Clinical transition to model-based dose calculation algorithm: A retrospective analysis of high-dose-rate tandem and...

To retrospectively compare clinical dosimetry of CT-based tandem-ring treatment plans using a model-based dose calculation algorithm (MBDCA) with the standard TG-43-based dose formalism.

A cohort of 10 cervical cancer cohorts treated using the tandem and ring high-dose-rate applicators were evaluated. The original treatment plans were created using the department CT-based volume optimization clinical standards. All plans originally calculated with TG-43 dose calculation formalism were recalculated using the MBDCA algorithm. The gross target volume and organs at risk (OARs) were contoured on each data set along with significant heterogeneities like air in cavity and high-density plastic tandem and ring components. The patient tissue was modeled as homogenous liquid water. D₉₀, D₉₅, and D₁₀₀ for gross target volume, D_0.1cm, D_1.0cm, and D_2.0cm for bladder, rectum, and sigmoid were extracted from dose-volume histograms for TG-43 and MBDCA calculated plans. Mean absolute difference ± 2σ in the above metrics was calculated for each plan.

Using the manual applicator contouring method, MBDCA plans (n = 10) showed 2.1 ± 1.1% reduction in dose to Point A average, 2.6 ± 0.9% reduction in Target D₉₀ dose, and 2.1 ± 0.3% dose reduction to OARs. Results from plans using vendor supplied solid applicator models (n = 5) showed 2.2 ± 1.10% reduction in dose to Point A average, 2.7 ± 0.2% reduction in Target D₉₀ dose, and 2.7 ± 1.0% dose reduction on average to OARs.

For unshielded plastic gynecologic applicators, minimal dosimetric changes (<5%) were found using MBDCA relative to standard TG-43. Use of solid applicator model is more efficient than manual applicator contouring and also yielded similar MBDCA dosimetric results. Currently, TG-186 dose calculations should be reported along TG-43 until we obtain studies with larger cohorts to fully realize the potential of MBDCA dosimetry.