Seventy-two patients with nonmetastatic head and neck cancer treated with dose painting were compared with 72 control patients matched on tumor site and T classification. Either 18 F-FDG-PET-guided dose painting by contour (DPBC) or voxel intensity-based dose painting by number (DPBN) was performed; control patients underwent standard IMRT. A total median dose to the dose-painted target was 70.2-85.9 Gy/30-32 fractions versus 69.1 Gy/32 fractions with conventional IMRT. In 31 patients, dose painting was adapted to per-treatment changes in the tumor and organs-at-risk (OAR).
Median follow-up in living dose-painting and control patients was 87.7 months (range 56.1-119.3) and 64.8 months (range 46.3-83.4), respectively. Five-year local control rates in the dose-painting patients were 82.3% against 73.6% in the control (P = .36); in patients treated to normalized isoeffective doses >91 Gy (NID2Gy) local control reached 85.7% at 5 years against 73.6% in the control group (P =.39). There was no difference in regional (P = .82) and distant control (P = .78). Five-year overall and disease-specific survival rates were 36.3% versus 38.1% (P = .50) and 56.5% versus 51.7% (P = .72), respectively. A half of the dose-painting patients developed acute grade ≥3 dysphagia (P = .004). Late grade 4 mucosal ulcers at the site of dose escalation in 9 of 72 patients was the most common severe toxicity with dose painting versus 3 of 72 patients with conventional IMRT (P = .11). Patients in the dose-painting group had increased rates of acute and late dysphagia (P = .004 and P = .005).
Dose-painting strategies can be used to increase dose to specific tumor subvolumes. Five-year local, regional, and distant control rates are comparable with patients treated with conventional IMRT. Volume and intensity of dose escalation should be further tailored, given the possible increase in severe acute and chronic toxicity. Adapting treatment and decreasing dose to the swallowing structures might contribute to lower toxicity rates when applied in smaller tumor volumes. Whether adaptive DPBN can significantly improve outcomes is currently being investigated in a novel clinical trial.