Urinary biomarkers for bladder cancer diagnosis fall short of mark

Nov 4 2015

By Shreeya Nanda, Senior medwireNews Reporter

Authors of a meta-analysis question the accuracy of urinary biomarkers for diagnosing the initial incidence or recurrence of bladder cancer.

“Urinary biomarkers miss a substantial proportion of patients with bladder cancer and are subject to false-positive results in others”, they write in the Annals of Internal Medicine.

The meta-analysis included 56 studies that assessed the accuracy of six biomarkers approved by the US Food and Drug Administration and one that evaluated a test categorised as laboratory developed, thus rendering it exempt from approval.

The biomarkers were quantitative and qualitative nuclear matrix protein 22 (NMP22); quantitative and qualitative bladder tumour antigen (BTA); fluorescent in situ hybridisation (FISH) to detect chromosome 3, 7 and 17 aneuploidy and 9p21 deletion; and fluorescent immunohistochemistry to detect the bladder cancer-specific antigens M344, LDQ10 and 19A211 (ImmunoCyt). The remaining test was Cxbladder, which quantifies the IGFBP5, HOXA13, MDK, CDK1 and CXCR2 messenger RNAs.

Sensitivity values ranged from 57% for qualitative NMP22 to 82% for Cxbladder, and specificity values from 74% for quantitative BTA to 88% for qualitative NMP22.

Positive and negative likelihood ratios ranged from 2.52 for quantitative BTA to 5.53 for Cxbladder and 0.21 for Cxbladder to 0.48 for qualitative NMP22, respectively, report Roger Chou (Oregon Health & Science University, Portland, USA) and colleagues.

Biomarkers that were more sensitive for the initial evaluation of symptoms versus for disease recurrence included qualitative BTA (76 vs 60%), FISH (73 vs 55%) and ImmunoCyt (85 vs 75%), while qualitative NMP22 was more sensitive for disease recurrence (47 vs 70%). The sensitivity of quantitative NMP22 did not differ by testing indication.

Although there were relatively few studies directly comparing biomarkers, say the researchers, analysis of the seven that compared quantitative NMP22 (cutoff >10 U/mL) and qualitative BTA showed comparable overall sensitivities (69 vs 66%) and specificities (73 vs 76%). And three studies that assessed FISH and ImmunoCyt found that the latter had higher sensitivity (71 vs 61%), but lower specificity (71 vs 79%).

The sensitivity of the studied biomarkers was higher when used in conjunction with cytological assessment than when used alone, but specificity was comparable for both scenarios.

“Urinary biomarkers miss 18% to 43% of patients with bladder cancer and are falsely positive in 12% to 26% of patients without bladder cancer”, write Chou et al.

They conclude: “The value of urinary biomarkers and whether they are sufficiently accurate to reduce the need for cystoscopy depends on the ability of clinicians to estimate the pretest probability of disease, the importance to patients and clinicians of relatively small changes in the probability of bladder cancer, and the acceptable threshold and clinical consequences of missed or delayed diagnoses and false-positive results.”

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.