False-positive results of the NMP22 test due to hematuria

Purpose: The problem with available markers for bladder cancer is their low specificity and low positive predictive value due to false-positive results. False-positive results of the NMP22 nuclear matrix protein test (Matritech, Cambridge, Massachusetts) are usually observed in some clinical categories that are usually associated with hematuria and pyuria. This problem is especially serious in bladder cancer since 85% of patients present with hematuria. We investigated the effect of the degree of hematuria and pyuria on NMP22 results in an experimental model and human subjects. Materials and Methods: This study was performed in 202 urine samples from 30 healthy individuals (group 1), 20 with symptomatic urinary tract infection (group 2) and 32 with known bladder carcinoma (group 3). In the first group to achieve 0, 10, 100, 1,000 and 5,000 red blood cells per high power field the blood obtained from each patient was added to test tubes at 0.02, 0.2, 2 and 10 mul., respectively. Results: In the first group median urinary NMP22 in healthy individuals was 4 units per ml. (range 1.6 to 9.5). When blood was added to the urine sample, the NMP22 increase paralleled the increase in the amount of rod blood cells in the sediment. When greater than 2 mul./ml. blood or 1,039.5 red blood cells per high power field (range 278 to 1,438) were added to the urine of a healthy individual, the NMP22 level reached and surpassed the level in patients with bladder carcinoma. The leukocyte count in the urine sediment also had a significant impact on urinary NMP22 in group 2. The degree of hematuria and pyuria did not significantly effect NMP22 in group 3. The sensitivity, specificity, positive and negative predictive values of NMP22 were 78.1%, 66%, 59.5% and 82.5%, respectively. Test sensitivity increased as grade and stage progressed. Conclusions: In an experimental model pyuria and hematuria significantly affected urinary NMP22. The effect of white blood cells was more pronounced than that of red blood cells. The source of NMP22 in isolated hematuria remains to be elucidated. On the other hand, in group 3 the tumor was the main source of NMP22, and urinary erythrocytes and/or leukocytes had a negligible effect.