Do We Need Yet Another Hybrid Imaging Device?

The arrival of another hybrid imaging modality fills us with excitement and despair. Excitement at the prospect of simultaneously using different ways of interrogating pathophysiology non-invasively in vivo; despair at the invariable increase in costs associated with its use. Is it any use, we ask, is it worth it? The accompanying article by Fraum et al (1) provides a broad and sweeping overview of the various ways in which whole body positron emission tomography/magnetic resonance imaging (PET/MRI) can identify and characterize numerous malignancies. The article provides comparisons of PET/MRI and positron emission tomography/computed tomography (PET/CT)—the latter the current standard for hybrid imaging—and concludes that this novel hybrid imaging modality can provide comparable and sometimes clinical information in a variety of cancers. The new modality does seem to be of use. We, however, need to determine its incremental utility. There are now two classes of PET/MRI devices: nontime of flight (non-ToF) devices with avalanche photodiodes (APD), and time of flight (ToF) devices with silicon photodiodes. Attenuation correction using magnetic resonance (MR) remains challenging for bone as well as air (2), with consequent impact on PET quantification. Although acquisition of the MR signal (for attenuation correction as well as for MRspecific sequences) occupies more of the acquisition time necessary for PET, it is indeed simultaneous, in contrast to the sequential nature of PET/CT. Moreover, additional MR sequences—of the same area, at the same time—can be obtained during the PET acquisition. PET/MRI thus enables simultaneous acquisition of linked imaging characteristics—for example, kinetics of exogenous choline (administered as a PET tracer) with changes in endogenous choline (measured by MR proton spectroscopy). CT provides structural information (as a radiodensity map); MR provides much more pathophysiologic information: water content, oxygenation, molecular concentrations. The marriage of PET and MRI may, by enabling the interrogation of multiple pathophysiologic characteristics, potentially provide far greater information than currently available with PET/CT (3). PET/MRI can certainly provide comparable and complementary clinical information to PET/CT. Data regarding its utility are being gathered in prospective trials, as recommended recently at a workshop, where reimbursement considerations were also discussed (4). What only time will answer, however, is whether imaging groups will commit to the high investments necessary for PET/MRI, especially given current health-care cost constraints (not to mention the investments already made in PET/CT). What is more certain is the potential for PET/MRI to uniquely enhance our understanding of pathophysiology by providing complementary functional information. For those of us fortunate enough to be able to acquire such a device, many questions can now be explored; the device is worth it. Glutamine uptake measured by PET could be compared to endogenous hydroxyglutarate levels measured by MR, glucose uptake with tissue oxygenation. Such correlative assessments would yield new knowledge into the functional states of disease, enabling better drug discovery, development, and eventually selection. This last item will demonstrate the incremental utility of PET/MRI, enabling its optimal use in our clinical practice, and making it truly worth it.