This is the first paper to describe the incidence, amount, and duration of transient PSA elevations which subsequently fall without further treatment in patients treated with HDR brachytherapy and EBRT, with or without hormonal therapy. Multiple definitions of biochemical failure were used and all produced some false positives. The ASTRO definition of biochemical failure, one of the most common, had a 33% false positive rate.
Demanes et al, performed a review of 209 patients, without prior androgen suppression, treated with HDR-BT plus EBRT. Using a definition of failure with endpoints of local failure (determined by DRE or positive biopsy >2 years after treatment associated with PSA progression), distant failure, hormonal therapy, or a post-treatment PSA level >25 ng/mL, the authors reported 10 of 209 patients (4.8%) had false positive diagnoses of failure using the ASTRO definition. As in this study, Demanes et al found that the nadir plus 2 ng/ml definition correlated better with actual clinical outcome than the ASTRO definition. However, the authors did not further describe the nature of the amount or duration of PSA rise [24
Rising PSA values after radiation therapy are worrisome for both patients and treating physicians. There is no clear evidence that very early initiation of salvage hormone therapy improves survival [27
]. However, survival is better when hormones are initiated prior to development of distant metastases or when PSA is less than 20 [28
]. Thus, there is an understandable desire to initiate salvage hormone therapy quickly prior to further potential advancement of disease.
Transient PSA elevations, which resolve without therapy, further complicate the decision to initiate salvage hormone therapy. Such transient elevations, or benign PSA bounces, after EBRT and/or LDR brachytherapy have been described [13
]. The prognostic significance of PSA bounces following EBRT remains unclear [29
]. PSA bounces do not appear to negatively impact long term outcome following LDR brachytherapy [13
In a cohort of patients treated with LDR brachytherapy, using definition 1, Ciezki et al reported a 46% incidence of PSA bounce with a median time to occurrence of 15 months [13
]. Using bounce definition 1, we found a similar 55% incidence but all patients had experienced bounce by 12 months. Therefore, time to bounce following HDR may be faster than following LDR.
The present study is the first to report the details of transient PSA rises in a population treated with HDR brachytherapy and EBRT. The optimal PSA-based definition to predict ultimate failure remains elusive necessitating evaluation of new definitions. Such evaluations will be facilitated if, as done here, authors graphically report the duration and magnitude of the PSA elevation.
Some patients in our study also received limited duration androgen deprivation therapy which is known to produce a high incidence of PSA bounce[14
]. The small size of the present cohort combined with brief follow-up limits our ability to interrogate the causes and consequences of this PSA bounce.
We hope that our data will encourage reviews of larger databases of patients treated with HDR brachytherapy which will illuminate more optimal management thereby reducing both patient and physician anxiety.