Serum and urinary endocan levels for bladder cancer diagnosis

Dear Editor

I read with interest the paper from Laloglu et al. ¹ published recently in Annals of Clinical Biochemistry. The researchers enrolled 50 patients with bladder cancer, 50 with urinary tract infection (UTI) and 51 healthy volunteers, and measured concentrations of endothelial cell-specific molecule 1 (ESM-1 or endocan) in serum and urine. They found that in patients with bladder cancer, serum and urine endocan concentrations were significantly higher than those of healthy volunteers. However, no significant difference in serum and urine endocan concentrations was observed between bladder cancer patients and patients with UTI. At a cut-off of 630 pg/mL, serum endocan concentrations exhibited 50% sensitivity and 77% specificity in differentiating bladder cancer from healthy individuals. At a cut-off of 1100 pg/mL, the sensitivity and specificity of urine endocan concentrations for differentiating bladder cancer patients from healthy individuals were 62% and 71%, respectively.

This is the first study investigating urine endocan in bladder cancer patients, and the topic is novel and interesting. However, based on the results of this study, neither serum nor urine concentrations of endocan can reliably be used to diagnose bladder cancer. Crucially, endocan concentrations in serum and urine were not significantly different in patients with bladder cancer and those with UTI. Another point relates to the design of this study. This study was a ‘two-gate’ design using healthy volunteers as controls ² ; this design has been shown to produce inflated estimates of diagnostic accuracy of an index test. ³ Healthy volunteers have no symptoms and signs of bladder cancer, and clinicians do not need a laboratory test to distinguish between them and patients with bladder cancer. By contrast, a ‘one-gate’ study design applies a single set of inclusion and exclusion criteria to all subjects, who are typically enrolled consecutively and in a way defined by the clinical presentation. ² In this case, the target population in which the diagnostic test would be used should ideally be selected by the presence of symptoms and signs of bladder cancer. The work by Laloglu et al. is a significance advance to this field, but further well-designed studies are needed to rigorously evaluate the diagnostic value of endocan for bladder cancer.