Bin Zhu, Ph.D., a tenure-track investigator in the Biostatistics Branch, seeks to puzzle out how the vast constellation of germline risk variants, individually or in combination, affect the number and type of somatic mutations in the resulting tumor. His work to understand the relationship between the “two genomes” has depended upon the convergence of robust analytic techniques and availability of samples and detailed treatment and survival data from tumor banks, which have catalyzed him to develop methods to integrate these resources. Ultimately, he hopes this work will lead to meaningful improvements to understanding cancer etiology, and estimating patient survival and prognosis.
In September 2016, Dr. Zhu and DCEG colleagues reported on the relationship between germline and point somatic mutations in more than 600 breast cancer cases. The team examined data from paired tumor and non-cancerous lymphocyte samples from The Cancer Genome Atlas. They found an inverse association between the genetic risk score for cancer, as determined by which germline risk variants (single nucleotide polymorphisms or SNPs) a woman carried, and the number of somatic mutations in the tumor—the higher her risk, the fewer somatic mutations in her tumor tissue. There was an especially strong inverse association between a particular SNP in the DNA repair gene RAD51B and number of somatic mutations. This finding presents an exciting opportunity to explore the molecular pathways that relate this mutation to cancer initiation and progression.
“We would like to understand how tumors initiate and progress through somatic mutations in the context of germline variants and environmental factors,” said Dr. Zhu. He and his collaborators are conducting similar studies on a wide variety of cancer types, as well as investigations into how particular germline SNPs cause specific somatic mutation patterns.
The team is also exploring the influence of other “-omics” features on survival. Clinical information from cancer registries, such as age, race, and cancer type, is traditionally used to estimate patient survival, but preliminary results from Dr. Zhu’s project indicate that multiple layers of “-omics” data for each patient can improve the accuracy of such prognoses. Leveraging the capacity of DCEG to gather detailed data on the genome (DNA), epigenome (chemicals that regulate gene expression), and transcriptome (RNA), he and others are creating a framework for combining molecular data with clinical data. The plan takes into account the accumulated effects of many small molecular features that can influence survival.
Dr. Zhu’s research represents part of the underpinnings of the personalized medicine movement, a technology-driven shift towards using increasingly specific patient characteristics to influence diagnosis and treatment. Learning how particular mutations interact with each other and how those mutations can predict survival are important steps towards improving cancer care.
Read more articles in the spring 2017 issue of Linkage newsletter.