目的 探讨患者主观整体评估(PG-SGA)在结直肠癌中的预后价值,并构建一个联合患者营养状况与临床病理参数 的列线图模型。 方法 选取 2016 年 1 月至 2020 年 1 月期间于南京中医药大学南通附属医院接受根治性手术的 202 例结直肠 癌患者进行回顾性分析,使用 PG-SGA 量表评估患者术前营养状况,评估其对患者总生存(OS)率的影响。 通过单、多变量 Cox 回归分析确定结直肠癌患者的独立预后因素,并构建列线图模型,C 指数和校准曲线用于评价模型的预测能力。 结果 根 据 PG-SGA 量表,33. 7%的患者存在中度营养不良,26. 7%为严重营养不良。 严重营养不良患者肿瘤直径更大(P = 0. 044),术 后接受辅助化疗的比例明显更低(P = 0. 029)。 营养良好、中度营养不良及严重营养不良患者的 5 年 OS 率分别为 81. 6%、 63. 2%和 46. 5%,组间差异均有统计学意义(P<0. 05)。 肿瘤大小(P = 0. 046)、pT 分期(P = 0. 002)、淋巴结转移(P = 0. 001)、 脉管侵犯(P<0. 001)与 PG-SGA 量表评估结果(P<0. 001)是结直肠癌患者的独立预后因素。 基于上述预后参数构建列线图 模型,C 指数为 0. 781 (95%CI = 0. 671~ 0. 890),校准曲线显示模型预测概率与实际生存概率具有良好一致性。 结论 PG-SGA 量表评分是结直肠癌患者一项有用的生存预测工具,联合临床病理参数可进一步优化患者的预后风险分层。

Objective To investigate the prognostic value of patient - generated subjective global assessment PG - SGA for colorectal cancer patients and construct a nutrition - related nomogram integrated with nutritional status with clinicopathologic parameters for survival prediction. Method A retrospective analysis was conducted on 202 colorectal cancer patients who underwent curative resection in Nantong Affiliated Hospital of Nanjing University of Chinese Medicine from January 2016 to January 2020. Preoperative nutritional status was evaluated by PG-SGA and its impact on overall survival OS of colorectal cancer patients was analyzed. The univariate and multivariate Cox regression analysis was used to determine independent prognostic factors. A nomogram model was constructed and its predictive performance was tested by the consistency index C-index and calibration plots. Result According to the PG-SGA 33. 7% of patients had moderate malnutrition and 26. 7% were severely malnourished. Patients with severe malnutrition had a larger tumor diameter P = 0. 044 and a lower proportion of adjuvant chemotherapy following the surgery P = 0. 029 than other patients. The 5-year OS rates of patients with well nutrition moderate malnutrition and severe malnutrition were 81. 6% 63. 2% and 46. 5% respectively. There were significant survival differences between the groups P< 0. 05 . Tumor size P= 0. 046 pT stage P= 0. 002 lymph node metastasis P = 0. 001 vascular invasion P<0. 001 and PG-SGA P<0. 001 were independent prognostic factors for colorectal cancer patients. A nomogram model was developed according to the above prognostic parameters and its C-index was 0. 781 95% CI = 0. 671 - 0. 890 . The calibration plots showed a good consistency between the predicted survival probability and actual survival probability. Conclusion PG-SGA is a useful tool for survival prediction of colorectal cancer patients and it might further optimize the prognostic risk stratification of patients along with clinicopathologic parameters.