Gene Expression Signature In Urine For Diagnosing And Assessing Aggressiveness Of Bladder Urothelial Carcinoma

Main Category: Urology / Nephrology
Also Included In: Cancer / Oncology
Article Date: 02 Aug 2010 - 8:00 PDT

email to a friend   printer friendly   opinions  

The current diagnostic systems of bladder cancer are based upon a combination of urine cytology and direct observation of the bladder by cystoscopy. The latter is, in actual fact, the main diagnostic and tumour follow-up technique. It is performed transurethrally, and consequently it is an invasive approach, quite uncomfortable for most patients.

The sensitivity and specificity of this technique were believed to be quite high, but some improvements to the technique itself (fluorescence cystoscopy) suggest that probably this is not the case, and that some of the recurrences of non muscle invasive tumours could be due to a lack of complete resection of some nonvisible parts of the tumours. As to urine cytology, it is a noninvasive diagnostic technique with low sensitivity and specificity regarding low grade tumours. Moreover cytology interpretation is highly observer-dependent, which makes the existence of intra-observer differences possible, particularly in low grade tumours.

All these limitations have prompted a search for more reliable noninvasive markers of bladder cancer. Finding a noninvasive marker of bladder cancer, with high sensitivity and specificity regarding bladder cancer, would be enormously useful to clinical practice. It would also be extremely interesting that this same marker allowed us to establish the true aggressiveness of an individual patient's tumour, since this could enable us to administer a much more customised and effective treatment.

A great many works have focused on studying urinary tumour markers in search of a noninvasive diagnostic method for bladder cancer. In fact, different tests have been commercialised with such purpose (NMP22, UroVysion, ImmunoCyt, Accu-Dx, etc.), but even though most of them are more sensitive than urine cytology, the latter continues to be more specific. An option not yet commercialised is bladder cancer detection in urine samples by determination of the bladder cancer markers gene expression. In fact, there are some studies that suggest the usefulness of such method, despite having been carried out with just one or very few marker genes. Since the nature of these tumours is very heterogeneous, it does not seem very probable that all or most carcinomas can be identified with a single marker. Thus, it seems essential to somehow combine several of the most effective markers for characterising the majority of tumours. The aim of this study was to identify these markers and combine them in order to develop a method for the noninvasive diagnosis and aggressiveness assessment of bladder cancer.

The experimental approach of the work was based on initially testing a few tissue samples with a large number of genes using microarray technique to identify genes differentially expressed in bladder cancer (Mengual et al. 741-48). Afterwards, we analyzed selected differentially expressed genes in urine samples by reverse transcription quantitative real time PCR (qRT-PCR). This approach has allowed us to select the best performing genes and test them in a large number of samples.

The present study is divided into two phases. In the first phase, a gene set expression signature was obtained. During this phase, 3 consecutive steps were performed wherein the number of genes analyzed progressively decreased while the number of samples evaluated was progressively increased. In this phase 365 samples were analyzed; 244 urines from bladder washings from patients with histologically confirmed tumors and 121 control voided urine samples. These samples were consecutively analyzed by RT-qPCR in TaqMan Arrays (TA): 39 were randomly chosen for analysis in TA that contained 384 genes selected from microarray data, 35 for analysis in TA that contained 96 genes selected from the former 384 and 291 for analysis in TA that contained a final group of 48 genes. After selecting the optimal and minimal combination of genes for diagnosis and aggressiveness assessment purposes from 48-TA (12+2 genes), the selected genes were combined using a Naive Bayes classifier method.

In the second phase, we translated the gene set panel obtained to a routine clinical scenario by testing it in an independent series of voided urine samples. In this phase, 211 voided urine samples were analyzed; 97 voided urines from bladder cancer patients with histologically confirmed tumors and 114 control voided urine samples. A cDNA pre-amplification technique was added to the protocol in this second phase to improve the sensitivity of the method.

These samples were analyzed in TAs that contained 48 genes.

The most significant finding of this study is the description of a 12 gene set expression signature in urine capable of identifying patients suffering from bladder cancer with a high sensitivity and specificity. Our test presents a sensitivity (89%) and a specificity (95%) in the high range, improving the results of currently available noninvasive bladder cancer markers and tests. Moreover, our test achieves a sensitivity of 80% in low grade non muscle invasive bladder tumours, 93% in high grade non muscle invasive bladder cancer and 100% in muscle invasive bladder cancer, while maintaining a specificity of 95% in all 3 groups. This data represents an improvement of the currently available noninvasive bladder cancer tests, especially in low grade bladder tumours. Furthermore, the final model including two additional genes can differentiate low grade from high grade tumours with a sensitivity of 79% and a specificity of 91%.

Currently we are performing an international, multicenter, prospective validation study of the gene set that will provide us with the actual precision values of the test. This is a collaborative trial in which 5 European centres are participating, and this fact unifies results and homogenizes any potential population or performance protocol biases.

Lourdes Mengual, PhD

Reference List
Mengual, L. et al. "DNA Microarray Expression Profiling of Bladder Cancer Allows Identification of Noninvasive Diagnostic Markers." J.Urol. 182 (2009): 741-48.

UroToday - the only urology website with original content written by global urology key opinion leaders actively engaged in clinical practice. To access the latest urology news releases from UroToday, go to: www.urotoday.com

Copyright © 2010 - UroToday

• Additional
• References
• Citations