Reconstruction of Organ Doses of Patients Treated Historically for Hodgkin's Lymphoma with Cobalt-60 Teletherapy

Abstract

Approximately half of all long-term cancer survivors treated with radiotherapy experience radiation related side effects, including secondary malignancies and cardiovascular disease. To better understand how absorbed dose impacts normal tissues volumes, studies are required to correlate secondary effects with mean organ doses. Since the latency period of radiation-induced secondary cancers ranges between 5 to 25 years, analyzing patients with numerous years of medical record follow-up is necessary for determining a dose-response relationship. Patients treated historically for Hodgkin’s lymphoma (HL) are ideal for studying late effects, since they present at an early age and have long-term survival following treatment.

Calculating mean organ doses of patients treated historically with radiotherapy presents a dosimetric challenge, since 3D imaging was not available for treatment planning prior to the 1990s. Whole-body computational phantoms coupled with Monte Carlo based models of radiotherapy devices permit estimates of in-field and out-of-field organ doses as needed for studies associating radiation organ dose and late tissue toxicities.

In this study, a computational model generated in MCNPX (v2.70) of the Theratron-1000 cobalt-60 teletherapy unit is presented. Model validation is based on infield commissioning data collected at the University of Florida, published out-of-field data from the BJR Supplement 25, and extensive out-of-field measurements recently performed at the UWADCL. By changing field size and position, and adding patientspecific field shaping blocks, more complicated set-ups are simulated to recreate the dose distributions actually produced by historical treatments. In-house Python™ scripts which interface with Rhinoceros™ are created to aid with historical treatment planning.

In order to test the accuracy of using the UF/NCI hybrid computational phantom library for historical dose reconstructions, modern day patients treated for HL with CT images sets are selected and matched to UF/NCI phantoms based on gender, height, and weight. Both patient specific voxel and UF/NCI phantoms receive historical treatment plans, and a dosimetric comparison is performed within the Monte Carlo framework.

Using this methodology, patient-specific dose reconstructions were preformed for patients (n=29) treated for HL at the UF from 1964 to 1972. Mean organ doses are calculated and correlated with long term follow-up. Special attention was given to cardiac doses.