综合全身免疫炎症指数和预后营养指数的化疗前列线图的开发和验证，以预测结直肠癌癌症的严重佐剂毒性
Development and validation of a pre-chemotherapy nomogram integrating systemic immune-inflammation index and prognostic nutritional index to predict severe adjuvant toxicity in colorectal cancer.
摘要(中文译文)
II/III期结直肠癌（CRC）辅助化疗常伴随严重毒性反应，导致剂量降低和治疗结局受损。尽管全身免疫炎症指数（SII）和预后营养指数（PNI）是公认的预后标志物，但其在预测化疗耐受性（反映生理储备）方面的价值仍研究不足。本研究评估化疗前基线SII和PNI，构建并验证用于预测严重化疗诱导毒性（CIT）的列线图。我们回顾性分析了306例接受奥沙利铂为基础辅助化疗（CAPOX或mFOLFOX6方案）的II/III期CRC患者。主要终点为3-4级不良事件。通过ROC分析和约登指数确定化疗前最佳截断值。将多因素logistic回归中的独立预测因子整合构建列线图。通过区分度（表观AUC及经500次Bootstrap重抽样校正的乐观度校正AUC）、校准度（校准曲线和Brier评分）及临床实用性（决策曲线分析，DCA）评估模型性能。78例（25.5%）患者发生严重CIT。基线高SII（校正OR=8.56，P<0.001）和低PNI（校正OR=4.69，P<0.001）为独立危险因素。列线图（年龄、SII、PNI）显示出优异的区分度（表观AUC：0.868；乐观度校正AUC：0.867）和良好的校准度（Brier评分：0.1114）。与低风险组相比，高风险组（概率≥37.4%）患者的中位治疗持续时间显著缩短（4.1个月 vs. 4.8个月），奥沙利铂相对剂量强度显著降低（中位数76.7% vs. 100.0%）（均P<0.001）。DCA证实该列线图较单一标志物或传统评估方法提供更大的净临床获益。化疗前SII升高和PNI降低是严重毒性反应的稳健预测因子。所提出的列线图是一种可靠、经济有效的风险分层工具，可助力高风险CRC患者实施风险适应性策略，如营养预康复、剂量调整和预防性生长因子支持。
Abstract
Severe toxicities frequently complicate adjuvant chemotherapy for stage II/III colorectal cancer (CRC), leading to dose reductions and compromised outcomes. While the Systemic Immune-Inflammation Index (SII) and Prognostic Nutritional Index (PNI) are established prognostic markers, their value in predicting chemotherapy tolerance, reflecting physiological reserve, remains understudied. This study evaluated baseline pre-chemotherapy SII and PNI to construct a validated nomogram for predicting severe chemotherapy-induced toxicity (CIT). We retrospectively analyzed 306 patients with stage II/III CRC receiving oxaliplatin-based adjuvant chemotherapy (CAPOX or mFOLFOX6). The primary endpoint was grade 3-4 adverse events. Optimal pre-chemotherapy cut-off values were determined via ROC analysis and the Youden index. Independent predictors from multivariate logistic regression were integrated into a nomogram. Performance was assessed via discrimination (apparent and optimism-corrected AUC via 500 bootstrap resamples), calibration (plots and Brier score), and clinical utility (Decision Curve Analysis, DCA). Severe CIT occurred in 78 (25.5%) patients. Baseline high SII (adjusted OR = 8.56, P < 0.001) and low PNI (adjusted OR = 4.69, P < 0.001) were independent risk factors. The nomogram (Age, SII, PNI) showed excellent discrimination (apparent AUC: 0.868; optimism-corrected AUC: 0.867) and strong calibration (Brier score: 0.1114). Patients in the high-risk group (probability ≥37.4%) experienced a significantly shorter median therapy duration (4.1 vs. 4.8 months) and a markedly lower oxaliplatin relative dose intensity (median 76.7% vs. 100.0%) compared to the low-risk group (both P < 0.001). DCA confirmed the nomogram offered greater net clinical benefit than single markers or traditional assessments. Pre-chemotherapy elevated SII and decreased PNI are robust predictors of severe toxicity. The proposed nomogram is a reliable, cost-effective tool for risk stratification, enabling risk-adapted strategies like nutritional prehabilitation, dose adjustment, and prophylactic growth factor support for high-risk CRC patients.
如何引用
AMA
Wenjing Li, Hua Zhang, Jie Wu, Liyan Jin. Development and validation of a pre-chemotherapy nomogram integrating systemic immune-inflammation index and prognostic nutritional index to predict severe adjuvant toxicity in colorectal cancer.. Frontiers in oncology. 2026; doi:10.3389/fonc.2026.1799706.
APA
Wenjing Li, Hua Zhang, Jie Wu, Liyan Jin (2026). Development and validation of a pre-chemotherapy nomogram integrating systemic immune-inflammation index and prognostic nutritional index to predict severe adjuvant toxicity in colorectal cancer.. Frontiers in oncology. https://doi.org/10.3389/fonc.2026.1799706
GB/T 7714
Wenjing Li, Hua Zhang, Jie Wu, Liyan Jin. Development and validation of a pre-chemotherapy nomogram integrating systemic immune-inflammation index and prognostic nutritional index to predict severe adjuvant toxicity in colorectal cancer.[J]. Frontiers in oncology, 2026 doi:10.3389/fonc.2026.1799706.
在 PubMed增强版 阅读全文
翻译全文 · AI 文献问答 · PDF 原文下载 · 引用导出