The impact of different optimization strategies on the agreement between planned and delivered doses during volumetric modulated arc therapy for total marrow irradiation

Introduction
To evaluate the agreement between planned and delivered doses and its potential correlation with the plans’ complexity subjected to dosimetric verification.

Material and methods
Four isocentre volumetric modulated arc therapy for total marrow irradiation plans optimized simultaneously with (P1) and without (P2) MU reduction were evaluated dosimetrically by γ method performed in a global mode for 4 combinations of -index criteria (2%/2 mm, 2%/3 mm, 3%/2 mm, and 3%/3 mm). The evaluation was conducted for 4 regions (head and neck, chest, abdomen and upper pelvis, and lower pelvis and thighs) that were determined geometrically by the isocentres. The Wilcoxon test was used to detect significant differences between γ passing rate (GPR) analysis results for the P1 and P2 plans. The Pearson correlation was used to check the relationship between GPR and the plans’ complexity.

Results
Except for the head and neck region, the P2 plans had better GPRs than the P1 plans. Only for hard combinations of -index criteria (i.e. 2%/3 mm, 2%/2 mm) were the GPRs differences between P1 and P2 clinically meaningful, and they were detected in the chest, abdomen and upper pelvis, and lower pelvis and thighs regions. The highest correlations between GPR and the indices describing the plans’ complexity were found for the chest region. No correlation was found for the head and neck region.

Conclusions
The P2 plans showed better agreement between planned and delivered doses compared to the P1 plans. The GPR and the plans’ complexity depend on the anatomy region and are most important for the chest region.