Comparison of the nuclear matrix protein 22 with voided urine cytology and BTA stat test in the diagnosis of transitional cell carcinoma of the bladder.

Abstract

To assess sensitivity, specificity, accuracy, positive predictive value and negative predictive value of nuclear matrix protein 22 (NMP22) test, BTA stat test and cytology in the urine of patients with a spectrum of urologic conditions, including bladder cancer. A total of 140 patients (40 with bladder cancer) provided a urine sample which was divided into appropriate aliquots for each of the tests cited above. The endoscopist, pathologist, cytologist and the person performing BTA stat test and NMP22 test were blinded as to the results of the other tests. Receiver-operating characteristics curve interpretation determined that 12.0 U/ml was an optimal reference value for NMP22 to detect transitional cell carcinoma of the bladder in this patient group. Comparative results demonstrate a clear superiority of NMP22 and BTA stat tests in sensitivity in bladder cancer detection (p < 0.01), while cytology and NMP22 were better than BTA stat test in specificity (p < 0.05). NMP22 and BTA stat test results represented significant improvement over urinary cytology for detection of transitional cell carcinoma. The sensitivities of NMP22 and BTA stat tests for detection of transitional cell carcinoma in this group of patients were as much as twice that of cytology. When the cutoff value of urinary NMP22 was set at 12.0 U/ml, NMP22 was more accurate than the other tests (p < 0.05).