Treatment plan quality comparison between ZAP‐Axon ® Krylov optimizer and Leksell Gamma Knife ® Lightning optimizer for brain metastasis

Purpose This study compares stereotactic radiosurgery (SRS) treatment plan quality between the recently launched ZAP‐Axon ® Krylov optimizer and the Leksell Gamma Knife ® (LGK) Lightning optimizer for brain metastasis (BM). Methods Thirty ( n = 30) patients who were previously treated with LGK SRS or fractionated SRS were selected and divided into three groups: large post‐operative resection cavities of BM (BM lg, postop ), large intact BM (BM lg, intact ) and small intact BM (BM sm, intact ). Based on our institutional planning guidelines, the 30 clinical cases were replanned using ZAP‐Axon Krylov and LGK Lightning optimizers without manual adjustments. Statistical analysis was performed to compare plan quality metrics (Paddick conformity index [PCI], gradient index [GI], target coverage [TC], prescribed isodose line [Rx IDL], and treatment time/scaled beam‐on‐time [sBOT] [min]) between the two optimizers. Results For the BM lg, postop group, mean values of PCI, GI, TC, Rx IDL, and treatment time/sBOT (min) for Krylov versus Lightning plans were 0.91 versus 0.91, 2.48 versus 2.56, 99.5% versus 99.7%, 51.1% versus 53.7% and 37.4 min versus 36.5 min, respectively. All metrics except for the Rx IDL (%) were not statistically significantly different. For the BM lg, intact group, corresponding values were 0.93 versus 0.93, 2.42 versus 2.57, 99.5% versus 99.7%. 51.5% versus 53.0% and 35.9 min versus 36.3 min. While GI was statistically significantly different ( p <0.05) between the two optimizers, the rest of the metrics were not. For the BM sm, intact group, corresponding values were 0.71 versus 0.64, 3.10 versus 3.26, 99.5% versus 99.5%, 61.9% versus 53.0% and 8.1 min versus 19.8 min. In this group, PCI and treatment time/sBOT were statistically significantly different ( p <0.05), whereas GI and Rx IDL were not. Conclusion Without manual adjustments, ZAP‐Axon Krylov plans were comparable to LGK Lightning plans for large BM lesions, but the Krylov optimizer outperformed the Lightning optimizer for small BM lesions.