Purpose: To investigate whether radiomic features computed on high b-value DWI sequences referring to tumour cellularity correlate with the 68GA-PSMA PET/CT ligand, highly specific for the diagnosis of prostate cancer (PCa). Materials and Methods: This study retrospectively enrols 17 patients belonging to a multi-cohort investigation for clinical impact of 3T-mpMRI and 68GA-PSMA PET/CT in PCa diagnosis and staging. PCa lesions were contoured in consensus by two-experienced radiologists in either DWI or T2w sequences, depending on where they were more visible. 40% of SUVmax was used as the threshold to contour lesions on PET images and, on these regions, the median of last decile of SUV (SUVM90th) was computed. Instead, 84 radiomic features were computed on b-2000 DWI lesions and their value was correlated to SUVM90th through absolute Spearman index (ρ). Results: Several radiomic features showed excellent correlations with 68GA-PSMA-SUVM90th. In particular, the radiomic feature performing as the best is related to local tumour heterogeneity and showed ρ≥0.7 in 82% of patients, ρ≥0.5 in just two cases, and one-only patient yielded ρ=0.3. Conclusions: The outcome reveals a rank correlation between degrees of PCa cellularity heterogeneity and 68GA- PSMA -SUVM90th. In other words, a wider expression of membrane receptors for PSMA seems corresponding to an over-proliferation of cells, which theoretically is suggestive of tumour onset and malignancy progression. Limitations: A wider cohort of patients is needed to better understand this correlation and to deepen the physiological and biomolecular causes of such behaviour.
Titolo: | Prostate cancer heterogeneity in high b-value DWI correlates with [68Ga]-PSMA PET/CT: preliminary results | |
Autore/i: | Margherita Mottola; Fabio Ferroni; Domenico Barone; Alice Turci; Monica Celli; Federica Matteucci; Giampaolo Gavelli; Giovanni Paganelli; Alessandro Bevilacqua | |
Autore/i Unibo: | ||
Anno: | 2020 | |
Rivista: | ||
Abstract: | Purpose: To investigate whether radiomic features computed on high b-value DWI sequences referring to tumour cellularity correlate with the 68GA-PSMA PET/CT ligand, highly specific for the diagnosis of prostate cancer (PCa). Materials and Methods: This study retrospectively enrols 17 patients belonging to a multi-cohort investigation for clinical impact of 3T-mpMRI and 68GA-PSMA PET/CT in PCa diagnosis and staging. PCa lesions were contoured in consensus by two-experienced radiologists in either DWI or T2w sequences, depending on where they were more visible. 40% of SUVmax was used as the threshold to contour lesions on PET images and, on these regions, the median of last decile of SUV (SUVM90th) was computed. Instead, 84 radiomic features were computed on b-2000 DWI lesions and their value was correlated to SUVM90th through absolute Spearman index (ρ). Results: Several radiomic features showed excellent correlations with 68GA-PSMA-SUVM90th. In particular, the radiomic feature performing as the best is related to local tumour heterogeneity and showed ρ≥0.7 in 82% of patients, ρ≥0.5 in just two cases, and one-only patient yielded ρ=0.3. Conclusions: The outcome reveals a rank correlation between degrees of PCa cellularity heterogeneity and 68GA- PSMA -SUVM90th. In other words, a wider expression of membrane receptors for PSMA seems corresponding to an over-proliferation of cells, which theoretically is suggestive of tumour onset and malignancy progression. Limitations: A wider cohort of patients is needed to better understand this correlation and to deepen the physiological and biomolecular causes of such behaviour. | |
Data stato definitivo: | 17-mag-2020 | |
Appare nelle tipologie: | 1.06 Abstract in rivista |
File in questo prodotto:
File | Descrizione | Tipo | Licenza | |
---|---|---|---|---|
2020_Article_ECR2020BookOfAbstracts_Extract.pdf | Versione (PDF) editoriale | Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY) | Open Access Visualizza/Apri | |
2020_Article_ECR2020BookOfAbstracts.pdf | Versione (PDF) editoriale | Licenza per Accesso Aperto. Creative Commons Attribuzione (CCBY) | Open Access Visualizza/Apri |