T classification and clivus margin as risk factors for determining locoregional control by radiotherapy of nasopharyngeal carcinoma.

BACKGROUND: The purpose of this study was to determine risk factors that affect locoregional control of nasopharyngeal carcinoma (NPC) after radiotherapy. Computed tomography (CT) is utilized for radiotherapy planning and for identifying high risk anatomic areas. METHODS: Between April 1990 and December 1993, 40 consecutive patients (1 in Stage I, 3 in Stage II, 5 in Stage III, and 31 in Stage IV) who had locoregional NPC were given definitive radiotherapy at the Koo Foundation Sun Yat-Sen Cancer Center in Taipei, Taiwan. All patients had individualized CT treatment planning. The dimension of each tumor as shown on the treatment planning CT were mapped on conventional simulation films. The extent of each tumor was further affirmed by magnetic resonance imaging (MRI) and the tumor map revised as necessary. The primary radiation fields were designed to include the primary tumor and potential spread areas with appropriate margins. Concurrent chemotherapy was also given to 35 patients (87.5%) who had positive cervical lymph nodes or primary tumors extending beyond the nasopharynx. RESULTS: By the end of December 1995, after a median follow-up of 42 months and minimal follow-up of 24 months, the locoregional control rate at 4 years was 84.8% (95% confidence interval [CI], 72.3-97.3), disease free survival 68.4% (95% CI, 52. 1-84.7), and overall survival 76.7% (95% CI, 63.4-90.0). The radiation field margin near the sphenoid sinus averaged 1.9 cm, the clivus margin 1.1 cm, the pterygoid fossa margin 2.0 cm, and the oral cavity margin 1.7 cm. Risk factor analysis revealed that T classification and the radiation field margin at the clivus were the most important factors for locoregional control of the tumor. The locoregional control rates were 92.6% (25/27) for T1-T3 patients and 76.9% (10/13) for T4 patients (P = 0.03). The locoregional control rates were 71.4% (5/7) for patients with a clivus margin < 1 cm and 90.6% (29/32) for patients with a clivus margin > or = 1 cm (P = 0.08). CONCLUSIONS: The excellent locoregional control observed in this series may be attributed to the concurrent chemotherapy and radiotherapy as well as meticulous treatment planning with CT and MRI. The precise delineation of the involved area with the aid of CT, which is taken while the patient is in the position for irradiation, serves to define the necessary safety margin of the radiation field. T classification and clivus margin are the most important factors in determining locoregional control of radiotherapy of NPC. The statistical trend observed in this study indicated that the clivus margin should be adequate to reduce the failure around the clivus, as all local recurrences were observed in this area.

Duke Scholars

Author Andrew T. Huang Medicine, Hematologic Malignancies and Cellular Therapy