免疫炎症指标对食管鳞癌新辅助放化疗疗效的预测价值

林光一; 梁伟成; 唐汉; 谭黎杰

doi:10.3760/cma.j.cn115610-20230221-00075

免疫炎症指标对食管鳞癌新辅助放化疗疗效的预测价值

The value of immune inflammatory index in predicting the therapeutic efficacy of neoadju-vant chemoradiotherapy for esophageal squamous cell carcinoma

摘要

摘要:
目的探讨免疫炎症指标对食管鳞癌新辅助放化疗疗效的预测价值。
方法采用回顾性病例对照研究方法。收集2015年12月至2020年12月复旦大学附属中山医院收治的163例食管鳞癌患者的临床病理资料；男135例,女28例；年龄为（62±8）岁。163例患者均行新辅助放化疗+食管癌根治术。观察指标：（1）免疫炎症指标与患者临床特征的关系。（2）免疫炎症指标与新辅助放化疗疗效的关系。（3）新辅助放化疗后病理学完全缓解和肿瘤退缩分级（TRG）反应良好的影响因素分析。（4）免疫炎症指标预测新辅助放化疗疗效的效能。正态分布的计量资料以x±s表示；计数资料以绝对数表示,组间比较采用χ²检验。等级资料比较采用秩和检验。采用受试者工作特征（ROC）曲线判断最佳截断值。单因素和多因素分析采用Logistic回归模型。以ROC曲线的曲线下面积（AUC）评价预测模型效能。
结果（1）免疫炎症指标与患者临床特征的关系。①全身免疫炎症指数（SII）、中性粒细胞和淋巴细胞比值（NLR）、血小板和淋巴细胞比值（PLR）的最佳截断值。ROC曲线分析结果显示：SII、NLR、PLR预测食管鳞癌患者新辅助放化疗疗效的AUC分别为0.7095%可信区间（CI）为0.61~0.77、0.78（95%CI为0.69~0.84）、0.79（95%CI为0.70~0.85）,约登指数最大值分别为0.25、0.32、0.52,最佳截断值分别为446 $\times$ 10⁹/L、2.09、138。②SII、NLR、PLR与患者临床特征的关系。将163例患者分别按SII、NLR、PLR最佳截断值分类；其中SII<446×10⁹/L 99例、SII≥446×10⁹/L 64例,NLR<2.09 107例、NLR≥2.09 56例,PLR<138 88例、PLR≥138 75例。SII<446×10⁹/L和SII≥446×10⁹/L患者肿瘤临床N分期比较,差异有统计学意义（P<0.05）。NLR<2.09和NLR≥2.09患者肿瘤临床N分期、肿瘤临床TNM分期比较,差异均有统计学意义（P<0.05）。（2）免疫炎症指标与新辅助放化疗疗效的关系。163例患者行新辅助放化疗后,54例达到病理学完全缓解,109例未达到病理学完全缓解；94例TRG反应良好,69例TRG反应较差。54例达到病理学完全缓解患者新辅助放化疗前SII（<446 $\times$ 10⁹/L、≥446 $\times$ 10⁹/L）,NLR（<2.09、≥2.09）,PLR（<138、≥138）分别为42、12例,47、7例,48、6例；109例未达到病理学完全缓解患者上述指标分别为57、52例,60、49例,40、69例；两者上述指标比较,差异均有统计学意义（χ²=9.83,16.39,39.60,P<0.05）。94例TRG反应良好患者新辅助放化疗前SII（<446 $\times$ 10⁹/L、≥446 $\times$ 10⁹/L）,NLR（<2.09、≥2.09）,PLR（<138、≥138）分别为59、35例,78、16例,56、38例；69例TRG反应较差患者上述指标分别为40、29例,29、40例,32、37例；两者新辅助放化疗前SII和PLR比较,差异均无统计学意义（χ²=0.38,2.79,P>0.05）；新辅助放化疗前NLR比较,差异有统计学意义（χ²=29.59,P<0.05）。（3）新辅助放化疗后病理学完全缓解和TRG反应良好的影响因素分析。多因素分析结果显示：新辅助放化疗前PLR<138和NLR<2.09分别是食管鳞癌患者新辅助放化疗后病理学完全缓解和TRG反应良好的独立保护因素（优势比=1.98,2.50,95%CI为1.56~2.51,1.40~4.46,P<0.05）。（4）免疫炎症指标预测新辅助放化疗疗效的效能。新辅助放化疗前PLR<138预测食管鳞癌患者新辅助放化疗后病理学完全缓解的AUC为0.79（95%CI为0.64~0.87,P<0.05）,灵敏度为0.89,特异度为0.63,约登指数为0.52。新辅助放化疗前NLR<2.09预测食管鳞癌患者新辅助放化疗后TRG反应良好的AUC为0.76（95%CI为0.64~0.81,P<0.05）,灵敏度为0.83,特异度为0.58,约登指数为0.41。
结论新辅助放化疗前PLR<138和NLR<2.09分别是食管鳞癌患者行新辅助放化疗后病理学完全缓解和TRG反应良好的独立保护因素,均可良好预测新辅助放化疗疗效。

Abstract:
Objective To investigate the value of immune inflammatory index in predic-ting the therapeutic efficacy of neoadjuvant chemoradiotherapy for esophageal squamous cell carci-noma (ESCC).
Methods The retrospective case‑control study was conducted. The clinicopatholo-gical data of 163 patients with ESCC who were admitted to Zhongshan Hospital of Fudan University from December 2015 to December 2020 were collected. There were 135 males and 28 females, aged (62±8)years. All 163 patients underwent neoadjuvant chemoradiotherapy and radical resection for ESCC. Observation indicators: (1) relationship between immune inflammatory index and clinical characteristic in patients; (2) relationship between immune inflammatory index and efficacy of neoadjuvant chemoradiotherapy in patients; (3) influencing factor analysis for pathologic complete response and good response of tumor regression grade after neoadjuvant chemoradiotherapy; (4) efficiency of immune inflammatory index in predicting efficacy of neoadjuvant chemoradiotherapy. Measurement data with normal distribution were represented as Mean±SD. Count data were described as absolute numbers, and comparison between groups was conducted using chi‑square test. Comparison of ordinal data was conducted using the rank sum test. The receiver operating characteristic (ROC) curve was used to determine the optimal cut‑off value. Univariate and multi-variate analyses were conducted using the Logistic regression model. The area under the curve (AUC) of ROC curve was used to evaluate the efficiency of predictive model.
Results (1) Relationship between immune inflammatory index and clinical characteristic in patients. ① Optimal cut‑off value of systemic immune‑inflammation index (SII), neutrophil‑lymphocyte ratio (NLR), platelet-lympho-cyte ratio (PLR). Results of ROC curve analysis showed that the AUC of SII, NLR, PLR in predicting efficacy of neoadjuvant chemoradiotherapy for patients with ESCC was 0.70(95% confidence interval as 0.61‒0.77), 0.78(95% confidence interval as 0.69‒0.84), 0.79(95% confidence interval as 0.70‒0.85), respectively, with the maximum value of Youden index and the optimal cut‑off value as 0.25, 0.32, 0.52 and 446×10⁹/L, 2.09, 138. ② Relationship between SII, NLR, PLR and clinical charac-teristic in patients. According to the optimal cut‑off value of SII, NLR, PLR, all 163 patients were divided into cases with SII <446×10⁹/L as 99, cases with SII ≥446×10⁹/L as 64, cases with NLR <2.09 as 107, cases with NLR ≥2.09 as 56, cases with PLR<138 as 88, cases with PLR ≥138 as 75, respectively. There was a significant difference in clinical N staging of tumor in patients with SII <446×10⁹/L and SII ≥446×10⁹/L (P<0.05). There were significant differences in clinical N staging and clinical TNM staging of tumor in patients with NLR<2.09 and NLR≥2.09 (P<0.05). (2) Relationship between immune inflammatory index and efficacy of neoadjuvant chemoradiotherapy in patients. Of 163 patients undergoing neoadjuvant chemoradiotherapy, there were 54 cases with pathologic complete response and 109 cases without pathologic complete response, 94 cases with good response of tumor regression grade and 69 cases with poor response of tumor regression grade. Of the 54 patients with pathologic complete response, cases with SII <446×10⁹/L and SII ≥446×10⁹/L, cases with NLR <2.09 and NLR ≥2.09, cases with PLR <138 and PLR ≥138 before neoadjuvant chemoradiotherapy were 42 and 12, 47 and 7, 48 and 6, respectively. The above indicators were 57 and 52, 60 and 49, 40 and 69 in the 109 cases without pathologic complete response. There were significant differences in the above indicators between patients with pathologic complete response and without pathologic complete response (χ²=9.83, 16.39, 39.60, P<0.05). Of the 94 cases with good response of tumor regression grade, cases with SII <446×10⁹/L and SII ≥446×10⁹/L, cases with NLR <2.09 and NLR ≥2.09, cases with PLR <138 and PLR ≥138 before neoadjuvant chemoradiotherapy were 59 and 35, 78 and 16, 56 and 38, respectively. The above indicators were 40 and 29, 29 and 40, 32 and 37 in the 69 cases with poor response of tumor regression grade. There was no significant difference in the SII and PLR (χ²=0.38, 2.79, P>0.05) and there was a significant difference in the NLR (χ²=29.59, P<0.05) between patients with good response of tumor regression grade and poor response of tumor regre-ssion grade. (3) Influencing factor analysis for pathologic complete response and good response of tumor regression grade after neoadjuvant chemoradiotherapy. Results of multivariate analysis showed that PLR <138 before neoadjuvant chemoradiotherapy was an independent protective factor for pathologic complete response in ESCC patients undergoing neoadjuvant chemoradiotherapy (odds ratio=1.98, 95% confidence interval as 1.56‒2.51, P<0.05) and NLR <2.09 before neoadjuvant chemo-radiotherapy was an independent protective factor for good response of tumor regression grade (odds ratio=2.50, 95% confidence interval as 1.40‒4.46, P<0.05). (4) Efficiency of immune inflam-matory index in predicting efficacy of neoadjuvant chemoradio-therapy. The AUC of PLR <138 before neoadjuvant chemoradiotherapy in predicting pathologic complete response of ESCC patients undergoing neoadjuvant chemoradiotherapy was 0.79(95% confidence interval as 0.64‒0.87, P<0.05), with the sensitivity, specificity and Youden index as 0.89, 0.63 and 0.52, respectively. The AUC of NLR <2.09 before neoadjuvant chemoradiotherapy in predic-ting good response of tumor regression grade of ESCC patients undergoing neoadjuvant chemoradio-therapy was 0.76 (95% confidence interval as 0.64‒0.81, P<0.05), with the sensitivity, specificity and Youden index as 0.83, 0.58 and 0.41, respectively.
Conclusion The PLR<138 and NLR <2.09 before neoadjuvant chemoradiotherapy are independent protective factors for the pathologic complete response and good response of tumor regression grade, respectively, of ESCC patients undergoing neoadjuvant chemoradiotherapy, and both of them can predict the curative effect of neoadjuvant chemoradiotherapy well.

HTML全文

参考文献(33)

施引文献

资源附件(0)