Objectives: To compare the results of the BTA Trak test with voided urine cytology (VUC) in the diagnosis and follow-up of bladder tumors.
Patients and methods: Urine samples were obtained from 53 patients with bladder tumor (77 samples) and 53 patients treated for bladder tumor with no evidence of disease on the basis of cystoscopic evaluation (88 samples). Urine samples were collected prior to cystoscopy. The BTA assay was performed by the BTA Trak test according to the manufacturer's recommendations. A value >14 U/ml was considered abnormal.
Results: There was a statistically significant increase in median BTA value with increasing stage of tumor: 11.9, 57.9 and 391.0 U/ml respectively for stages pTa, pT1 and pT2/3 (p<0.0001, Kruskal-Wallis test). There was also a correlation between increasing grade and median BTA values measured at 6.9, 13.1 and 235.0 U/ml in grades 1, 2 and 3 tumors respectively (p<0.0001, Kruskall-Wallis test). The overall sensitivity of the BTA Trak test was 58.4% compared to 46.7% for VUC, a difference of 11.7%, which was statistically significant (McNemar test, p<0.005). The sensitivity of both tests combined was 63.6%. The specificity of the VUC (94.3%) was significantly higher than that of the BTA Traktrade mark (75.0%) (p<0.005, McNemar test). The accuracy of the Bard Trak test (67.3%) was similar to that of VUC (66.9%).
Conclusion: The BTA Trak test is more sensitive than urinary cytology in the detection of bladder tumors but the improvement involved is insufficient to consider decreasing the frequency of endoscopic examinations in the follow-up of superficial bladder tumor.