2021 年第 8 期 第 16 卷

中性粒细胞/淋巴细胞比值的变化对失代偿期肝硬化合并自发性腹膜炎患者不良预后的预测价值

Predictive value of neutrophil to lymphocyte ratio change in poor prognosis of patients of decompensated cirrhosis complicated with spontaneous peritonitis

作者：王京京1苏云娟2曹颖1王琦1谢雯1

英文作者：Wang Jingjing1 Su Yunjuan2 Cao Ying1 Wang Qi1 Xie Wen1

单位：1首都医科大学附属北京地坛医院肝病科100015；2首都医科大学附属北京地坛医院心内科100015

英文单位：1Department of Liver Diseases Beijing Ditan Hospital Capital Medical University Beijing 100015 China; 2Department of Cardiology Beijing Ditan Hospital Capital Medical University Beijing 100015 China

关键词：失代偿期肝硬化；自发性腹膜炎；中性粒细胞/淋巴细胞比值；死亡率

英文关键词：Decompensatedcirrhosis;Spontaneousperitonitis;Neutrophiltolymphocyteratio;Mortality

• 摘要：

• 目的 探讨血清中性粒细胞/淋巴细胞比值（NLR）的变化对失代偿期肝硬化合并自发性腹膜炎患者90 d死亡的预测价值。方法 回顾性分析首都医科大学附属北京地坛医院肝病科2018年1月至2019年12月收治的201例失代偿期肝硬化合并自发性腹膜炎患者的临床资料。根据90 d预后情况分为死亡组（67例）和存活组（134例）。比较2组的性别、年龄、既往病史、入院当天或次日清晨实验室检查指标水平、NLR差值水平及肝硬化相关失代偿期并发症情况。根据患者入院1周后血清NLR水平及入院当天或次日清晨NLR水平计算NLR差值。分析患者90 d死亡的危险因素，评价NLR差值单独与联合其他危险因素对患者90 d死亡的预测能力。计算NLR差值最佳截断值，根据NLR差值最佳截断值分组，比较不同NLR差值组患者的90 d病死率。结果 死亡组年龄、白细胞计数、中性粒细胞计数、NLR差值、C反应蛋白、降钙素原、终末期肝病模型（MELD）评分以及天冬氨酸转氨酶（AST）＞40 U/L、脓毒血症、肝性脑病、肝肾综合征(HRS)、入住重症监护病房比例均高于存活组（均P＜0.05）。2组其他基线临床特征比较差异均无统计学意义（均P＞0.05）。多因素Cox回归分析结果显示，NLR差值、AST>40 U/L、MELD评分、HRS均为预测患者90 d死亡的独立危险因素［风险比（95%置信区间）分别为1.026(1.008~1.045)、1.792(1.039~3.090)、1.077(1.032~1.123)、5.767(3.189~10.430)，均P＜0.05］。NLR差值+HRS预测患者90 d死亡的曲线下面积最大，为0.891，然后是NLR差值+AST>40 U/L和NLR差值，曲线下面积分别为0.850、0.831。NLR差值的最佳截断值为1.32，201例患者中NLR差值≥1.32有114例（高NLR差值组），NLR差值＜1.32有87例（低NLR差值组），高NLR差值组90 d肝病相关病死率高于低NLR差值组［50.9%（58/114）比10.3%（9/87）］（P＜0.001)。结论 NLR差值可用于预测失代偿期肝硬化合并自发性腹膜炎患者的死亡风险，NLR差值联合HRS的预测价值最高。

• Objective To investigate the predictive value of serum neutrophil to lymphocyte ratio (NLR) change in 90 d death in patients of decompensated cirrhosis complicated with spontaneous peritonitis. Methods Clinical data of 201 patients of decompensated cirrhosis complicated with spontaneous peritonitis admitted to Department of Liver Diseases, Beijing Ditan Hospital, Capital Medical University from January 2018 to December 2019 were retrospectively analyzed. According to 90 d prognosis, they were divided into death group (67 cases) and survival group (134 cases). The gender, age, previous medical history, levels of laboratory examination indexes on admission or next day in the morning, the NLR change and the decompensated complications related to cirrhosis were compared between the two groups. The NLR change was calculated according to the serum NLR level 1 week after admission and that on admission or next day in the morning. Risk factors for 90 d death were analyzed, and the predictive ability of NLR change alone and in combination with other risk factors for 90 d death of patients were evaluated. The optimal cut-off value of NLR change was calculated, and the 90 d mortality of patients in different NLR change groups were compared. Results Age, white blood cell count, neutrophils count, NLR change, C-reactive protein, procalcitonin, model of end-stage liver disease (MELD) score and the proportions of aspartate aminotransferase (AST) >40 U/L ,sepsis, hepatic encephalopathy, hepatorenal syndrome (HRS) and hospitalization in intensive care unit in the death group were higher than those in the survival group (all P<0.05). There were no significant differences in other baseline clinical characteristics between the two groups (all P>0.05). Multivariate Cox regression analysis showed that NLR change, AST>40 U/L, MELD score and HRS were independent risk factors for 90 d death ［hazard ratios (95% confidence interval) were 1.026(1.008-1.045), 1.792(1.039-3.090), 1.077(1.032-1.123), 5.767(3.189-10.430), all P<0.05］. The area under the curve of NLR change +HRS for predicting 90 d death was the largest, which was 0.891, followed by NLR change+AST>40 U/L and NLR change, which were 0.850 and 0.831. The optimal cut-off value of NLR change was 1.32. Among the 201 cases, 114 cases had NLR change ≥1.32 (high NLR change group) and 87 cases had NLR change<1.32(low NLR change group). The 90 d mortality associated with liver disease in the high NLR change group was higher than that in the low NLR change group ［50.9%(58/114) vs 10.3% (9/87)］(P<0.001). Conclusion  The NLR change can be used to predict the risk of death in patients with decompensated cirrhosis complicated with spontaneous peritonitis, and NLR change combined with HRS has the highest predictive value.