KINETICS OF TUMOR GROWTH OF PROSTATE CARCINOMA ESTIMATED USING PROSTATE-SPECIFIC ANTIGEN ALEXANDER M. TRUSKINOVSKY, ALAN W. PARTIN, AND MARTIN H. KROLL ABSTRACT Objectives. To investigate whether the second-order kinetics model of prostate-specific antigen (PSA) increase would be more appropriate in some cases than the traditionally assumed first-order model. PSA levels are used to detect tumor recurrence after radical prostatectomy, and the rate of PSA increase is used to predict patients’ prognosis. Methods. Sequential PSA values from 147 patients showing detectable PSA after radical prostatectomy were analyzed by nonlinear regression analysis. The best-fitting model was chosen using the lowest sum of squares residual error. Results. Of the 147 cases, 90 (61.2%) followed the first-order kinetics, and 57 (38.8%) were better described by second-order kinetics. The order of PSA increase correlated with the Gleason score of the tumor in the radical prostatectomy specimens. The first-order kinetics were associated with tumors with Gleason scores of 5 to 6, 7, and 9, and the second-order kinetics were associated with tumors with a Gleason score of 8 (P 0.01). Our data did not show a correlation between the order of PSA increase and the clinical prognosis. Conclusions. The presence of a group of tumors in which the increase in PSA follows second-order kinetics is significant. Such an increase may be associated with an autocatalytic mechanism assisting tumor growth. The existence of such a mechanism requires additional investigation. UROLOGY 66: 577–581, 2005. © 2005 Elsevier Inc. S ince the 1980s, prostate-specific antigen (PSA) has been the most important tumor marker in urology. 1 PSA is used to detect and stage prostate carcinoma and to monitor disease after radical prostatectomy (RP) or radiotherapy. 1 After defini- tive radiotherapy, the serum PSA concentration falls to very low, although detectable, values. 2 After RP, the serum PSA level declines to undetectable levels. The finding of measurable serum PSA con- centrations greater than 0.2 ng/mL after RP or of a rising PSA level after radiotherapy constitutes bio- chemical disease recurrence. 2–6 Mathematical modeling of PSA kinetics after radiotherapy is complex, because of the variable sources of PSA in the irradiated prostate, including the regenerating non-neoplastic prostatic tissue, recurrent prostate carcinoma, and distant metastases. 7 In contrast, biochemical recurrence is a virtually tumor-spe- cific finding after RP, because no residual benign prostatic tissue producing PSA is present. There- fore, biochemical recurrence is a direct sign of car- cinoma recurrence. 3 Changes in the serum PSA levels have been used as a surrogate marker for the rate of tumor growth. 8 The PSA doubling time is the most widely used parameter to assess the rate of PSA increase after biochemical recurrence. 4 A meaningful use of the PSA doubling time is possible only if one assumes an exponential, or first-order, tumor growth, with a corresponding exponential increase in the PSA level. Therefore, many researchers have fitted the sequential PSA values from the same patient to an exponential curve using linear and nonlinear re- gression analysis. 2 The exponential growth model assumes that the net sum of positive and negative influences on the From the Department of Laboratory Medicine and Pathology, University of Minnesota Medical School, Minneapolis, Minne- sota; Department of Urology, Johns Hopkins Medical Institutions, Baltimore, Maryland; and Department of Pathology and Labora- tory Medicine, University of Texas Southwestern School of Med- icine, Dallas, Texas Reprint requests: Alexander M. Truskinovsky, M.D., De- partment of Laboratory Medicine and Pathology, University of Minnesota Medical School, 420 Delaware Street Southeast, Mayo Mail Code 76, Minneapolis, MN 55455. E-mail: atruski@ atruski.cnc.net Submitted: November 23, 2004, accepted (with revisions): March 31, 2005 ADULT UROLOGY CME ARTICLE © 2005 ELSEVIER INC. 0090-4295/05/$30.00 ALL RIGHTS RESERVED doi:10.1016/j.urology.2005.03.085 577