预测前列腺穿刺阳性风险的列线图模型的建立

Development of a nomogram for predicting the risk of positive prostate biopsy

目的 基于单中心数据资料建立预测前列腺穿刺阳性风险的列线图模型,以提高早期预测前列腺癌的能力.方法 收集2009年7月至2015年3月行前列腺穿刺活检患者的临床资料,包括年龄、前列腺体积、游离PSA（fPSA）和总PSA （tPSA）等.选取PSA＜ 100 ng/ml、资料完整者纳入研究,共958例.中位年龄为70岁（25 ～ 89岁）,中位tPSA为11.30 ng/ml（0.09 ～98.23 ng/ml）.采用电脑随机分组法选择767例（80％）为建模组,余191例（20％）为验证组.在建模组中利用单因素和多因素Logistic分析筛选出预测前列腺癌的独立性影响因素,构建回归方程,并以此为基础建立预测前列腺穿刺结果的列线图模型.利用ROC曲线评估该模型对前列腺癌的诊断价值,并与临床常用的tPSA、f/tPSA和PSAD相比较.结果 本研究958例的穿刺结果：前列腺增生494例,前列腺癌464例.前列腺癌组与前列腺增生组的年龄分别为（71.6±8.8）岁和（67.3 ±9.7）岁,血清tPSA水平分别为（24.3±21.3） ng/ml和（12.4±12.4）ng/ml,f/tPSA分别为0.13±0.08和0.18±0.13,PSAD分别为0.68±0.47和0.23±0.47,前列腺体积分别为（69.2±50.6）cm3和（46.0±34.0）cm3,差异均有统计学意义（P＜0.05）.单因素和多因素Logistic回归分析显示,年龄（OR=1.094）、fPSA（OR=1.551）、tPSA（OR =0.958）、f/tPSA（OR=0）、PSAD（OR=13.647）和前列腺体积（OR=0.985）是前列腺癌独立的预测因素（P＜0.05）.利用上述指标构建列线图模型,该模型预测前列腺癌的ROC曲线下面积为0.854,高于tPSA（0.709）、f/tPSA（0.667）和PSAD（0.807）,差异均有统计学意义（P＜0.05）.结论 本研究构建的列线图模型较单独应用PSA及tPSA、f/tPSA和PSAD具有更高的诊断价值,该模型可在不增加患者检查项目的前提下,以更加直观、简洁的形式为患者提供个体化的前列腺穿刺阳性风险预测.

Objective We develop a nomogram for predicting the risk of positive prostate biopsy based on clinic indexes of patients.Methods We retrospectively analyzed the clinical data of patients who underwent prostate biopsy for prostate cancer in our hospital from July 2009 to March 2015.Variables including age,prostate volume,free PSA （fPSA） and total PSA （tPSA） were collected.958 patients with integrity data,which PSA was under 100 ng/ml,were enrolled in this study.Of these patients,we randomly selected 20% as validation group （191）,and the other 80% as development group （767）.We performed Logistic regression analysis to identify the independent risk factors related to positive prostate biopsy in development group.Then we constructed a prediction model according to regression equation to predict the result of prostate biopsy.Receiver operating characteristic （ROC） analysis was performed to verify the diagnostic value of the novel prediction model,which was furtherly compared with the diagnostic values of PSA alone,f/t PSA,PSAD respectively.Result The age of benign prostatic hyperplasia was （67.3 ± 9.7） years,tPSA （12.4 ± 8.8） ng/ml,the prostate volume （69.2 ± 50.6） cm3.The age of prostate cancer was（71.6 ± 9.7） years,tPSA （24.3 ± 21.3） ng/ml,the prostate volume was （46.0 ± 34.0） cm3.The fPSA,tPSA,PSAD,f/tPSA,age and prostate volume were independent risk factors by Logistic regression analysis,the ORs were 1.094,1.551,0.958,0,13.647 and 0.985 respectively （P 〈 0.05）.The nomogram based on all variables was established.The area under the ROC curve of the model （0.854） was greater than those of tPSA （0.709）,f/tPSA （0.667） and PSAD （0.807） （P 〈 0.05）.Conclusions The nomogram established in this study has better diagnostic value for prostate cancer compared with PSA and other PSA related parameters,which is a simple and convenient method without additional medical spending.