Tissue heterogeneity in IMRT dose calculation for lung cancer

Area of Science:

• Medical Physics
• Radiation Oncology

Background:

• Accurate dose calculation is critical for effective intensity-modulated radiation therapy (IMRT).
• Common algorithms like Pencil Beam (PB), Anisotropic Analytical Algorithm (AAA), and Collapsed Cone Convolution Superposition (CCCS) have varying accuracies.

Purpose of the Study:

• To compare the dose calculation accuracy of PB, AAA, and CCCS algorithms in IMRT.
• To evaluate the impact of dose calculation differences on dose-volume histograms (DVHs) and clinical outcomes (tumor control and normal tissue complication probability - NTCP).

Main Methods:

• 2D dose distributions from PB, AAA, and CCCS were compared pixel-by-pixel using MATLAB.
• Agreement was assessed using the gamma (γ) function.
• Effects on DVHs, tumor control probability (TCP), and NTCP were evaluated using nonparametric tests.

Main Results:

• The PB algorithm generally overestimated dose in lung and tumor regions.
• Discrepancies were observed in lung regions, confirmed by γ analysis and significant in 2 of 4 patients.
• Dose calculation inaccuracy is patient-dependent, influenced by beam geometry and tumor location.

Conclusions:

• Dose calculation inaccuracies in IMRT, particularly with the PB algorithm, can significantly affect NTCP.
• The impact of dose calculation errors is patient-specific and may not be easily predictable with complex beam arrangements.