The goal of finding a “cure for cancer” was eclipsed long ago by the need to identify and understand the wide variety of cancers and their subtypes, and the need to focus on developing therapies specific to different cancer mechanisms and challenges. With this understanding and the rise of immunotherapy and genomic sequencing, a vast chunk of current cancer research now focuses on finding biomarkers that can predict disease response to certain drugs and guide therapy protocols.
But single biomarkers — genomic, molecular, or otherwise — can only go so far in predicting responses given the complexity and heterogeneity of individual malignancies and their microenvironments.
The exploding field of genomics is advancing, and researchers are starting to examine constellations of features that may better characterize disease subtypes on the path to precision medicine, but genomics alone cannot always distinguish differing phenotypes within cancer subtypes.
Hence, the rise of radiomics and pathomics, which are fields that take a similar approach to genomics — using technology to better understand features of solid tumors.
“We have been moving in the latest 15 years from an organ-based cancer treatment to [a] histology-based one, to the most recent precision medicine, which means that we are going to treat the specific alteration of the tumor independently by the site where [it] arose,” Giuseppe Luigi Banna, MD, of the United Lincolnshire Hospital Trust in Lincoln, United Kingdom, told Cancer Therapy Advisor.
Dr Banna and his colleagues recently published a paper exploring the “promise of digital biopsy” for predicting immunotherapy outcomes based on radiomics and pathomics.1 Although molecular determinants such as PD-1 or PD-L1 expression, and tumor mutational burden (TMB), are already used in clinical practice, these “fail in consistency, applicability, or reliability to precisely identify the responding patients mainly because of their spatial intratumoral heterogeneity,” they wrote.
Dr Banna elaborated: “The following 3 main problems with the current assessment of PD-L1 could be overcome by pathomics and radiomics: the different platforms used to test it, the possible interobserver variability, and the dynamic changes in PD-L1 expression.”