Constructing a Planning Model for Volumetric Modulated Arc Therapy in the Treatment of Breast Cancer

Researchers from the Department of Medical Physics at Instituto Zunino – Fundación Marie Curie in Córdoba, Argentina, set out to construct a knowledge-based volumetric modulated arc therapy (VMAT) planning model to enable the treatment of breast cancer without lymph node irradiation. Their article, published in the Journal of Medical Physics, reported that they successfully implemented two RapidPlan™ (RP) models for breast cancer treatment that reduced planning time and improved treatment planning efficiency while ensuring high-quality treatment plans.

According to the study’s lead author, Oscar Abel Apaza Blanco, the researchers found that “the RP plans performed are dosimetrically equivalent to [manual plans (MP)] generated by expert physicists and dosimetrists,” and further suggested that “the same procedure could be used to implement VMAT RP models with different dose prescription protocols.”

The study initially selected 100 VMAT-based breast plans to construct a breast left and a breast right knowledge-based VMAT model using RP. The RP models were verified “based on goodness-of-fit statistics using the coefficients of determination (R²) and Chi-square (χ²), and the goodness-of-estimation statistics through the mean square error (MSE).” The two RP models were then validated by reoptimizing 20 plans that integrated the models and 20 plans that did not. Dosimetrical parameters were used to compare MP and RP plans for treatment locations including heart, homolateral lung, contralateral lung, and contralateral breast.

The least favorable R² results for the RP models were 0.51 for contralateral lung in RP-right breast (RP-RB) and 0.60 for heart in RP-left breast (RP-RB); While the most unfavorable χ² results were 1.02 for contralateral breast in RP-RB and 1.03 for heart in RP-LB. Additionally, the study’s authors reported that no overfitting occurred, no unfavorable mean square error (MSE) results were found, and estimation power was strong in both RP models.

Given their validation findings, the study’s collaborators advanced that “the use of RP models generates high-quality plans, without differences from the planner experience” in planning volumetric modulated arc therapy treatments for patients with breast cancer. They did acknowledge, however, that “longer time and experience in the use of RP are necessary to confirm the results shown in this study.”