Author: Nina Burbure, Tawfik G. Giaddui, Shidong Li, Curtis Miyamoto, Jeremy Price, Bin Wang 👨🔬
Affiliation: FCCC at Temple University Hospital 🌍
Purpose: Our GammaKnife(GK) has been upgraded from 4C to ICON recently. In this study we evaluated the performance of Lightning IPO for ICON.
Methods: 10 cases with single brain metastasis previously treated with GK-4C were selected. GTV of the 10 cases are on average 7.047+/-2.702cc, ranging from 0.035 to 23.997cc. GK-ICON plans were generated with the Lightning IPO using various relative importance settings for low dose(LD) and beam-on time(BOT), without critical structure constraints. BOT of all plans were scaled to the same dose rate of 3 Gy/min. Note that our comparison of ICON plans with the treated GK-4C plans is for benchmarking purposes only, as the plan quality is also affected by the machine designs (different collimation/shot clearance, etc.).
Results: For one case (GTV=17.316cc), 25 IPO plans were generated, with V12Gy ranging from 29.446 to 38.325cc and BOT ranging from 32.8 to 511.3 minutes. In comparison, V12Gy and BOT of the GK-4C clinical plan were 42.953cc and 26.1 minutes. The IPO plans have Paddick Conformity Index (PCI) ranging from 0.82 to 0.92 as compared with 0.76 for the GK-4C plan. With fixed parameters (LD=0.5, BOT=1), IPO plans were generated and compared with GK-4C plans for all cases. PCI was improved for 10 cases, BOT was improved for 4 cases with relatively small target volumes, LD conformity was improved for 9 cases. The only case with a worse LD conformity has a tiny GTV(0.035 cc). Re-optimizing the case with lower BOT importance gave us a plan with good LD conformity and reasonable BOT.
Conclusion: GK Lightning IPO is very efficient in generating plans with preferrable PCI and LD conformity. However, for large targets, IOP plans could require very long BOT depending on the optimization parameters. Users need to use clinical judgement to strike a balance between treatment time and plan quality.