Abstract
Advanced urothelial carcinoma (UC) has a poor prognosis and new treatments are needed. In our study, multiplatform molecular profiling of UC identified unconventional treatment options in most cases. Different molecular profiles are exhibited between bladder and nonbladder UCs.
Background: Infiltrating UC represents the second most common genitourinary malignancy. Advanced UC has a poor prognosis and new treatments are needed. Molecular profiling of UC might identify biomarkers associated with targeted therapies or chemotherapeutics, providing physicians with new treatment options. Materials and Methods: Five hundred thirty-seven cases of locally advanced or metastatic UC of the bladder, 74 nonbladder, and 55 non-urothelial bladder cancers were profiled using mutation analysis, in situ hybridization, and immunohistochemistry assays for biomarkers predictive of therapy response. Results: Molecular profiling of UC showed high overexpression of topoisomerase 2 alpha, common phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha and/or phosophatase and tensin homolog (PTEN) alterations in nonbladder (27%) and bladder UC (21%), and rare gene mutations across subtypes. Compared with nonbladder, bladder UC consistently exhibited more frequent abnormal protein expression, including HER2 (10% vs. 3%; P = .04), tyrosine protein c-Kit receptor kinases (11% vs. 5%), c-Met protooncogene, receptor tyrosine kinases (25% vs. 8%), androgen receptor (16% vs. 6%), O(6)-methylguanine-methyltransferase (63% vs. 43%), ribonucleotide reductase M1 (32% vs. 11%), Serum protein acidic and rich in cysteine (SPARC) (69% vs. 33%), and topoisomerase 1 (63% vs. 39%). Bladder UC also exhibited increased amplification of HER2 (12% vs. 2%; P = .06). Conclusion: Comprehensive molecular profiling of UC identified a large number of biomarkers aberrations that might direct treatment in conventional chemotherapies and targeted therapies, not currently recommended in this population. As a group, bladder UC exhibited higher levels of actionable biomarkers, suggesting that UC from different primary sites and non-UC are driven by different molecular pathways. These differences could have clinical implications resulting in different treatment regimens depending on the site of origin of UC.