The Role of PET-CT Fusion in Head and Neck Cancer

In their article, Rusthoven and colleagues highlight the utility of combined positron-emission tomography/ computed tomography (PET-CT) imaging for diagnosing primary and recurrent head and neck carcinoma, and for defining tumor target volumes for radiotherapy treatment planning in the head and neck. PET offers noninvasive measures of tumor biology yet suffers from limited spatial resolution; the physiologic information obtained with PET is complementary to the high-resolution structural information obtained with CT or magnetic resonance imaging (MRI). The authors provide us with an up-to-date review of the literature on fluorine-18 fluorodeoxyglucose (FDG)- PET used in conjunction with CT in the head and neck cancer patient. FDG-PET has been shown to be effective for staging and restaging head and neck cancer, and monitoring therapeutic response.[1] However, the highly detailed anatomy and variable physiologic uptake of FDG in the head and neck-both of which are altered by surgery and other cancer therapies- emphasize the special advantages of combining structural and functional imaging in head and neck cancer.[2] Combined Scan vs Post Hoc Fusion
Rusthoven et al note that PET and CT data can be coregistered by a post hoc fusion of images acquired on separate scanners, or can be coregistered after acquisition on a single combined PET-CT scanner. The authors do not differentiate between studies based on data from a single combined scanner and studies based on post-hoc fusion. However, we feel it is appropriate to emphasize that retrospective fusion of image data acquired on separate devices is fraught with pitfalls in the head and neck. Since the relative spatial relationships among head and neck structures vary from scan to scan, and fixing the neck in a hard collar is cumbersome, registration of individually acquired PET and CT images is difficult to achieve without the introduction of complex nonlinear algorithms. Reports on the accuracy of rigid-body approaches do not adequately reflect the real life setting.[3] Patient considerations- including the need for two imaging procedures rather than one, and potential changes in tumor and nontumoral tissue characteristics if the time delay between acquisition of the PET and CT images is weeks or months-further strengthen the favorability of combined scanners over post hoc techniques. Shifting Clinical Protocols
Until recently, serial imaging with CT or MRI was the predominant approach to monitoring the head and neck cancer patient, with PET imaging reserved (when available) for difficult cases. However, the increasing prevalence of combined PET-CT scanners at both academic medical centers and in the community has shifted clinical protocols at some institutions to involve or substitute PETCT studies at regular intervals. Recently, Schoder and colleagues evaluated 68 head and neck cancer patients with PET-CT.[4] FDG-PET images were initially evaluated by consensus and lesions graded as benign, equivocal, or malignant. Then the CT data were made available, and the incremental benefit of PET-CT over PET alone was assessed. The accuracy of PET significantly increased from 90% to 96%, and the fraction of equivocal lesions was reduced by 53%. Branstetter et al prospectively demonstrated (in 65 patients) improved detection of malignancy in the head and neck with combined PET-CT images relative to both FDG-PET and contrast-enhanced CT alone when each modality was interpreted separately by expert readers.[5] Rusthoven et al note that PET is less reliable immediately after cancer treatment, both surgical and nonsurgical. Although false-negative results from microscopic deposits of viable tumor may never be obviated, we believe that false-positive results are diminished with the use of PET-CT instead of PET, and with greater experience of the interpreting radiologists. Impact on Radiation Therapy
Literature on the application of PET-CT as a basis for radiation treatment planning is limited, yet its potential benefit is evident, particularly for intensity-modulated radiotherapy, where accurate delineation of tumor and spared normal structures may limit patient morbidity. Work by Scarfone and colleagues demonstrated the modification of tumor target volume definitions when FDG-PET information was added to CT simulation data.[6] Ciernik et al specifically evaluated the utility of combined-scanner PETCT images for target volume definition in 39 patients with solid tumors of the lung, pelvis, and head and neck.[7] The addition of overlay PET data either reduced or increased the target volume more than 25% in 56% of cases, including 6 of the 12 head and neck tumor cases studied. Furthermore, the use of integrated PETCT information for treatment planning for three-dimensional conformal radiation therapy decreased volume delineation variability between two oncologists who independently conducted treatment planning first with CT alone, then using PET-CT. Rusthoven et al review their own preliminary work as well as these and other recent studies that suggest PETCT imaging may affect radiation therapy planning in a substantial and positive way. We agree that PET-CT will have a strong impact on radiation planning in the head and neck. We also appreciate their note of caution that awareness of CT-based attenuation artifacts and reconstruction of non-attenuation-corrected images is warranted.[8] It is particularly important that referring ENT surgeons become aware of the utility of PET-CT in radiation planning. Patients who are likely to be treated nonsurgically benefit from the application of a customized radiation planning mask during PET-CT scanning. This mask can dramatically improve the reproduceability of patient position between diagnostic and treatment scans. Thus, nonsurgical patients should be seen by a radiation oncologist before the PET-CT is obtained, to ensure that the scan is performed properly.