Applying narrow-band imaging in complement with white-light imaging cystoscopy in the detection of urothelial carcinoma of the bladder ☆

Abstract Objectives To investigate the value of narrow-band imaging (NBI) flexible cystoscopy in the detection of urothelial carcinoma (UC) of the bladder. Materials and methods Clinical data of 179 patients with suspected UC, who presented with gross hematuria, were collected at China PLA General Hospital from January 2009 to August 2010. These patients underwent white-light imaging (WLI) cystoscopy followed by NBI. The tumors were visualized, imaged, and recorded. Suspected UCs were biopsied or treated by transurethral resection, and then sent for pathologic examination. Detection results for NBI and WLI were compared. Results WLI and NBI confirmed UC in 143 patients; a total of 285 tumors were detected. The patient-level detection rates for NBI and WLI were 97.9% (140/143) and 88.8% (127/143), respectively ( P = 0.002). The patient-level false-positive detection rates for NBI and WLI were 21.8% (39/179) and 29.1% (52/179), respectively ( P = 0.12). NBI detected a total of 59 additional tumors (17.2%; 34pTa, 17pT1, 3pT2, and 5pTis) in 44 of 143 patients (30.8%). NBI found 1 additional tumor in 34 cases, 2 additional tumors in 6 cases, 3 additional tumors in 3 cases, and 4 additional tumors in 1 case. The mean ± SD (range) number of identified UCs per patient was 1.97 ± 0.67 (1–5) for NBI and 1.78 ± 0.53 (1–4) for WLI ( P = 0.01). The tumor-level detection rates for NBI and WLI were 96.8% and 79.3%, respectively ( P Conclusions Compared with WLI, NBI improves UC detection. It has a higher rate of detection and a comparative rate of false-positive detection. NBI is simple and requires no dyeing. It can be conveniently applied to complement WLI.