Department of Urology, Tianjin Institute of Urology, The Second Hospital of Tianjin Medical University, Tianjin, China.
Department of Pulmonary & Critical Care Medicine, 8th Medical Center, Chinese PLA General Hospital, Beijing, China.
BMC Cancer. 2024 Jan 30;24(1):148. doi: 10.1186/s12885-024-11870-1.
We aimed to identify preoperative predictors of aggressive pathology for cT1 solid renal cell carcinoma (RCC) by combining clinical features with qualitative and quantitative CT parameters, and developed a nomogram model.
We conducted a retrospective study of 776 cT1 solid RCC patients treated with partial nephrectomy (PN) or radical nephrectomy (RN) between 2018 and 2022. All patients underwent four-phase contrast-enhanced CT scans and the CT parameters were obtained by two experienced radiologists using region of interest (ROI). Aggressive pathology was defined as patients with nuclear grade III-IV; upstage to pT3a; type II papillary renal cell carcinoma (pRCC), collecting duct or renal medullary carcinoma, unclassified RCC or sarcomatoid/rhabdoid features. Univariate and multivariate logistic analyses were used to determine significant predictors and develop the nomogram model. To evaluate the accuracy and clinical utility of the nomogram model, we used the receiver operating characteristic (ROC) curve, calibration plot, decision curve analysis (DCA), risk stratification, and subgroup analysis.
Of the 776 cT1 solid RCC patients, 250 (32.2%) had aggressive pathological features. The interclass correlation coefficient (ICC) of CT parameters accessed by two reviewers ranged from 0.758 to 0.982. Logistic regression analyses showed that neutrophil-to-lymphocyte ratio (NLR), distance to the collecting system, CT necrosis, tumor margin irregularity, peritumoral neovascularity, and RER-NP were independent predictive factors associated with aggressive pathology. We built the nomogram model using these significant variables, which had an area under the curve (AUC) of 0.854 in the ROC curve.
Our research demonstrated that preoperative four-phase contrast-enhanced CT was critical for predicting aggressive pathology in cT1 solid RCC, and the constructed nomogram was useful in guiding patient treatment and postoperative follow-up.
本研究旨在通过结合临床特征与定性和定量 CT 参数，识别 cT1 期肾细胞癌（RCC）的侵袭性病理预测因子，并建立列线图模型。
我们回顾性分析了 2018 年至 2022 年间接受部分肾切除术（PN）或根治性肾切除术（RN）治疗的 776 例 cT1 期实体 RCC 患者。所有患者均接受了四期增强 CT 扫描，并由两名经验丰富的放射科医生使用感兴趣区域（ROI）获取 CT 参数。侵袭性病理定义为核分级 III-IV 级；分期至 pT3a 期；II 型乳头状 RCC、集合管或肾髓质癌、未分类 RCC 或肉瘤样/横纹肌样特征。采用单因素和多因素逻辑回归分析确定显著预测因子，并建立列线图模型。为了评估列线图模型的准确性和临床实用性，我们使用了受试者工作特征（ROC）曲线、校准图、决策曲线分析（DCA）、风险分层和亚组分析。
776 例 cT1 期实体 RCC 患者中，250 例（32.2%）存在侵袭性病理特征。两名阅片者评估的 CT 参数的组内相关系数（ICC）范围为 0.758 至 0.982。逻辑回归分析显示，中性粒细胞与淋巴细胞比值（NLR）、距集合系统距离、CT 坏死、肿瘤边界不规则、肿瘤周新生血管和 RER-NP 是与侵袭性病理相关的独立预测因子。我们使用这些显著变量构建了列线图模型，ROC 曲线下面积（AUC）为 0.854。
本研究表明，术前四期增强 CT 对预测 cT1 期肾细胞癌的侵袭性病理具有重要意义，构建的列线图有助于指导患者治疗和术后随访。
