Purpose: Bone scintigrams of patients with increasing serum prostate specific antigen (PSA) after radical prostatectomy are only rarely positive. We identify clinical parameters that would improve our ability to select patients for this imaging study.
Materials and methods: We reviewed all bone scintigrams done at our institution between 1991 and 1996 in patients with persistently increasing serum PSA after radical prostatectomy. What prompted the clinician to obtain the bone scintigram was trigger PSA (tPSA). The rate of increase in PSA to tPSA was measured by tPSA/time from radical prostatectomy (slope 1) and tPSA/time from last undetectable PSA (slope 2). These parameters were evaluated together with standard clinicopathological data in univariate and multivariate analyses to determine the ability to predict the bone scintigram result.
Results: In univariate analysis tPSA (p = 0.003), slope 1 (p = 0.005) and slope 2 (p = 0.004) were useful in predicting the bone scintigram result but pathological stage, Gleason score, preoperative PSA and time to recurrence were not. In multivariate analysis the single most useful parameter in predicting the bone scintigram result was tPSA (p = 0.01). Based on a logistic regression model the probability of a positive bone scintigram was less than 5% until tPSA increased to 40 to 45 ng./ml.
Conclusions: In patients with increasing serum PSA after radical prostatectomy current serum PSA is the best predictor of the bone scintigram result. Furthermore, there is limited usefulness of bone scintigraphy until PSA increases above 30 to 40 ng./ml.