Abstract
3065
Introduction: [68Ga]-PSMA-11 positron emission tomography (PSMA PET) is a robust tool for detecting prostate cancer extent for initial staging (IS) and biochemical recurrence (BCR). Whether this tool can confidently determine the presence or absence of prostate cancer in equivocal lesions seen on conventional imaging studies has not been formally evaluated. We hypothesized that PSMA PET would increase confidence in determining the presence or absence of prostate cancer in such equivocal lesions, leading to a clinically significant changes in management (CIM).
Methods: We prospectively recruited patients with prostate cancer at any stage of diagnosis or treatment who had any equivocal lesion(s) for prostate cancer identified on computed tomography (CT), magnetic resonance imaging (MRI), bone scan, or [18F]-fluciclovine PET. Patient demographics, prostate cancer characteristics, and number and sites of equivocal lesions identified on all imaging modalities are summarized. CIM using PSMA PET information was confirmed by comparing final management decisions in the medical record with planned management without PSMA PET captured in referring provider questionnaires. PSMA PET imaging was acquired on a GE Discovery MI PET/CT scanner at 60 min after intravenous injection of 5.7 ± 1.0 mCi [68Ga]-PSMA-11. Any lesion with uptake above background at a typical metastatic location was interpreted as positive for prostate cancer by a single experienced PET reader (DLC). Total PSMA-expressing tumor volume (TTV) was calculated from all lesions with uptake above blood pool using MIM Encore (version 7.1.2, Cleveland, OH). All values were represented as the mean ± standard deviation or median [interquartile range, IQR]. Comparison of mean and proportions was performed using independent sample t-tests on MedCalc software version 20.023. Coefficient of determination (R2) determined by linear regression characterized dependence of TTV on PSA and PSAdt. P-value <0.05 determined statistical significance.
Results: 23 patients ages 69 ± 8.7 years were imaged with PSMA PET: 8 undergoing IS (Gleason score range 7 to 9) and 15 with BCR (Gleason score range at diagnosis 6 to 9) with 8 previously treated by radical prostatectomy and 7 by radiation therapy). Patients had conventional imaging at median of 54 [Q1 17, Q3 96, IQR 79] days prior to PSMA PET. Equivocal lesions were described at 79 sites, including prostate (9.3%), lymph nodes (34.3%), bones (31.2%) and other organs (25%). Sixty-one sites (77%) were identified on bone scan or [18F]-fluciclovine PET and 28 (35%) sites on CT or MRI, including 10 sites in 3 patients seen on both [18F]-fluciclovine PET and CT.
PSMA PET confidently identified presence (14 patients, 49 sites) or absence (7 patients, 25 sites) of prostate cancer in equivocal lesions in 21 patients (91.30%, p=0.002) at 74 sites (93.67%, p<0.0001). PSMA PET identified unsuspected metastases in 11 of 23 (47.82%) cases. Two patients had equivocal lesions after PSMA PET. This led to CIM in 18 of 20 patients (90%). When excluding 6 patients with conventional imaging > 90 days prior to PSMA PET, CIM rate remained high (13 out of 15 with final management decisions, 87%). TTV depended moderately on PSA (R2 = 0.79) and not on PSAdt (R2 = 0.01). After removing a single outlier from the IS group that may have driven the PSA dependency, TTV still depended moderately on PSA (R2 = 0.62). Within the IS group, TTV depended strongly on PSA (R2 = 0.87) and not PSAdt, R2 0.01. TTV did not depend on PSA or PSAdt in the BCR group.
Conclusions: In this cohort of patients with prostate cancer, PSMA PET was shown to increase confidence in determining diagnosis of metastases equivocal on conventional imaging. This led to a nearly 90% rate of change in management. Total PSMA-expressing tumor volume also appeared to be associated with PSA level at the time of imaging and not the PSA doubling time. Accrual is ongoing to confirm these initial results.
