For more than a century, the human fingerprint has been a symbol of individual identity, a plot device for mystery writers and moviemakers, and a tool for criminal investigators hoping to prove that only one person out of several billion could be the culprit.
That’s the simple version of a complicated story. But it’s also the basic premise of a question researchers at the University of Washington are now asking about one of the most prolific killers in history: cancer.
As the analogy goes, what if a patient’s cancer had a unique fingerprint defined, not by a pattern of tiny ridges, but by the coding in a person’s genes, the genetic nature of the cancer-causing mutation, and their cancer’s sensitivity to hundreds of possible drugs? And what if that fingerprint could be used to help physicians treat patients, faster and more effectively than ever?
The answer could be game-changing for cancer patients.
This year in the United States, more than 60,000 children and adults will be diagnosed with leukemia, a family of blood cancers that claim more than 22,000 lives annually. While the outlook is relatively promising for younger people, five-year survival rates for adult patients with a more aggressive form of cancer, acute myeloid leukemia (AML), is only about 25%. The most effective treatments, including chemotherapy and blood or marrow transplants, can be physically harmful and lead to serious side effects.
Now, researchers at the UW Institute for Stem Cell and Regenerative Medicine (ISCRM) are extending the lives of leukemia patients by harnessing the combined might of technology, collaboration, and innovation.
At the center of the effort is Dr. Pamela Becker, Professor of Medicine in the Division of Hematology at UW Medicine, Physician at Seattle Cancer Care Alliance, and ISCRM faculty member. For more than a decade, Becker has been developing groundbreaking precision medicine protocols to improve treatments for blood cancers, significantly broadening the range of potential drugs available to patients.
Broadly speaking, precision medicine is an approach to disease treatment in which care plans are based on factors like patient genetics, drug sensitivities and resistances, environment, and lifestyle. Becker and her team are particularly interested in the interplay between the molecular makeup of a patient’s cancer and the drugs to which their cancer shows sensitivity or resistance. Where on the genome are mutations occurring? How are those mutations being expressed? And which drugs are most effective for which patient profiles? Answering those questions requires turning large amounts of data into patterns, applying those patterns to an individual, and using that information to continually validate and improve this process. The result is a feedback loop that is only possible because of advances in artificial intelligence and machine learning.
Becker is working tirelessly with collaborators at UW Medicine, Fred Hutchinson Cancer Research Center, and Seattle Cancer Care Alliance to answer those questions. In multiple clinical trials, more than 100 patients have been tested as candidates for treatment. To date, 75 (and counting) have received drugs. Survival in some instances is up to more than 800 days.
Recently, Becker presented new data from three ongoing clinical trials that show these game-changing approaches to treating cancer are working and extending the lives of very sick patients who have failed many other treatments. “We’re looking at data in ways we never have before,” says Becker. “We’re learning from it. We’re treating patients. And people are living longer.”
Results from these trials were highlighted in December 2019 at the annual meeting of the American Society for Hematology (ASH), where Becker also chaired a special symposium highlighting machine learning advances in precision medicine. Becker presented two abstracts that detailed the feasibility of using genomics, gene expression, and drug screening data to guide treatment plans for acute leukemia and multiple myeloma, and a third abstract that potentially validates the drug screening capabilities of ISCRM’s Quellos High Throughput Screening Core with the help of Secura Bio.
All three trials are attempting to continuously refine groundbreaking approaches to cancer treatment that allow physicians to choose drugs based on a patient’s cancer fingerprint, an impressive feat of data crunching that accounts for more than 150 drugs and 17,000 genes.
At the ASH meeting, Becker and her team shared results from three trials.
One trial is being conducted in partnership with the study’s principal investigator, Dr. Andrew Cowan of the UW Division of Medical Oncology, and Dr. David Coffey, of the Fred Hutchinson Cancer Research Center. In that trial, the researchers are finding new drug options for relapsed and refractory multiple myeloma and plasma cell leukemia patients who have undergone multiple rounds of treatments, including chemotherapy and stem cell transplants. For his part, Coffey is examining the gene mutations and gene expression observed in patients to determine if there are any drugs that target specific molecular abnormalities.
As of December 2019, 15 patients have been enrolled in the study. By running sophisticated tests on a participant’s cancer cells, including a drug screen performed in ISCRM’s high throughput screening core, Becker, Cowan, and Coffey have identified a drug, or combination of drugs, that appear to be a unique match for each individual patient. In some cases, a patient’s cancer cells show sensitivity to a drug that he or she has not been given yet. This information can prove to be of tremendous value to both the patient and the treating physician when other options are limited.
Of the 15 patients enrolled in this study, nine received individualized therapy guided by the test. Among those patients, six had an evaluable response so far and five achieved effective disease control (stable disease or better).
At the same time, Becker is running a second-generation trial for patients with acute leukemia. She has collected samples from 60 patients, roughly half of whom have been treated with drugs that her screening protocol identified as promising treatments. The results are heartening. Approximately a third of the treated patients have responded positively, leading to extended lives for people with advanced stages of cancer who might otherwise be in hospice or receiving only supportive care.
While the details are critical, Becker is ultimately focused on the big picture. “The hope for a trial like this is for a leukemia patient to walk through the door and we have enough information to say, ‘oh the fingerprint of your cancer looks a lot like the fingerprint in this other case, so let’s do what worked before.’ We’re not there yet. But that’s the goal. That’s what we’re trying to validate.”
In a third trial, Becker is testing cancer cells from multiple myeloma patients’ sensitivity to a triple drug combination. (The results of this screening are documented but not shared with individual patients, because the researchers do not yet know what this data means.) The patients are then treated with the triple drug combination and their clinical outcomes are cross-referenced with the laboratory testing results in the study record. “Many of the patients are showing overall responses, despite having had many prior treatments,” says Becker. “If we can predict how patients will respond in real life by running a drug sensitivity approach in the beginning, that’s huge. Then we can combine drug sensitivity insights with what we know about gene expression to create even more accurate predictive models.”
Becker and her collaborators hope to grow the enrollment of the trials. However, while the third trial is underwritten by a pharmaceutical company, the others require private support for the clinical research to move forward.
An expansion of the scope of Becker’s screening protocol into solid cancers and a broader range of blood cancers has allowed some patients, often from overseas, to pay for the drug screening out-of-pocket, giving them and their treating physicians new hope for promising treatment.
The protocols used in the ongoing clinical trials are based on principals of multiomics, essentially a holistic analytical approach that combines data derived from genomics, epigenetics, transcriptomics, proteomics, metabolomics and functional testing.
“Conventional precision medicine methods for cancer treatment target specific gene mutations,” says Becker. “That strategy leads to improvement in only a small percentage of cases. To get a more complete picture, we are looking at gene expression, mutational data, bioinformatics, and drug sensitivity data so we can make the best treatment choices possible for individual patients.”
Becker uses rapid, automated screening methods and machine learning technologies to design individualized treatment plans for relapsed patients with several types of cancer, including acute myeloid leukemia (AML), acute lymphocytic/lymphoblastic leukemia (ALL), acute leukemia of ambiguous lineage, multiple myeloma (MM), and other cancers.
To help with data analysis, Becker has turned to Ka Yee Yeung-Rhee, a Professor in the School of Engineering and Technology at the University of Washington-Tacoma and an Adjunct Professor in the UW Department of Microbiology. Yeung-Rhee has created a cloud-based software tool, called BioDepot-workflow-builder, that enables researchers like Becker to identify statistical patterns in data, for example, how mutations affect drug sensitivity or resistance.
Another important resource for Becker is the Quellos High-Throughput Screening Core, an ISCRM facility that allows her team to efficiently study how tissue samples derived from a patient’s malignant cells respond to potential drugs, and the role genetics play in drug susceptibility, revealing telling patterns that physicians can use to form treatment plans and better understand stem cells and their relationship with malignancies.
“This is truly a team effort,” Becker emphasizes. “None of these advances would be possible without the enormous dedication of my lab members Sylvia Chien and Jin Dai, PhD, who work tirelessly day and night to process patient samples as soon as we receive them from the clinic or hospital, and the Hematology research study coordinators who collect and transport those samples and who are on the front lines with patients, helping with both consent and monitoring treatment and response.”
According to Dr. Becker, the real key is not any single tool or technology, but the integrated use of multiple approaches to better diagnose each patient’s cancer, identify targets (such as mutated genes or sensitive cells), and choose the best possible care plan from the beginning. In a major breakthrough, this approach is extending lives and even leading to remission in cases where hope had been lost. “We are moving from a one-size-fits-all model to customized, individual treatment approaches. We’re already helping very sick people live longer. But that’s not enough. We want it to be an option for the newly diagnosed – to get it right the first time. That’s a paradigm shift in cancer treatment.”