Reader Comments

Post a new comment on this article

The challenge for classifiers for bladder cancer is the confounding pathologies

Posted by katewilliamson on 18 Jun 2013 at 11:32 GMT

The accuracy of the combination of IL-8, VEGF and APOE to predict bladder cancer in the cohort of patients described in this manuscript is very impressive. The cohort examined comprised 64 patients with bladder cancer and 63 patients without bladder cancer. The authors excluded as controls patients with characteristics that usually confound diagnostic classifiers. Further, as the authors discuss, the cancers evaluated were predominantly high stage; they include only 9 patients with high grade bladder cancer and only 15 pTa tumours. Detection of high stage bladder cancer is not the challenge. Most classifiers detect all aggressive bladder cancer. The latter may explain why the authors achieved such a high accuracy?

Competing interests declared: I am named on patents
Methods for the detection of, or risk of, bladder cancer
JWJ01571US January 2013
JWJ01571GB August 2009

A method of defining the likelihood of a subject having bladder cancer
JWJ01839GB October 2012

RE: The challenge for classifiers for bladder cancer is the confounding pathologies

crosser replied to katewilliamson on 26 Jun 2013 at 00:46 GMT

We thank Dr Williamson for comments on our recent article. As we alluded to in the discussion, our cohort was predominantly high stage because that’s the nature of our clinic which operates as a tertiary care center. The performance of the 3-biomarker assay for detection of bladder cancer of all stages and grades remains to be seen, and studies using samples from multiple sites are under way in order to further evaluate the signature in this context.

Competing interests declared: Officer for Nonagen Bioscience Corp.