Wang Changming, Dong Qifei, Yuan Lei, Zhang Zheng, Xu Shengjun, Gao Yukui, Guo Yuanyuan, Chen Mengjie, Wang Sheng, Zhuo Dong, Xiao Jun
Department of Urology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei, China.
Department of Radiology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei, China.
Int J Surg. 2025 Sep 11. doi: 10.1097/JS9.0000000000003269.
To evaluate the diagnostic performance of apparent diffusion coefficient (ADC) values for the detection of clinically significant prostate cancer (csPCa) in patients with equivocal Prostate Imaging-Reporting and Data System (PI-RADS) 3 lesions.
In this multicenter retrospective study, data from 460 eligible patients meeting predefined inclusion criteria were analyzed. Following the establishment of a standardized region of interest (ROI) delineation protocol, ADC measurements were obtained for all PI-RADS 3 lesions. Univariate and multivariate logistic regression analyses were performed to identify independent predictors. Diagnostic performance was evaluated using receiver operating characteristic (ROC) curves with calculation of the area under the curve (AUC). The multivariate model's discriminative ability was assessed through ROC analysis, while calibration was evaluated using calibration plots. Clinical utility was quantified via decision curve analysis. A risk stratification system was subsequently developed to optimize clinical decision-making.
For the 460 patients with PI-RADS 3 lesions, 108 (23.5%) were diagnosed with any grade prostate cancer, and 62 (13.5%) were diagnosed with csPCa. The results of the multivariate analysis indicated that prostate volume (OR = 0.957, 95% CI: 0.931-0.984, P = 0.002), minimum ADC (ADCmin) (OR = 0.009, 95% CI: < 0.001-0.381, P = 0.014), and lesions in the peripheral zone (OR = 6.269, 95% CI: 2.332-16.850, P < 0.001) were independent predictors of csPCa. Among the ADC parameters, ADCmin demonstrated superior diagnostic accuracy with an AUC of 0.773 (95%CI: 0.717-0.823) for csPCa. The multivariate prediction model incorporating prostate volume, ADCmin and lesion location showed good discrimination and satisfactory calibration in validation cohorts. Applying the threshold prostate volume < 50 mL or ADCmin <0.65 × 10-3 mm2/s as the diagnostic criteria of csPCa achieved very high sensitivity (93.5%) and negative predictive value (98.3%).
Among the ADC parameters, ADCmin exhibits the highest diagnostic accuracy for identifying csPCa in patients presenting with PI-RADS 3 lesions. Furthermore, we developed a prediction model and a risk stratification system to aid in clinical decision-making regarding prostate biopsy.
评估表观扩散系数（ADC）值对前列腺影像报告和数据系统（PI-RADS）3类病变患者中具有临床意义的前列腺癌（csPCa）的诊断效能。
在这项多中心回顾性研究中，分析了460例符合预定纳入标准的合格患者的数据。在建立标准化感兴趣区（ROI）勾画方案后，对所有PI-RADS 3类病变进行ADC测量。进行单因素和多因素逻辑回归分析以确定独立预测因素。使用受试者操作特征（ROC）曲线评估诊断效能，并计算曲线下面积（AUC）。通过ROC分析评估多变量模型的判别能力，同时使用校准图评估校准情况。通过决策曲线分析量化临床实用性。随后开发了一种风险分层系统以优化临床决策。
对于460例PI-RADS 3类病变患者，108例（23.5%）被诊断患有任何分级的前列腺癌，62例（13.5%）被诊断患有csPCa。多因素分析结果表明，前列腺体积（OR = 0.957，95%CI：0.931 - 0.984，P = 0.002）、最小ADC（ADCmin）（OR = 0.009，95%CI：<0.001 - 0.381，P = 0.014）和外周带病变（OR = 6.269，95%CI：2.332 - 16.850，P < 0.001）是csPCa的独立预测因素。在ADC参数中，ADCmin对csPCa的诊断准确性更高，AUC为0.773（95%CI：0.717 - 0.823）。纳入前列腺体积、ADCmin和病变位置的多变量预测模型在验证队列中显示出良好的判别能力和令人满意的校准情况。将前列腺体积<50 mL或ADCmin<0.65×10-3 mm2/s作为csPCa的诊断标准，可获得非常高的敏感性（93.5%）和阴性预测值（98.3%）。
在ADC参数中，ADCmin在识别PI-RADS 3类病变患者中的csPCa方面表现出最高的诊断准确性。此外，我们开发了一种预测模型和风险分层系统，以协助前列腺活检的临床决策。
