METHODS: Twenty patients with stage III-IV head and neck cancer were prospectively accrued. A computed tomography simulation was done prior to treatment and repeated at weeks 3 and 6 of concurrent intensity-modulated radiotherapy and chemotherapy. The final plan was coregistered with the subsequent computed tomography images, and dosimetric/volumetric changes at weeks 1 (baseline), 3, and 6 were quantified in high-risk clinical target volumes, low-risk clinical target volumes , right parotid , left parotid , and spinal cord . An event to trigger ART was defined as spinal cord maximum dose >45 Gy, parotid mean dose >26 Gy, and clinical target volume coverage <95%.
RESULTS: Comparing the 2 groups, the proportion of patients with at least 1 event triggering ART was higher in bulky disease than in postoperative group: 72.7% versus 18.2% (P = .03) overall; 54.6% versus 1.8% (P = .064) at week 3; and 63.6% versus 18.2% (P = .081) at week 6. In the bulky disease group, 8 of 11 patients had events at week 3 and/or 6 as follows: overdose in spinal cord (n = 2), right parotid (n = 3), left parotid (n = 5), coverage < 95% seen in low-risk clinical target volumes (n = 3), and high-risk clinical target volumes (n = 5). In the postoperative group, 2 of 11 patients had events: spinal cord (n = 1) and low-risk clinical target volume (n = 1).
CONCLUSIONS: Our study confirmed the need for ART in patients with head and neck cancer having bulky disease due to target under dosing and/or spinal cord/parotids overdosing in weeks 3 and 6. In contrast, the benefit of ART in postoperative patients is less clear.