[The consistency of skeletal muscle mass measured by CT at L1 and L3 levels and the correlation of skeletal muscle density at L1 level with prognosis in dialysis patients].

Objective: To investigate the consistency of skeletal muscle mass by CT at 1st lumbar vertebrae (L1) and 3rd lumbar vertebrae (L3) levels and the correlation of skeletal muscle density (SMD) at L1 level with prognosis in dialysis patients. Methods: A total of 1 020 patients who underwent initial dialysis and had CT examination data in four centers (Zhongda Hospital Affiliated to Southeast University, the Third Affiliated Hospital of Soochow University, Taizhou People's Hospital Affiliated to Nanjing Medical University and the Affiliated Hospital of Yangzhou University) from January 2014 to December 2019 were retrospectively collected. The skeletal muscle index (SMI) and SMD at L1 and L3 CT images were measured and calculated in patients with both L1 and L3 level CT images. The consistency of SMI and SMD at L1 and L3 levels was analyzed, and the cut-off value of SMI and SMD at L1 level for predicting all-cause mortality and their correlation with the prognosis of dialysis patients were studied. Cox regression model was used to analyze the risk factors for all-cause death and cardiac death. Results: A total of 383 patients had both L1 and L3 level images, including 233 males and 150 females. The average SMD value of 16 samples (4.2%) exceeded the 95% consistency limit range (-8.71 to 7.75 HU), and the average SMI value of 15 samples (3.9%) exceeded the 95% consistency limit range (-20.45 to 9.53 HU). The optimal cut-off value of SMD at L1 level for predicting all-cause mortality was 36.46 HU and the area under curve (AUC) of receiver operating characteristic (ROC) curve was 0.658 (95%CI: 0.596-0.721, P<0.001), with the sensitivity and specificity of 83.8% and 57.5%, respectively. SMI at L1 level was not significantly associated with all-cause mortality (P=0.299). Multivariate Cox regression analysis showed that low SMD at L1 level was associated with all-cause mortality (HR=2.861, 95%CI: 1.576-5.193, P=0.001) and cardiac death (HR=3.771, 95%CI:1.462-9.724, P=0.006). Conclusions: SMD at L1 levelis consistent with SMD at L3 level and can be used to evaluate muscle mass. Low SMD is a risk factor for mortality in dialysis patients.目的： 探讨CT测量第1腰椎（L1）与第3腰椎（L3）层面透析患者骨骼肌质量的一致性及L1层面骨骼肌密度（SMD）与预后的相关性。 方法： 回顾性收集2014年1月至2019年12月在东南大学附属中大医院、苏州大学附属第三医院、南京医科大学附属泰州人民医院及扬州大学附属医院4家中心初始透析且同时有CT检查结果的1 020例患者资料，筛选出同时具有L1和L3层面CT图像的患者，测量计算骨骼肌指数（SMI）和SMD。分析L1和L3层面SMI、SMD的一致性，探讨L1层面SMI和SMD预测全因死亡的最佳临界值及与透析患者预后的相关性。应用Cox回归模型研究透析患者全因死亡和心源性死亡的危险因素。 结果： 共纳入同时具有L1和L3层面CT图像的透析患者383例，男233例，女150例。有16例样本（4.2%）SMD平均值超出了95%的一致性界限范围（-8.71~7.75 HU），15例样本（3.9%）SMI平均值超出了95%的一致性界限范围（-20.45~9.53 cm2/m2）。L1层面SMD预测全因死亡的受试者工作特征（ROC）曲线下面积为0.658（95%CI：0.596~0.721，P<0.001），最佳临界值为36.46 HU，灵敏度83.8%，特异度57.5%；L1层面SMI与全因死亡无明显相关性（P=0.299）。多因素Cox回归分析结果显示，L1层面低SMD是全因死亡（HR=2.861，95%CI：1.576~5.193，P=0.001）和心源性死亡（HR=3.771，95%CI：1.462~9.724，P=0.006）的危险因素。 结论： CT测量L1与L3层面透析患者SMD具有良好一致性，可通过L1层面SMD评估肌肉质量，且低SMD是透析患者死亡的危险因素。.