肿瘤免疫微环境与营养状态生物标志物预测局部晚期结肠癌全结肠切除术后复发：列线图的开发与内部验证

Post-colectomy recurrence affects 25%-45% of patients with locally advanced colon cancer despite adjuvant chemotherapy. The tumor-immune microenvironment (TIME)-encompassing mismatch repair (MMR) status, tumor-infiltrating lymphocyte (TIL) density, PD-L1 expression, and systemic indices such as the neutrophil-to-lymphocyte ratio (NLR)-critically determines recurrence risk and immunotherapy response. Perioperative nutritional status (preoperative albumin; NRS 2002) further modulates immune competence. No validated clinical tool has integrated these multi-dimensional biomarkers for post-colectomy risk stratification.

This retrospective dual-centre cohort enrolled 341 patients (stage II-III colon cancer; total colectomy, 2020-2024). Tumour-immune biomarkers (MMR/MSI status, PD-L1 CPS, TIL density, NLR, PLR) and nutritional indicators (albumin, NRS 2002) were assessed preoperatively. Primary endpoint: 3-year disease-free survival (DFS). A nomogram was constructed from multivariable Cox regression and internally validated by 1,000-resample bootstrapping; performance was assessed by AUC, calibration, and decision curve analysis.

Among 341 patients (median follow-up 36.0 months), 94 (27.6%) recurred; 3-year DFS was 72.4%. Independent predictors of inferior DFS: pMMR/MSS status (aHR = 2.84), advanced stage (aHR = 2.52), high lymph node ratio (aHR = 2.13), low TIL density (aHR = 2.21), elevated CEA (aHR = 1.88), and high NLR (aHR = 1.78; all p ≤ 0.008). Preoperative albumin (aHR = 0.94 per 1 g/L increase, p = 0.003) and NRS 2002 >=4 (aHR = 1.58, p = 0.035) also remained independently associated with DFS. Adjuvant ICI use was associated with improved DFS (aHR = 0.48, p = 0.005). The nomogram achieved AUC = 0.841 (optimism-corrected: 0.824), good calibration (Hosmer-Lemeshow p = 0.557), and net clinical benefit across threshold probabilities 0.10-0.80.

A nomogram integrating MMR status, TIL density, PD-L1 CPS, NLR, CEA, tumour stage, lymph node ratio, and perioperative nutritional indicators robustly predicts 3-year recurrence after total colectomy for locally advanced colon cancer (AUC = 0.841), substantially outperforming clinical-only models. These hypothesis-generating findings require prospective randomised validation before clinical implementation.