Knowledge Bank Approach Makes Precision Oncology a Reality

January 17, 2017
Knowledge Bank Approach Makes Precision Oncology a Reality
Source: NASA

An international collaboration led by researchers at the Wellcome Trust Sanger Institute has shown that truly personalized therapy will be possible in the future for people with cancer. Details of how a knowledge bank could be used to find the best treatment option for people with acute myeloid leukemia (AML) were published recently in Nature Genetics through an article entitled “Precision oncology for acute myeloid leukemia using a knowledge bank approach.”

AML is an aggressive blood cancer that develops in bone marrow cells. Earlier this year, the researchers reported that there are 11 types of AML, each with distinct genetic features. Now they indicate how a patient's individual genetic details can be incorporated into predicting the outcome and treatment choice for that patient.

The investigators assembled a knowledge bank using data from 1,540 patients with AML who participated in clinical trials in Germany and Austria, combining information on genetic features, treatment schedules, and outcome for each person. From this, the team developed a tool that shows how the experience captured in the knowledge bank could be used to provide personalized information about the best treatment options for a new patient.

"The knowledge bank approach makes far more detailed and accurate predictions about the likely future course of a patient with AML than what we can make in the clinic at the moment,” explained senior study investigator Peter Campbell, M.D., Ph.D., head of cancer, ageing, and somatic mutation, and senior group leader at Sanger. “Current guides use a simple set of rules based on only a few genetic findings. For any given patient, using the new tool we can compare the likely future outcomes under a transplant route versus a standard chemotherapy route - this means that we can make a treatment choice that is personally tailored to the unique features of that particular patient."

There are currently two primary treatment options for young patients with AML—stem cell transplant or chemotherapy. Stem cell transplants cure more patients overall, but up to one in four people die from complications of the transplant and a further one in four experience long-term side effects. Weighing the benefits of better cure rates with transplant against the risks of worse early mortality is a harrowing decision for patients and their clinicians. The team showed that these benefits and risks could be accurately calculated for an individual patient, enabling therapeutic choices to become personalized.

The researchers estimated that up to one in three patients could be prescribed a different treatment regimen using the tool compared with current practice. In the long-term, they hope the tool could spare one in ten young AML patients from a transplant while maintaining overall survival rates. The tool is currently available for scientists to use in research but needs further testing before it can be used to prescribe treatments in AML clinics.

"It has long been recognized that cancer is a complex genetic disease,” noted lead study investigator Moritz Gerstung, Ph.D., computational cancer biology group leader at EMBL-EBI. “Our study provides an example of how detailed genetic and clinical information can be rationally incorporated into clinical decisions for individual patients. We tested this philosophy in one type of leukemia, but the concept could theoretically be applied to other cancers with difficult clinical decisions as well. Our analysis reveals that knowledge banks of up to 10,000 patients would be needed to obtain the precision needed for a routine clinical application."

The authors believe this paper is a step towards validation of genetic techniques as a route to personalized medicine.

"Building knowledge banks is not easy,” remarked co-senior study author Hartmut Döhner, M.D., professor in the department of internal medicine at the University of Ulm. “To get accurate treatment predictions, you need data from thousands of patients and all tumor types. Furthermore, such knowledge banks will need continuous updating as new therapies become approved and available. As genetic testing enters routine clinical practice, there is an opportunity to learn from patients undergoing care in our health systems. Our paper gives the first real evidence that the approach is worthwhile, how it could be used and what the scale needs to be."