Background: The standard treatment for non-metastatic muscle-invasive bladder cancer (stages T2-T4a) is radical cystectomy with lymphadenectomy. However, patients undergoing cystectomy show metastatic spread in 25% of cases and these patients will have limited benefit from surgery. Identification of patients with high risk of lymph node metastasis will help select patients that may benefit from neoadjuvant and/or adjuvant chemotherapy.
Methods: RNA was procured by laser micro dissection of primary bladder tumors and corresponding lymph node metastases for Affymetrix U133 Plus 2.0 Gene Chip expression profiling. A publically available dataset was used for identification of the best candidate markers, and these were validated using immunohistochemistry in an independent patient cohort of 368 patients.
Results: Gene Set Enrichment Analysis showed significant enrichment for e.g. metastatic signatures in the metastasizing tumors, and a set of 12 genes significantly associated with lymph node metastasis was identified. Tumors did not cluster according to their metastatic ability when analyzing gene expression profiles using hierarchical cluster analysis. However, half (6/12) of the primary tumor clustered together with matching lymph node metastases, indicating a large degree of intra-patient similarity in these patients. Immunohistochemical analysis of 368 tumors from cystectomized patients showed high expression of GEM (P = 0.033; HR = 1.46) and EDNRA (P = 0.046; HR = 1.60) was significantly associated with decreased cancer-specific survival.
Conclusions: GEM and EDNRA were identified as promising prognostic markers for patients with advanced bladder cancer. The clinical relevance of GEM and EDNRA should be evaluated in independent prospective studies.