Data-Driven: Paul Spellman, PhD

Dr. SpellmanA computational biologist deduces cancer

It's one of the biggest questions in cancer research: How exactly does a tumor develop? To try to answer it, researchers collect a sample from a patient's tumor biopsy early in the disease, and if the patient recurs, they collect additional samples to study changes over time. Yet obtaining "serial" samples can be difficult for many reasons, and for some cancers, it's nearly impossible. Almost all patients with high grade serous ovarian cancer, for example, come to their providers in advanced stages of the disease.

Working at the intersection of computational biology, genomics and cancer biology, Paul Spellman, PhD, Associate Professor in the Department of Molecular & Medical Genetics and member of the OHSU Knight Cancer Institute, developed a breakthrough methodology last year that reconstructs the development of key parts of a tumor genome from the final-stage tumor. Bypassing serial samples, Dr. Spellman, an internationally recognized scientist, his research team and collaborators created a statistical approach that maps the biological evolution of a high grade serous ovarian cancer tumor genome.

Building on this methodology, Dr. Spellman and his team are now investigating the genome to learn if there are selective pressures that always occur in precise ways during the tumor's evolution. The goal over the next six to 10 years is to identify those processes and exploit them for early diagnostics or preventive therapies. "Ideally, there will be a urinebased test that will detect when you have a very early lesion of high grade serous ovarian cancer, not full blown metastatic disease," he said.

The breakthrough is also part of a much larger National Institutes of Health-sponsored effort called The Cancer Genome Atlas Project (TCGA), which is cataloging the molecular diversity of 20 tumor types.

Dr. Spellman, one of 30 TCGA principal investigators, coordinated the analysis of the ovarian cancer genome sequencing project. Now his charge is to perform bioinformatics analyses on those data to understand processes and answer questions such as why different people get different types of tumors.

spellmangray_webDr. Spellman describes himself as a "molecular historian"—someone who studies the biology of cancer by surveying and analyzing its molecular remains and applying the results. His single sample tumor method in the TCGA is just one of several major projects that harness the tools of molecular biology, genomic and proteomic assays, and bioinformatic analyses to better understand cancer. In another major project, Dr. Spellman and his team are using cell line model systems to understand cancer signaling with the end goal of developing predictors of therapeutic response.

"Dr. Spellman has that ideal combination of deep scientific training and cutting-edge technical expertise that will help us unravel the intricacies of cancer biology so we can end cancer as we know it," said Brian Druker, MD, Director of the OHSU Knight Cancer Institute.

Dr. Spellman joined the OHSU faculty in 2011 after spending more than seven years at Lawrence Berkeley National Laboratory in Berkeley, Calif. working on cancer biology with Joe Gray, PhD (also now at OHSU and directing the OHSU Center for Spatial Systems Biomedicine; see the OHSU media release). At Oregon's academic health center, Dr. Spellman has found natural synergies with clinical faculty, a supportive research environment and plenty of opportunities to collaborate with other great minds across the institution. Other than the gloomy winter weather, Dr. Spellman said he's enjoying Portland, and appreciates a shorter commute that allows him to spend more time with his children: ages six, four and eight months.

Cancer has dominated Dr. Spellman's scientific career, but it's not just an abstraction. His fatherin-law battled a neuroendocrine carcinoma for six years before dying of it in 2005. "It was a horrible disease," said Dr. Spellman. "My father-in-law suffered terribly the last two years." Dr. Spellman said he uses that experience and others to inform and motivate his science.

Though it might mean unemployment for this scientist one day, Dr. Spellman hopes to see pain and death from cancer disappear entirely. "I think it has a shot in 30 years," he said. "Not 10, but 30 years is plausible."


Pictured: (top) In his wet lab, Paul Spellman, PhD, demonstrates how to make and distribute a compound library in 96-well plates. This compound library is designed to discover new drug candidates which preferentially kill cancer cells while sparing normal cells; (bottom) Joe Gray, PhD (left) with Paul Spellman, PhD