Simplified prostatitis test is put to the test

"It's too expensive, it's too difficult, and, in most patients, it doesn't make any difference," J. Curtis Nickel, MD, told Urology Times.

Before prescribing antibiotics for a man with prostatitis symptoms, some idea of whether bacteria are present in the lower urinary tract is important. Knowing that is all the more important in light of the Chronic Prostatitis Collaborative Research Network (CPCRN) study showing antibiotic therapy to be ineffective in heavily pretreated chronic prostatitis patients (Arch Intern Med 2004; 141:581-9). For these patients, "you have to do something," said Dr. Nickel.

Now he and other prostatitis experts in the National Institutes of Health's multicenter CPCRN have examined whether the test could be useful in clinical practice. They presented the results of their analysis at the AUA annual meeting.

Results compared

To gauge the accuracy of this approach, CPCRN investigators compared test data on 353 chronic prostatitis patients in the CPCRN database who had had complete Meares-Stamey four-glass test results. They compared results of cultures and leukocyte counts of midstream and postprostatic massage urine specimens, known as VB2 and VB3 samples, with the complete results from VB1(initial stream), VB2, expressed prostatic fluid (EPS), and VB3 specimens.

For analysis, the investigators carried out the Chi-square test to assess the association of leukocyte counts in EPS (used in the classic test) and VB3 specimens and constructed a receiver-operating-characteristic (ROC) curve to determine the optimal cutoff count of leukocytes in VB3 specimens that would predict the presence of leukocytes in EPS. Analysts also used 2x2 contingency tables to calculate the sensitivity and specificity of VB3 cultures in predicting EPS results and in predicting positive overall Meares-Stamey results.

The comparison of leukocyte counts in EPS and VB3 showed that leukocytes were highly likely (p<.0001) to be present in EPS when any leukocytes were in the VB3 specimen. A cutoff of 3 leukocytes per high-power field in VB3 specimens predicted a count of 5⁺ in EPS with a sensitivity of 76% and a specificity of 70%. A cutoff of 4 leukocytes per HPF in VB3 specimens predicted a count of 10 in EPS with a sensitivity of 62% and a specificity of 75%.

Finding bacteria in the VB3 specimen predicted bacteria in EPS with a sensitivity of 67.6% and a specificity of 96.5%. Localizing actual uropathogens to either EPS or VB3 confirmed a positive Meares-Stamey test.

Finding bacterial pathogens in the VB3 culture only predicted a positive Meares-Stamey test with a sensitivity of 75% and a specificity of 100%.

Although the modified test has a lower sensitivity than the Meares-Stamey test, "it's better than doing no test at all," said Dr. Nickel. Overall, the results are 96% to 98% accurate.

"We feel that we can actually say that this test is accurate enough for clinical practice," he concluded.