Purpose: The noninvasive detection of urothelial carcinoma remains challenging. We prospectively evaluated urine markers for bladder carcinoma. We compared the NMP22 (Matritech, Cambridge, Massachusetts) and BTA Stat (Bard Diagnostics, Redmond, Washington) tests with immunocytology using mAbs 486p3/12 and BG7 against Lewis X antigen.
Materials and methods: The NMP22 and BTA Stat tests were performed in urine samples and immunocytology with mAbs 486p3/12 and BG7 staining were performed in bladder washing specimens in 146 samples of 115 patients undergoing transurethral resection for suspected bladder carcinoma (70) or 45 undergoing followup cystoscopy for a history of bladder carcinoma (76). Bladder carcinoma was detected in 54 patients, including stages pTa in 25, pT1 in 20, pT2 in 8 and carcinoma in situ in 1, while 61 had no evidence of bladder carcinoma. The cutoff was 10 units per ml. for the NMP22, 30% positive cells for 486p3/12 and 5% positive cells for the Lewis X tests.
Results: BTA Stat was positive in 65 samples (44.5%) and NMP22 was positive in 69 (47.3%). Immunocytology with mAbs 486p3/12 and BG7 against Lewis X was positive in 52 (35.6%) and 109 (74.7%) samples, respectively. Sensitivity was 70.3% for BTA Stat, 68.5% for NMP22, 68.5% for 486p3/12 and 94.4% for Lewis X. Specificity was 70.6% for BTA Stat, 65.2% for NMP22, 83.6% for 486p3/12 and 36.9% for Lewis X. Area under the receiver operating characteristics curve was 0.6804 for NMP22, 0.7226 for Lewis X and 0.8002 for 486p3/12. False-positive results on BTA Stat in 2 of 22 patients (9%), on NMP22 in 2 of 25 (8%), on 486p3/12 in 3 of 11 (27%) and on Lewis X in 4 of 43 (9.3%) were associated with tumor recurrence. Furthermore, negative results on BTA Stat in 2 of 39 patients (2%), on NMP22 in 2 of 36 (0.5%), on Lewis X in 0 of 18 (0%) and on 486p3/12 in 1 of 50 (2%) was associated with tumor recurrence during followup.
Conclusions: Immunocytology with mAbs against Lewis X showed higher sensitivity than all commercially available tests evaluated. Because of its high sensitivity and high negative predictive value, it may be useful for screening in a high risk population. Patients with a false-positive 486p3/12 test results are at increased risk for tumor recurrence compared with those with negative results.