Curative Therapy Choice and Lesion Detection by 18F-DCFPyL and Multiparametric MRI in Biochemically Recurrent Prostate Cancer

1322 Introduction: Lesion detection in men with biochemically recurrent prostate cancer (BCR) can be both challenging and important in deciding management options, especially at low PSA levels. This study compared lesion detection by multiparametric (mp) MRI and a PSMA-targeting PET agent, 18F-DCFPyL, with regards to therapy to determine if curative treatment choices impact recurrent tumor findings on imaging. Methods: This prospective analysis included men with rising prostate specific antigen (PSA) after prostatectomy and/or radiation therapy and normal conventional imaging (99mTc-bone scan and CT). Men underwent mpMRI of the pelvis on a 3-Tesla scanner. Imaging occurred within a 13 day mean interval of 18F-DCFPyL PET/CT (range 0-112 days, median 2 days). Patients were injected with a mean dose of 298.96 MBq [8.08 mCi] 18F-DCFPyL and were imaged from head to feet two hours post-injection on PET/CT. MRI and PET/CT were read independently and blinded to findings from the other modality. This is a subgroup analysis of a larger, ongoing HIPAA compliant, single institution approved trial (ClinicalTrials.gov NCT03181867). Concordance in lesion detection between 18F-DCFPyL PET/CT and mpMRI was estimated by the proportion of specific agreement defined as the conditional probability of a randomly selected modality detecting a lesion in the same location as the other modality. Concordance was defined as the rate of positive agreement between 18F-DCFPyL PET/CT and mpMRI for lesion detection. Results: A total of 85 eligible participants were evaluated. Lesion analysis included 77 men with abnormal findings detected on at least one imaging modality and excluded 8 men with negative findings on both mpMRI and 18F-DCFPyL. Prior prostatectomy occurred in 38% (29/77 of men; median time from surgery to imaging of 2.4 years (range 2 months - 20 years) and median PSA 0.44 ng/mL (range 0.2 - 7.42 ng/mL). Radiation therapy (external beam radiation therapy or brachytherapy) transpired in 33% (25/77) of men; median time from therapy to imaging of 3.3 years (range 27 days - 8.6 years) and median PSA 4.51 ng/mL (range 0.57 - 27.45 ng/mL). Men underwent both therapies in 30% (23/77) of the population; median time from last treatment to imaging of 3.4 years (range 8.8 months - 14 years) and median PSA 3.23 ng/mL (range 0.41 - 22.37 ng/mL). Median initial Gleason score was 7 (3+4) in the surgery and radiation groups and Gleason 7 (4+3) in the dual therapy group. Only lesions seen by either modality within the pelvic MRI field of view were included (282 lesions). Total lesion counts by group were surgery = 86, radiation = 95 and both =101. In the surgical group, 18F-DCFPyL detected 61/86 lesions (71%) while MRI detected 35/86 lesions (41%) with a positive concordance rate of 21% (20/96; 95% CI 9.5,41). In the radiation group, 18F-DCFPyL detected 78/95 lesions (82%) and MRI detected 36/95 lesions (38%) with positive concordance of 33% (38/114; 95% CI 18,55). In the dual therapy group, 18F-DCFPyL detected 89/101 lesions (88%) and MRI detected 24/101 lesions (24%) with positive concordance of 21% (24/113; 95% CI 8.8,37). There were no significant differences in concordance rates among the groups (p>0.05). MRI detected more prostate bed lesions than PET in all groups. Conclusions: 18F-DCFPyL detected more lesions than mpMRI within the same FOV across all therapy groups with the most lesions in those who had undergone both prostatectomy and radiation (89 lesions). MRI detected more lesions than 18F-DCFPyL in the prostate bed, primarily in the surgical and radiotherapy groups (35 and 36 lesions, respectively) with the least number of lesions seen in the dual therapy group (24 lesions). Concordance rates between the two modalities was comparably modest among the treatment groups (20.8% for surgery, 33.3% for radiation and 21.2% for both.) Curative therapy method (surgery, radiation or both) appears to influence lesion detection for 18F-DCFPyL and MRI in patients with BCR.