Detection of Failure Patterns Using Advanced Imaging in Patients With Biochemical Recurrence Following Low Dose Rate Brachytherapy for Prostate Cancer

Abstract

Low dose rate (LDR) brachytherapy alone or with external beam radiation (EBRT) results in excellent rates of biochemical control in patients (pts) with prostate cancer (PCa). Understanding failure patterns in pts with biochemical recurrence (BCR) may inform implant technique or pt selection. The primary objective of this study was to describe locations of recurrence on multiparametric MRI (mpMRI) and prostate specific membrane antigen (PSMA) imaging in pts with BCR after LDR brachytherapy.Pts with BCR after definitive LDR (± EBRT, ± androgen deprivation therapy (ADT)) referred for IRB approved imaging protocols were included. All pts underwent 3T mpMRI (T2 weighted, dynamic contrast-enhanced, and diffusion-weighted imaging) prospectively reviewed by prostate-dedicated radiologists. A subset of pts underwent PSMA-PET/CT imaging. Pathologic confirmation was obtained unless contraindicated. Prostate lesions were categorized as left (L)/right (R)/midline (ML); transitional zone (TZ)/peripheral zone (PZ); apical/mid/base; and anterior/posterior/central. Locations of recurrence (prostate, seminal vesicles (SV), lymph nodes (LN) and distant metastases (DM)) were recorded and reported descriptively.Between Jan 2011 and Feb 2021, 57 pts with BCR after LDR underwent prostate mpMRI. Treatment was LDR (n = 39) or EBRT + LDR (n = 18), ± ADT (n = 12). At time of LDR, initial Gleason score was ≤6 (n = 29), 7 (n = 19), 8-9 (n = 8), unknown (n = 1) and median PSA was 6 ng/mL (3.4-49). Of dosimetry available (n = 27), 2 had D90 < 90%. Median time from LDR to mpMRI was 6.7 yrs (0.8-18.3). At time of mpMRI, median pt age was 68 yrs (51-85), median PSA was 5.64 ng/mL (0.17-37.47). Lesions (n = 67) in the prostate/SV were detected in 45/57 pts on mpMRI, with a median lesion size of 1.2 cm (0.3-3.6). SV involvement was present in 20/45 patients with mpMRI detected lesions. Locations of mpMRI prostate lesions were L (n = 19) vs R (n = 16) vs ML (n = 15); TZ (n = 22) vs PZ (n = 14); apical (n = 17) vs mid (n = 25) vs base (n = 23); and anterior (n = 13) vs posterior (n = 0) vs central (n = 8). In pts with PSMA (n = 32), 31 had uptake in the prostate (n = 16, 14 detected on mpMRI) and SV (n = 12, 10 detected on mpMRI). One pt had uptake in the prostate+SV detected only on PSMA. Extraprostatic uptake on PSMA included LN (n = 15), DM (n = 5) and penile bulb (n = 1). Of 37 pts who had prostate/SV biopsy, tumor recurrence was confirmed in 31 (prostate (n = 25) and SV (n = 12)). Three of 6 pts with negative biopsy had no lesion on mpMRI. SV recurrence (imaging ± biopsy) occurred in 18/39 pts with LDR alone and in 5/18 with LDR + EBRT.In this cohort of pts with BCR following LDR, the predominant locations of local recurrence were in the anterior/central TZ and SV. PSMA corroborated the majority of mpMRI findings and often demonstrated LN and DM. These findings may inform LDR technique and impact patient selection or pre-treatment imaging evaluation; and should be confirmed in larger subsets and correlated with post-implant dosimetry.