Looking for PET/MR’s Role
By David Yeager
Vol. 14 No. 9 P. 20
Primarily a research tool at the moment, radiologists are investigating where this imaging tool will find clinical utility.
Researchers and clinicians are investigating many uses for the pairing of PET’s whole-body functional imaging with MR’s local anatomic detail and morphological information. It’s not clear how the combination ultimately will fit into clinical practice, but PET/MR trends are beginning to form. Researchers are exploring potential uses from oncology to neurology and cardiology, but more data are needed to establish and validate PET/MR’s clinical role.
In June 2011, the FDA approved Siemens’ Biograph mMR PET/MR system, and in November of that year, Philips’ Ingenuity TF PET/MR also was approved. GE Healthcare’s Trimodality system utilizes previously approved technology: the Discovery PET/CT, Discovery 1.5 T MR450w, or 3.0T MR750w MRI systems. These varying approaches present different advantages and challenges to PET/MR researchers.
More Than One Way to…
Although it’s not an integrated system, GE’s Trimodality costs less because it uses scanner technology that many facilities already have in place. Some consider it a stepping-stone to an integrated system.
Patients typically are given an FDG injection 10 to 15 minutes before going into the MR scanner, according to Gustav K. von Schulthess, MD, PhD, MD hon, the chairman of the department of medical radiology, professor and director of nuclear medicine at UniversityHospital Zurich in Switzerland who uses the Trimodality system. This allows the FDG, which requires about 60 minutes of uptake time, to be taken up during the MR exam, which lasts about 40 to 50 minutes. Therefore, the patient can be immediately moved to the PET/CT machine following the MR exam.
Patients are moved from the MR machine to the PET/CT machine with a forced air system that floats the patient between tables in a way similar to a hovercraft. The arrangement allows side-by-side docking and requires less space to site the two machines. When the PET/CT is finished, the dual exam is complete unless contrast enhancement is needed. The MR and PET/CT images are then combined using image fusion software.
“First of all, it’s easier for the patient because the FDG uptake takes place during the MR exam rather than the patient waiting around and having the PET/CT and then the MR,” von Schulthess says. “It also forces us to have an efficient MR protocol. We want to identify a workable and cost-effective MR protocol so that, once we fully integrate PET and MR and run them simultaneously, we can basically immediately transfer the protocols that we have.”
Philips’ Ingenuity TF PET/MR two gantries located on opposite ends of a rotating table. Patients receive one exam and then are rotated 180 degrees for the other. Philips chose this method because an MR system’s magnetic field makes it difficult to simultaneously acquire a PET image.
Norbert Avril, MD, a professor and research scholar in the department of radiology at University Hospitals Case Medical Center in Cleveland, an affiliate of Case Western Reserve University, says Philips’ time-of-flight PET is highly reproducible, allowing for a reduction in gating activity, and MR offers better soft tissue visualization than CT. Separating the PET and MR exams obtains the best possible image quality, but it makes the overall exam take longer, which can be problematic. MR also costs significantly more than CT. For these reasons, identifying which patients can best benefit from a PET/MR exam is a research priority.
“So what we are currently discussing as a protocol is that we perform a PET scan and then evaluate where there are potential abnormalities that we can—right away, while the patient is still on the scanner and we have the coregistered information—acquire a localized MR scan to avoid having the patient come back a second time to get the MR scan,” Avril says.
Siemens combined MR and PET in a single scanner for the Biograph mMR. MRI exams can take 20 to 60 minutes, depending on the area being imaged and the pulse sequences used, such as diffusion-weighted imaging (DWI) or dynamic contrast-enhancement (DCE) imaging. Performing PET concurrently with MR saves time and could reduce the amount of contrast given to a patient.
Despite the challenge of optimizing the timing of the concurrent exams to make them work efficiently together, progress is being made. Siemens uses gating techniques to synchronize the images. Alexander R. Guimaraes, MD, PhD, medical director at the Martinos Center for Biomedical Imaging in the radiology department at Massachusetts General Hospital in Boston and an assistant professor of radiology at Harvard Medical School, says finding ways to make PET/MR clinically viable without adding too much time to the exam is an ongoing area of research.
“One area of active research is utilization and workflow. How do you perform an exam that benefits the patient in the least amount of time but doesn’t encumber them with two exams, potentially two doses of contrast, and added inconvenience?” Guimaraes says. “[We’re looking at] areas where throughput is important, and we can come up with an examination that’s not too much longer than a PET/CT, especially where radiation sensitivity, [such as with pediatric cases or oncology treatment response], is an issue and/or where MRI already plays a large role.”
There have been questions about how PET/MR will be reimbursed in clinical practice. Currently, clinicians can pursue PET/MR reimbursement by billing for each exam separately. On June 11, the Centers for Medicare & Medicaid Services issued a memo expanding FDG PET coverage from one initial and one follow-up scan to three follow-up scans. Beyond three, local Medicare administrative contractors will determine whether additional scans are covered. The memo included FDG PET/CT and FDG PET/MRI “unless context indicates otherwise” under its definition of FDG PET. It also eliminated reporting rules that required FDG PET scans for certain cancer types to be registered with the National Oncologic PET Registry to be eligible for reimbursement. These changes may alleviate some concerns about PET/MR reimbursement.
One key issue with PET/MR is determining when it provides a clear advantage over PET/CT. The jump from PET to PET/CT was large, but there may be less of a gap between PET/CT and PET/MR. There are fewer exam options available with PET/CT compared with the range of pulse sequences available with MR. The number of sequences, such as DWI and DCE, makes it more challenging to determine how it is best used.
Part of the reason for that is MR may provide information that already is known, such as reconfirming a lesion already seen by PET, making it less clinically relevant, according to von Schulthess. The goal is to get the greatest amount of nonoverlapping information. To do that, he says it’s important to consider the relative strengths of each modality and follow a targeted approach to exam protocols. For example, MR provides better contrast of soft tissue, such as in brain imaging, while CT already is commonly used for lung imaging.
“In lung cancer, for instance, many patients come with a CT. Then you can then do a PET/MR because you have the CT data already,” von Schulthess says. “You could do a brief, dedicated brain MR to rule out brain metastases involved, which are relatively frequent in lung cancer, even in early stages.”
Although still mainly used for research, PET/MR is working its way into clinical practice. Later this year, the Cleveland Clinic will become the first US site to use PET/MR strictly for clinical purposes. Nearly all of PET/MR’s current clinical use is for oncology, particularly looking at cancers of the abdomen and pelvis, such as endometrial, cervical, and rectal cancers, as well as metastases in the liver and brain.
One benefit of PET/MR is its potential to reduce patients’ radiation exposure. A study published earlier this year in Pediatric Radiology found that PET/MR delivers only 20% of the effective radiation dose of PET/CT. Because pediatric patients especially are sensitive to radiation and may be clinically followed for their entire lives, many researchers believe PET/MR will be useful in pediatric cases.
Much of the excitement around PET/MR involves how it may be used with non-FDG biomarkers. For example, FDG is not particularly useful for prostate cancer because it’s excreted through the kidneys and lights up the bladder on PET images, obscuring the prostate. A biomarker that targets the prostate would allow clinicians to use MR’s contrast-enhancement ability to examine the prostate. Avril says new biomarkers can capitalize on MR’s ability to look at tissue parameters, such as inflammation and contrast uptake, allowing researchers and clinicians to gather significantly more information than they can get from PET/CT.
“Since FDG-PET is quite established, it is my personal impression that a significant strength of PET/MR is the use of non-FDG tracers, such as [18F] fluorocholine, [18F] fluorothymidine, and other newly developed tracers, which would then make most use of the functional information that we get from MR,” Avril says.
PET/MR also may prove useful in cardiac imaging and neurodegenerative diseases. Clinicians could use MR to look at anatomy and heart wall motion and PET to look at blood perfusion. For neurological imaging, new biomarkers, such as Pittsburgh compound B, commonly known as PiB, target plaques associated with Alzheimer’s disease. More neurological biomarkers would allow clinicians to investigate a broader range of central nervous system conditions.
In addition, new biomarkers could improve the effectiveness of treatment monitoring. Typically, FDG PET/CT is used for this purpose, but PET/MR potentially could provide more detailed information sooner because of its added capabilities and potentially lower radiation dose. Guimaraes says it may be possible to tailor an examination to specific biomarkers while taking advantage of MR’s ability to measure changes in tumor permeability with DCE, providing an earlier measurement of treatment response than what currently is possible. One hurdle, however, is the lack of attenuation correction capability on PET/MR systems, which is necessary for treatment monitoring. Guimaraes says it will also be a challenge to devise an exam that doesn’t take too long, but he adds that new applications such as this could be the gateway to personalized medicine.
Avril says a series of coordinated pilot exams are needed to determine the best uses of PET/MR, but obtaining funding from the National Institutes of Health (NIH) to investigate personalized medicine may be challenging. Because personalized medicine is a very broad topic and the NIH favors small, targeted studies, he says researchers will need to produce innovative study designs to satisfy the NIH’s criteria.
However, there will be no shortage of scientific inquiries. With so many possibilities, it’s not hard to see why PET/MR has generated so much interest. The work over the coming years will narrow those possibilities.
“I don’t know if the killer app—for lack of a better term—exists to be quite honest. It hasn’t really shown itself,” Guimaraes says. “But, on the other hand, the data are quite provocative in terms of the equivalence of the [PET/CT and PET/MR] examinations and the superiority of the soft tissue contrast in the MRI.”
— David Yeager is a freelance writer and editor based in Royersford, Pennsylvania.