PET/MRI: Radiologists Are Exploring the Clinical Role of What Is Still Seen Primarily as a Research Tool
By Beth W. Orenstein
Vol. 16 No. 4 P. 12
Since its introduction in the early 2000s, PET/CT has replaced stand-alone PET as the tool for many cancer diagnoses, staging, and treatments. Since 2001, there have been more than 400 installations of PET/CT scanners annually worldwide. Over the past four years, PET/MRI has been making inroads as well and some believe it could eventually supplant PET/CT as the modality of choice in some areas of oncology, neurology, and cardiology.
PET/MR is still largely research based, says Frank DiLalla of Philips Imaging, which is one of the three manufacturers that make the hybrid modality machines, along with Siemens Healthcare and GE Healthcare. (A fourth is in development.) "We are still trying to determine which clinical applications will emerge as significantly better than its brethren PET/CT," DiLalla says. Currently, he says, "It's a mixed bag. Honestly, it depends on whom you talk to." The early adopters are working to determine where and when PET/MRI should be the hybrid modality of choice. "The next level of adopters are watching them closely," DiLalla says.
Eric Stahre, president and CEO of global MRI for GE Healthcare, says PET/MR is something that has certainly captured the imagination of academic researchers and is slowly moving into the clinical realm. "It is not something that will be ubiquitous, but it is starting to get some really nice traction in clinical settings," Stahre says.
Abe Voorhees, PhD, a business manager for Siemens, says when its Biograph mMR was introduced in 2011, the first installations were all for research purposes. But Siemens is beginning to see more adoption in the clinical realm. "Four of the last five scanners we installed will have a heavy focus, if not exclusive, on clinical imaging," Voorhees says.
At this point, Stanford University is using its GE whole body PET/MR scanner for research only, says Andrei Iagaru, MD, who practices in Stanford's division of nuclear medicine and molecular imaging. "Our efforts are focused on identifying which diseases and which patients are most appropriate for PET/MRI and where we can use it to see early if a treatment is working or not. We are also making progress developing efficient workflows for the various indications for the PET/MR," Iagaru says.
Likewise, the University of California, San Francisco (UCSF) is using its GE hybrid scanner for research only at this time. UCSF has 16 research projects involving its use, says Spencer Behr, MD, an assistant professor of clinical radiology. That includes Behr's project on the use of PET/MR for prostate cancer. He's hoping to show that one PET/MR scan could replace the multiple tests (MR of the prostate, a bone scan, and CT of the abdomen for lymph nodes) currently needed to stage prostate cancer and determine if it's localized and surgically resectable or a systemic disease that requires a different treatment plan. The issue, Behr says, is finding the tracer that works for the prostate. FDG is nonfunctional in patients with prostate cancer. "If we can find that one agent, you could do one scan and you're done," Behr says. However, he's not sure "where PET/MR is going because medicine is changing so fast."
Others have begun to use their hybrid scanner for clinical applications. Stony Brook Medicine in Long Island, New York, was among the first sites in the United States to offer simultaneous PET/MRI technology for clinical use. It installed the Siemens Biograph mMR hybrid imaging system in its Lisa and Robert Lourie Imaging Suite at the Stony Brook University Cancer Center in 2013. Mark Schweitzer, MD, FRCPSC, chief of diagnostic imaging and chair of the department of radiology at Stony Brook, believes the jury is still out on whether PET/MRI is better than PET/CT for staging and restaging cancer. It's more a matter of time than anything else, he says. "You need a body of research to show you how it's better and in what situations," he says. "That's the natural history. To expect otherwise is not realistic. PET/MRI has only been out for a few years, and you couldn't expect us to know more than we know now."
Radiologists at Cleveland Clinic are finding that they are using their PET/MRI resource, which they installed in July 2013, for clinical applications much more than they anticipated, says Shetal N. Shah, MD, an abdominal imager and director of its PET/MR program. To make it financially feasible, the department estimated it needed to do 128 PET/MRI scans and 600 MRIs (without PET) in the first year. Within the first 10 months, it had performed 185 PET/MRI scans and close to 1,000 MRIs without PET. Shah is convinced PET/MRI is not a fad but a modality that is here to stay when fully integrated into clinical care paths. The physicians at Cleveland Clinic are finding it particularly useful for cancers of the head, neck, pelvis, rectum, and liver; for epilepsy care; and in the pediatric population.
Shah recalls one patient with ovarian cancer where her tumor markers were rising but her doctors could find no metastases. She underwent a PET/MRI and sure enough she had a 6-cm mass in her brain that had not been detected to date, he says. The finding was surprising because the patient did not have neurological symptoms and ovarian cancer doesn't typically metastasize to the brain, Shah says.
Will PET/MRI ever replace PET/CT? Shah doesn't think so. "But the synergism of MRI and PET can offer vastly superior anatomic detail and biologic data that are extremely valuable in advanced evaluation of cancer, particularly with respect to surgical and radiation planning as well as in assessing therapy response," he adds. Like Behr, Shah believes it will take some time, education, and experience to build up usage and until physicians are comfortable ordering a PET/MRI. "In the meantime, early adopters ought to focus on collaborating to discover robust applications," Shah says.
Others are more encouraged by the results they have seen in the last two years. Steven Mendelsohn, MD, CEO and medical director of Zwanger-Pesiri, is perhaps the most enthusiastic of PET/MRI users. Zwanger-Pesiri acquired Siemens Biograph mMR in November 2014 for its Lynbrook, New York, facility. Zwanger-Pesiri, which has numerous locations on Long Island, is the first private radiology clinic in the United States to acquire a PET/MRI. Mendelsohn believes "PET/MRI is the biggest game changer in radiology in the last 25 years." His facility advertises its hybrid scanner as: "Radiology's answer to cancer."
A huge advantage to PET/MR over PET/CT is that patients have much less radiation exposure, Mendelsohn says. MR involves no radiation and the tracer dose with PET/MR is one-half what it is with PET/CT. "You're saving patients approximately 100 chest X-rays' worth of radiation," Mendelsohn says. Lowering radiation exposure is a huge concern in the field of radiology these days, particularly for children and women, Mendelsohn says. Many times cancer patients must undergo repeat studies, and being able to do so with less radiation could be a key to their long-term survival, he says.
Patient convenience is another advantage, Mendelsohn says. With Siemens Biograph mMR, it takes about 45 minutes to acquire a whole body PET/MRI. That's far easier for patients than having to go for a PET scan and MR separately and far less time-consuming. But perhaps the most important advantage, Mendelsohn says, is the superior information that can be acquired during a PET/MR study. "It's profoundly scary how much metastatic disease we're picking up on this machine," Mendelsohn says. A PET/MR study can find metastases, especially in the brain, that would not otherwise be detected until they grew larger and symptomatic, he says.
Mendelsohn's theory is that a PET/MR may prove to be the only study an oncology patient needs in many cases. "My contention is this one test will replace a huge number of tests and condense the whole workup into one visit and you will get the answer you're looking for 80% to 90% of the time, saving patients weeks of anxiety coming back and forth for additional testing, additional testing," he says. Also, Mendelsohn says, detecting cancer earlier before it is symptomatic means treatment can be preemptive and, perhaps, more successful. Eventually, Mendelsohn expects to have enough cases to be able to publish the center's results, but its radiologists have found they have not needed to suggest a follow-up study in any patients who have undergone a PET/MR for cancer staging or restaging.
The one application where Mendelsohn sees a slight weakness of the hybrid modality is for lung nodules. "If the nodules are smaller than 5 mm, the FDG is not particularly good," he says. "We might miss tiny lung nodules on PET/MR." Currently, those patients would need a PET/CT scan. However, Mendelsohn is optimistic that Siemens can solve this weakness. "Siemens has a phenomenal breath-hold sequence that over the next few months or a year could migrate to its PET/MR platform and show tiny pulmonary nodules," he adds.
Shah agrees that advanced tools and an adjustable field of view on PET/MRI can be of tremendous value in assessing small volume and nodal disease in cancer patients, often a limitation with PET/CT.
Voorhees says that there's no doubt the quality of the PET/MR scan is enabling clinicians who use it to "call their cases with greater confidence."
PET/MR is also proving useful in cardiac imaging and neurological imaging. Much of the work is still in the research phase, but radiologists are looking at PET/MR as the best modality for examining heart wall motion and blood perfusion. It also shows potential for detecting a broad range of central nervous system conditions including Alzheimer's disease. Mendelsohn says his radiologists are very comfortable with cardiac PET/MR and use it where appropriate. Schweitzer says using PET/MR to determine myocardial viability is on Stony Brook's radar for later this year. Behr says the cardiac surgeons and neurologists at UCSF are excited about the possibilities of PET/MR "because they see it has the potential to change the way we treat patients."
As PET/MR makes inroads in clinical settings, some issues remain to be resolved. Who reads the hybrid studies? What will insurance companies reimburse for them? Iagaru thinks this hybrid modality adds a level of complexity to interpreting exams. "We will need one MR specialist and one PET specialist for the specific indication of the exam (eg, neuro, cardiac, MSK, pediatric)," Iagaru says. At Zwanger-Pesiri, every PET/MR scan is read by at least three radiologists, each specializing a modality or discipline pertinent to the exam, and it takes at least three hours of aggregated time. "I wish I could tell you that one person is capable of reading all those modalities, but I can't," Mendelson says. Admittedly, Mendelsohn says, having three, and sometimes four, readers for each exam is time-consuming and not cost-effective, especially given that insurance does not reimburse accordingly. The facility bills the patient's insurer for the PET and the MRI, but may only be reimbursed for one or the other depending on which study or studies the patient was authorized to have. "It is economic suicide getting paid the same as a PET scan in most cases," Mendelsohn says. His job, he adds, is to convince physicians and insurance companies that it's better and more economical to order this one exam than several in sequence. And, he believes, that day will be here eventually.
Mendelsohn remembers that when MRI machines first came on the market in the late '80s and early '90s, it was a challenge to get them in use. PET/MR is no different, he says. Still, he believes, that within five to 10 years, PET/MR will become the gold standard. "People will know about it," he says. "There will be enough publications, and it will be a no-brainer to order a PET/MR vs a PET/CT and save patients exposure to radiation, especially youngsters in need of PET scans."
— Beth W. Orenstein is a freelance writer based in Northampton, Pennsylvania. She is a frequent contributor to Radiology Today.