Purpose: To compare intracranial SRS/SRT plan quality among CyberKnife (CK), Multi-leaf collimator (MLC), and Cone techniques.
Methods: Three groups of patients treated with CK were selected; group one (14 patients) with gross target volume (GTV) of 0.4 – 4.7 cc was re-planned with Eclipse Cone & MLC, group two (12 patients) with GTV of 0.1-1.3 cc was re-planned with Cone, and group three (12 patients) with GTV of 0.9 - 74.4 cc was re-planned with MLC. Plan quality was evaluated with Paddick’s conformity index (CI), Gradient index (GI: the ratio of isodose volume V50%/V100%) to represent the normal tissue dose. One isocenter with the coplanar beam and the same GTV dose coverage were used for plan comparison.
Results: For CK/MLC/Cone comparison in group one, three Cone-plans with GTV (>2.5cc) could not meet the GTV coverage and were eliminated from the comparison. The median CI was 1.16 (1.06-1.39), 1.18 (1.03-1.55) and 1.46 (1.06-2.9), and the median GI was 3.67 (2.68-4.38), 4.68(4.25-5.33) and 2.51(2.26-4.72), respectively. MLC-plans were not attempted for tumor volume of < 0.4 cc due to the MLC limitation. CK and MLC-plans were comparable in CI and were superior to Cone-plans. Cone-plans showed the lowest GI, and MLC-plans had the highest GI among all three techniques. Further comparison on MLC/CK and Cone/CK plans showed a good agreement with the group one results using patients with more different tumor size ranges.
Conclusion: The Cone-plans were highly affected by tumor shape and size, and showed the most inferior CI and superior GI than the other two methods. The MLC-plans had some limitations for small size targets and also received the highest normal tissue doses in low dose areas. The CK-plans did not have any tumor size restriction, and the plan quality was either comparable or in between the other two methods.