Objectives: To evaluate various Prostate-Specific Antigen (PSA) thresholds at which a F-fluciclovine PET scan could be considered in the setting of biochemical recurrent prostate cancer after definitive treatment.
Methods: We analyzed available records of men who underwent a F-fluciclovine PET scan after definitive therapy at a single academic institution between November 2016 to May 2018. The primary outcome was the rate of positive imaging findings at specific PSA thresholds. We then employed empiric strategies including a ROC curve and decision curve analysis to identify a specific threshold for which obtaining a positive result would be optimized.
Results: A total of 115 men underwent imaging with F-fluciclovine PET. No concerning lesions were identified in 25 (21.7%) patients, 32 (27.8%) had a solitary lesion identified, 45 (39.1%) had 2 to 5 lesions, and 13 (11.3%) had greater than 5 suspicious lesions identified. At PSA thresholds of less than 0.5, 0.5 to 2.0, and greater than 2, lesions were detected in 55.5% (12/22), 70.6% (24/34), and 91.5% (54/59) of patients respectively [P < 0.001]. Our ROC analysis yielded a PSA threshold of 2.10 while our decision curve analysis provided a PSA cutoff of 1.38.
Conclusion: This study constitutes an early single institution series evaluating the use of F-fluciclovine PET scans in the assessment of biochemically recurrent prostate cancer after definitive treatment. The probability of having positive imaging findings and increasing numbers of suspicious lesions rises with increasing PSA. Utilization of a lower PSA threshold of 0.5 may allow earlier intervention with salvage therapies in biochemical recurrence. However, using a threshold below 1 carries a higher risk of negative scans. Employing a higher PSA threshold of 1 to 2 carries greater sensitivity and specificity and may maximize identifying individuals with early BCR who may benefit from early intervention, while minimizing negative scans.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.urolonc.2020.03.021 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!