Four Dimensional Target Volume Generation in Pulmonary Stereotactic Body Radiotherapy

Article Outline

• Purpose/Objective(s)
• Materials/Methods
• Results
• Conclusions
• Copyright

Back to Article Outline

Purpose/Objective(s) 

To evaluate the influence of breathing motion on four dimensional dose distributions in pulmonary stereotactic body radiotherapy (SBRT) and to calculate appropriate margins for compensation of breathing motion.

Back to Article Outline

Materials/Methods 

For seven patients with eight pulmonary tumors a respiratory correlated 4D-CT study was acquired. Eight CT series equally distributed over the breathing cycle were reconstructed for each 4D-CT study (2 mm slice thickness). The GTV (macroscopic tumor including spiculae) was delineated in the CT pulmonary window in the CT series of end-exhalation and propagated into all other CT series using automatic image segmentation (Pinnacle treatment planning system v8.1s). No margin for generation of the CTV was applied. The CT series showing the lesion in the time-weighted mean tumor position was selected for treatment planning: a 3D-conformal treatment plan was generated with a dose prescription of 3 × 12.5 Gy to the CTV enclosing 80% isodose (plan-static). 4D dose distributions with consideration of tumor motion were calculated based on non-rigid model-based image registration (plan-4D). The treatment fields were increased symmetrically in steps of 4 mm (2 mm in each direction) to compensate for breathing motion.

Back to Article Outline

Results 

Predominant direction of tumor motion was superior-inferior (SI) in all lesions. Peak-to-peak tumor motion and SD of motion in SI-direction was 12.7 mm ± 4.8 mm and 5.0 mm ± 1.8 mm, respectively. The D95 to the CTV in plan-static was 38.1 Gy ± 2 Gy and the minimum dose (D_min) 34 Gy ± 2.9 Gy. Consideration of tumor motion in plan-4D decreased the dose to the CTV to 30.2 Gy ± 2 Gy (D95) and to 25 Gy ± 1.3 Gy (D_min), a reduction of 22% and 27% compared to plan-static. For complete compensation of 4D breathing motion of the D95 and the D_min level, symmetrical margins of 3.7 mm ± 2.1 mm and 3.9 mm ± 2.4 mm around the CTV based on the mean tumor position were calculated, respectively. These margins are about 40% smaller than the margins used in the internal target volume concept (ITV; the hull of the CTVs in all phases of the breathing cycle). Margins of 2.5 mm ± 1.3 mm (D95) and 2.9 mm ± 1.7 mm (D_min) were necessary if a 5% decrease of the dose to the CTV was accepted; the ITV concept resulted in margins more than twice as large.

Back to Article Outline

Conclusions 

If the mean tumor position from a 4D-CT is selected for planning and if no margins are added to the CTV, breathing motion decreased the dose to the CTV by about 25%. Using 4D dose calculation the margins for compensation of breathing motion were significantly smaller compared to the ITV concept. Inter- and intra-fractional uncertainties are additional factors, which need to be considered in the process of target definition or addressed by image-guidance.