Author: Jimmy Caudell, Eduardo G. Moros, Nour Nasser, Gage H. Redler, George Yang 👨🔬
Affiliation: Moffitt Cancer Center 🌍
Purpose: Dosimetric evaluation of PTV margin reduction for Head and Neck(H&N) sequential boost plans using the Ethos online adaptive radiotherapy(oART) system to investigate the tradeoff in decreased organ-at-risk(OAR) dose with potentially decreased target coverage.
Methods: Retrospective clinical data was from ten H&N patients treated with sequential boost, which received offline adaptation due to significant changes in anatomy. Prescribed PTV doses were: PTVHigh/PTVMed 20Gy/NA(n=6) and 20Gy/18Gy(n=4). Patients were re-planned within the oART system with CTV=GTV+5mm and PTV=CTV+3mm, followed by simulation of a physician-driven online adaptive session. In the initial planning workspace, using physician-drawn contours from this simulated online adaptive session, PTV margin was varied (5/3/2/1/0mm) and plans were reoptimized to evaluate the effect on target coverage and OAR doses. To study the uncertainty introduced by the accelerated oART process, dose from online adapted plans with different PTV margins was evaluated both on structures drawn during oART(“Contoured Anatomy”) and during clinical offline-ART(“True Anatomy).
Results: Adapted plans reoptimized to varying PTV margins successfully maintained GTV coverage (meeting all goals for both Contoured and True Anatomy) while decreasing Contoured/True OAR doses by 6%/11%, 10%/17%, 13%/21%, and 22%/28% on average when going from 5mm to 3mm, 2mm, 1mm, and 0mm PTV margins, respectively. Coverage for true targets when using different PTV margin(5/3/2/1/0mm) was: CTVHigh Dmin,0.03cc=99%/97%/95%/91%/90%, CTVHigh V100%=99%/98%/98%/96%/91%, CTVMed Dmin,0.03c=99%/97%/95%93%86% and CTVMed V100%=99%/99%/98%/98%/95%, respectively. The average decrease in CTVHigh V100% between True and Contoured Anatomy for 5/3/2/1/0mm margins was 0%/1%/1%/3%/5%, with coverage goal(CTVHigh Dmin,0.03cc) failing 0%/20%/20%/30%/60% of the time, respectively.
Conclusion: oART can reduce treatment uncertainties, enabling PTV margin reduction, which is shown to improve OAR sparing. However, the accelerated online adaptive process and/or autocontouring tools may introduce uncertainty in structure delineation, as demonstrated by the decrease in doses evaluated on True versus Contoured Anatomy. Therefore, care must be taken when reducing margins to avoid underdosing targets.