Results from the OSPREY trial: A PrOspective Phase 2/3 Multi-Center Study of ¹⁸F-DCFPyL PET/CT Imaging in Patients with PRostate Cancer - Examination of Diagnostic AccuracY

Objectives: Prostate-specific membrane antigen (PSMA) is a transmembrane protein that is overexpressed by prostate cancer (PCa) cells. PSMA-based imaging is considered highly promising for PCa detection. Among these promising agents for PSMA-targeted PET imaging is the novel radiotracer 18F-DCFPyL. This prospective multicenter trial was designed to determine the safety and diagnostic performance of ¹⁸F-DCFPyL PET/CT for detecting pelvic lymph node metastases and sites of distant metastases in men with PCa.

Methods: ¹⁸F-DCFPyL PET/CT was evaluated in 385 men with high-risk PCa who were planned for radical prostatectomy with lymphadenectomy (Cohort A) or with radiological evidence of recurrent or metastatic PCa feasible for biopsy (Cohort B). 9 mCi (333 MBq) of ¹⁸F-DCFPyL was administered 1-2 hours prior to PET/CT. Co-primary endpoints of specificity and sensitivity of ¹⁸F-DCFPyL PET/CT for detecting prostate cancer metastases in pelvic lymph nodes were evaluated in Cohort A. Key secondary endpoints included incidence of adverse events, positive predictive value (PPV) and negative predictive value (NPV) of ¹⁸F-DCFPyL PET/CT imaging to detect pelvic lymph node metastases in Cohort A, and sensitivity and PPV of 18F-DCFPyL PET/CT imaging to detect prostate cancer within sites of metastasis or recurrence in Cohort B. Three central, blinded, and independent readers evaluated the ¹⁸F-DCFPyL scans. Imaging findings were compared to histopathology as the truth standard.

Results: For detection of pelvic lymph node metastases in Cohort A (n=252 evaluable), ¹⁸F-DCFPyL PET/CT imaging demonstrated a sensitivity among the three readers ranging from 30.6-41.9% (lower bound of 95% CI: 19.2-29.7%), specificity of 96.3-98.9% (lower bound of 95% CI: 93.6-96.0%), and PPV and NPV ranging from 78.1-90.5% (lower bound of 95% CI: 63.8-69.9) and 81.4-83.8% (lower bound of 95% CI: 76.4-78.9%), respectively. In Cohort B, n=93 evaluable subjects underwent biopsy. Among the three readers, a total of 74-85 subjects were ¹⁸F-DCFPyL PET/CT positive and 7-18 were ¹⁸F-DCFPyL PET/CT negative. Sensitivity and PPV ranged among the three readers from 92.9-98.6% (lower bound of 95% CI: 84.0-91.6%) and 81.2-87.8% (lower bound of 95% CI: 72.9-80.4%), respectively, with a range of 1-5 (1.4-7.1%) false negative and 9-16 (12.2-18.8%) false positive subjects. Sensitivity and PPV of ¹⁸F-DCFPyL PET/CT were also evaluated in different lesion locations on a region level (prostatic, pelvic and extra-pelvic) in Cohort B. Sensitivity and PPV ranged from 93.3-100% (lower bound of 95% CI: 68.0-76.0) and 75.0-93.8% (lower bound of 95% CI: 52.0-69.7%) for the pelvic region, respectively; and 90.9-98.2% (lower bound of 95% CI: 80.0-89.0%) and 83.1-86.2% lower bound of 95% CI: 74.0-77.0%) for the extra-pelvic region, respectively. There were no evaluable subjects with a local recurrence to the prostatic region for analysis. Overall, no drug-related serious adverse events were observed. Twenty-seven (7.0%) men experienced ≥1 drug-related adverse event, with dysgeusia (2.1%) and headache (2.1%) being most frequent.

Conclusions: ¹⁸F-DCFPyL PET/CT has a favorable toxicity profile and is generally well tolerated in patients with PCa. ¹⁸F-DCFPyL PET/CT demonstrated high sensitivity regarding distant metastatic prostate cancer with excellent specificity regarding pelvic lymph node metastases. The associated strong PPV and NPV of ¹⁸F-DCFPyL imaging in these disease settings suggest its high clinical utility. ¹⁸F-DCFPyL is currently under investigation in a phase 3 study in patients with biochemically recurrent PCa. ClinicalTrials.gov identifier NCT02981368.