Radiology’s Role in Precision Medicine

During his State of the Union Address this year, President Obama called on the nation to lead the world into a new era of medicine-one that delivers the right treatment at the right time. He was referring to precision medicine, which he said will cure diseases like cancer and diabetes, while at the same time giving patients access to personalized information to keep them healthy.

With precision medicine now part of the national consciousness, all fields of medicine have an opportunity to step up to the plate. Many of the tools are already in place to understand the correlations between genome, habits, and disease.

“Physicians think they have been doing precision medicine forever,” said James Thrall, MD, chief of radiology at Massachusetts General Hospital. “But they don’t understand it in the sense that we are talking about it-in the sense of phenotyping and genotyping people, identifying precise cohorts of people.”

Normally people associate precision medicine with genotyping-identifying an individual’s DNA sequence. But that’s not the complete picture. According to Thrall, just looking at an individual’s genome won’t necessarily create cohorts, mainly because of epigenetics, or the cellular trait variations that are caused outside of changes to the DNA sequence and other sources of variation in gene expression.

Imaging is one way to observe-or phenotype-the physical manifestations of disease in order to establish these cohorts for clinical trials.

A 2011 publication from the National Research Council (NRC) defined precision medicine as medical diagnosis and treatment based on the individual characteristics of a person. They go on to say that this can be achieved by identifying small cohorts of people who have a similar clinical presentation and biology of their disease, and are likely to respond to the same treatments, as well as have similar prognosis.

By identifying which subpopulations respond best to a drug therapy or certain follow-up treatment, much more efficient clinical trials can be done with the ultimate goal of treating patients based on the precise characteristics of their disease.

“In the era of precision medicine, instead of having large cohorts of people that may have different characteristics, you look for people with similar characteristics,” said Thrall.

It turns out there are only four ways to phenotype a patient. Imaging is one way, along with clinical history and a physical exam, lab testing, and histopathology and immuno pathology.

“A good phenotyping system segments patients into different prognostic categories,” said Thrall.

For example, a colleague of Thrall’s developed a scoring system for patients exhibiting intracerebral hemorrhage based on the number of spots of contrast seen on CT angiograms. Five different scores, categorized by levels of bleeding or not bleeding, segment the patient into various risk categories.

Thrall said there are at least five clinical trials going on currently where this scoring system is being used.

“I’ve come to think of it like this: When a radiologist renders a report, the report is actually a description of the imaging phenotype of disease manifestation, but we’ve just never thought of it in those terms,” said Thrall.

He urges radiologists to start talking about imaging biomarkers and phenotypes instead of rank and signs.

Molecular Imaging-Precise All Along[[{"type":"media","view_mode":"media_crop","fid":"34722","attributes":{"alt":"precision medicine","class":"media-image media-image-right","id":"media_crop_1890595446844","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"3698","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 249px; width: 251px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"©Lonely/Shutterstock.com","typeof":"foaf:Image"}}]]
Imaging has the capability to influence precision medicine in many ways, especially with molecular imaging. For instance, Hodgkin’s lymphoma patients receive an FDG PET scan after a few rounds of chemotherapy. In one case of imaging phenotypes, where there’s no abnormal uptake of tracer, patients end up having a normal life expectancy. But in cases where there’s a continued presence of abnormal uptake, those patients have significantly lowered life expectancy.

“As a predictor of prognosis, the use of imaging to create these two phenotypes is far better than something called the International Prognosis Scoring System,” said Thrall.

Going forward, molecular imaging will likely play the biggest role in broadening precision medicine outside of oncology and into metabolic disorders and neurological diseases. In fact, the theme at this year’s annual World Molecular Imaging Society (WMIS) Congress is precision medicine.

“We’ve been using imaging to characterize the cellular and molecular underpinning of tumors and defining tumor genetics-that’s the very heart of precision medicine and that’s been molecular imaging,” said H. Charles Manning, PhD, World Molecular Imaging Congress program chair.

Tools, Databases, and Challenges
Dr. Eliot Siegel, professor and vice chair of research informatics at the University of Maryland School of Medicine, sees another opportunity for radiology to participate in precision medicine. The more accurate identification of follow-up treatment for the millions of smokers aged 55-77 who will receive low-dose lung CT screening now that the Centers for Medicare and Medicaid Services will cover the screening.

“What if we could perform personalized screening not just based on age and smoking history, but based on index of likelihood of developing cancer or other disease?” said Siegel.

Current recommendations for follow up and management of lung nodules have been established by the Fleischner Society. The criteria are mostly based on a patient’s smoking history and nodule size. Siegel said there is an opportunity to personalize these recommendations-make them more precise-by adding additional parameters, such as the lobe in which the nodule is found, nodule shape, patient age, ethnicity, and other factors using National Lung Screen Trial Data.

According to Siegel, there is another database to help determine the a priori probability that a patient will develop lung cancer. He said the Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) Screening Trial dataset, published in 2009, is a resource that can be used for matching patients with the outcomes of demographically or diagnostically comparable patients. PLCO enrolled 155,000 patients to determine whether certain screening exams reduced mortality from prostate, lung, colorectal, and ovarian cancer.

“So rather than just creating a set of recommendations based on the size of the nodule [and smoking history], you can use multiple parameters and a large database and then make the recommendations for follow up very precise,” said Siegel.

At the same time, the precision criteria derived from the database can be used to determine which patients should be screened in order to create the highest yield in the most cost-effective way.

“We use these databases and others like them in making decisions about who to screen in the first place but also [we use them for] what I should do when there’s a positive finding-how aggressive should I be as far as a follow-up imaging, biopsy, or surgery,” said Siegel.

Right now, data from large clinical trials guide many treatment decisions, often in a one-size-fits-all manner. But in the future, decisions will increasingly be made based on patient-specific characteristics, according to Siegel.

While some of the tools might be there, not all of the capabilities are. Siegel would like to see databases that are more accessible for searching.

Variability in medical imaging data also presents a challenge if radiology is going to have an impact on precision medicine.

“Standardization in medical imaging helps enable smaller, faster, and smarter clinical trials for providing the evidence needed to bring new patient-specific treatments to the public,” said Daniel Krainak, PhD, from the FDA’s Magnetic Resonance and Electronic Products Branch in the Office of In Vitro Diagnostics and Radiological Health.

A number of research groups, including the FDA, are currently identifying the major sources of variation and developing techniques to reduce uncertainty so that image-based measurements and assessments are more accurate and precise.

Other challenges include clinical validation of image analysis tools and the development of novel techniques that combine images with other sources of data, such as in vitro diagnostic tests, family history, patient demographics, and genomics.

With these challenges met, and radiologists eager to make their mark, imaging’s role in precision medicine will only expand as new therapies and prevention techniques are developed.