METHODS: Clinicopathologic information was retrospectively collected from 162 patients with high-risk non-metastatic PCa receiving NCHT and radical prostatectomy at our center. The postoperative pathological findings were used as the gold standard for evaluating the efficacy of NCHT. The least absolute shrinkage and selection operator (LASSO) was conducted to develop radiomics signature. Multivariate logistic regression analyses were conducted to identify the predictors of a positive pathological response to NCHT, and a nomogram was constructed based on these predictors.
RESULTS: Sixty-three patients (38.89%) experienced positive pathological response to NCHT. Receiver operating characteristic analyses showed that the area under the curve (AUC) of periprostatic fat (PPF) radiomics signature was 0.835 (95% CI, 0.754-0.898), while the AUC of intratumoral radiomics signature was 0.822 (95% CI, 0.739-0.888). Multivariate logistic regression analysis revealed that PSA level, PPF radiomics signature and intratumoral radiomics signature were independent predictors of positive pathological response. A nomogram based on these three predictors was constructed. The AUC was 0.908 (95% CI, 0.839-0.954). The Hosmer-Lemeshow goodness-of-fit test showed that the nomogram was well calibrated. Decision curve analysis revealed the favorable clinical practicability of the nomogram. The nomogram was successfully validated in the validation cohort. Kaplan-Meier analyses showed that nomogram and positive pathological response were significantly related with survival of PCa.
CONCLUSIONS: The radiomics-clinical nomogram based on mpMRI radiomics features exhibited superior predictive ability for positive pathological response to NCHT in high-risk non-metastatic PCa.
方法:回顾性收集了在我们中心接受NCHT和前列腺癌根治术的162例高危非转移性PCa患者的临床病理信息。术后病理结果作为评价NCHT疗效的金标准。进行最小绝对收缩和选择操作员(LASSO)以开发影像组学签名。进行多变量逻辑回归分析以确定NCHT阳性病理反应的预测因子。并根据这些预测因子构建了列线图。
结果:63例患者(38.89%)出现NCHT阳性病理反应。受试者工作特征分析显示,前列腺周围脂肪(PPF)影像组学特征曲线下面积(AUC)为0.835(95%CI,0.754-0.898),而肿瘤内影像组学特征的AUC为0.822(95%CI,0.739-0.888)。多因素logistic回归分析显示,PSA水平,PPF影像组学特征和肿瘤内影像组学特征是阳性病理反应的独立预测因子。构建了基于这三个预测因子的列线图。AUC为0.908(95%CI,0.839-0.954)。Hosmer-Lemeshow拟合优度测试表明,列线图得到了很好的校准。决策曲线分析显示列线图具有良好的临床实用性。在验证队列中成功验证了列线图。Kaplan-Meier分析表明,列线图和阳性病理反应与PCa的生存显着相关。
结论:基于mpMRI影像组学特征的影像组学临床列线图对高危非转移性PCa患者NCHT阳性病理反应表现出优越的预测能力。