PSA and PSA Kinetics as Predictors for 18F-Fluciclovine PET/CT Positivity in Biochemically Recurrent Prostate Cancer

Abstract

<b><i>Introduction:</i></b> <sup>18</sup>F-Fluciclovine PET/CT is one of the imaging techniques currently employed to restage prostate cancer (PCa). Due to the conflicting results reported in the literature, it is not yet known at what PSA threshold <sup>18</sup>F-fluciclovine PET/CT could reliably demonstrate the presence of recurring disease. We explored the association between <sup>18</sup>F-fluciclovine PET/CT positivity and prescan PSA, PSA doubling time, and PSA velocity in patients with biochemical recurrence (BCR) of PCa after curative-intent treatment. <b><i>Methods:</i></b> Data from 59 patients who underwent <sup>18</sup>F-fluciclovine PET/CT for BCR after radical prostatectomy or radiotherapy were retrieved from a single institution database. Patients already undergone salvage treatments at the time of PET/CT, with newly diagnosed PCa or with initial diagnosis of metastatic PCa were excluded. A 2-sided independent samples Bayesian <i>t</i> test and Bayesian Mann-Whitney U test were used to assess the association between PET/CT and prescan PSA, PSA doubling time, and PSA velocity. <b><i>Results:</i></b> Evidence for no difference between PET/CT-positive and -negative patients for log-transformed PSA was found (BF<sub>01</sub> 3.61, % error: 0.01). Robustness check and sequential analysis showed stability across a wide range of prior distribution specifications. The hypothesis of no difference in terms of PSA-dt and for PSA-vel between groups was found to be more likely compared to the alternative hypothesis (BF<sub>01</sub> of 3.44 and 3.48, respectively). <b><i>Conclusion:</i></b> PSA and PSA kinetics are unlikely to be associated with <sup>18</sup>F-fluciclovine PET/CT positivity in patients with BCR, and none of these serum biomarkers might be used as single predictors of PET/CT detection. Larger studies might be needed to evaluate the role of different predictors.