The Importance of Multi-Parametric MRI, PET/CT, and Biopsy for Identifying and Delineating the Extent of Locally Radiorecurrent Prostate Cancer: A Multi-institutional Analysis of the F-SHARP Clinical Trial

Up to 50% of clinical recurrences after curative-intent radiation are intraprostatic local radiorecurrences (LRR), with improved detection through the recent incorporation of multi-parametric MRI and PET/CT in workup. Salvage local therapy (SLT) is increasingly being offered, particularly focal SLT to try to reduce toxicity due to prior radiation. Limited data exist on the incremental value of each imaging modality and biopsy in defining LRR. The objective of this study is to compare the findings of MRI, PET/CT and biopsy in patients with LRR prostate cancer, and the impact each modality has on identifying recurrence and defining the extent of prostate involvement.This is a secondary analysis of 58 patients enrolled on the ongoing F-SHARP phase I/II clinical trial of salvage HDR brachytherapy from 3 institutions who underwent PSMA or fluciclovine PET/CT, MRI, and biopsy prior to enrollment. Recurrent tumor was delineated on each imaging modality and by inclusion of involved regions on biopsy. Descriptive statistics were used to compare the imaging-defined tumor with biopsy findings to assess the congruence between the imaging modalities and generate the percentage of patients with disease involvement on biopsy outside of the image-defined targets.Initial therapy was conventional/moderately hypofractionated photons in 35 patients, LDR in 13, proton therapy in 7, SBRT in 2, and neutrons in 1. Recurrence Gleason grade groups included 1 (n = 3), 2 (17), 3 (12), 4 (8), 5 (9), and uninterpretable (9). MRI/TRUS sextant + fusion biopsy was performed in 40 patients, TRUS saturation biopsy in 4, and TRUS systematic biopsy in 14. The median number of cores involved and obtained were 6 and 14. The median number of discrete lesions on biopsy in different quadrants of the prostate was 3 (1-6). The median number of discrete lesions seen on MRI was 1 (0-4). MRI did not identity a discrete lesion in 4 patients. The sensitivity of MRI for detection of the LRR was 92.8%. The false negative rate for not detecting the focus of LRR on MRI was 7.2%. 68.4% of patients had biopsy-proven cancer outside of the MRI-defined target. Fluciclovine PET/CT was used in 45 patients, and 13 had PSMA PET/CT. The median number of lesions on PET/CT was 1 (0-2). PET/CT did not identify a discrete lesion in 8 patients. The pooled sensitivity of PET/CT in detecting the focus of LRR was 86.2% (Fluciclovine: 82.2%, PSMA: 100%). PET/CT false negative rate of PET/CT for not detecting the focus of LRR was 13.8% (Fluciclovine: 17.8%, PSMA 0%). 72.41% of patients had biopsy-proven cancer outside of the PET/CT-defined target (Fluciclovine: 77.8%, PSMA: 53.8%).Although mpMRI and PET/CT are valuable tools for identifying LRR and delineating the extent of prostate/SV involvement, a thorough biopsy is mandatory if pursuing focal SLT. Such treatment should optimally be performed on a clinical trial with robust integration of all imaging and histopathologic data.