On the PTV homogeneity objective in the era of photon advanced dose calculation algorithms: Bridging robust and PTV-based planning

Abstract

Aim

The planning target volume (PTV) homogeneity objective was developed for previous-generation dose calculation algorithms. Advanced algorithms report doses to medium-in-medium (D_m,m) and their values depend on the medium considered, breaking the link between uniform irradiation and dose homogeneity. This work revises the PTV homogeneity objective when high-density heterogeneities are involved. We evaluated robust against PTV-based planning, and a dose reporting method that removes composition dependencies to express doses to muscle in muscle-like medium (D_{muscle,muscle*}).

Methods

Four cases featuring bone or metal within the PTV were selected and planned in RayStation with Monte Carlo. Three plans were created for each case: robust optimization for D_m,m (Robust-D_m,m), and PTV-based optimization for D_m,m (PTV-D_m,m) and D_{muscle,muscle*} (PTV-D_{muscle,muscle*}). The plans were reported in D_m,m and D_{muscle,muscle*}, and their dosimetric parameters, robustness, complexities, and optimization times were assessed.

Results

Robust-D_m,m and PTV-D_{muscle,muscle*} plans presented similar D_m,m distributions with inhomogeneous PTV doses due to cold spots in high-density regions. PTV-D_m,m plans achieved homogeneous PTV doses but required local fluence compensations that impaired robustness, with significant hot spots, and increased complexity. Robust optimization was 7 to 11 times slower. Reporting in D_{muscle,muscle*} restored consistency between PTV-based and robust evaluations.

Conclusions

Robust optimization proves that PTV homogeneity should not be prioritized when advanced algorithms reporting D_m,m are used with high-density heterogeneities. PTV-D_{muscle,muscle*} optimization offers a practical alternative for maintaining PTV-based planning and the PTV homogeneity objective while ensuring consistency with robust optimization. D_{muscle,muscle*} reporting simplifies plan evaluation and aligns with clinical practice, facilitating decision-making in treatment planning.