Room: Exhibit Hall
Purpose: To use NTCP analysis to assess the clinical difference from assigning different materials to high density pixels when using the Acuros algorithm.
Methods: A total of four patients previously treated at our clinic were selected - two left-sided breast patients with tissue expanders with metallic components and two head and neck (HN) patients with extensive dental work. In each CT dataset, pixels with an HU>3069 were segmented within Eclipse by using the "Segment high density artifact" function. Multiple structure sets were created per patient with different materials (water, aluminum, stainless steel and titanium) assigned for the high density contour. The clinical plan used was then applied to each dataset and the resulting dose calculated. Eclipse's NTCP analysis functionality was used to assess the risk of the following endpoints: pneumonitis and pericarditis for breast patients and xerostomia, TMJ dysfunction, cord myelitis, and esophagitis for the HN patients.
Results: For all cases, the dose distribution was virtually the same for every assigned material, with DVHs overlaying almost perfectly except for small changes seen when using aluminum. Even for those cases, plans met all planning criteria. For the breast patients, there was no change found for risk of pericarditis for the four materials used. For the pneumonitis endpoint, the maximum risk change was 1.05%. For HN patients, the maximum change in xerostomia risk was 2.14%. There was almost no change in risk for TMJ dysfunction and esophagitis. The risk for myelitis was zero in all cases.
Conclusion: Assignment of widely different materials to the high density pixels when using Acuros does not seem to have much effect on the calculated dose distribution. The biological effect also does not seem to change very much. We also intend to present expanded results to include data for cases with spine hardware and hip replacements.