[PSA bounce after permanent implant prostate brachytherapy may mimic a biochemical failure]

To assess the frequency of the PSA "bouncing" phenomenon after a significant follow-up in a series of patients treated by permanent implant brachytherapy for a prostate cancer. To look for the clinical and dosimetric parameters possibly linked to this transitory secondary PSA increase. To evaluate in which percentage of cases this bouncing could have mimicked a biochemical relapse according to the ASTRO consensus criteria.

From January 1999, to December 2001, 295 patients were treated by a permanent prostate implantation (real-time technique, with free (125)I seeds- Isoseed Bebig-) by the Institut Curie-Hôpital Cochin-Hôpital Necker Paris group. The mean follow-up is 40.3 months (9-66 months). The PSA level was regularly checked, at least every 6 months. We defined as a "bouncing" all increase in PSA, starting at 0.1 ng/ml, subsequently followed by a spontaneous (without any treatment) decrease, with return to the previous level or lower. We particularly focused on the patients fulfilling the criteria for a biochemical relapse according to the ASTRO consensus (Three successive increases in PSA). A multivariate analysis tried to identify independent factors among the usual clinical and dosimetric parameters.

In our series, 161 patients (55%) showed a transitory PSA increase (bouncing) of at least 0.1 ng/ml; 145 patients (49%) a bouncing of 0.2 ng/ml, 93 patients (32%) a bouncing of 0.4 ng/ml and 43 patients (15%) a bouncing of at least 1 ng/ml. Mean PSA bounce was 0.8 ng/ml (0.1-4.1), and mean time to bounce was 19 months. Thirty-two patients (11% of the total number) presented three successive PSA increases with a significant (3 months) interval between the dosages, and therefore were to be considered as being in biochemical relapse according to the ASTRO consensus criteria. Actually, among those 32 patients, 18 (56%) subsequently showed a complete normalization of their PSA, without any treatment. Ten patients went on increasing their PSA, and were considered to be really in biochemical relapse. For the last 4 patients, the situation still remains ambiguous. In multivariate analysis, age<70 years (P<0.00001) and D90>200 Gy (P<0.003) were identified as independent factors for a PSA bouncing of at least 0.4 ng/ml.

The observed rate of 32% of patients showing a PSA bouncing of at least 0.4 ng/ml in our series is in good agreement with what has been previously reported in the literature. Age<70 years and D90>200 Gy were found to be independent factors predicting for such a secondary transitory increase in PSA. Interestingly, among 32 patients fulfilling the classical criteria of the ASTRO for a biochemical relapse, 18 (56%) subsequently showed a spontaneous PSA decrease, demonstrating that the ASTRO consensus is not well adapted to the biochemical follow-up of our patients undergoing permanent implant prostate Brachytherapy.