The hypopharynx extends from the level of the hyoid bone to the inferiormost level of the cricoid cartilage. Anteriorly, the border is formed by the connection of the two pyriform sinuses at the postcricoid region and with the posterior extent defined by the posterior pharyngeal wall.

The three subdivisions and their distribution of tumor involvement according to the SEER database, from 2000 to 2008, are the pyriform sinuses (83 %), posterior pharyngeal wall (9 %), and postcricoid region (4 %).

Dosimetric coverage of the primary site can be a challenge because hypopharyngeal cancers have a propensity for submucosal spread. Approximately 60 % microscopic spread of about 10 mm (superiorly), 20 mm (laterally and inferiorly), and 25 mm (medially) (Ho et al. 1993).

The rich lymphatic network of the hypopharynx drains primarily into the jugulodigastric lymph nodes and middle jugular chains (see Fig. 1).

Clinical examination with a thorough laryngoscopic evaluation is critical in defining the mucosal extent of hypopharyngeal cancers, particularly for those arising in the pyriform sinus.

MRI may better delineate primary tumor borders, inform regarding potential cartilage or esophageal invasion, and may also facilitate identification of abnormally enhancing lymph nodes (Castelijns et al. 1988; Roychowdhury et al. 2000; Rumboldt et al. 2006; Wenig et al. 1995).

The sensitivity and specificity of PET are superior to CT alone (Di Martino et al. 2000). FDG-PET provides metabolic information and may identify hypermetabolic cancer cells within morphologically normal-appearing lymph nodes (Adams et al. 1998; Schwartz et al. 2005), which should be included within the high-dose PTV.

CT simulation using 3-mm slice thickness with IV contrast is used (unless medically contraindicated) to delineate the primary tumor, gross lymphadenopathy, and additional regions at risk. The simulation scan should extend from the top of the skull to the carina with the isocenter generally placed immediately above the arytenoids.

The patient should be immobilized with a 5-point thermoplastic head and neck mask or with a 3-point thermoplastic head and neck mask and a shoulder pull board. The neck should be maximally hyperextended to pull as much of the oral cavity and mandible out of the field as possible, and the shoulders should be pulled inferiorly to minimize the risk of beam interference.

Due to concerns regarding treatment-related late toxicities, hypofractionation and/or simultaneous integrated boost with IMRT is not recommended for this disease site.

We utilize dose-painting IMRT (Lee et al. 2007) and prescribe to 70 Gy in 2 Gy fractions to sites of gross disease, 59.5 Gy in 1.7 Gy fractions to high-risk subclinical regions, and 56 Gy in 1.6 Gy fractions to lower-risk subclinical regions. Additional fractionation regimens as well as cone-down IMRT approaches are also appropriate.

We routinely use all-in-one IMRT for hypopharynx cancers, as a match line to a low anterior neck field would frequently fall in the region of the primary tumor or involved lymph nodes.

Suggested treatment target volumes for gross disease (Table 1), high-risk subclinical (Table 2), and lower-risk subclinical regions (Table 3) are described in the following tables. The principles of target delineation are similar for early and advanced stages (Figs. 2 vs 3, 4, and 5). Shown here are examples of each subsite: pyriform sinus (Figs. 2 and 3), posterior pharyngeal wall (Fig. 4), and postcricoid region (Fig. 5).