Intensity-modulated radiation therapy for head-and-neck cancer: the UCSF experience focusing on target volume delineation.

PURPOSE: To review the University of California-San Francisco (UCSF) experience of using intensity-modulated radiation therapy (IMRT) to treat head-and-neck cancer focusing on the importance of target volume delineation and adequate target volume coverage. METHODS AND MATERIALS: Between April 1995 and January 2002, 150 histologically confirmed patients underwent IMRT for their head-and-neck cancer at our institution. Sites included were nasopharynx 86, oropharynx 22, paranasal sinus 22, thyroid 6, oral tongue 3, nasal cavity 2, salivary 2, larynx 2, hypopharynx 1, lacrimal gland 1, skin 1, temporal bone 1, and trachea 1. One hundred seven patients were treated definitively with IMRT +/- concurrent platinum chemotherapy (92/107), whereas 43 patients underwent gross surgical resection followed by postoperative IMRT +/- concurrent platinum chemotherapy (15/43). IMRT was delivered using three different techniques: 1) manually cut partial transmission blocks, 2) computer-controlled auto-sequencing segmental multileaf collimator, and 3) sequential tomotherapy using dynamic multivane intensity-modulating collimator. Forty-two patients were treated with a forward plan, 102 patients with an inverse plan, and 6 patients with both forward and inverse plans. The gross target volume (GTV) was defined as tumor detected on physical examination or imaging studies. In postoperative cases, the GTV was defined as the preoperative gross tumor volume. The clinical target volume (CTV) included all potential areas at risk for microscopic tumor involvement by either direct extension or nodal spread including a margin for patient motion and setup errors. The average prescription doses to the GTV were 70 Gy and 66 Gy for the primary and the postoperative cases, respectively. The site of recurrence was determined by the diagnostic neuroradiologist to be either within the GTV or the CTV volume by comparison of the treatment planning computed tomography with posttreatment imaging studies.
RESULTS: For the primary definitive cases with a median follow-up of 25 months (range 6 to 78 months), 4 patients failed in the GTV. The 2- and 3-year local freedom from progression (LFFP) rates were 97% and 95%. With a median follow-up of 17 months (range 8 to 56 months), 7 patients failed in the postoperative setting. The 2-year LFFP rate was 83%. For the primary group, the average maximum, mean, and minimum doses delivered were 80 Gy, 74 Gy, 56 Gy to the GTV, and 80 Gy, 69 Gy, 33 Gy to the CTV. An average of only 3% of the GTV and 3% of the CTV received less than 95% of the prescribed dose. For the postoperative group, the average maximum, mean, and minimum doses delivered were 79 Gy, 71 Gy, 37 Gy to the GTV and 79 Gy, 66 Gy, 21 Gy to the CTV. An average of only 6% of the GTV and 6% of the CTV received less than 95% of the prescribed dose.
CONCLUSION: Accurate target volume delineation in IMRT treatment for head-and-neck cancer is essential. Our multidisciplinary approach in target volume definition resulted in few recurrences with excellent LFFP rates and no marginal failures. Higher treatment failure rates were noted in the postoperative setting in which lower doses were prescribed. Potential dose escalation studies may further improve the local control rates in the postoperative setting.