Valer Algarabel J, Agüera Fernández L, Adot Zurbano JM, Virseda Chamorro M, Marcos Díez J, Gómez López A, Alpuente Román C. [Diagnostic algorithm based on a multivariate model to reduce the percentage of negative prostatic biopsies in patients with changed PSA or suspicious rectal palpation].
Actas Urol Esp 2000;
24:549-59. [PMID:
11011445 DOI:
10.1016/s0210-4806(00)72502-2]
[Citation(s) in RCA: 1] [Impact Index Per Article: 0.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022]
Abstract
OBJECTIVES
To determine what tests have a better diagnosis utility in patient with suspected prostate cancer due to PSA equal or greater than 4 ng/ml or abnormal digital rectal examination in order to reduce the number of negative prostate biopsies.
MATERIAL AND METHODS
We carried out a ultrasound-guided sextant transrectal biopsy in a series of 400 patient with prostate-specific antigen (PSA) levels equal or greater than 4 ng/ml or abnormal digital rectal examination. All patients had also transrectal rectal ultrasonography to value the echographic prostatic characteristics and the prostatic volume, and a free PSA determination.
RESULTS
The free/total PSA ratio (PSAl) and the PSA density (DPSA) were the most powerful predictors of prostate cancer, both with a 66% sensitivity and a 70% specificity (at a 0.15 cutoff), followed by the total PSA (PSAt), the digital rectal examination and the hypoechogenic prostatic nodules. We constructed a logistic multivariate model with these data. The outcome variable of logistic model was the probability of having prostate cancer. The significant predictive variables of the model were the PSAl, the DPSA, the digital rectal examination and the presence and extension of hypoechogenic prostatic. This model had a 81% sensitivity and 79% specificity at a 0.24 probability cutoff. We considered a 0.1 probability cutoff to reduce the number of false negative. With this strategy the sensitivity was 94% and the specificity 54%. If we had applied this model to the patient of our series then, we would be able to avoid prostate biopsy in 164 cases (the 41% of the patient), leaving only 7 cases without diagnosis of prostate cancer.
CONCLUSIONS
The most effective combination parameters were the digital rectal examination, PSAl, presence and extension of prostatic hypoechogenic zones and DPSA. This combination allows us to diminish the percentage of negative prostate biopsy in patient with prostate-specific antigen (PSA) levels equal or greater than 4 ng/ml or abnormal digital rectal examination without significantly descending the number of detected prostate cancers.
Collapse