While we’ve made much progress in clinical oncology over the last decade, truly personalized medicine eludes most cancer patients. Too many cancers cannot be diagnosed at the molecular level when clinical tests (e.g., clinical NGS panels) fail to identify genetic drivers of disease and, as a result, therapeutic targets.
This is a multi-dimensional problem. One aspect is that today’s clinical panels are predominantly focused on coding regions of the genome, meaning that any potential driver outside of the coding region will be missed. Another challenge is the availability of sufficient high-quality archival tissue specimens. While frozen tissue has the highest quality, the vast majority of tumor samples are formalin-fixed and paraffin-embedded (FFPE), which negatively affects nuclei acid quality. Even RNAseq on new FFPE samples can miss critical cancer drivers, either due to the panel design (because the driver is uncommon), the RNA quantity, RNA degradation over time, or the low level of fusion transcripts in low grade tumors.
In our pathology lab at NYU Langone Health, where my research focuses on diagnosing and understanding brain tumors, we’ve found that using a 3D genomics approach with Hi-C technology can help identify novel clinically actionable aberrations and provide insight into the molecular drivers of brain tumors, which previously had no known drivers.
3D genomics can overcome many of the technical challenges described above and offers a much more detailed view of a cancer genome—including a three-dimensional view of gene fusions and rearrangements. It also performs well with older FFPE samples, making it an ideal solution for clinical research because it gives us the ability to unlock archival samples of rare cancers for the discovery of novel gene fusions inaccessible by RNAseq.
I am currently leading a study with a pilot cohort of more than 120 patients whose tumors had no detectable drivers using NGS of both DNA and RNA and inconclusive whole-genome DNA methylation tests. Using Hi-C technology, we found additional molecular information for 70 percent of these cases, and for many of them, potential therapeutic targets for current approved therapies or diagnostic and prognostic biomarkers that have resulted in a changed course of patient management. We also found novel genetic drivers and discovered new insights into 3D chromatin architecture.
These initial findings are so powerful that we’ve decided to launch an open-access 3D Cancer Genomics hub at NYU, enabling collaborators to send us clinical or research samples for 3D profiling to share unclassifiable cases to further cancer research and confirm clinical actionability.