How successful is high-dose (> or = 60 Gy) reirradiation using mainly external beams in salvaging local failures of nasopharyngeal carcinoma?

PURPOSE: To evaluate the efficacy of high-dose (> or = 60 Gy) reirradiation using mainly external beams in salvaging local failures of nasopharyngeal carcinoma (NPC) after modern primary radical radiotherapy that delivered radical dose-to-target volumes defined by CT scan.

METHODS AND MATERIALS: Nine hundred and three patients with nondisseminated NPC whose primary radical radiotherapy was administered between 1984 and 1989 inclusive were studied. One hundred and seventy-six had local failures comprising 9 persistences and 167 recurrences. In 10 patients the local failures were preceded or accompanied by (within 2 months) distant metastases, and these were given supportive treatment or palliative radiotherapy in low dose (< 60 Gy) if symptomatic. Most of the rest (123 of 166) were treated with either reirradiation to high dose (> or = 60 Gy) using mainly external photon beams (n = 103) or nasopharyngectomy with/without radical neck dissection with/without postoperative radiotherapy (n = 20). The remainder (n = 43) received only palliative treatments because of poor general condition and/or patients' refusal of radical treatments. The primary radiotherapy was planned on the basis of target volumes defined by CT scan and given to a standard nasopharyngeal dose of 62.5 Gy/29 fractions/6 weeks. In the presence of parapharyngeal involvement, an additional boost of 20 Gy/10 fractions/2 weeks was given via a posterior oblique photon beam. If local residual tumors were diagnosed at 4-6 weeks after the completion of external radiotherapy, an additional boost of 24 Gy/3 fractions/15 days was given by intracavitary intubation. For the local failures given high-dose reirradiation, the target volume was defined by CT scan and treated by a two-field or a three-field photon arrangement with or without additional dose supplement by intracavitary intubation. Nasopharyngectomy was performed via the transcervico-mandibulo-palatal approach or the maxillary swing approach. Radical neck dissection was only performed for the clinically evident nodal failures.

RESULTS: With a median follow-up of 20 months (range 2.5-81 months) since the diagnosis of local failure, the actuarial 5-year overall survival, further relapse-free survival and free-from-local-tumor rates were 9.4, 11.5, and 18.7%, respectively, for the 123 patients treated by either high-dose reirradiation (n = 103) or nasopharyngectomy (n = 20). Palliatively treated patients (n = 43) had a survival comparable to patients whose local failures were preceded or accompanied by distant metastasis (n = 10). Reirradiation to high dose (> or = 60 Gy) mainly by external photon beams achieved a 5-year overall survival of 7.6% and 5-year local control of 15.2% with significant complications. Radiation-induced temporal lobe encephalopathy was radiologically evident in 21 patients (20.4%), and 13 of these 21 patients were symptomatic. It could have been the cause of death in three patients who also suffered from uncontrolled local tumor. Significant morbidity was also associated with the other frequent radiation complications, including xerostomia, trismus, and deafness. Uni- and multivariate analyses indicated that brief initial disease-free interval between completion of primary radiotherapy and diagnosis of local failures and advanced recurrent T-stage and recurrent N-stage were significant prognosticators predicting poor survival and/or further local failure after reirradiation. These patients were unlikely to benefit from the treatment. Nasopharyngectomy (+/- neck dissection +/- postoperative radiotherapy) was associated with earlier recurrent T-stages (mostly rT1 and rT2) and better survival and local control than reirradiation. However, restricting the comparison to rT1 and rT2 still demonstrated the superior results in favor of nasopharyngectomy, which could not be explained by the selection of less advanced lesions or patients with better performance status for surgery. (ABSTRACT TRUNCATED)