showed lower 1-year survival (69.7%) than group H patients (95.8%; P = 0.0002). Moreover, group H patients died with a functioning graft (n = 3), whereas group L patients died of graft failure (n = 6). After multivariate analysis, LAI alone remained independently associated with 1-year mortality (P = 0.014; odds ratio = 0.845; 95% confidence interval, 0.739-0.967). The sensitivity and specificity of LAI were 84.6% and 75%, respectively, and LAI outperformed MELD score as a predictor of 1-year mortality after LDLT by receiver operating characteristic curve analysis. In conclusion, LAI, as determined by unenhanced CT 10 days after LDLT, well predicts 1-year patient survival after LDLT.
活体肝移植（LDLT）术后早期对移植肝功能进行评估对于预测患者和移植物的预后至关重要。计算机断层扫描（CT）通常用于评估LDLT术后的各种并发症和实质异常情况。在此，我们试图确定移植肝CT衰减变化对预测患者1年生存率的预后价值。在2002年9月至2004年8月期间，对62例接受LDLT的成年患者，于术后10天获取的平扫CT图像上计算肝脏衰减指数（LAI），该指数由肝脏与脾脏衰减值的差值得出。根据术后第10天的LAI值将患者分为2组，如下：L组（LAI≤5，n = 14）或H组（LAI＞5，n = 48）。术后第10天还使用肝活检标本评估实质功能障碍评分，这是包括汇管区和小叶中心特征在内的组织学功能障碍的综合参数。在L组中，组织学实质功能障碍，尤其是小叶中心区域的胆汁淤积、小叶中心坏死性炎症、中央静脉纤维化、脂肪变性、单核细胞浸润和肝细胞气球样变，比H组更显著，而两个研究组在汇管区的情况相似。LAI与实质功能障碍评分之间存在显著的负线性相关（r = 0.486，P＜0.001）。L组患者的1年生存率（69.7%）低于H组患者（95.8%；P = 0.0002）。此外，H组患者中有3例死于有功能的移植物，而L组患者中有6例死于移植物衰竭。多因素分析后，单独的LAI仍与1年死亡率独立相关（P = 0.014；比值比 = 0.845；95%置信区间，0.739 - 0.967）。通过受试者工作特征曲线分析，LAI预测LDLT术后1年死亡率的敏感性和特异性分别为84.6%和75%，且LAI作为预测指标优于终末期肝病模型（MELD）评分。总之，LDLT术后10天通过平扫CT测定的LAI能够很好地预测患者的1年生存率。
