Combined positron emission tomography (PET) and computed tomography (CT) has several advantages over either CT or PET alone in the evaluation of patients with head and neck cancer. The accurate coregistration of images and improved localization achieved with a combined modality are particularly important in the neck, where there is not only anatomical complexity, but several structures that demonstrate variable physiologic fluorodeoxyglucose (FDG) uptake.

In order to maximize the utility of PET/CT, it is important to optimize the several scan protocol options available when performing PET/CT for the evaluation of patients with head and neck cancer. A typical baseline diagnostic PET/CT for a patient with head and neck cancer is increasingly common to include a full-power CT scan with intravenous contrast in addition to an FDG PET scan acquired during the same imaging session on a hardware combined PET/CT device. Oral contrast may or may not be utilized, depending on the extent of area being scanned. A typical protocol for patients with head and neck cancer might include the skull base through the abdomen; therefore, oral contrast is not routinely utilized in the authors’ protocol, but oral contrast is routinely utilized in some centers. FDG must be injected at a consistent time interval prior to the start of patient imaging; a relatively standard protocol would involve injection of FDG 1 hour prior to the commencement of the scan. Longer times from injection until imaging are used by some groups as the tumor/background uptake ratios typically rise with time postinjection; however, about 60 minutes postinjection of tracer is very common to commence imaging.

Neck immobilization is an important protocol consideration with the evaluation of patients with head and neck cancer, as this is the area of interest. A variety of soft collars and external immobilization devices are available to help minimize or eliminate neck movement.

Arm positioning for the examination is also an important consideration. The two most commonly used positions for arm positioning during PET/CT imaging are standard “arms up” and “arms down.” In general, the most important determinant of arm positioning in PET/CT is the indication of malignancy that is being evaluated and subsequent areas of interest being scanned. The goal of any arm positioning protocol should be to minimize CT beam-hardening artifacts in the area of interest. Therefore, for patients with head and neck cancer it is usually preferable to scan patients with arms down, at least at the level of the neck. Alternatively, two scans may be performed, one with data acquisition through the chest and abdomen (with or without the pelvis) while the arms are positioned up followed by a more limited second scan acquisition through the neck area with arms positioned down. Care should be taken to ensure adequate overlap of scans at the thoracic inlet, as lower neck structures can shift substantially with the change in arm position.

The most important CT consideration for patients with head and neck cancer is whether, when, and how to use intravenous contrast and whether to use low- or high-dose CT. Because of the anatomical complexity of the neck, if logistically possible, a diagnostic CT with contrast is typically recommended. This will make it easier to differentiate the vascular from nonvascular structures, as well as to distinguish pathology from normal structures more clearly. Vascular structures can contain FDG, and such uptake can sometimes be confused with small nodes in noncontrast studies.

Other options include performing a low-dose, noncontrast CT (40 to 80 mA), which reduces radiation exposure to the patient but may have a lower sensitivity and specificity for the detection of malignancy on its own, although this has not been rigorously studied in a variety of clinical settings. It can be used for simple localization purposes, but is typically more difficult to interpret due to the lack of structural differentiation. The quality of the scan is too poor
to charge for reimbursement in most centers for the PET alone. However, since most of the tumor detection is being done with the FDG, the quality of the CT scan may not be as critical to the total diagnostic accuracy as the PET plus the localizing information from the CT. It is possible that for follow-up PET studies, such as those used for tumor surveillance, that there is little benefit from the use of a full-contrast CT scan. Not all noncontrast CT scans are performed the same way. Full-dose, noncontrast CT is another option that may be the best for patients with a significant contrast allergy.

Contrast-enhanced CT offers the most advantages to the interpreting physician, but it does require medical personnel immediately available to respond to unexpected contrast reactions. One potential disadvantage of using contrast in PET/CT imaging is the presence of attenuation-correction artifacts that are unique to PET/CT. These are discussed in detail at the end of this chapter. Although attenuation-correction artifacts generally are easy to recognize and are a valid reason to avoid intravenous contrast for qualitative PET assessments, the interpreting physician must be aware of these artifacts in order to avoid confusion with malignancy. Further, there can be some modest alterations in apparent standardized uptake value (SUV) in the presence of high concentrations of intravenous contrast.

Another approach involves a low-dose noncontrast CT prior to PET imaging (for the purpose of artifact-free attenuation correction) followed by a full-dose contrast CT of diagnostic quality. This is ideal from an imaging point of view but exposes the patient to potentially unnecessary radiation. However, this approach may be acceptable if contrast CT is only needed for a small area of the patient such as the head and neck.

Optimizing arm positioning and intravenous contrast into a single algorithm is a convenient approach. A full CT scan algorithm might include a noncontrast low-dose CT with arms up performed first (this CT will be utilized for attenuation correction) followed by CT performed with contrast after the PET acquisition. During the contrast-enhanced CT acquisition, the arms would be positioned at the side to minimize beam hardening through the area of interest. Another approach is to perform the neck study first with no contrast and the arms down, followed by the CT with contrast. This practice varies somewhat by center, but it is suggested that the use of intravenous contrast be seriously considered in any protocol designed to detect cancer in the head and neck region.

It is not clear whether all patients with a history of head and neck cancer will benefit from a diagnostic CT as part of their PET/CT workup. In general, when patients are suspected of having active disease during their initial evaluation, during therapy, and in the interim following therapy, contrast-enhanced CT is recommended. An exception can be made for patients who have had a contrast-enhanced CT performed recently. Asymptomatic patients with successfully treated malignancies undergoing surveillance with low suspicion of viable cancer and those with a recent diagnostic CT are good candidates for low-dose CT as part of their PET/CT examination.

Several normal structures in the head and neck may demonstrate physiological FDG activity on PET or PET/CT. It is important to be familiar with the appearance of normal, as well as atypical, patterns of physiological FDG uptake. Among the structures that take up FDG normally are the salivary glands, thyroid gland, brown fat, lymphoid tissue, mucosal tissue, and muscles of the neck, pharynx, and face. FDG is taken up by salivary glands and excreted into the saliva to a limited extent (1). The parotid, submandibular, and thyroid glands can demonstrate mild to intense symmetric uptake (Figs. 8.7.1 and 8.7.2).
Brown fat may be found throughout the neck and thorax as well as outside the neck (Fig. 8.7.3) (2,3,4). FDG is taken up by macrophages and lymphocytes, therefore, lymphoid structures are also commonly appreciated on FDG PET scans (Fig. 8.7.4). Examples of lymphoid structures include all the structures of Waldeyer’s ring (adenoids, palatine, and lingual tonsils), as well as lymph nodes (5,6,7,8). Mucosal tissue such as that found in the nasal turbinates and seen throughout the oro- and nasopharynx, commonly shows FDG uptake. Muscles are among the many normal structures commonly seen on FDG PET or PET/CT studies (Fig. 8.7.5). Muscles of mastication, including the pterygoid muscles, and muscles of the oral floor may also be seen (Fig. 8.7.6), as well as the pharyngeal constrictor muscles, muscles controlling the vocal cords, and the vocal cords themselves (Fig. 8.7.7) (9). An in-depth review of the more atypical physiologic FDG uptake patterns and pitfalls is provided at the end of this chapter.