The role of neck surgery in patients with primary oropharyngeal cancer treated by radiotherapy

Purpose: The role of neck surgery in node- positive patients whose primary tumours are treated by definitive radiotherapy is controversial. A planned neck dissection following radiotherapy is frequently recommended regardless of response of the neck nodes to treatment. This analysis was undertaken to assess the risk of withholding planned neck dissection in patients who obtain a complete nodal response to irradiation. Materials and Methods: The analysis is based on all 100 patients treated using the concomitant boost protocol described below who presented between 1984 and 1993 with primary squamous cell carcinomas of the oropharynx and clinically positive cervical lymphadenopathy. There were 73 males and 27 females with a median age of 59. Primary disease site was base of tongue 39, tonsil 40, soft palate 14 and pharyngeal wall 7. Nodal stages were N1: 35, N2: 51 and N3: 15. Nodal size varied from 1 - 9 cm with a median of 3 cm. Radiotherapy consisted of 54 Gy in 30 fractions over 6 weeks to large fields with a boost to gross disease of 18 Gy in 12 fractions being delivered as a second daily fraction during the last 2.4 weeks of treatment. Seventy-five patients had their nodal disease treated definitively by radiotherapy; those who had complete clinical resolution of all nodal disease (62) had no planned surgery, while the remaining 13 underwent neck dissection for presumed residual disease. Twenty-five patients had either node excision (8) or neck dissection (17) prior to radiotherapy. Results: There were 8 cases of isolated neck failure of which 3 occurred in the 62 patients who had no planned neck surgery, 0 in the 13 patients who were operated for presumed residual disease (pathologically negative in 7) and 5 in the 25 patients who had initial neck surgery. Of the 62 patients who had a complete response to radiotherapy, the two year probability of neck control was 87% if the initial nodal size was ≤ 3 cm versus 85% for nodes > 3 cm. However the likelihood of a complete response was less with larger nodes: 12 of the 13 patients who failed to achieve a complete response to radiotherapy (and underwent a planned neck dissection) had nodes initially > 3cm. Fifteen patients had documented failures at the primary site. Of these, 5 were associated with ultimate neck failure compared with only 8 neck failures in the 85 patients who were controlled at the primary site. Conclusions: The policy of observation of the neck after complete nodal response high dose radiation is safe and cost effective. While planned neck dissection is a very effective treatment strategy, the low frequency of isolated neck failure in patients who respond completely to definitive radiotherapy does not justify its routine use. The probability of obtaining a complete response is reduced with large nodes but the risk of relapse after a documented complete response is unrelated to initial nodal size. Ultimate neck failure is highly correlated with failure at the primary site