ABSTRACT
OBJECTIVE: To analyze if free PSA percentage can help to predict a potential surgical failure (PSF) in patients undergoing radical prostatectomy. MATERIAL AND METHODS: Analysis of serum PSA concentration and free PSA percentage in 92 patients undergoing retropubic radical prostatectomy. In 38 cases, the carcinoma was organ-confined, 26 had capsule penetration, 20 had positive margins, 6 seminal vesicle invasion and 2 lymph nodes. PSF was demonstrated in 28 patients (30.4%) and in 64 (69.6%) the carcinoma was organ-confined. RESULTS: No significant relationship was found between PSA serum concentration or free PSA percentage to the pathological stage. The logistic regression analysis where the clinical status, Gleason sum, and free PSA percentage were included as predictive variables, showed that the latter was the only factor with capacity for PSF prediction. Over all, the probability of a carcinoma being confined in the surgical specimen when percentage of free PSA was greater than 10 was 83.8% and 60% when it was lower or equal, p < 0.03. However, the distribution was only significant when PSA concentration ranged between 4.1 and 10 ng/mL, p < 0.008. In this range of PSA, the relative risk of PSF was 5.5 (95% CI 1.4-21.8) when free PSA percentage was equal or lower than 10, the probability being 50% versus 9.1% when it was greater than 10. CONCLUSIONS: Free PSA percentage can help to predict PSF. PSA serum concentration lower than 10 ng/mL and free PSA percentage greater than 10 allows to detect a subgroup of patients with good prognosis and with less than 10% probability of having positive margins, seminal vesicles invasion or lymph nodes.
