Purposes
To assess cumulative dose delivered for head-and-neck (H&N) IMRT/IGRT by employing tighter margins for CTV 2 and 3 throughout the entire treatment course for better normal tissue sparing.
Methods
Ten IMRT cases with oropharyngeal cancer treated with simultaneous integrated boost on a TomoTherapy unit (Accuray Inc.) were analyzed. CTVs were delineated on the planning CT by adhering to the principle of respecting anatomic boundaries. Dose-limiting critical structures included brainstem, spinal cord, cochleas, parotids and mandible. A CTV-to-PTV margin of 3 mm was given for PTV1, while zero margins were employed for CTV2 and 3 (plan_0margin). As a comparison, a reference plan using standard margin expansion of 3 mm for PTV1-3 was also generated. The prescription doses were 69.3, 62.7 and 56.1 Gy in 33 fractions for PTV1 (gross disease volume), PTV2 (next echelon nodal regions) and PTV3 (areas harboring subclinical disease), respectively. All patients went through daily pre-treatment megavoltage CT (MVCT) and a long weekly MVCT scan. A GPU-based 3D image deformation/visualization tool was developed to register the long MVCT scan with their planning CT scan. The deformation of each contoured structures was computed to account for non-rigid change in the patient setup. Calculation of the dose accumulation was performed to determine the delivered mean/minimum/maximum dose, dose volume histograms (DVHs), etc.
Results
Under MVCT guidance on TomoTherapy, for the reference plans, the averaged cumulative mean dose ratios during the entire treatment course were consistent within 5% and 1.5% of the planned mean doses for PTVs and CTVs, respectively. For the tighter margin plans, the cumulative mean dose ratios were consistent within 4.3% and 2.3% of the planned mean doses for CTV2 and CTV3, respectively. No significant changes in D95 and D90 for the CTV2/3 cumulative doses in both reference and tighter margin plans were observed. While interfraction anatomical changes introduced minor variations in delivered target doses, improved normal structure sparing was observed in plan_0margin. The planned averaged maximum cord doses in Plan_0margin was 7.6% lower, and the parotid mean doses were 18.9% lower than the corresponding reference plans. Similar dose variations of the delivered dose were seen for the reference and tighter margin plans. The delivered maximum and mean doses for the cord were 20% and 10% higher than the planned doses; a 3.6% higher cumulative mean dose for the parotids was also observed for the delivered dose than the planned doses in both plans.
Conclusions
The GPU-based image deformation/visualization tool enables real-time dose verification, accumulation and documentation. By imposing tighter PTV margins for level II and III targets for H&N irradiation, decent cumulative doses are achievable with daily MVCT guidance for CTV2 and 3 while improving normal structure sparing.