Estimation of radiation-induced, organ-specific, secondary solid-tumor occurrence rates with total body irradiation and total marrow irradiation treatments

Abstract

PurposeWe aimed to predict and compare radiation-induced, organ-specific, secondary solid-tumor occurrence risks from conventional total body irradiation (TBI) and total marrow irradiation (TMI) for patients undergoing hematopoietic cell transplant. Methods and MaterialsWe retrospectively selected 20 patients who received TMI treatments before hematopoietic cell transplant. Ten patients (5 men; 5 women) received 12 Gy to the skeletal bones, lymph nodes, and spleen, and the other 10 patients (5 men; 5 women) received an escalated dose of 20 Gy to the same targets and 12 Gy to the brain and liver. A conventional TBI treatment plan was generated for each patient with a prescription dose of 12 Gy, using anterior-posterior and posterior-anterior photon beams with lung shielding and a chest wall boost with electron beams. Secondary cancer risks were estimated using linear-exponential and plateau models for major organs. ResultsAt the 12 Gy dose level, using the linear-exponential model, the total radiation-induced secondary solid-tumor risks for major organs were 159.3 ± 8.7 for men and 221.5 ± 14.4 for women per 10,000 people per year with the TMI plans, which is a reduction of 38.8% and 32.9%, respectively, compared with those with the TBI plans. At the 20 Gy dose level, the risks were 220.3 ± 8.3 for men and 298.5 ± 9.3 for women with the TMI plans, which is a reduction of 14.6% and 9.2%, respectively, compared with those with the 12 Gy TBI plans. Significant risk reductions were also found with the TMI plans using the plateau risk model. ConclusionsAt both the 12 Gy and 20 Gy prescription dose levels, a conditioning regimen using TMI could significantly lower overall radiation-induced secondary solid-tumor risks for major organs compared with a conditioning regimen with standard 12 Gy TBI. Clinical data from long-term follow-up studies are needed to verify the model predictions.