NMP 22, BTA stat test and cytology in the diagnosis of bladder cancer: A comparative study

Objective: To evaluate the usefulness of the NMP 22 and BTA stat test in th e diagnosis and follow-up of bladder cancer and to compare these tests to c ytology and cystoscopy, routine diagnostic methods. Methods: 150 patients f ollowed up for bladder cancer or symptoms suggestive of bladder cancer unde rwent cystoscopy after cytology, NMP 22 and BTA stat test using a recently voided urine sample. In suspect cases, TUR and histopathological analysis w ere performed. Results: Bladder cancer was proven in 76 patients and exclud ed in 74. For NMP 22 we have used the cutoff value recommended by the manuf acturer (10 U/ml) and that obtained by our receiver-operating characteristi c curve (6 U/ml). Sensitivity was 84.21% for NMP 22 at the cutoff value of 6 U/ml and 76.32% with 10 U/ml; 72.37% for BTA stat test; 69.74% for cytolo gy, and 100% for cystoscopy. Specificity was 86.49% for NMP 22 at a cutoff value of 6 U/ml and 90.54% at 10 U/ml; 89.19% for the BTA stat test; 93.24% for cytology and 89.19% for cystoscopy. NMP 22 sensitivity for grades 1, 2 , and 3 was 68.75, 75.86 and 100%, respectively, at a cutoff value of 6 U/m l, and 50, 68.97 and 96.77%, respectively, at a cutoff level of 10 U/ml; fo r BTA stat the sensitivity was 56.25% in G1, 62.07% in G2 and 90.32% in G3, and for cytology the sensitivity was 43.75, 62.07 and 90.32%, respectively . The sensitivity of NMP 22 was 68.75% in stage Ta, 84.78% in T1 and 100% i n T2-T4 at a cutoff level of 6 U/ml and 50, 80.43 and 92.86%, respectively, at a cutoff level of 10 U/ml; BTA stat sensitivity was 50% in Ta, 73.91% i n T1 and 92.86% in T2-T4; and in cytology the results were 37.50, 73.91 and 85.71%, respectively. Using the McNemar test, there was only a significant difference between the sensitivity of NMP 22 at a cutoff level of 6 U/ml a nd cytology in the overall sample. Conclusions: The high sensitivity of the NMP 22 and BTA stat test in combination with the data obtained from the pa rameters used for the evaluation of the test demonstrate their usefulness i n the diagnosis and follow-up of bladder cancer. NMP 22 at a cutoff value o f 6 U/ml is significantly more sensitive than cytology and consequently a t horoughly valid diagnostic tool in the diagnosis of bladder cancer which ma y substitute voided urine cytology. Copyright (C) 2001 S. Karger AG, Basel.