Clinical Research Center, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, 197 Ruijin Er Road, Huangpu District, Shanghai, 200025, China.
BMC Cancer. 2019 Jul 23;19(1):726. doi: 10.1186/s12885-019-5958-9.
The aim of this study was to identify the determinants of overall survival (OS) within patients over 40 years old with surgically resected pancreatic carcinoma (PC), and to develop a nomogram with the intention of OS predicting.
A total of 6341 patients of 40 years of age or later with surgically resected PC between 2010 and 2015 were enrolled from the Surveillance, Epidemiology, and End Results (SEER) program and randomly assigned into training set (4242 cases) and validation set (2099 cases). A nomogram was constructed for predicting 1-, 2- and 3-years OS based on univairate and multivariate Cox regression. The C-index and calibration plot were adopted to assess the nomogram performance.
Our analysis showed that age, location of carcinoma in pancreas, tumor grade, TNM stage, size of carcinoma together with lymph node ratio (LNR) were considered to be independent overall survival predictors. A nomogram based on these six factors was developed with C-index being 0.680 (95%CI: 0.667-0.693). All calibration curves of OS fitted well. The OS curves stratified by nomogram-predicted probability score (≥20, 10-19 and < 10) demonstrated statistically significant difference not only within training set but also in validation set.
The present nomogram for OS predicting can serve as the efficacious survival-predicting model and assist in accurate decision-making for patients over 40 years old with surgically resected PC.
本研究旨在确定 40 岁以上接受手术切除胰腺癌 (PC) 治疗的患者总生存 (OS) 的决定因素,并开发一个列线图来预测 OS。
本研究从监测、流行病学和最终结果 (SEER) 计划中纳入了 2010 年至 2015 年间年龄在 40 岁及以上且接受手术切除 PC 的 6341 例患者,并将其随机分为训练集 (4242 例) 和验证集 (2099 例)。基于单变量和多变量 Cox 回归构建了用于预测 1、2 和 3 年 OS 的列线图。采用 C 指数和校准图评估列线图的性能。
我们的分析表明,年龄、胰腺癌位置、肿瘤分级、TNM 分期、肿瘤大小以及淋巴结比值 (LNR) 被认为是独立的总生存预测因素。基于这六个因素开发的列线图具有 0.680 的 C 指数 (95%CI:0.667-0.693)。所有 OS 校准曲线拟合良好。在训练集和验证集中,根据列线图预测概率评分 (≥20、10-19 和 <10) 分层的 OS 曲线均存在统计学差异。
本研究提出的 OS 预测列线图可以作为有效的生存预测模型,有助于为 40 岁以上接受手术切除 PC 的患者做出准确的决策。