In Translation: Easing PET/MR into Clinical Practice

PET/MR is slowly carving out a space with an increasing number of major institutions taking advantage of the system for a widening range of research applications. However, when the technology will debut in mainstream clinical practice remains up in the air. Some experts contend PET/MR is gaining support, including from payers, as the next big hybrid imaging system.

The future of PET/MR in practice is riding on further developments in research that present substantial evidence of patient benefit. Translating this technology into clinical practice could be facilitated by fleshing out niches for the technology, ironing out technological kinks and addressing the current lack of reimbursement for integrated PET/MR procedures.

Applications that could benefit from translation are those related to neurology, neuro-oncology, pediatric medicine and especially pediatric oncology, hybrid imaging of prostate cancer and staging of thyroid cancer patients, notes Florian C. Gaertner, MD, a researcher and PET/MR expert from the department of nuclear medicine at Technische Universität München, München, Germany. “However, its definitive clinical benefit over established imaging modalities, like for example PET/CT, and its cost-efficiency in the routine setting still remains to be established,” he adds.

Cutting-edge brain imaging

One of the most promising applications is in the realm of neurology, especially with the use of novel biomarkers and labeled amino acids for assessment of neurodegenerative diseases and dementia, epilepsy and cancers of the central nervous system, Gaertner notes.

A study of brain imaging applications for PET/MR published in the December 2012 issue of the Journal of Nuclear Medicine reviewed several neurological and neuropsychiatric studies. Simultaneous PET/MR could become the technology of choice for these disciplines, as dynamics in anatomy and brain chemistry can change from moment to moment and sequential imaging may not suffice.

“In the study of physiologic brain function, simultaneous acquisition may allow improved in vivo assessment and cross-correlation of several neuropsychologic events, such as changes in hemodynamics, including quantitative assessment of cerebral blood flow, volume, and oxygenation; neurovascular coupling; inflammation and microglial activation; ischemia; necrosis and apoptosis,” notes Ciprian Catana, MD, PhD, assistant professor of radiology at Massachusetts General Hospital (MGH) and Harvard Medical School in Boston, and colleagues.

“Simultaneous PET/MR allows us to change the way we do brain imaging,” says Catana, who estimates he is involved in as many as 10 ongoing brain studies involving PET/MR that have the potential to have an impact on clinical practice, especially in terms of dementia imaging. New techniques in MR imaging—functional MRI, quantitative MRI morphology and diffusion tensor imaging—and techniques that tap into the default mode network and other resting-state functional connectivity neural networks may form the core of next-generation neurodegenerative imaging.

New view on oncology

MR has emerged as the optimal choice, over CT, for successful staging of tumors in the head and neck region due to tumor characterization and MR’s ability to image a diverse range of tissue heterogeneity, including variations in soft tissue structures and lymph- node metastases.

“It is widely assumed that patients with tumors in regions with adjacent different soft-tissue types benefit from the better intrinsic soft-tissue contrast delivered by MRI compared to CT, which includes for example evaluation of the brain, head and neck, female breast, upper abdominal organs, pelvis and the musculoskeletal system,” explains Gaertner.

Well-matched radiopharmaceuticals, including F-18 FDG, greatly improve the specificity of PET/MR. At MGH, an estimated 20 percent of head and neck cancer patients require both metabolic F-18 FDG PET imaging as well as MR scanning. Prostate cancer detection could potentially get a boost with PET/MR when used with radiotracers C-11 acetate (Eur J Radiol. 2012) and C-11 choline (J Nucl Med. 2012;53:546–551).

Attention to attenuation

Recent reports indicate that attenuation correction derived from MR-based data is comparable to conventional transmission-based attenuation correction (Med. Phys. 0094-2405/2012–/39(10)/6443/12/). Not only that, but it promises to reduce radiation dose by approximately 50 percent (J Nucl Med 2013; 54:1–10). Accounting for photon attenuation in the context of PET/MR, with the lack of CT and the addition of radiofrequency coils, has necessitated advancements in MR sequences. Specialized ultra-short time echo (UTE) sequences have been developed specifically for musculoskeletal imaging and do a better job than conventional pulse sequences.

“Using conventional MR sequences, it is almost impossible to image bone,” says Catana. “When it comes to attenuation correction, bone is the most significant part of the anatomy because it has the highest attenuation coefficient. With MR, it is easy to identify soft tissues, but segmenting bone from air cavities is very challenging. The new sequences allow just that—imaging of bone.”

PET/MR attenuation correction, especially for whole-body MR, is reasonably effective and shows very little difference in SUV and visibility of lesions, says Pamela Woodard, MD, of the Mallinckrodt Institute of Radiology at Washington University School of Medicine in St. Louis, Mo. “However, for the head it is more of a problem, both with standard MR attenuation correction and even with some of the ultra-short UTE sequences that people have looked to.” Electronic data stores could one day fill in the blanks. “Many of the bugs and concerns about attenuation correction for bone will be worked out with attenuation correction libraries.” 

Rally for reimbursement

Currently there are only about six or seven PET/MR systems in North America, estimates Woodard, with more planned, but all are presently used for research applications. Washington University School of Medicine in St. Louis may be first in line to receive a certificate of need for clinical use of PET/MR in the U.S., says Woodard.

Use of PET/MR in a clinical context is still very much in its infancy. In fact, it is almost nonexistent, even in Europe where there are greater numbers of research sites for the technology. There are 25 to 30 systems worldwide and no promises of dedicated reimbursement, says Catana, who suggests that the pattern is likely to continue until more research is conducted to prove its clinical appropriateness.

In the meantime, the first institutions using simultaneous PET/MR for clinical applications in the U.S. will have to bill individually, due to a lack of CPT codes from CMS, according to Woodard. 

However, CMS language is starting to include PET/MR, bit by bit. “It’s encouraging that CMS mentioned PET/MR in one of their recent memos for PET in December,” says Catana. “It was a memo on a different topic, but in the background they made a statement about PET and PET/CT and integrated PET/MR. This seems to suggest that they would be willing to enter PET/MR as a new modality and this also suggests that they will introduce it as they did PET/CT, but it’s too early yet to tell.”

Despite the challenges and the tremendous expense of the technology, there appears to be a place for clinical PET/MR. Continued clinical research in neurology, oncology, cardiac and pediatric applications as well as technological advancement in attenuation correction, segmentation and MR sequencing in the next five to 10 years should provide a litmus test for how this hybrid will fare outside of the realm of research.