Molecular subtyping overhauls conventional cancer classification

A powerhouse of genomic information has led to a revamped tumor classification system that reclassifies as many as one out of every 10 cancer cases, according to a meta-analysis published Aug. 7 in the journal Cell.

Previous analyses have built up an arrangement of multiple cancer subtypes within a given tissue, but researchers have not been able to understand just how that genetic information is spread across tissues. Here, researchers including Katherine A. Hoadley, PhD, of Oregon Health Sciences University in Portland, Ore., examined 3,527 specimens from 12 cancer types using one proteomic platform and five genome-wide platforms compiling an integrated classification of 11 major cancer subtypes.

The researchers found that classifying cancers according to molecular traits was superior to identifying cancers by their tissue type. In fact, molecular characterization led to a significant number of reclassifications.

“This integrated multiplatform analysis of 12 cancer types provides independent and clinically relevant prognostic information above and beyond tumor stage and primary tissue-of-origin,” wrote the authors. “Based on this study, one in 10 cancer patients would be classified differently by this new molecular taxonomy versus our current tissue-of-origin tumor classification system.”

Results of the study showed that five subtypes were almost exactly matching their tissue-of-origin, but other cancer types were seen converging with common cancer subtypes. TP53 alterations, TP63 amplifications and overexpression of genes in immune and proliferation pathways led to the convergence of head and neck, lung squamous, and a select type of bladder cancer. Bladder cancers as a whole ended up having three differentiated cancer types.

The reclassification offers new information that could have a strong impact on prognosis and health outcomes. These datasets are available for further research and could inform new therapeutic investigations.