Prospective evaluation of prostate specific antigen (PSA), PSA density, free-to-total PSA ratio and a new formula (prostate malignancy index) for detecting prostate cancer and preventing negative biopsies in patients with normal rectal examinations and intermediate PSA levels

Objective: To improve the specificity and sensitivity of prostatic cancer detection, we prospectively evaluated total prostate specific antigen (PSA) level, PSA density, free-to-total PSA ratio and a new formula called prostate malignancy index (PMI) as a discriminator of prostate cancer in patients with intermediate PSA levels and normal digital rectal examinations. Materials and methods: Between November 995 and October 1997, 95 patients who had serum PSA levels of 4.0 to 10.0 ng/ml with normal digital rectal examinations were prospectively evaluated. All patients underwent one or two times transrectal ultrasound guided prostate biopsies. Based on age specific reference range of PSA, PSA density and % free PSA ratio, PMI was calculated for each patient. The free and total serum PSA concentrations were determined by an Immulite assay system. (Diagnostic Product Corp., Los Angeles, California). Results: Overall 20 of 95 (21%) patients had prostate cancer. There were no significant differences in patient mean age and mean total PSA between those with benign and those with malignant biopsies (p>0.05). However, there were significant differences in mean PSAD, mean free-to-total PSA ratio and mean PMI (p<0.01, p<0.05, p<0.01, respectively). Benign condition specificities for PM index, percent free PSA, PSA density and total PSA at a 90% sensitivity for prostate cancer were 48%, 10.6%, 8% and 4%, respectively. Of 95 patients, 27 (28.4%) had a PMI of equal or more than 3.1, including 12 of 75 (16%) with negative biopsy and 15 of 20 (75%) with positive biopsy. Furthermore a cutoff MI 0.86 P correctly identified 24% of benign cases without missing any prostate cancer cases. The comparison of receiver operating characteristic (ROC) curve areas showed that PMI was better than total PSA (p<0.01). Although, the area under the ROC curve of % free PSA and PSAD were higher than the area of total PSA, these differences were not statistically significant (p> 0.05). Conclusions: We concluded that the prostate malignancy index could be utilized to differentiate benign conditions from prostate cancer in patients with intermediate PSA levels and normal digital rectal examination. Also significant numbers of negative biopsies can be prevented in these patients.