摘要
目的 探讨第三腰椎骨骼肌指数(L3-SMI)预测食管癌患者总生存(OS)期的最佳截止点与临床应用价值。 方法 对
2016 年 1 月至 2020 年 10 月于南通大学附属海安医院接受手术或根治性放化疗的 150 例食管癌患者进行回顾性分析,通过分
析每例患者的腹部计算机断层扫描(CT)影像,计算 L3-SMI。 通过受试者操作特征(ROC)曲线认定 L3-SMI 预测食管癌患者预
后的最佳截止点,并将所有患者分为低 L3-SMI 组与高 L3-SMI 组进行生存分析。 采用 Kaplan-Meier 曲线计算两组患者的 3
年 OS 率,单、多因素 Cox 回归分析确定 L3-SMI 在食管癌患者中的独立预后价值。 结果 研究队列中,男性患者 L3-SMI 平均
值为(43. 8±6. 3)cm
2
/ m
2
,女性为(38. 7±6. 7)cm
2
/ m
2
,L3-SMI 预测食管癌患者预后的理想截止点为:男性<42. 8 cm
2
/ m
2
,女性<
39. 2 cm
2
/ m
2
,曲线下面积(AUC)值分别为 0. 628(95%CI = 0. 515~ 0. 742)和 0. 654(95%CI = 0. 513 ~ 0. 796)。 根据 L3-SMI 的
最佳截止点,将所有患者分为高 L3-SMI 组(n = 79)与低 L3-SMI 组(n = 71)。 结果显示,两组患者 3 年 OS 率分别为 33. 5%和
54. 6%,差异有统计学意义(χ
2 = 5. 358,P = 0. 021)。 亚组分析表明,低 L3-SMI 与男性患者不良预后显著相关( 3 年 OS 率:
24. 8%比 53. 6%,χ
2 = 4. 752,P= 0. 029)。 对于女性食管癌患者,低 L3-SMI 组患者较高 L3-SMI 组有更低的 3 年 OS 率(39. 8%
比 65. 4%),但差异无统计学意义(χ
2 = 1. 835,P= 0. 176)。 单、多因素 Cox 回归分析证实,影响食管癌患者预后的独立因素包
括低 L3-SMI(HR= 1. 669,95%CI = 1. 062 ~ 2. 625,P = 0. 027)、预后营养指数(PNI) <50. 2(HR = 1. 700,95%CI = 1. 033 ~ 2. 796,
P= 0. 037)与 TNM 分期Ⅲ期(HR= 1. 503,95%CI = 1. 077~ 2. 479,P = 0. 024)。 结论 L3-SMI 是一项预测食管癌患者预后的重
要影像学参数。
Abstract
Objective This study aimed to explore the optimal threshold and predictive value of skeletal muscle index at the third
lumbar vertebra level L3-SMI for the prognosis of esophageal cancer patients. Method 150 esophageal cancer patients who were
treated with surgical resection or curative chemoradiotherapy in Haian Hospital Affiliated to Nantong University from January 2016 to
October 2020 were analyzed retrospectively. The L3-SMI was measured by analyzing abdominal computed tomography CT images.
The optimal cutoff values of L3 - SMI for survival prediction of male and female patients were identified by the receiver operation
characteristics ROC curves respectively. All patients were divided into high and low L3-SMI groups and Kaplan-Meier curve was
used to estimate the 3-year overall survival OS rate of the two patient groups. In addition the independent prognostic factors of L3-SMI
for esophageal cancer patients were further determined by the univariate and multivariate Cox regression analysis. Result In this study
cohort the average value of L3-SMI for male and female patients were 43. 8±6. 3 cm
2
/ m
2
and 38. 7±6. 7 cm
2
/ m
2
respectively.
The best cutoff values of L3-SMI for survival prediction of esophageal cancer patients were <42. 8 cm
2
/ m
2
for males and <39. 2 cm
2
/ m
2
for females respectively. The AUC value of the former was 0. 628 95% CI = 0. 515- 0. 742 and the latter had an AUC value of
0. 654 95%CI = 0. 513-0. 796 . All patients were divided into high L3-SMI n = 79 and low L3-SMI groups n = 71 based on the
identified cutoff values. The results of survival analysis indicated that the 3-year OS rate of low L3-SMI and high L3-SMI group were
33. 5% and 54. 6% with a significant statistical difference
χ
2 = 5. 358 P= 0. 021 . The subgroup analysis showed that low L3-SMI
was significantly associated with worse OS in male patients 3-year OS rate 39. 8% vs 65. 4%
χ
2 = 4. 752 P= 0. 029 . For female
patients with esophageal cancer low L3-SMI group had a lower 3-year OS rate than high L3-SMI group 39. 8% vs 65. 4% but the
survival difference was not statistically significant
χ
2 = 1. 835 P= 0. 176 . The results of the univariate and multivariate Cox regres-sion analysis demonstrated that low L3-SMI HR= 1. 669 95%CI = 1. 062-2. 625 P= 0. 027 prognostic nutritional index PNI <
50. 2 HR=1. 700 95%CI = 1. 033-2. 796 P = 0. 037 and Ⅲ stage of the TNM stage HR = 1. 503 95%CI = 1. 077-2. 479 P =
0. 024 were independent prognostic factors for esophageal cancer patients. Conclusion L3-SMI is a valuable imaging parameter of
prognostic assessment for esophageal cancer patients.
关键词
食管癌 /
第三腰椎骨骼肌指数 /
肌肉减少症 /
预后 /
预测
吴 洁,周晶晶,陈海生.
第三腰椎骨骼肌指数对食管癌患者生存结局的预测
价值[J]. 肿瘤代谢与营养电子杂志. 2022, 9(5): 602-608
Wu Jie, Zhou Jingjing,Chen Haisheng.
Predictive value of L3-skeletal muscle index for survival outcome in esophageal cancer[J]. Electronic Journal of Metabolism and Nutrition of Cancer. 2022, 9(5): 602-608
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}