目的：

探讨系统性免疫-炎症评分（Systemic immune-inflammation score，SIS）在不可切除的III期非小细胞肺癌（Non-small cell lung cancer, NSCLC）接受根治性同步放化疗（definitive chemoradiotherapy，dCRT）和dCRT联合免疫检查点抑制剂（Immune checkpoint inhibitors，ICIs）巩固治疗患者中的预后预测价值。

方法：

回顾性分析两组人群：2014年至2018年于本中心接受dCRT的227例，以及2018年至2022年于本中心接受dCRT+ICIs的183例III期不可切除NSCLC患者的单中心临床资料。收集患者治疗前至dCRT结束后1个月、3个月、出现疾病复发时的外周血细胞计数并计算SIS，其定义为“SIS=P*N/L”，主要研究终点为总生存（Overall survival，OS），次要研究终点包括无进展生存（Progression-free survival，PFS）、无局部区域复发生存（local-regional recurrence-free survival，LRFS）、无远处转移生存（Distant metastasis-free survival，DMFS）。

第一部分在dCRT队列患者中验证SIS的预后预测价值。使用X-Tile软件确定治疗前、治疗后1个月及3个月SIS对OS的最佳预测截断值（cut-off value），并依据截断值将患者分为高炎症评分组（High SIS level，HIS）和低炎症评分组（Low SIS Level，LIS）。分析SIS在区分两组的主要和次要研究终点指标时是否有统计学差异；对 SIS 预后预测值的影响因素进行评估，并将SIS预测值与其他预后判断指标进行对比。

第二部分探索SIS在dCRT+ICIs队列患者中的预后预测价值。使用X-Tile软件确定治疗前、治疗后1个月及3个月SIS对OS的最佳cut-off值并分组。分析 SIS 在区分两个组别的主次研究终点指标时有无统计学上的不同；对 SIS 预后预测值的影响因素进行评估，对 SIS 的预测值与其他预后判断指标进行比较。

以上2组队列均使用SPSS 30.0软件进行Kaplan-Meier生存分析，使用R 4.2.2绘制生存曲线。使用多因素Cox回归分析探索SIS预后预测价值的可能影响因素。通过绘制连续型时间依赖受试者工作特征曲线（time-dependent ROC）并计算曲线下面积（AUC）比较SIS及其他预后指标的预测效能。

结果：

第一部分中，dCRT队列共入组患者227例，其中108例患者出现死亡事件，161例患者出现疾病进展或复发。人群的中位OS 36.5月（95%CI 28.41-44.65）,3年OS、PFS、LRFS、DMFS率分别为50.5%、25.5%、52.7%、46.8%。dCRT队列患者基线SIS的cut-off值为590Í10⁹，疗后1个月和疗后3个月的cut-off值为970Í10⁹和1200Í10⁹。根据cut-off值对患者进行分组。生存分析结果表明疗前（SIS≥590Í10⁹, P＜0.001）、疗后1个月（SIS≥970Í10⁹, P＜0.001)和疗后3个月（SIS≥1200Í10⁹, P=0.008）较高的SIS提示OS预后不佳。但疗后3个月SIS对PFS、LRFS、DMFS不具有预测价值。多因素回归分析显示疗前SIS中，dCRT是OS预后协变量，N分期是PFS、DMFS的预后协变量；疗后1个月时SIS中，吸烟是OS预后协变量。ROC评估SIS的AUC值0.598（95%CI 0.523-0.674, P=0.011），高于除年龄（0.602, 95%CI 0.528-0.677, P=0.007）以外的其他预后预测指标。

第二部分中，dCRT+ICIs组共入组患者182例，其中35例患者出现死亡事件，101例患者出现疾病进展或复发。人群的中位OS未达到，3年OS、PFS、LRFS、DMFS率分别为76.1%、43.5%、62.6%、69.5%。dCRT+ICIs队列患者基线SIS的cut-off值为530Í10⁹，疗后1个月和疗后3个月的cut-off值为470Í10⁹和1570Í10⁹。根据cut-off值对患者进行分组。生存分析显示，疗前（SIS≥530Í10⁹, P＜0.001）和疗后3月（SIS≥1570Í10⁹, P＜0.001）较高的SIS提示OS预后不佳。疗后1个月SIS对OS、PFS、LRFS、DMFS不具备预测价值。多因素回归分析显示疗前SIS中，辅助免疫治疗是OS预后协变量，同步化疗是LRFS的预后协变量；疗后3个月时SIS中，辅助免疫治疗是OS预后协变量。ROC评估SIS的AUC值0.622（95%CI 0.534-0.709, P=0.007），高于除T分期（0.647, 95%CI 0.548-0.746, P=0.004）以外的其他预后预测指标。

结论：

1、SIS是接受同步放化疗和序贯免疫巩固治疗的不可切除III期非小细胞肺癌患者预后的独立预测因素，根治性治疗前后HIS提示患者相对不良的预后，且相对NLR、PLR等传统指标有更好的预测效能。

2、在加用免疫治疗后，系统性炎症指标仍然能够较好地预测患者的预后，且随时间变化该指标仍具有显著预测价值。

3、患者复发时SIS升高，且高于未复发患者的SIS，提示此类患者在随访期间出现SIS升高，需要警惕转移，增加随访频率和项目。

Objectives:

To investigate the prognostic value of the Systemic Immune-Inflammation Score (SIS) in patients with unresectable stage III non-small cell lung cancer (NSCLC) undergoing definitive concurrent chemoradiotherapy (dCRT) or dCRT combined with immune checkpoint inhibitors (ICIs).

Methods:

A retrospective analysis was conducted on two cohorts: unresectable stage III NSCLC patients underwent dCRT at our center between 2014 and 2018, and those treated with dCRT combined with ICIs (dCRT+ICIs) between 2018 and 2022. A total of 227 dCRT patients and 183 dCRT+ICIs patients were included. Peripheral blood counts were collected before definitive radiotherapy to calculate SIS (excluding cases with high inflammation due to non-cancer-related factors such as infection or obstructive pneumonia), one month after dCRT, and at the time of disease recurrence. Primary endpoint is overall survival (OS). Secondary endpoint includes progression-free survival (PFS), local-regional recurrence-free survival (LRFS), and distant metastasis-free survival (DMFS).

In the first part, the prognostic value of SIS was evaluated in the dCRT cohort. Optimal cutoff values for pre-treatment SIS, 1-month post-treatment SIS, and 3-month post-treatment SIS in predicting overall survival (OS) were determined using X-Tile software. Patients were stratified into a high systemic inflammation score group (HIS) and a low systemic inflammation score group (LIS) based on these cutoff values. Statistical analyses were conducted to assess whether SIS significantly distinguished primary and secondary endpoints between the two groups. Influencing factors that could affect the prognostic value of SIS were evaluated, and its predictive performance was compared with that of other established prognostic indicators.

In the second part, the prognostic value of SIS was explored in the definitive chemoradiotherapy plus immune checkpoint inhibitors (dCRT+ICIs) cohort. Optimal cutoff values for pre-treatment SIS, 1-month post-treatment SIS, and 3-month post-treatment SIS in predicting OS were determined using X-Tile software, with patients stratified into high and low SIS groups based on these values. Statistical analyses were performed to assess whether SIS significantly distinguished primary and secondary endpoints between the two groups and confounders of prognostic value of SIS were evaluated, and its predictive performance was compared with that of other established prognostic indicators.

Kaplan-Meier survival analysis was conducted using SPSS 30.0 software for both cohorts. Multivariate Cox regression analysis was applied to explore potential influencing factors of SIS prognostic value. Predictive efficacy of SIS and other prognostic indicators was compared by construction of time-dependent receiver operating characteristic curves (ROC) and calculation of the area under the curve (AUC).

Results

In the first part, a total of 227 patients were enrolled in the dCRT cohort, among whom 108 experienced death events and 161 had disease progression or recurrence. The median overall survival (OS) was 36.5 months (95%CI: 28.41–44.65), with 3-year OS, PFS, LRFS and DMFS rates of 50.5%, 25.5%, 52.7%, and 46.8%, respectively. Optimal cutoff values for baseline SIS, 1-month post-treatment SIS, and 3-month post-treatment SIS were 590×10⁹, 970×10⁹, and 1200×10⁹, respectively. Patients were then stratified into high and low SIS groups based on cutoff values. Survival analysis demonstrates that higher SIS at baseline (SIS≥590×10⁹, P<0.001), 1 month post-treatment (SIS≥970×10⁹, P<0.001), and 3 months post-treatment (SIS≥1200×10⁹, P=0.008) was associated with worse OS. However, 3-month post-treatment SIS had no predictive value for PFS, LRFS, or DMFS. Multivariate regression analysis indicated that baseline SIS and N stage were prognostic covariates for OS and PFS/DMFS in the dCRT cohort, respectively; smoking status was a covariate for OS in the 1-month post-treatment SIS model. Time-dependent ROC analysis showed that the AUC of SIS was 0.598 (95%CI: 0.523–0.674, P=0.011), higher than most prognostic indicators except age (0.602, 95%CI: 0.528–0.677, P=0.007).

In the second part, 182 patients were enrolled in the dCRT+ICIs cohort, with 35 death events and 101 disease progression/recurrence events. The median OS was not reached, with 3-year OS, PFS, LRFS, and DMFS rates of 76.1%, 43.5%, 62.6%, and 69.5%, respectively. Optimal cutoff values for baseline SIS, 1-month post-treatment SIS, and 3-month post-treatment SIS were 530×10⁹, 470×10⁹, and 1570×10⁹, respectively. Survival analysis demonstrated that higher baseline SIS (SIS≥530×10⁹, P<0.001) and 3-month post-treatment SIS (SIS≥1570×10⁹, P<0.001) were associated with worse OS, while 1-month post-treatment SIS had no predictive value for OS, PFS, LRFS, or DMFS. Multivariate regression analysis revealed that baseline SIS and adjuvant immunotherapy were prognostic covariates for OS, and concurrent chemotherapy was a covariate for LRFS in the dCRT+ICIs cohort; adjuvant immunotherapy remained a covariate for OS in the 3-month post-treatment SIS model. Time-dependent ROC analysis showed an AUC of 0.622 (95%CI: 0.534–0.709, P=0.007) for SIS, higher than most indicators except T stage (0.647, 95%CI: 0.548–0.746, P=0.004).

Conclusions

SIS is an independent prognostic factor for unresectable stage III NSCLC patients receiving concurrent chemoradiotherapy with adjuvant immune consolidation therapy. High systemic inflammation scores (HIS) before and after definitive treatment were associated with relatively poor prognosis, and SIS demonstrated better predictive efficacy compared with traditional indicators.

Systemic inflammatory markers maintained robust prognostic predictive value when combined with immunotherapy, and their significance persisted over time across different treatment phases.

Elevated SIS at the time of recurrence, which was significantly higher than that in non-relapsed patients, suggested that an increase in SIS during follow-up should alert clinicians to the risk of metastasis, warranting more frequent follow-up and expanded monitoring protocols.