ABSTRACT
INTRODUCTION: Transurethral resection of the prostate (TURP) is commonly performed as the surgical management of lower urinary tract symptoms due to clinically benign disease. However, prostate cancer is not uncommonly diagnosed after such a procedure. We, therefore, determined in a retrospective study the incidence and factors that might predict the detection of prostate cancer after TURP. PATIENTS AND METHODS: Between June 2005 and June 2007, a total of 476 men underwent TURP at our department. Of these, 411 men (86%) were thought to have benign disease and were included in the study. Univariate and multivariate logistic regression analyses using age, serum prostate-specific antigen (PSA), urinary retention status, prostate resection weight, FBC and U&E were performed to determine whether prostate cancer could be predicted. RESULTS: A total of 47 men (11.4%) were diagnosed with prostate cancer [24/47 with pT1a (51%) and 23/47 with pT1b (49%)]. Furthermore, the Gleason scores ranged from 5 to 9. Univariate logistic regression analyses revealed that only age (mean: 76 years, range: 54-90 vs. mean: 71 years, range: 49-91 for prostate cancer and non-cancer cases, respectively) and serum PSA (mean: 14.9 ng/ml, range: 0.4-78.0 vs. mean: 7.4 ng/ml, range: 0.2-90.0 for prostate cancer cases and non-cancer cases, respectively) were able to distinguish between cancer and non-cancer cases. In addition, using multivariate logistic regression, age and serum PSA were also the only variables that separated the two groups with a ROC-AUC of 70%. CONCLUSIONS: Our retrospective study has demonstrated that a substantial percentage of men are unexpectedly found to have prostate cancer after TURP. In addition, age and serum PSA were independent predictors of those who are likely to have prostate cancer.