Perioperative high-dose-rate brachytherapy (PHDRB) in previously irradiated head and neck cancer: Initial results of a Phase I/II reirradiation study.

Background: This study was undertaken to determine the feasibility of salvage surgery and perioperative high-dose-rate brachytherapy (PHDRB) at the dose/fractionation schedule proposed in patients with previously irradiated, recurrent head and neck cancer or second primary tumors arising in a previously irradiated field.

Methods And Materials: Twenty-five patients were treated with surgical resection and PHDRB. The PHDRB dose was 4 Gy b.i.d. x 8 (32 Gy) for R0 resections and 4 Gy b.i.d. x 10 (40 Gy) for R1 resections. Further external beam radiotherapy or chemotherapy was not given.

Results: Resections were categorized as R0 (negative margins of at least 10 mm) in 3 patients (12.0%) and R1 (negative margins of less than 10 mm or microscopically positive margins) in 22 (88.0%). Twelve patients with R1 resections had microscopically positive margins (48%), and 10 patients had close margins (40%), with a median of 2.0 mm. Ten patients (40.0%) developed Radiation Therapy Oncology Group Grade 3 or greater toxicity. Seven patients (28%) presented complications requiring a major surgical procedure. Four of these complications appeared in the immediate postoperative period and were surgical in nature (flap failure, n = 2; fistula, n = 2), and the other three were mainly related to the brachytherapy procedure (n = 2) or the radiation dose delivered (n = 1). One patient died on postoperative day 11 due to bleeding. After a median followup of 14 months, the 4-year local control rate and overall survival were 85.6% and 46.4%, respectively.

Conclusions: Surgical salvage and PHDRB at the dose/fractionation proposed are feasible in this high-risk population. Toxicity is high, but not substantially different from other reirradiation series. Four-year local control results are encouraging taking into account that 22 of 25 patients (88%) had either close or microscopically positive margins.