Encrypted login | home

Program Information

Spine SBRT Planning for Patients with Spinal Hardware

J Meyer

J Meyer*, K Hendrickson , S Bowen , M Nyflot , S Patel , L Halasz , University of Washington, Seattle, WA


SU-E-T-302 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

To investigate the effects of different imaging artifact correction strategies and susceptibility to setup errors for multiple spinal SBRT planning approaches in treating a patient with titanium hardware.

Two-step-IMRT, IMRT and two arc VMAT plans were created for a C-spine SBRT case with titanium hardware. The plans were normalized to the maximum point dose to the thecal sac. Density overrides for the planning CT were applied to mitigate imaging artifacts. All plans were recalculated on CTs with no density overrides and with the density of all structures set to 1g/cm^3. Setup errors up to 2mm toward the spinal cord were simulated.

The 2-step IMRT plans achieved a mean dose to the cord of 6.5Gy compared to 8.0 and 8.7Gy for the IMRT and VMAT plans. The PTV coverage was comparable for both the 2-step-IMRT and the IMRT plan. The VMAT plan achieved a higher mean dose, but had hot spots >130%. If no density overrides were applied for the same MUs the increase in maximum dose to the cord was 0.3, 1.2 and 0.0% for 2-step-IMRT, IMRT and VMAT plans, respectively, compared to an increase of 2.3, 2.0 and 1.0% for unit density override.
An isocenter shift of 2mm toward the cord resulted in an increase in maximum dose to the cord of 10.7, 8.0 and 5.0% for the 2-step IMRT, IMRT and VMAT plan, respectively, with the absolute dose to the cord still being lowest for the 2-step-IMRT plan compared to the other plans without simulated setup errors.

The 2-step-IMRT plan achieved the clinically most acceptable dose distribution. The plan quality was relatively insensitive to density overrides. The 2-step-IMRT plan was the most sensitive to setup errors due to the steep gradient but this was offset by the initial improvement in sparing of the cord.

Contact Email: