Radiotherapy is often given as an adjuvant to surgery for moderately advanced but resectable tumors. For most head and neck sites, giving adjuvant radiotherapy improves local control for T3 or T4 primaries or in situations where there is pathologic involvement of cervical lymph nodes.

Radiotherapy can be given either preoperatively or postoperatively. The aims of preoperative radiotherapy are to sterilize microscopic disease outside the resection field and to shrink the tumor bulk making surgery more tractable. Theoretically, preoperative radiotherapy should also reduce the risk of disseminating viable tumor cells at the time of surgery. Dosages of 5,000 cGy over 5 to 5.5 weeks is generally given, with no significant wound healing problems occurring at this dosage.

In the postoperative setting, the surgical procedure has disrupted the regional blood supply. Conventional wisdom suggests that higher dosages of radiation are needed because of the increased likelihood of hypoxic tumor cells which would be more radioresistant than their well-oxygenated counterparts. Generally, 5,500 to 6,000 cGy in 180 to 200 cGy fractions are given for microscopic residual disease. If the surgical margins are grossly positive or if there is a high likelihood of macroscopic residual disease, then higher dosages areused. Peters and colleagues have shown that at least 6,300 cGy should be given if extracapsular nodal extension is found in the operative specimen.

Postoperative radiotherapy has the advantage of being given only to those patients thought to be at significant risk for locoregional recurrence based upon a review of the pathologic data. It has the additional advantage of not delaying the surgical procedure which is the most important treatment modality for patients with advanced, operable tumors.

Preoperative and postoperative radiotherapy were compared in a randomized clinical trial by the RTOG. A total of 277 patients with tumors of the oral cavity, oropharynx, supraglottic larynx, or hypopharynx were entered into the study. Patients in the preoperative arm received 5,000 cGy followed by surgery in 4 to 6 weeks while patients in the postoperative arm received 6,000 cGy starting 2 to 4 weeks after the surgical resection. A higher percentage of patients in the postoperative arm completed the combined course of treatment within protocol guidelines (74% vs 56%). The 4-year competing risk, locoregional tumor control, was better on the postoperative arm than on the preoperative arm (65% vs 48%, p = .04). For the subgroup of 194 patients who completed their overall treatment within protocol guidelines, the locoregional control rates were improved: 74% on the postoperative arm and 56% on the preoperative arm. There were no significant differences between the two arms in terms of complications or survival.

Although it is generally felt that there is little use for a debulking surgical procedure unless one can achieve microscopic disease levels, there may be situations where a gross total resection followed by high-dose radiotherapy is preferable to treatment with radiotherapy alone. An analysis by the Head and Neck Intergroup (IG0034) showed that excluded patients with positive surgical margins had improved locoregional tumor control compared to matched cohorts from the RTOG databases of patients treated with radiotherapy alone. At 4 years, respective locoregional control rates were 44% vs 24% (p = .007). However, there was no difference in survival. Since this was not a randomized study, the authors do not argue for changing traditional resectability criteria but rather testing this concept in the context of a controlled clinical trial.

To date there has been limited work exploring the role of accelerated or hyperfractionated radiotherapy or concomitant chemotherapy and radiotherapy in the surgically adjuvant setting. Ang and colleagues report on a multi-institutional randomized trial in which high-risk patients were assigned to either 6,300 cGy in 7 weeks using a conventional fractionation schema or in 5 weeks using a concomitant-boost technique. Risk for recurrence was assigned on the basis of primary tumor site, number and location of positive nodes, presence of extracapsular nodal extension, and presence of perineural invasion. There was a nonsignificant trend (p = .11) for improved locoregional control at 5 years (72% vs 62%) and also a nonsignificant trend (p = .08) in survival (47% vs 30%) using the concomitant-boost technique. Unfortunately, there was also a higher incidence of acute confluent mucositis with the concomitant boost (62% vs 36%, p = .001).