2. UroToday Home
  3. Recent Abstracts
  4. Urologic Oncology
  5. Bladder Cancer

Identification and Validation of an 18-gene Signature Highly-predictive of Bladder Cancer Metastasis - Beyond the Abstract

The present study provides important insight into the relationship between gene expression profiles and lymph node metastasis. We took advantage of the large cohort in the SEER database to construct an accurate prediction model of lymph node metastasis. Using the gene expression profile in the TCGA database, we identified the two deviant groups (G and P) with a completely different genetic background, as shown by the gene expression heat map. Furthermore, we narrowed down the identified genes from 183 to 18 by examining their predictive ability in a more common population. In the development of our 18-gene signature, by incorporating the variable ‘number of lymph nodes examined’ in the logistic regression model, the probability of a positive lymph node gets higher with the increasing number of lymph nodes examined and vice versa. To some extent, we are looking for patients with ‘extremely adequate’ sampling but have pN0 and patients with ‘extremely inadequate’ sampling but have pN+. In this manner, we have the best chance to get their true lymph node status which would minimize the influence of including the whole cohort instead of patient only with adequate sampling.

To the best of our knowledge, for the first time, we combined the two most widely used external databases and studied the different gene expression status between extreme populations whose outcomes could not be predicted by clinicopathological factors. With external validation with RT-qPCR in a consecutive FUSCC cohort, we first give a clue that NECTIN2 might be a trigger for metastasis in bladder cancer pending larger cohorts and basic research. Most importantly, the 18-gene signature we proposed that is highly predictive of bladder cancer metastasis outperformed three other published signatures in another two GEO datasets. 

Written by: Wang Beihe,  Clinical Medicine (8-year M.D.program) Shanghai Medical College of Fudan University

Read the Abstract

Read More:

Update on The Cancer Genome Atlas (TCGA)