Researchers discover common markers of tumor hypoxia across 19 cancer types

Landmark pan-cancer study analyzes mutation signatures of low oxygen in more than 8,000 tumors
Paul Boutros
Paul Boutros
3 min read

Unlike healthy tissues, tumors thrive in low-oxygen environments, often acquiring the ability to resist treatment and spread to other sites in the body. Despite being a well-known cause of therapy resistance and metastasis, the impact of low oxygen, known as hypoxia, on tumor cells is poorly understood. As published online in Nature Genetics, researchers have discovered molecular hallmarks of hypoxia in the first-ever pan-cancer analysis of low oxygen in human tumors, with a special focus on prostate cancer.

The study investigated more than 8,000 human tumors across 19 different cancer types, including prostate tumors from the Canadian Prostate Cancer Genome Network (CPC-GENE). The authors discovered common markers of hypoxia that could help predict cancer aggressiveness and inform treatment decisions.

These findings, which include several genes more commonly mutated in hypoxic cancers and new information about hypoxia-related patterns of tumor evolution, make up the largest resource available for hypoxia research.

“If we look at any single aspect of cancer, we only gain a partial understanding of this complex disease. But here we’ve exploited a wealth of human tumor data to gain a more comprehensive understanding,” said Vinayak Bhandari, lead author of this study and Ph.D candidate at the University of Toronto, who is conducting research at the Ontario Institute for Cancer Research (OICR). “By tying together our new understanding of the environment in which tumous develop with detailed evaluation of genetic changes, we created a biological signature that highlights patients who may benefit from more therapy.”

The markers observed in this study also open new opportunities for researchers to develop therapies that target hypoxia-related treatment resistance and metastasis across many types of cancer, including prostate cancer.

“Understanding common genomic traits across cancer types is critically important to the future of cancer diagnosis and treatment,” said senior investigator of the study Paul Boutros, director of cancer data science at the UCLA Jonsson Comprehensive Cancer Center and professor of urology and human genetics at the David Geffen School of Medicine at UCLA. “We were initially motivated by the inability to differentiate between aggressive and non-aggressive prostate cancers, but our findings provide insights into how treatments might be developed for many tumor types.”

“Hypoxia was previously associated with aggressive disease, but the mechanisms by which it drives this process in human tumors was poorly understood from a genetic angle,” said professor Robert Bristow, a director at the Manchester Cancer Research and CRUK Manchester Centres at the University of Manchester, formerly at Princess Margaret Cancer Centre, University Health Network. “We can now start to exploit these findings into novel clinical trials to target hypoxia and abnormal genetics at the same time,” explained Bristow, who with Boutros, were former Co-Leads of the CPC-GENE project.

“This remarkable study was enabled by leveraging OICR’s unique computational biology expertise, rich prostate cancer data from CPC-GENE, and clinical expertise from collaborators across the province,” said Dr. Laszlo Radvanyi, president and scientific director of OICR. “The team’s relentless focus on clinically-relevant research questions has brought us new insights that can be translated back to the clinic to apply this molecular profiling approach to guide better treatment decisions to manage prostate cancer.This can also be applied to other cancer types.”

The study was supported by the Movember Foundation, Prostate Cancer Canada and OICR through CPC-GENE – the largest prostate cancer genomics project in the world. The project was also funded by the Terry Fox Research Institute.Tumor sample data was provided by CPC-GENE, the International Cancer Genome Consortium and The Cancer Genome Atlas project.