Intensity-modulated radiation therapy (IMRT) for nasopharynx cancer: update of the Memorial Sloan-Kettering experience.

PURPOSE

We previously demonstrated that intensity-modulated radiation therapy (IMRT) significantly improves radiation dose distribution over three-dimensional planning for nasopharynx cancer and reported positive early clinical results. We now evaluate whether IMRT has resulted in improved outcomes for a larger cohort of patients with longer follow-up.

METHODS AND MATERIALS

Since 1998, all 74 patients with newly diagnosed, nonmetastatic nasopharynx cancer were treated with IMRT using accelerated fractionation to 70 Gy; 59 received a hyperfractionated concomitant boost, and more recently 15 received once-daily treatment with dose painting. With the exception of Stage I disease (n = 5) and patient preference (n = 1), 69 patients received concurrent and adjuvant platinum-based chemotherapy similar to that in the Intergroup 0099 trial.

PATIENT CHARACTERISTICS

median age 45; 32% Asian; 72% male; 65% World Health Organization III; 6% Stage I, 16% Stage II, 30% Stage III, 47% Stage IV. Median follow-up is 35 months. The 3-year actuarial rate of local control is 91%, and regional control is 93%; freedom from distant metastases, progression-free survival, and overall survival at 3 years are 78%, 67%, and 83%, respectively. There was 100% local control for Stage T1/T2 disease, compared to 83% for T3/T4 disease (p = 0.01). Six patients failed at the primary site, with median time to local tumor progression 16 months; 5 were exclusively within the 70 Gy volume, and 1 was both within and outside the target volume. There is a trend for improved local control with IMRT when compared to local control of 79% for 35 patients treated before 1998 with three-dimensional planning and chemotherapy (p = 0.11). Six months posttherapy, 21%, 13%, 15%, and 0% of patients with follow-up audiograms (n = 24 patients) had Grade 1, 2, 3, and 4 sensorineural hearing loss, respectively. For patients with >1 year follow-up (n = 59), rates of long-term xerostomia were as follows: 26% none, 42% Grade 1, 32% Grade 2, and zero Grade 3.

CONCLUSIONS

The pattern of primary site failure within the target volume suggests locally advanced T stage disease may require a higher biologic dose to gross tumor. Rates of severe (Grade 3-4) ototoxicity and xerostomia are low with IMRT as a result of normal-tissue protection. Distant metastases are now the dominant form of failure, emphasizing the need for improved systemic therapy.