SUMMARY:
Researchers from UCLA’s Departments of Human Genetics and Urology have developed a novel method to predict disease susceptibility and aggression of multiple hereditary cancer types in individuals which will inform more personalized and effective therapies.
BACKGROUND:
The American Cancer Society reports that approximately 40% of people will develop cancer in their lifetime. While environmental factors, such as diet and smoking, can be controlled to decrease risk, inherited genetic factors contribute strongly to an individual’s propensity to develop cancer. It is estimated that heritable factors account for up to a third of the risk associated with cancer diagnosis. While many of the mutations that convert normal genes to their oncogenic form have been characterized, exactly how the inherited germline genome can increase risk of developing these oncogenic transformations remains largely unknown. Current therapeutics treat cancer only after diagnosis, and with varying degrees of effectiveness, in part due to the vast genetic and molecular landscape even within specific cancer types. Information on patient-specific germline genetics could help to inform clinicians on the molecular signature of cancers before they even develop, opening the door to the development of therapies that reduce risk of cancer using precision medicine. Furthermore, a more in-depth understanding of the relationship between germline genetics and cancer could vastly aid in designing tumor-specific treatment plans rather than the ineffective one-size-fits-all approaches currently employed in the clinic. Elucidation of this relationship would help to predict the molecular features and risk of a cancer decades prior to diagnosis, allowing for more personalized and effective treatment of a deadly disease.
INNOVATION:
UCLA researchers in the Department of Human Genetics have identified 62 regions of association between germline polymorphisms and somatic mutational features in prostate cancer called driver quantitative trait loci (dQTLs). The research team led by Dr. Paul Boutros developed a method utilizing whole genome sequencing to predict the somatic mutational features and clinical outcomes across multiple cancer types based on germline polymorphisms. The researchers assembled a discovery cohort of more than 400 patients with localized prostate cancer, one of the most heritable cancers, each with whole-genome sequencing of blood and tumor, and subsequently replicated their findings in a cohort of over 1,000 tumors from multiple countries around the world. Using novel data science and machine-learning approaches they identified 62 dQTLs that modulate tumorigenesis by affecting the tumor epigenome, transcriptome, and proteome. Amongst those dQTLs, a subset predicts clinical outcomes, and some individual dQTLs affect multiple cancer types, significantly enhancing their actionability. Taken together, the researchers revealed intricate crosstalk between the germline and somatic genome of primary tumors that can underlie a powerful tool for cancer diagnostics and therapies.
POTENTIAL APPLICATIONS:
● Machine-learning strategies to create sensitive biomarkers for drug-sensitivity and toxicity
● Predict cancer aggression and inform more personalized therapy
● Identify prognostic germline loci that might be minimally invasive biomarkers
● Aid the triage of patients to more expensive tissue- or radiology-based assays
● Improve understanding of ancestral differences in cancer genomic landscapes
ADVANTAGES:
● Accurately predicts whether a tumor will harbor a specific gene mutation, allowing for prediction of cancer susceptibility and aggression decades prior to diagnosis
● Hugely expands the utility of genomic testing with hundreds of dQTLs yet to be discovered
DEVELOPMENT-TO-DATE:
Researchers have successfully developed a method that identified 62 dQTLs in a discovery cohort, from which a subset was accurately able to predict clinical outcomes in a replication cohort.
Related Papers (from the inventors only):
Houlahan, K.E.; et al. Nature Medicine 2019, 25, 1615-1626.