The technique of ^18FFDG PET is based on the detection of coincident annihilation photons released during decay of ^18FDG [4] (Fig. 1). Neovascularization is essential to the proliferation of malignant cells; however, even in the presence of oxygen, these cells have a high rate of glycolysis. The glycolysis phenomenon was first described by Otto Warburg in 1924 and is known as the Warburg effect [7]. Metabolically active cells use a facilitated transport system for glucose uptake. ^18FFDG is a glucose analogue that competes with glucose for the same transport system. Unlike glucose, however, after initial phosphorylation into FDG-6-phosphate, FDG cannot undergo further metabolism and thus it accumulates in metabolically active tumor cells [8] (Fig. 2). Because glucose competes with ^18FFDG for uptake into metabolically active cells, it is important to have adequate blood glucose control to improve the sensitivity of the examination.

The standardized uptake value (SUV) is a semiquantitative method for assessing ^18FFDG uptake in tissues and is determined using the following formula:

At our institution, we consider an SUV of greater than 3.0 (Fig. 3A, 3B) suggestive of malignancy in the appropriate clinical setting. SUV has also been used to follow up response to therapy. Because SUV is dependent on a patient's body weight and the radiotracer injected, corrections for residual activity in the syringe and tubing and for the dose of ^18FFDG at time of injection are required to prevent incorrect results [1]. However minor errors in SUV determination are unlikely to affect patient treatment significantly.

In certain situations, it may be difficult to accurately locate an area of increased activity on PET alone due to the absence of identifiable anatomic structures, particularly in the abdomen and sometimes in the neck (Fig. 4A, 4B, 4C). Investigators recognized this limitation in oncology imaging, and Beyer et al. [6] at the University of Pittsburgh designed and built the first prototype PET/CT scanner to be used in clinical imaging. The software used by the combined scanner precisely coregisters the PET and CT images obtained during a single study on the hybrid PET/CT scanner (Fig. 5). The overall combination allows focal FDG uptake on PET to be located with greater confidence because the relative PET and CT “weighting” of the coregistered images can be altered, thus enabling FDG uptake to be precisely mapped to an anatomic structure (Fig. 5). However, the precision of coregistration is only as good as the patient's ability to remain immobile between the CT and PET portions of the examination. Movement between the two examinations may result in misregistration artifact. Motion due to involuntary activity is not as much a problem in ^18FFDG PET/CT of the head and neck as in examinations of the chest or abdomen. Adequate patient instruction and immobilization during scanning can prevent misregistration due to voluntary movements.

All ^18FFDG PET/CT scans are obtained on a dedicated hybrid PET/CT system (CTI PET, Siemens Medical Solutions) and viewed after attenuation correction on a fusion workstation using the Syngo software platform (Siemens Medical Solutions). Before scanning is performed, the patient is injected IV with 370 MBq (10 mCi) ^18FFDG and then is instructed to lie quietly in a room to decrease muscle and vocal cord uptake. Approximately 60 min after the injection, a contrast-enhanced CT scan from the skull base to iliac crests is obtained on a single-detector helical scanner (which is a component of the hybrid PET/CT scanner) with a collimator width of 5.0 mm and a pitch of 1.5. A PET scan is obtained immediately after the CT scan without allowing the patient to move on the gantry table between the two acquisitions so that the images can be accurately coregistered.

Numerous structures in the neck show physiologic uptake of FDG. These include muscles (tongue and paraspinal muscles, vocal cords), lymphoid tissue (mucosa-associated lymphoid tissue, palatine and lingual tonsils), brown fat, and salivary glands [9, 10] (Fig. 6A, 6B, 6C, 6D, 6E, 6F, 6G, 6H, 6I, 6J). A high SUV (in the malignant range) may be seen in these normal tissues as a result of physiologically increased metabolic activity (Fig. 6A, 6B, 6C, 6D, 6E, 6F, 6G, 6H, 6I, 6J), and sometimes it may not be possible to distinguish benign from malignant disease on the basis of SUV alone. In such cases, the pattern of FDG uptake in the neck may be more helpful. For example, in a patient with metastasis of unknown primary cancer, asymmetry in uptake in the tonsils or tongue base may help guide biopsy (Fig. 3A, 3B). Occasionally, differentiation between normal and abnormal ^18FFDG uptake may be difficult on PET alone; PET/CT may be valuable in such situations by accurately localizing increased ^18FFDG uptake (Fig. 4A, 4B, 4C).

Accurate initial assessment of the primary site of head and neck cancers, nodal involvement, and metastasis evaluation is crucial in staging, treatment planning, and establishing the prognosis of patients with these cancers. Although most primary malignancies of the oral cavity, larynx, and pharynx are readily accessible for clinical examination and biopsy, clinical examination often understages the extent of disease [11]. Five percent of patients with head and neck squamous cell carcinoma present with metastatic cervical nodes without an identifiable primary site at clinical examination [12, 13]. Also, evaluation of nodal status may be limited by a patient's body habitus or lack of accessibility due to the location of the node. Therefore, almost all patients with head and neck cancers need some form of cross-sectional imaging for assessment of the extent of disease.

^18FFDG PET is superior to CT or MRI for detecting lymph node metastasis, with sensitivity and specificity of approximately 90% and 94%, respectively, as compared with 82% and 85% for CT, 80% and 79% for MRI, and 72% and 70% for sonography, respectively [14]. However, ^18FFDG PET has poor spatial resolution and cannot accurately assess lymph node morphology, which is relevant for nodal staging. Number (single or multiple), distribution (ipsilateral, contralateral, or bilateral), and lymph node size (smaller than 3 cm, between 3–6 cm, and greater than 6 cm) are all important for nodal staging of head and neck cancers. Although PET is better for the assessment of metastasis in lymph nodes that appear morphologically normal according to size criteria, CT is more accurate for assessing the level and size of nodes (Fig. 7A, 7B), the number of nodes in conglomerate nodal masses, and the presence of extracapsular spread—factors that are important for determining the prognosis of patients [15]; macroscopic extranodal spread carries a 10 times greater risk of recurrence and reduces survival by 50% compared with nodes that have either no or only microscopic extracapsular spread. Also, intense ^18FFDG uptake by the primary tumor may obscure uptake by adjacent enlarged lymph nodes, thereby resulting in false-negative results. Therefore, a combination of PET/CT is likely to result in more accurate nodal staging than PET or CT alone. Size and local extent of the primary tumor, which are important for staging, prognostication, and treatment planning, are more accurately assessed on CT or MRI (Figs. 8A, 8B, 8C and 9A, 9B).

PET with ^18FFDG has been shown to be able to reveal unknown primary tumors in approximately 30–50% of patients with metastatic disease to the lymph nodes in the neck who had no detectable primary tumor at clinical examination [16] (Figs. 10A, 10B, 10C and 11A, 11B, 11C), whereas CT and MRI in combination with endoscopy and random biopsy of likely primary sites such as tonsils, nasopharynx, and tongue base reveal the primary site in 10–20% of these patients [13]. Because normal structures such as lymphoid tissue in the tonsils, tongue base, and nasopharynx show variable physiologic activity, they may potentially hinder detection of such unknown primaries. Abnormally high or asymmetric ^18FFDG uptake helps in guiding direct examination and biopsy.

PET/CT may also aid in the detection of a second primary tumor in patients with head and neck carcinoma who are at an increased risk for synchronous or metachronous carcinomas of the head and neck, esophagus, and lung (Fig. 12A, 12B). The risk of a second primary tumor is approximately 4% per year, with 80% of the second primary lesions located in the oral cavity, 40% in the larynx or pharynx, 31% in the lungs, and 9% in the esophagus [17]. Presence of distant metastasis greatly affects the treatment (palliative vs curative) and prognosis of patients with head and neck cancers. Metastasis to the lungs, liver, and bones is likely in patients with advanced-stage and recurrent head and neck cancer [12]. PET with ^18FFDG has a sensitivity and specificity of 90% and 94%, respectively, for detection of distant metastasis [18] (Figs. 12A, 12B and 13A, 13B). In the presence of distant metastases, palliative therapy would be more appropriate than aggressive therapy.

Accurate initial staging is critical in the treatment of patients with head and neck cancer. Although CT and MRI are superb cross-sectional imaging techniques, ^18FFDG PET/CT combines the excellent anatomic detail of CT with metabolic activity data from ^18FFDG PET in a single imaging session, thereby increasing the accuracy of initial staging in these patients.

Address correspondence to V. Kapoor ([email protected]).