Dose uncertainty in the spinal cord from setup errors in intensity-modulated radiation therapy

Objective: To investigate the physical dose and biologically effective dose uncertainty in the spinal cord from inter-fraction setup errors in intensity-modulated radiation therapy. Methods: Simulated actual setups on treatment plan system, compared with primary plan without setup errors to evaluate the difference in the physical dose and biologically effective dose between plan and delivery was carried out. Eight patients in the test, everyone had 5 setups, S0-S4. S0 was the setup of primary plan without setup errors. S1, S2, S3 and S4 were actual clinical setups simulated. S1, S2 and S3 had the same system setup errors, S1 showed no random setup errors, and the random setup error of S2 was smaller than S3. The setup errors of S2 and S3 were normally distributed. There was no random setup in S4 and its system setup errors were greater than those of S1, S2 and S3. Results: Compared with S0, the total dose of the reference point in the cord in S1, S2, S3 and S4 were added 6.9%±1.3%, 7.0%±1.4%, 7.8%±1.5% and 11.6% ±1.0%, respectively; the total biologically effective dose of the reference point of cord in S1, S2, S3 and S4 were added 9.2%±1.8%, 9.7%±1.9%, 11.2%±2.2% and 15.7%±1.4%, respectively; the standard deviation of fraction dose in S2 and S3 were 11.4%±2.6% and 16.6%±2.8%, respectively. Conclusions: Tiny setup errors could give rise to a big change of the dose in the spinal cord, with the change of total dose determined by system errors, the random errors would lead to dose fluctuation. As found in the biologically effective dose, the change were little more than the physical dose. (authors)