18F-DCFPyL PSMA PET imaging compared to conventional imaging in the detection of pelvic nodal metastases in patients with locally advanced or oligometastatic prostate cancer.

Abstract

36 Background: Interest has arisen in the use of prostate specific membrane antigen (PSMA) PET/CT imaging to detect prostate cancer at metastatic sites using different tracers. Here, we examined the ability of 18F-DCFPyL (DCFPyL) PSMA-based PET imaging to detect nodal disease in comparison to conventional imaging in a cohort of men with locally advanced or oligometastatic prostate cancer (PC). Methods: UW17009 is an IRB-approved open-label, single-arm trial that enrolled 26 patients with newly diagnosed advanced PC. Patients received androgen deprivation therapy and docetaxel for 3 months followed by radical prostatectomy (RP) and pelvic lymph node dissection (PLND). Exploratory interventions include PSMA PET/CT and MRI imaging as a method for determining treatment response and heterogeneity in primary PC and metastatic lesions performed before and after chemohormonal therapy. Prior to randomization, patients received DCFPyL PET/CT and PET/MR imaging as well as CTs and Bone Scans. A mean dose of 7.86 mCi DCFPyL was administered. Whole-body PET/CT images were acquired starting at approximately 60 minutes after radiotracer injection followed by dedicated pelvic PET/MR and whole-body PET/MR. PET imaging findings were compared to conventional dedicated CT imaging and were correlated to the results of final pathologic examination of each pelvic nodal dissection. Results: 26 patients underwent conventional and exploratory imaging with subsequent neoadjuvant treatment, RP and PLND. The mean diagnostic PSA was 32.1 ng/dl and 88.5% had Gleason 9 PCa. Using conventional imaging, pelvic nodal disease was identified in 6/26 patients. Pelvic lymph node uptake was identified in 12/26 patients using DCFPyL-based PSMA PET. Initial correlation of the pathologic specimens with pretreatment PSMA PET imaging revealed pelvic nodal metastatic PC in 10/12(83%) patients. On a per-lymph node packet basis (6 per patient), there were 156 evaluable regions, including 65 from patients with positive nodes. PSMA detected 14 packets that were positive for PC and 102 packets that were negative on imaging and final pathology. PC was missed in 5 packets. The mean tumor size in the missed nodes was 2.3 mm(range 1-4 mm). Calculated sensitivity was 73.7%(95% CI [48.8, 90.8]), 85.7 % specificity(95% CI[78.1, 91.4]), and 95.3 % negative predictive value(95% CI[90.5, 97.7]). Conclusions: In comparison to conventional imaging, in this cohort, DCFPyL PSMA-based PET imaging identified nodal positive disease at twice the rate and when evaluating on a per-packet basis, there was high negative predictive value. Ongoing analysis of post-chemohormonal therapy PET imaging may provide more information regarding tumor response in this cohort.