QIMP Team, Center for Medical Physics and Biomedical Engineering, Medical University of Vienna, Vienna, Austria.
Department of Biomedical Imaging and Image-guided Therapy, Division of Nuclear Medicine, Medical University of Vienna, Währinger Gürtel 18-20, 1090, Vienna, Austria.
Eur J Nucl Med Mol Imaging. 2021 Jun;48(6):1795-1805. doi: 10.1007/s00259-020-05140-y. Epub 2020 Dec 19.
Risk classification of primary prostate cancer in clinical routine is mainly based on prostate-specific antigen (PSA) levels, Gleason scores from biopsy samples, and tumor-nodes-metastasis (TNM) staging. This study aimed to investigate the diagnostic performance of positron emission tomography/magnetic resonance imaging (PET/MRI) in vivo models for predicting low-vs-high lesion risk (LH) as well as biochemical recurrence (BCR) and overall patient risk (OPR) with machine learning.
Fifty-two patients who underwent multi-parametric dual-tracer [F]FMC and [Ga]Ga-PSMA-11 PET/MRI as well as radical prostatectomy between 2014 and 2015 were included as part of a single-center pilot to a randomized prospective trial (NCT02659527). Radiomics in combination with ensemble machine learning was applied including the [Ga]Ga-PSMA-11 PET, the apparent diffusion coefficient, and the transverse relaxation time-weighted MRI scans of each patient to establish a low-vs-high risk lesion prediction model (M). Furthermore, M and M predictive model schemes were built by combining M, PSA, and clinical stage values of patients. Performance evaluation of the established models was performed with 1000-fold Monte Carlo (MC) cross-validation. Results were additionally compared to conventional [Ga]Ga-PSMA-11 standardized uptake value (SUV) analyses.
The area under the receiver operator characteristic curve (AUC) of the M model (0.86) was higher than the AUC of the [Ga]Ga-PSMA-11 SUV analysis (0.80). MC cross-validation revealed 89% and 91% accuracies with 0.90 and 0.94 AUCs for the M and M models respectively, while standard routine analysis based on PSA, biopsy Gleason score, and TNM staging resulted in 69% and 70% accuracies to predict BCR and OPR respectively.
Our results demonstrate the potential to enhance risk classification in primary prostate cancer patients built on PET/MRI radiomics and machine learning without biopsy sampling.
在临床常规中，原发性前列腺癌的风险分类主要基于前列腺特异性抗原（PSA）水平、活检样本中的 Gleason 评分以及肿瘤-淋巴结-转移（TNM）分期。本研究旨在通过机器学习探讨正电子发射断层扫描/磁共振成像（PET/MRI）体内模型预测低风险与高风险病变（LH）以及生化复发（BCR）和总体患者风险（OPR）的诊断性能。
2014 年至 2015 年期间，52 例患者接受了多参数双示踪剂[F]FMC 和[Ga]Ga-PSMA-11 PET/MRI 以及根治性前列腺切除术，作为一项随机前瞻性试验（NCT02659527）的单中心试点研究的一部分。对每位患者的[Ga]Ga-PSMA-11 PET、表观扩散系数和横向弛豫时间加权 MRI 扫描进行放射组学和集成机器学习分析，以建立低风险与高风险病变预测模型（M）。此外，通过结合患者的 M、PSA 和临床分期值，建立了 M 和 M 预测模型方案。通过 1000 倍蒙特卡罗（MC）交叉验证对建立的模型进行性能评估。结果还与传统的[Ga]Ga-PSMA-11 标准化摄取值（SUV）分析进行了比较。
M 模型的受试者工作特征曲线（ROC）下面积（AUC）（0.86）高于[Ga]Ga-PSMA-11 SUV 分析（0.80）。MC 交叉验证显示，M 和 M 模型的准确率分别为 89%和 91%，AUC 分别为 0.90 和 0.94；而基于 PSA、活检 Gleason 评分和 TNM 分期的标准常规分析预测 BCR 和 OPR 的准确率分别为 69%和 70%。
我们的结果表明，无需进行活检取样，基于 PET/MRI 放射组学和机器学习，有可能提高原发性前列腺癌患者的风险分类。
Zotero 插件
