(RxWiki News) You’ve seen the forecast models for the possible routes of hurricanes. Computers and humans work together to try and predict where a storm will land and how bad it will be. This same approach is being used to track brain tumors.
Scientists have developed a model that can predict how an individual patient’s glioblastoma multiforme (GBM) will grow. The system also shows how well the tumor is responding to treatment, and allows doctors to make treatment decisions based on this information.
This system, which is still under development, could provide much-needed personalized therapy for brain cancer patients. It will let physicians and patients distinguish between treatments that are working and those that aren’t.
"See a doctor if you have headaches that get worse with activity."
Researchers at Northwestern Medicine developed the model, which predicts brain tumor growth. Kristin Swanson, PhD, professor and vice chair of research for neurological surgery at Northwestern University Feinberg School of Medicine, was the senior author.
"When a hurricane is approaching, weather models tell us where it's going," Dr. Swanson said in a press release. “Our brain tumor model does the same thing. We know how much and where the tumor will grow. Then we can know how much the treatment deflected that growth and directly relate that to impact on patient survival."
The growth of brain tumors varies. Existing methods don’t take into account these variances in growth rate, shape and tumor density. As a result, physicians can’t tell if a patient has an aggressive tumor that will respond well to treatment or a slow-growing cancer that won't respond well to therapy.
“As we advance toward personalized medicine, tests that can predict a particular patient's response to a particular therapy will play a greater role in treatments for disease," said Keith L. Black, MD, chair and professor of Cedars-Sinai’s Department of Neurosurgery, director of the Cochran Brain Tumor Center and director of the Maxine Dunitz Neurosurgical Institute and the Ruth and Lawrence Harvey Chair in Neuroscience.
The Northwestern model considers growth rate, shape and density in its measurements.
For this study, researchers created a computer model of the individual GBM tumors from 33 different patients. The system predicted how the tumors would grow without treatment. The model was based on MRI (magnetic resonance imaging) scans taken on the day of diagnosis and the day of surgery.
The differences in the images, along with the density of tumor cells throughout the brain, enabled researchers to determine how fast the cancer was growing.
The scientists were able to judge treatment effectiveness by comparing tumor size after initial treatment to the size the model predicted the tumor would be without treatment.
Researchers then scored the effectiveness of the patient's treatment by comparing the size of the patient's tumor after treatment to the model-predicted size if untreated.
They used a “Days Gained” scoring system to predict both overall survival and progression-free survival (period during which the cancer does not get worse).
The researchers are hoping to turn this model into an iPad app or make it available on a website. Physicians then could simply enter the MRI data to determine treatment response.
The authors wrote, “Our study illustrates the potential of the emerging field of integrated, patient-specific modeling to impact clinical decision-making and patient outcomes. The next challenge is to translate our computational approach into the clinical setting.”
"How useful and precise this method ultimately is will require larger studies, but we should expect to see more techniques like this utilized to drive patient treatments, not just for cancer but many other disorders,” Dr. Black told dailyRx News.
This study was published in the January issue of PLOS ONE. The research was funded by the McDonnell Foundation, the Brain Tumor Funders Collaborative, the National Institutes of Health, the James D. Murray Endowed Chair and the University of Washington Academic Pathology Fund. The authors have declared that no competing interests exist.
