Use of Computational Human Phantoms Combined with a Treatment Planning System to Study the Sensitivity of Reconstructed Normal Tissue Dose to Patient Size and Assumptions On Second Tumor Location
C Lee1, C Lee2*, S Lamart1, SL Simon1, RE Curtis1, P Inskip1, (1) Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, MD, (2) Department of Radiation Oncology, Hospital of the University of Michigan, Ann Arbor, MISU-E-T-285 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: Epidemiologic studies of second cancer risk from radiotherapy require reconstruction of normal tissue doses for patients treated years ago and often rely on a single estimate, usually a mean tissue dose or less frequently, a point-dose at the site of second cancer development. However, uncertainties of dose estimates due to assumptions made in dose reconstruction, e.g., body size and location of the tumor, are rarely quantified. To assess the uncertainty of the estimated incidental breast dose from radiotherapy for Hodgkin Lymphoma, we investigated the sensitivity of the point-dose to dosimetric assumptions using computational human phantoms of different body sizes imported into a commercial treatment planning system (TPS).
Methods: We converted three adult female hybrid phantoms representing the 20th, 50th, and 80th weight percentiles to DICOM images that we imported to a TPS featuring a three-dimensional convolution/superposition algorithm. We simulated opposite mantle fields with 6 MV photons and typical lung blocks (transmission, 5%) to deliver 40 Gy at the phantom mid-plane. We located a point in the left breast 1.7 cm below the surface and in-field to reconstruct point-dose for each phantom. We retrieved the lateral dose profile starting at the point and crossing the block boundary.
Results: Point-doses were 43, 44, and 46 Gy for the three phantoms with 20th, 50th, and 80th weight percentiles, respectively. To deliver 40 Gy at mid-plane of the phantoms, a greater number of monitor units were required for thicker phantoms, resulting in a greater shallow dose. The lateral dose profile showed that the point-dose estimate can vary by a factor of 10 depending on whether the point is under the block or not.
Conclusion: Our new dose reconstruction strategy, combining computational human phantoms and a TPS, enabled us to examine the sensitivity of point-dose estimate to patient size and assumed second tumor location.