Response-adapted volume de-escalation (RAVD) in locally advanced head and neck cancer.

In this phase I/randomized phase II trial for patients with locally advanced head and neck cancer, the addition of everolimus to induction chemotherapy was not beneficial. However, the elimination of elective nodal coverage in patients with favorable response to induction chemotherapy did not appear to compromise outcomes and resulted in significantly improved late toxicity.Efforts to reduce the late toxicity associated with chemoradiation (CRT) for locally advanced head and neck squamous cell cancer (LA-HNSCC) have focused on radiotherapy (RT) dose de-escalation. In this phase I/II protocol investigating the addition of everolimus to induction chemotherapy (IC), we incorporated a novel response-adapted volume de-escalation (RAVD) approach using IC response to guide the extent of RT volume reduction. Patients with measurable LA-HNSCC received two cycles of IC (cisplatin, paclitaxel, cetuximab +/- everolimus). Patients with a parts per thousand yen50% reduction in the sum of tumor diameters [good response (GR)] received TFHX (paclitaxel, fluorouracil, hydroxyurea, and 1.5 Gy twice daily RT every other week) to a dose of 75 Gy with the single planning target volume (PTV1) encompassing exclusively gross disease. Patients with < 50% response [non-response (NR)] were treated with TFHX encompassing PTV1 and the next nodal station at risk (PTV2) to a dose of 45 Gy followed by a sequential boost to PTV1 to a dose of 75 Gy. Ninety-four patients were enrolled. Randomization to everolimus was discontinued on interim analysis after 50 patients due to futility. IC response was evaluable in 89 patients. Thirty-seven patients (41.6%) had GR and 52 (58.4%) had NR. There was a trend for improved progression-free (P = 0.086) but not overall survival (P = 0.94) for GR versus NR. The 2-year PFS and OS were 86.0% and 83.5% for GR and 68.7% and 85.4% for NR, respectively. NR were significantly more likely to undergo G-tube placement during treatment (50.0% GR versus 73.5% NR, P = 0.040) and be G-tube dependent at 6-month follow-up (5.7% GR versus 32.6% NR, P = 0.005). The addition of everolimus to IC was not beneficial. The elimination of elective nodal coverage in patients with GR to IC did not appear to compromise outcomes and resulted in significantly decreased late toxicity. Further investigation of RAVD is warranted. NCT01133678.