Evaluation of a radiomics nomogram derived from Fluoride-18 PSMA-1007 PET/CT for risk stratification in newly diagnosed prostate cancer

• Published in: Frontiers in oncology Vol. 12; p. 1018833
• Main Authors: Wang, Zhuonan, Li, Yunxuan, Zheng, Anqi, Gao, Jungang, Yuan, Wang, Shen, Cong, Bai, Lu, Duan, Xiaoyi
• Format: Journal Article
• Language: English
• Published: Frontiers Media S.A 15.11.2022
• Subjects: 18F-PSMA-1007 PET/CT; nomogram; Oncology; prostate cancer; radiomics; risk stratification
• ISSN: 2234-943X; 2234-943X
• DOI: 10.3389/fonc.2022.1018833

Objective The aim of this study was to evaluate the performance of Fluoride-18 ( 18 F)-PSMA-1007-PET/CT radiomics for the tumor malignancy and clinical risk stratification in primary prostate cancer (PCa). Materials and Methods A total of 161 pathological proven PCa patients in a single center were retrospectively analyzed. Prostate-specific antigen (PSA), Gleason Score (GS) and PET/CT indexes (SUVmin, SUVmax, and SUVmean) were compared according to risk stratification. Radiomics features were extracted from PCa 18 F-PSMA-1007-PET/CT imaging. The radiomics score integrating all selected parameters and clinicopathologic characteristics was used to construct a binary logistic regression and nomogram classifier. Predictors contained in the individualized prediction nomogram included radiomics score, PSA level and metastasis status. Results The radiomics signature, which consisted of 30 selected features, was significantly associated with PSA level and Gleason score (P < 0.001 for both primary and validation cohorts). Predictors contained in the individualized prediction nomogram included radiomics score, PSA level and metastasis status. The model showed good discrimination with an area under the ROC curve of 0.719 for the GS. Combined clinical-radiomic score nomogram had a similar benefit to utilizing the PET/CT radiomic features alone for GS discrimination. Conclusion The 18 F-PSMA-1007-PET/CT radiomics signature can be used to facilitate preoperative individualized prediction of GS; incorporating the radiomics signature, PSA level, and metastasis status had similar benefits to those of utilizing the PET/CT radiomics features alone.