Abstract：
OBJECTIVE: To explore and identify the relevant clinical and pathological predictors leading to biopsy Gleason score upgrading (GSU) in cognitive fusion targeted biopsy (COG-TB) in Chinese patients.
METHODS: Clinical and pathological information of 496 patients who underwent COG-TB and radical prostatectomy (RP) in our hospital from January 2020 to September 2023 were retrospectively compiled and analyzed. In this study, we screened valuable predictors through univariable and multivariable logistic regression analyses and then constructed predictive models. We draw nomograms to visualize the predictive models. In addition, the discriminatory power of the model was assessed using receiver operating characteristic (ROC) curves. Finally, calibration curves and decision curve analysis (DCA) were used to evaluate the predictive power of the model and the net benefits it could deliver.
RESULTS: Out of the 496 patients eligible for the study, 279 had a consistent Gleason score (GS) on biopsy and postoperative GS, 191 experienced GSU, and 26 experienced downgrading. Significant associations for GSU were identified for five risk factors through multivariable logistic regression analyses, which included age, prostate volume, BMI, tumor percentage in biopsy tissue, and tumor location. Our model had excellent discriminatory power through ROC analysis. Calibration curves and DCA showed that our model was well calibrated and provided certain benefits for patient treatment decisions.
CONCLUSIONS: Age, prostate volume, BMI, tumor percentage in biopsy tissue, and tumor location are risk indicators for predicting GSU in COG-TB. Our prediction model is more suitable for Chinese patients and can assist in accurately evaluating biopsy GS and developing effective treatment plans.
METHODS: Clinical and pathological information of 496 patients who underwent COG-TB and radical prostatectomy (RP) in our hospital from January 2020 to September 2023 were retrospectively compiled and analyzed. In this study, we screened valuable predictors through univariable and multivariable logistic regression analyses and then constructed predictive models. We draw nomograms to visualize the predictive models. In addition, the discriminatory power of the model was assessed using receiver operating characteristic (ROC) curves. Finally, calibration curves and decision curve analysis (DCA) were used to evaluate the predictive power of the model and the net benefits it could deliver.
RESULTS: Out of the 496 patients eligible for the study, 279 had a consistent Gleason score (GS) on biopsy and postoperative GS, 191 experienced GSU, and 26 experienced downgrading. Significant associations for GSU were identified for five risk factors through multivariable logistic regression analyses, which included age, prostate volume, BMI, tumor percentage in biopsy tissue, and tumor location. Our model had excellent discriminatory power through ROC analysis. Calibration curves and DCA showed that our model was well calibrated and provided certain benefits for patient treatment decisions.
CONCLUSIONS: Age, prostate volume, BMI, tumor percentage in biopsy tissue, and tumor location are risk indicators for predicting GSU in COG-TB. Our prediction model is more suitable for Chinese patients and can assist in accurately evaluating biopsy GS and developing effective treatment plans.
摘要：
目的：探讨中国认知融合靶向活检(COG-TB)中Gleason评分提升(GSU)的相关临床和病理预测因素。
方法：对2020年1月至2023年9月在我院行COG-TB和前列腺癌根治术(RP)的496例患者的临床病理资料进行回顾性分析。在这项研究中,我们通过单变量和多变量逻辑回归分析筛选了有价值的预测因子，然后构建了预测模型。我们绘制列线图来可视化预测模型。此外,使用受试者工作特征(ROC)曲线评估模型的判别能力.最后,校正曲线和决策曲线分析(DCA)用于评估模型的预测能力和它可以提供的净效益。
结果：在符合研究条件的496名患者中，279在活检和术后GS上有一致的Gleason评分（GS），191名经验丰富的GSU，和26经历降级。通过多变量逻辑回归分析，确定了GSU的五个危险因素的显着关联，其中包括年龄，前列腺体积,BMI,活检组织中的肿瘤百分比，和肿瘤的位置。通过ROC分析，我们的模型具有良好的判别能力。校准曲线和DCA显示我们的模型被很好地校准并且为患者治疗决策提供了某些益处。
结论：年龄，前列腺体积,BMI,活检组织中的肿瘤百分比，和肿瘤位置是预测COG-TB中GSU的风险指标。我们的预测模型更适合中国患者，可以帮助准确评估活检GS并制定有效的治疗计划。
方法：对2020年1月至2023年9月在我院行COG-TB和前列腺癌根治术(RP)的496例患者的临床病理资料进行回顾性分析。在这项研究中,我们通过单变量和多变量逻辑回归分析筛选了有价值的预测因子，然后构建了预测模型。我们绘制列线图来可视化预测模型。此外,使用受试者工作特征(ROC)曲线评估模型的判别能力.最后,校正曲线和决策曲线分析(DCA)用于评估模型的预测能力和它可以提供的净效益。
结果：在符合研究条件的496名患者中，279在活检和术后GS上有一致的Gleason评分（GS），191名经验丰富的GSU，和26经历降级。通过多变量逻辑回归分析，确定了GSU的五个危险因素的显着关联，其中包括年龄，前列腺体积,BMI,活检组织中的肿瘤百分比，和肿瘤的位置。通过ROC分析，我们的模型具有良好的判别能力。校准曲线和DCA显示我们的模型被很好地校准并且为患者治疗决策提供了某些益处。
结论：年龄，前列腺体积,BMI,活检组织中的肿瘤百分比，和肿瘤位置是预测COG-TB中GSU的风险指标。我们的预测模型更适合中国患者，可以帮助准确评估活检GS并制定有效的治疗计划。
