Different risks of symptomatic brain necrosis in NPC patients treated with different altered fractionated radiotherapy techniques

Purpose: To report our observation of excessive temporal lobe necrosis in nasopharyngeal carcinoma (NPC) patients treated with 160 cGy b.i.d. radiotherapy technique. During the same period, patients treated with 120 cGy b.i.d. have not shown a similar tendency. Our experience may be useful for designing unconventional radiotherapy regimens for NPC patients. Methods and Materials: During the period from October 1991 to January 1998, 81 M0, previously untreated NPC patients completed altered fractionated radiotherapy. Seventy patients were treated with the hyperfractionated technique, and 11 were treated using the accelerated-hyperfractionated scheme. Hyperfractionated radiotherapy was delivered using 120 cGy b.i.d. separated by 6-h intervals throughout the course. A minimum tumor dose of 8000 cGy was the standard dose over an 8-week period. With the accelerated-hyperfractionated scheme, 160 cGy was given twice daily, also with an interval of 6 h. The minimum tumor dose ranged between 6840 and 7640 cGy, with 7 of the 11 patients receiving 7000 cGy. The arrangement of portals was the same for both regimens. The follow-up period for patients alive was from 32 to 102 months with a median of 61 months for the hyperfractionated patients. For the accelerated-hyperfractionated group, it ranged from 67 to 82 months with a median of 72 months. No patient was lost to follow-up. Results: At the time of analysis, 49 of the 70 patients in the hyperfractionated group were alive. In the accelerated group, 8 of the 11 patients were alive. The estimated radiation dose to the temporal lobe for the hyperfractionated group was 6000-7440 cGy with a median of 7080 cGy. For the accelerated-hyperfractionated group, the dose range was 4480-6700 cGy with a median of 6400 cGy. Of the 70 patients treated with hyperfractionated radiotherapy, none developed symptomatic brain necrosis, despite the higher total dose to the temporal lobe in general. In contrast, 3 of the 11 (27%) patients irradiated using the accelerated-hyperfractionated regimen suffered from temporal lobe necrosis at 16, 19, and 40 months after completion of radiotherapy. Conclusion: An excessive incidence of temporal lobe necrosis was noted when an accelerated-hyperfractionated regimen with 160 cGy b.i.d. was used in NPC patients with a median brain dose of 6400 cGy. There has been no such event in patients treated using a hyperfractionated regimen with 120 cGy and a median brain dose of 7000 cGy. The real causes of this discrepancy are not known. However, a high sensitivity of the human brain to a change in fraction size may play a role.