Xu Xiao-Feng, Hu Xian-Hui, Zuo Qi-Ming, Zhang Jiao, Xu Hao-Yu, Zhang Yao
Department of Urology, the First Affiliated Hospital of Chongqing Medical University, Chongqing.
The Urinary Surgical Department of Chengdu First People's Hospital, Chengdu.
Medicine (Baltimore). 2020 May;99(20):e20167. doi: 10.1097/MD.0000000000020167.
The purpose of this study is to analyze the risk factors of sporadic renal hamartoma and establish a risk scoring system, and to intervene in patients with high-risk sporadic renal hamartoma who are prone to rupture and bleeding as soon as possible.Retrospective univariate and multivariate logistic analyzes were conducted for clinical data of 332 sporadic renal hamartoma patients to screen out independent risk factors of tumor rupture. Score of each independent risk factor was calculated. (Calculation formula: the risk coefficient of each factor = the beta regression coefficient of each factor/the minimum value of the beta regression coefficient of all factors, the value of the smallest beta regression coefficient corresponding to all the factors was assigned 1 point. The score of each factor was equal to the risk coefficient of each variable was taken as an integer value by rounding.) The total score was equal to the sum of all factors. Then the area under the receiver operating characteristics (AUC) curve was compared between high risk factors and scoring system. Finally, the scoring system was evaluated by the area under the curve (AUC) and the Hosmer-Lemeshow method in an independent cohort of 130 patients.Factors such as symptoms at presentation, tumor size, tumor blood supply, and tumor growth pattern were significant predictors of sporadic renal angiomyolipoma rupture in both the univariate and multivariate analyses; these predictors were included in the scoring system to predict sporadic renal angiomyolipoma rupture. There were no significant differences in AUCs between high risk factors and scoring system (z = 0.6434, P = .583, AUC = 0.913, and 0.903 for high risk factors and scoring system, respectively). The sporadic renal angiomyolipoma patients who scored >6 points were prone to rupture. AUROC of the scoring system in the validation set was 0.854(95%CI:0.779, 0.928). Using the Hosmer-Lemeshow method, the value of X was 2.916, P = .893, suggesting the scoring system fitted well.A scoring system based on clinical features is simple and effective in predicting sporadic angiolipoma rupture and hemorrhage. When the score is higher than 6 points, the probability of hamartoma rupture and hemorrhage is significantly increased and early intervention is needed.
本研究旨在分析散发性肾错构瘤的危险因素并建立风险评分系统，以便对易发生破裂出血的高危散发性肾错构瘤患者尽早进行干预。对332例散发性肾错构瘤患者的临床资料进行回顾性单因素和多因素逻辑分析，以筛选出肿瘤破裂的独立危险因素，并计算各独立危险因素的评分。（计算公式：各因素风险系数=各因素的β回归系数/所有因素β回归系数的最小值，将所有因素中最小β回归系数对应的值设定为1分，各因素得分等于各变量风险系数四舍五入取整后的值），总分等于所有因素得分之和。然后比较高危因素与评分系统的受试者工作特征（AUC）曲线下面积。最后，在130例患者的独立队列中，采用曲线下面积（AUC）和Hosmer-Lemeshow方法对评分系统进行评估。单因素和多因素分析结果均显示，首发症状、肿瘤大小、肿瘤血供及肿瘤生长方式等因素是散发性肾血管平滑肌脂肪瘤破裂的显著预测因素；这些预测因素被纳入评分系统以预测散发性肾血管平滑肌脂肪瘤破裂。高危因素与评分系统的AUCs差异无统计学意义（z=0.6434，P=0.583，高危因素和评分系统的AUC分别为0.913和0.903）。评分>6分的散发性肾血管平滑肌脂肪瘤患者易发生破裂。验证集中评分系统的AUROC为0.854（95%CI：0.779，0.928）。采用Hosmer-Lemeshow方法，χ²值为2.916，P=0.893，提示评分系统拟合良好。基于临床特征的评分系统在预测散发性血管脂肪瘤破裂和出血方面简单有效。当评分高于6分时，错构瘤破裂出血的概率显著增加，需要尽早干预。
