Unstable.

We all know oncology is in a watershed moment, transitioning from strictly anatomic to more biologic descriptors of disease. Beyond taking years off the lives of trainees memorizing convoluted staging algorithms that attempt to incorporate all of the above, there are clear imperfections in the ability of expression-based biomarkers to consistently predict response to therapies. One reason? Intratumor heterogeneity. Take breast cancer, for example, where mets are about as likely to not biologically match the primary tumor as they are to match. This secondary genomic analysis of non-small cell lung cancer (NSCLC) tumor specimens from 48 patients enrolled in the TACERx study confirms existing gene expression studies are vulnerable to such heterogeneity, largely born from chromosomal instability. But it's not all bad news. The investigators identify genes that are consistently expressed within tumors, but vary greatly across tumors, due to them commonly arising early in tumor evolution. These special genes, deemed “clonal transcriptomic biomarkers,” were confirmed to be independent predictors of survival. TBL: Newly-identified clonal transcriptomic biomarkers in NSCLC have uniquely low intratumor heterogeneity, minimizing tumor sampling bias. | Biswus, Nat Med 2019