Department of Gastrointestinal Surgery, Division of Colorectal & Anal Surgery, Guangxi Medical University Cancer Hospital, Nanning, Guangxi Zhuang Autonomous Region, P.R. China.
Guangxi Clinical Research Center for Colorectal Cancer, Nanning, Guangxi Zhuang Autonomous Region, P.R. China.
Cancer Med. 2020 Apr;9(7):2363-2371. doi: 10.1002/cam4.2896. Epub 2020 Feb 6.
The overall survival (OS) of patients diagnosed with stage II-III colorectal cancer (CRC) can vary greatly, even between patients with the same tumor stage. We aimed to design a nomogram to predict OS in resected, stage II-III CRC and stratify patients with CRC into different risk groups.
Based on data from 873 patients with CRC, we used univariate Cox regression analysis to select the significant prognostic features, which were subjected to the least absolute shrinkage and selection operator (LASSO) regression algorithm for feature selection. Cross-validation was used to confirm suitable tuning parameters (λ) for LASSO logistic regression. Then, the nomogram was used to estimate 3- and 5-year OS based on the multivariable Cox regression model. The survival curves of the two groups were produced using the Kaplan-Meier method. Risk group stratification was performed to assess the predictive capacity of the nomogram.
Preoperative mean platelet volume, preoperative platelet distribution width, monocytes, and postoperative adjuvant chemotherapy were identified as independent prognostic factors by LASSO regression and integrated for the construction of the nomogram. The nomogram provided good discrimination, with C-indices of 0.67 and 0.69 for the training and validation sets, respectively. Calibration plots illustrated excellent agreement between the nomogram predictions and actual observations for 3- and 5-year OS. Moreover, a significant difference in OS was shown between patients stratified into different risk groups (P < .001).
We constructed and validated an original predictive nomogram for OS in patients with CRC after surgery, facilitating physicians to appraise the individual survival of postoperative patients accurately and identify high-risk patients who need more aggressive treatment and follow-up strategies.
诊断为 II-III 期结直肠癌（CRC）的患者的总生存（OS）差异很大，即使是在肿瘤分期相同的患者之间。我们旨在设计一个列线图来预测接受手术治疗的 II-III 期 CRC 患者的 OS，并将 CRC 患者分层为不同的风险组。
基于 873 例 CRC 患者的数据，我们使用单因素 Cox 回归分析选择有意义的预后特征，并采用最小绝对收缩和选择算子（LASSO）回归算法进行特征选择。交叉验证用于确认 LASSO 逻辑回归的合适调整参数（λ）。然后，使用多变量 Cox 回归模型基于列线图来估计 3 年和 5 年的 OS。使用 Kaplan-Meier 方法生成两组的生存曲线。通过风险组分层评估列线图的预测能力。
术前平均血小板体积、术前血小板分布宽度、单核细胞和术后辅助化疗通过 LASSO 回归被确定为独立的预后因素，并整合到列线图的构建中。该列线图具有良好的区分能力，训练集和验证集的 C 指数分别为 0.67 和 0.69。校准图表明，列线图预测的 3 年和 5 年 OS 与实际观察结果之间具有极好的一致性。此外，不同风险组的患者之间的 OS 差异显著（P<.001）。
我们构建并验证了一个用于预测 CRC 患者手术后 OS 的原始预测列线图，有助于医生准确评估术后患者的个体生存情况，并识别需要更积极治疗和随访策略的高危患者。
