Purpose: To study the response of nonmalignant prostatic tissue to ionizing irradiation in terms of the resultant changes in serum prostate specific antigen (PSA) levels.
Methods and materials: Weekly serum PSA values were determined during radiotherapy (RT) in nine patients ("treatment group") without clinical evidence of prostate cancer (PC), and who received pelvic RT for other indications. Slopes for the rate of change in PSA was determined using model: log PSA = beta 0 + beta 1*week + beta 2*week2 + error. These results are compared with 17 normal volunteers ("control group") who were not exposed to ionizing irradiation. An attempt is made to compare any similarities and differences in subsets of 64 T1-T4N0M0 PC patients who received pelvic RT.
Results: An elevation in the serum PSA levels were noted in eight of nine patients in the "treatment group" with a median time of 4.2 weeks to reach the maximum serum PSA values. After an initial increase, PSA values declined. In some patients, manifold increase in PSA was noted, for example, from 1.8 to 13.5 ng/ml and 3.3 to 9.8 ng/ml in two patients. The PSA increase ranged from 50-650%. The median slope was 0.601 week-1 (range 0.192-3.045 week-1). No such increases were seen in the "control group" (median slope = 0.03 week-1; range, 0.18-0.13 week-1). When differences between the mean increase/decrease for each week compared to pretreatment values were analyzed, the irradiated group had statistically significant elevations in the PSA for weeks 3 (p = 0.034), 4 (p = 0.035), and 5 (p = 0.024). A similar trend of increasing PSA levels during radiotherapy was noted in prostate cancer patients whose initial PSA values were < or = 20 ng/ml: whereas positive slopes (i.e., increasing PSA levels during radiotherapy course) was seen in 7.1% of those with > 20 ng/ml preradiotherapy PSA values, such trends were seen in 52.7% of those with < or = 20 ng/ml preradiotherapy PSA values.
Conclusions: (a) Incidental exposure of noncancerous prostate to ionizing irradiation appears to lead to an initial elevation followed by a decline in serum PSA. (b) Similar elevations in serum PSA levels are seen in over 50% of prostate cancer patients with < or = 20 ng/ml pretreatment PSA. (c) Acinal cell death and sudden release of PSA into the circulation is the most likely explanation for our observations, although other mechanism cannot be excluded. (d) Our observations have to be considered in modeling PSA kinetics induced by RT and in correlating such kinetics to long-term outcomes. (e) Our findings in the control group indicate that there appears to be no significant variation in serum PSA over many weeks under physiological conditions in normal healthy ambulatory men.