A nomogram with coagulation markers for prostate cancer prediction in patients with PSA levels of 4–20 ng/mL

The global incidence of prostate cancer (PCa) is rising, necessitating improved diagnostic strategies. This study explores coagulation parameters’ predictive value for clinically significant PCa (csPCa) and develops a nomogram.
This study retrospectively analyzed data from 702 patients who underwent prostate biopsy at Shandong Provincial Hospital (SDPH) and 142 patients at Shandong Cancer Hospital and Institute (SDCHI). SDPH patients were randomly assigned at a 7:3 ratio for internal validation, while SDCHI data served as external validation. LASSO and logistic regression identified the best predictive factors for csPCa, which were used to construct a model. The model’s efficacy was tested using AUC, calibration curves, and decision curve analysis.
TPSA, age, D-dimer, prostate volume (PV), and digital rectal examination (DRE) were identified as independent risk factors for csPCa. A predictive model was constructed using a nomogram. The AUC for the training set was 0.841, for internal validation 0.809, and for external validation 0.814. Calibration and decision curves confirmed the model’s clinical utility.
The nomogram incorporating D-dimer, TPSA, age, PV, and DRE provides a highly accurate tool for assessing csPCa risk in individuals with PSA levels of 4–20 ng/mL, supporting personalized diagnostics and clinical decision-making.

前列腺癌（prostate cancer, PCa）的全球发病率持续攀升，亟需优化现有诊断策略。本研究探讨了凝血相关参数对临床显著性前列腺癌（clinically significant PCa, csPCa）的预测价值，并构建了列线图（nomogram）。
本研究回顾性分析了山东省立医院（Shandong Provincial Hospital, SDPH）702名接受前列腺穿刺活检患者以及山东省肿瘤医院与研究所（Shandong Cancer Hospital and Institute, SDCHI）142名患者的数据。其中山东省立医院的患者以7:3的比例随机分组用于内部验证，而山东省肿瘤医院与研究所的数据集则作为外部验证集。通过LASSO回归与logistic回归筛选出最优的csPCa预测因子，并以此构建预测模型。采用受试者工作特征曲线下面积（AUC）、校准曲线以及决策曲线分析对模型的效能进行评估。
研究筛选出总前列腺特异性抗原（TPSA）、年龄、D-二聚体（D-dimer）、前列腺体积（prostate volume, PV）以及直肠指检（digital rectal examination, DRE）为csPCa的独立危险因素。基于上述因子构建了列线图预测模型。训练集的AUC为0.841，内部验证集的AUC为0.809，外部验证集的AUC为0.814。校准曲线与决策曲线均证实了该模型的临床实用性。
该整合了D-二聚体、TPSA、年龄、前列腺体积以及直肠指检的列线图，对于前列腺特异性抗原（PSA）水平处于4~20 ng/mL的人群，可实现高精度的csPCa风险评估，可为个性化诊断及临床决策提供支持。