Large-scale dosimetric assessment of Monte Carlo recalculated doses for lung robotic stereotactic body radiation therapy. (August 2020)

• Record Type: Journal Article
• Authors: Heng, Veng Jean; Renaud, Marc-André; Zerouali, Karim; Doucet, Robert; Diamant, André; Bahig, Houda; DeBlois, François; Seuntjens, Jan
• Is Part Of: Physica medica. Volume 76(2020)
• Page Start: 7
• Page End: 15
• Publication Date: 2020-08
• Subjects: SBRT -- CyberKnife -- Non-small cell lung cancer -- Monte Carlo; Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2020.06.006
• Languages: English
• ISSNs: 1120-1797

Abstract

Highlights: Monte Carlo recalculation of 219 CyberKnife lung SBRT plans. Ray-Tracing algorithm is inaccurate and severely compromises target coverage. Recalculated near-maximum dose to OAR generally do not exceed planning limits. A higher target dose is associated with superior loco-regional control rate. Abstract: Owing to its short computation time and simplicity, the Ray-Tracing algorithm (RAT) has long been used to calculate dose distributions for the CyberKnife system. However, it is known that RAT fails to fully account for tissue heterogeneity and is therefore inaccurate in the lung. The aim of this study is to make a dosimetric assessment of 219 non-small cell lung cancer CyberKnife plans by recalculating their dose distributions using an independent Monte Carlo (MC) method. For plans initially calculated by RAT without heterogeneity corrections, target coverage was found to be significantly compromised when considering MC doses. Only 35.4% of plans were found to comply to their prescription doses. If the normal tissue dose limits were respected in the treatment planning dose, the MC recalculated dose did not exceed these limits in over 97% of the plans. Comparison of RAT and recalculated-MC doses confirmed the overestimation of RAT doses observed in previous studies. An inverse correlation between the RAT/MC dose ratio and the target size was also found to be statistically significant ( p < 10 - 4 ), consistent with other studies. In addition, the inaccuracy and variability in target coverage incurred from dose calculations using RAT without heterogeneity corrections was demonstrated. On average, no clinically relevant differences were observed between MC-calculated dose-to-water and dose-to-medium for all tissues investigated ( ⩽ 1%). Patients receiving a dose D 95 % larger than 119 Gy in EQD210 (or ≈ 52 Gy in 3 fractions) as recalculated by MC were observed to have significantly superior loco-regional progression-free survival rates ( p = 0.02 ) with a hazard ratio of 3.45 (95%CI: 1.14–10.5).