Xie Wenhao, Zhang Yida, Cao Runfu
Department of Urology, The First Affiliated Hospital of Nanchang University, Nanchang, China.
Front Surg. 2022 Oct 24;9:1025213. doi: 10.3389/fsurg.2022.1025213. eCollection 2022.
Primary adrenal malignant tumor is rare. The factors affecting the prognosis remain poorly defined. This study targeted to construct and corroborate a model for predicting the overall survival of adrenal malignant tumor patients.
We investigated the SEER database for patients with primary adrenal malignant tumor. 1,080 patients were divided into a construction cohort ( = 756) and a validation cohort ( = 324), randomly. The prognostic factors for overall survival were evaluated using univariate and multivariate Cox analyses. The nomogram was constructed and then validated with C-index, calibration curve, time-dependent ROC curve, and decision curve analysis in both cohorts. Then we divided the patients into 3 different risk groups according to the total points of the nomogram and analyzed their survival status by Kaplan-Meier curve with log-rank test.
The baseline characteristics of these two cohorts were not statistically different ( > 0.05). Using univariate and multivariate Cox analyses, 5 variables, including age, tumor size, histological type, tumor stage, and surgery of primary site, were distinguished as prognostic factors ( < 0.05). Based on these variables, we constructed a nomogram to predict the 3- year, 5- year, and 10-year overall survival. The C-indexes were 0.780 (0.760-0.800) in the construction cohort and 0.780 (0.751-0.809) in the validation cohort. In both cohorts, the AUC reached a fairly high level at all time points. The internal and external calibration curves and ROC analysis showed outstanding accuracy and discrimination. The decision curves indicated excellent clinical usefulness. The best cut-off values for the total points of the nomogram were 165.4 and 243.1, and the prognosis was significantly different for the three different risk groups ( < 0.001).
We successfully constructed a model to predict the overall survival of primary adrenal malignant tumor patients. This model was validated to perform brilliantly internally and externally, which can assist us in individualized clinical management.
原发性肾上腺恶性肿瘤较为罕见。影响其预后的因素仍不明确。本研究旨在构建并验证一个用于预测肾上腺恶性肿瘤患者总生存期的模型。
我们在监测、流行病学与最终结果（SEER）数据库中调查原发性肾上腺恶性肿瘤患者。1080例患者被随机分为构建队列（n = 756）和验证队列（n = 324）。使用单因素和多因素Cox分析评估总生存期的预后因素。构建列线图，然后在两个队列中用C指数、校准曲线、时间依赖性ROC曲线和决策曲线分析进行验证。然后根据列线图的总分将患者分为3个不同风险组，并通过Kaplan-Meier曲线和对数秩检验分析他们的生存状况。
这两个队列的基线特征无统计学差异（P>0.05）。通过单因素和多因素Cox分析，年龄、肿瘤大小、组织学类型、肿瘤分期和原发部位手术这5个变量被确定为预后因素（P<0.05）。基于这些变量，我们构建了一个列线图来预测3年、5年和10年总生存期。构建队列中的C指数为0.780（0.760 - 0.800），验证队列中的C指数为0.780（0.751 - 0.809）。在两个队列中，AUC在所有时间点都达到了相当高的水平。内部和外部校准曲线以及ROC分析显示出出色的准确性和区分度。决策曲线表明具有出色的临床实用性。列线图总分的最佳截断值分别为165.4和243.1，三个不同风险组的预后有显著差异（P<0.001）。
我们成功构建了一个预测原发性肾上腺恶性肿瘤患者总生存期的模型。该模型在内部和外部均被验证表现出色，可协助我们进行个体化临床管理。
