肉瘤样尿路上皮癌（SUC）是一种罕见且高度恶性的膀胱癌，预后不良。目前，关于膀胱 SUC 的影像学特征以及区分其与常规尿路上皮癌 (CUC) 的可靠指标的信息有限。我们研究的目的是确定膀胱 SUC 的独特影像学特征，并确定有助于其鉴别诊断的因素。这项回顾性研究招募了 22 名膀胱 SUC 参与者和 61 名 CUC 参与者。记录两组的临床、病理和CT/MRI数据，并通过单因素分析和多项logistic回归进行比较，以区分SUC和CUC。大多数SUC位于膀胱三角区，肿瘤较大大小、形状不规则、ADC 值低、膀胱成像报告和数据系统 (VI-RADS) 评分 ≥ 4、存在坏死和侵袭性。单变量分析显示肿瘤位置、形状、最大长轴直径 (LAD)、短轴直径 (SAD)、ADC 值、VI-RADS 评分、坏死、膀胱外扩展 (EVE)、盆腔方面存在显着差异SUC 和 CUC 之间的腹膜扩散 (PPS) 和肾积水/输尿管积水 (p < .001 ~ p = .037)。多项逻辑回归发现，只有 SAD (p = .014) 和坏死 (p = .003) 成为区分 SUC 和 CUC 的独立预测因子。基于这两个因素的模型在ROC曲线分析中获得了0.849的曲线下面积（AUC）。膀胱SUC具有多种明显的影像学特征，包括三角区受累率高、肿瘤体积大、侵袭性明显并伴有坏死。具有较大 SAD 和坏死证据的膀胱肿瘤更有可能是 SUC，而不是 CUC。© 2024。作者。

Sarcomatoid urothelial carcinoma (SUC) is a rare and highly malignant form of bladder cancer with a poor prognosis. Currently, there is limited information on the imaging features of bladder SUC and reliable indicators for distinguishing it from conventional urothelial carcinoma (CUC). The objective of our study was to identify the unique imaging characteristics of bladder SUC and determine factors that aid in its differential diagnosis.This retrospective study enrolled 22 participants with bladder SUC and 61 participants with CUC. The clinical, pathologic, and CT/MRI data from both groups were recorded, and a comparison was conducted using univariate analysis and multinomial logistic regression for distinguishing SUC from CUC.The majority of SUCs were located in the trigone of the bladder and exhibited large tumor size, irregular shape, low ADC values, Vesical Imaging-Reporting and Data System (VI-RADS) score ≥ 4, the presence of necrosis, and an invasive nature. Univariate analysis revealed significant differences in terms of tumor location, shape, the maximum long-axis diameter (LAD), the short-axis diameter (SAD), ADC-value, VI-RADS scores, necrosis, extravesical extension (EVE), pelvic peritoneal spread (PPS), and hydronephrosis/ureteral effusion (p < .001 ~ p = .037) between SUCs and CUCs. Multinomial logistic regression found that only SAD (p = .014) and necrosis (p = .003) emerged as independent predictors for differentiating between SUC and CUC. The model based on these two factors achieved an area under curve (AUC) of 0.849 in ROC curve analysis.Bladder SUC demonstrates several distinct imaging features, including a high incidence of trigone involvement, large tumor size, and obvious invasiveness accompanied by necrosis. A bladder tumor with a large SAD and evidence of necrosis is more likely to be SUC rather than CUC.© 2024. The Author(s).