Performance of 18F-DCFPyL PET/CT in Primary Prostate Cancer Diagnosis, Gleason Grading and D'Amico Classification: A Radiomics-Based Study

Yuekai Li; Fengcai Li; Shaoli Han; Jing Ning; Peng Su; Jianfeng Liu; Lili Qu; Shuai Huang; Shiwei Wang; Xin Li; Xiang Li

doi:10.1007/s43657-023-00108-y

Performance of ¹⁸F-DCFPyL PET/CT in Primary Prostate Cancer Diagnosis, Gleason Grading and D'Amico Classification: A Radiomics-Based Study

Phenomics. 2023 Jul 25;3(6):576-585. doi: 10.1007/s43657-023-00108-y. eCollection 2023 Dec.

Authors

Yuekai Li¹, Fengcai Li², Shaoli Han³, Jing Ning³, Peng Su¹, Jianfeng Liu¹, Lili Qu¹, Shuai Huang¹, Shiwei Wang³, Xin Li¹, Xiang Li⁴

Affiliations

¹ Department of Nuclear Medicine, Qilu Hospital of Shandong University, No. 107, Cultural West Road, Jinan, 250012 China.
² Department of Hepatology, Qilu Hospital of Shandong University, Wenhuaxi Road 107#, Jinan, 250012 China.
³ Evomics Medical Technology Co., Ltd, Shanghai, 201203 China.
⁴ Division of Nuclear Medicine, Department of Biomedical Imaging and Image-Guided Therapy, Vienna General Hospital, Medical University of Vienna, 1090 Vienna, Austria.

PMID: 38223686
PMCID: PMC10781655 (available on 2024-12-01)
DOI: 10.1007/s43657-023-00108-y

Abstract

This study aimed to investigate the performance of ¹⁸F-DCFPyL positron emission tomography/computerized tomography (PET/CT) models for predicting benign-vs-malignancy, high pathological grade (Gleason score > 7), and clinical D'Amico classification with machine learning. The study included 138 patients with treatment-naïve prostate cancer presenting positive ¹⁸F-DCFPyL scans. The primary lesions were delineated on PET images, followed by the extraction of tumor-to-background-based general and higher-order textural features by applying five different binning approaches. Three layer-machine learning approaches were used to identify relevant in vivo features and patient characteristics and their relative weights for predicting high-risk malignant disease. The weighted features were integrated and implemented to establish individual predictive models for malignancy (M_m), high path-risk lesions (by Gleason score) (M_gs), and high clinical risk disease (by amico) (M_amico). The established models were validated in a Monte Carlo cross-validation scheme. In patients with all primary prostate cancer, the highest areas under the curve for our models were calculated. The performance of established models as revealed by the Monte Carlo cross-validation presenting as the area under the receiver operator characteristic curve (AUC): 0.97 for M_m, AUC: 0.73 for M_gs, AUC: 0.82 for M_amico. Our study demonstrated the clinical potential of ¹⁸F-DCFPyL PET/CT radiomics in distinguishing malignant from benign prostate tumors, and high-risk tumors, without biopsy sampling. And in vivo ¹⁸F-DCFPyL PET/CT can be considered a noninvasive tool for virtual biopsy for personalized treatment management.

Supplementary information: The online version contains supplementary material available at 10.1007/s43657-023-00108-y.

Keywords: 18F-DCFPyL positron emission tomography/computerized tomography; High-risk tumor; Prostate cancer; Radiomics; Three layer-machine learning.

© International Human Phenome Institutes (Shanghai) 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.