One of the side effects of standard radiation therapy for cancer of the nasopharynx is a permanent lessening of normal mouth moisture (saliva). This reduction in saliva is important because it causes a feeling of dry mouth, and has been shown to increase the risk of dental cavities and infections; change or decrease the ability to taste certain foods; and make chewing and swallowing food more difficult. Recent technical advances in radiation therapy offer the possibility of shielding a portion of one of the major salivary glands (parotid gland) from receiving a dose of radiation that would eliminate its ability to produce saliva, while still treating all sites of known cancer effectively. Recently, cancer researchers in Ann Arbor, Michigan used this new treatment technique to treat patients with head and neck cancers (but none with nasopharyngeal cancer), and found that patients treated in this manner still had significant saliva production from the spared gland. This study will try to use the treatment planning technique called intensity-modulated radiation therapy to protect a portion of one parotid gland while treating all known and suspected areas of cancer to full radiation doses.

Radiotherapy is the primary treatment modality for patients with nasopharyngeal carcinoma, with local control being directly related to dose and to the technical accuracy with which the dose is delivered to the target volume. Treatment planning for nasopharyngeal carcinoma is complicated by the presence of many critical structures (spinal cord, brainstem, temporal lobe, optic chiasm and nerves, parotid glands, etc.) adjacent to the primary target volume (primary cancer), which itself requires a higher prescription dose for good probability of local control than the tolerance doses for any of these adjacent organs at risk. Conventional, traditional radiotherapy has therefore involved treating the adjacent organs at risk to tolerance (e.g. spinal cord) or even above tolerance (e.g. parotid gland) to achieve a reasonable chance of locoregional control, leaving the patient with significant acute and late normal tissue toxicity that impacts both ability to tolerate the actual treatment and on the patient's long-term quality of life. Recently, static conformal multisegmental intensity modulated radiotherapy for comprehensive irradiation of head and neck cancer with dose sparing of uninvolved tissues has been reported (Eisbruch et al., 1996). These authors have demonstrated substantial preservation of major salivary gland function using these techniques for primary sites in the oral pharynx, oral cavity, epiglottic larynx and pyriform sinus, which would be expected to significantly improve long-term quality of life, as a permanent xerostomia is the most prevalent late side effect of irradiation for head and neck cancers and is cited by patients as the major cause of decreased quality of life reported. To date, these techniques have not been described in the published literature for nasopharyngeal carcinoma. Such techniques offer the possibility of significant improvement in normal tissue toxicity, while maintaining possibly better rates of local control due to improved target localization. The primary objective of this study is to evaluate whether the intensity modulated radiotherapy can reduce the incidence of permanent xerostomia in N0 and N1 nasopharynx cancer patients by sparing the contralateral parotid gland, while delivering full dose to all known and suspected areas of disease. Patients on study will receive 70 Gy given in 2 Gy once daily fractionation five days a week to all grossly involved areas, and 50 Gy given in 2 Gy once daily fractionation five days a week to areas at 15% or more risk of subclinical microscopic involvement. Patients will have subjective and objective assessment of salivary gland function done prior to, during, and after radiotherapy. Toxicity evaluation using the acute and chronic RTOG grading scales will be done. The primary endpoint of this study is the frequency of grade 2 chronic salivary gland toxicity. Correlation between toxicity and quality of life will be assessed. The hypothesis that the use of IMRT can reduce grade 2 or worse chronic xerostomia by 50% relatively will be tested. The required sample size for 90% statistical power and one-sided type one error of 0.05 is 13 patients. This assumes that the contralateral parotid sparing approach will have 40% grade 2 or more chronic xerostomia, whereas the population of conventionally treated nasopharynx patients has 80% grade 2 or more chronic xerostomia. Additional patients are added for possible ineligibility and inevaluability. A total of 18 patients will be accrued for this study. A significant survival advantage to concurrent cisplatin-based chemoradiation plus adjuvant chemotherapy for nasopharyngeal cancer was demonstrated recently in the published results of Intergroup Study 0099 (Al-Sarrf et al., 1998) but this study used conventional radiotherapy techniques that did not incorporate recent refinements to radiation planning now available using image-guided beam's-eye-view modeling and 3-dimensional dose distributions. For the purposes of this study, concurrent chemoradiation and adjuvant chemotherapy as per the experimental arm of Intergroup Study 0099 will be considered standard management, so that patients who accepted this treatment would be eligible for this study.

Inclusion Criteria:

• Epithelial carcinoma of the nasopharynx, oral cavity, oropharynx, larynx, or hypopharynx

Exclusion Criteria:

• Previous therapy for carcinoma of head/neck