The performance characteristics of the bladder tumour antigen test

Abstract

Objective To evaluate the specificity and sensitivity of the Bard bladder tumour antigen (BTA) (Bard Inc, Covington, USA) test. Patients and methods The performance of the BTA test was assessed prospectively in 98 patients (30 women and 68 men, mean age 64 years, median 67, sd 14.8) undergoing cystoscopy for various indications. The urine of the patients was assessed using the BTA test, cytology, culture and a dipstick test for haematuria, and the results compared with those from cystoscopy as the ‘standard’. Results The overall specificity for the BTA test was 79% and the sensitivity 37%. The sensitivity was higher than for urine cytology (33%) but less than for the dipstick test (59%). The specificity was less than for cytology (100%) but higher than for the dipstick test (66%). Conclusion The BTA test is a rapid, non‐invasive qualitative test identifying degradation products of the basement membrane of the transitional cell epithelium. It represents an advance in the search for a ‘tumour marker’ for carcinoma of the bladder. The specificity is acceptable but the sensitivity relatively low, which limits its value in everyday practice. Further research and development are needed to improve sensitivity.

Diagnostic tests used in the investigation of adult haematuria: a systematic review

“…Two studies included cystoscopy with biopsy as the reference standard [46,47], one of which also used IVU [47]. Four studies evaluated the bladder tumour antigen (BTA) test; none of these reported the threshold that would define a positive test result [38,40,46,48]. Five studies evaluated nuclear matrix protein 22 (NMP22); one used a threshold of 10 units/mL [45], and the others used 12 units/mL [36,37,42,43].…”

“…Thirteen studies evaluated tumour markers for detecting malignancy, primarily of the bladder [36][37][38][39][40][41][42][43][44][45][46][47][48]. On the whole, these studies were poorly reported and/or were subject to numerous potential sources of bias.…”

“…In most of the studies, the reference standard was cystoscopy [36,37,40,44] or 'final diagnosis', however obtained [38,39,[41][42][43]45,48]. Two studies included cystoscopy with biopsy as the reference standard [46,47], one of which also used IVU [47].…”

“…Fifteen studies evaluated cytology as a test for detecting urinary tract malignancy [36,[38][39][40]42,43,[45][46][47][48][49][50][51][52][53]. In general, these studies were poorly reported and/or were subject to many potential sources of bias, and the influence of 'disease progression bias' was significant in both univariate ( P = 0.002) and multivariate regression models ( P = 0.02).…”

“…Most of the studies compared cytological findings to 'final diagnosis' [38,39,42,43,45,49,52,53]. In most studies where specific techniques were described, the reference standard consisted of cystoscopy [36,40], histology [48,51], or both [47,50]. One study used a reference standard of IVU and cystoscopy [46].…”