Normal Tissue Objective Parameter Optimization in Eclipse Treatment Planning System: Enhancing Stereotactic...

The purpose of this study was to identify the most effective manual normal tissue objective (mNTO) parameters for noncoplanar RapidArc (RA) stereotactic radiosurgery (SRS) plans and assess their performance against HyperArc (HA) plans.

This retrospective study included 28 patients with single brain metastasis prescribed 21 Gy. Noncoplanar RA (RA-mNTO) plans were generated using dose fall-offs of 0.1-5.0 mm^-1 and end doses of 50%, 25%, and 10%. In addition, two HA plans were created for comparison: one with SRS NTO (HA-sNTO) and the other with mNTO (HA-mNTO). Plans were evaluated using an integrated scoring approach that include Paddick Conformity Index (CI), gradient index (GI), homogeneity index (HI), monitor units (MUs), normal brain doses (V_{18 Gy}, V_{15 Gy}, and V_{12 Gy}), and delivery accuracy with aS1200 portal dosimetry applying 3%/1 mm gamma criteria. Statistical analysis was carried out using Wilcoxon signed-rank test.

RA-mNTO plan with a 25% end-dose and a 0.5 mm^-1 dose fall-off significantly outperformed HA plans (P < 0.05) in CI, GI, and HI values (0.92 ± 0.02, 3.0 ± 0.17, and 0.32 ± 0.05 vs. 0.91 ± 0.04, 3.41 ± 0.19, and 0.40 ± 0.04 for HA-sNTO and 0.90 ± 0.04, 3.17 ± 0.24, and 0.40 ± 0.05 for HA-mNTO). Furthermore, RA-mNTO significantly (P < 0.05) reduced brain doses at V18Gy (0.88 cc ± 0.40), V15Gy (1.80 cc ± 0.77), and V12Gy (3.19 cc ± 1.35) compared to HA-sNTO (1.13 cc ± 0.51, 2.30 cc ± 1.01, 3.96 cc ± 1.72) and HA-mNTO (1.02 cc ± 0.44, 2.06 cc ± 0.86, 3.53 cc ± 1.47). Moreover, RA-mNTO showed significantly (P < 0.05) lower MUs (8238 ± 948) compared to HA-sNTO (9505 ± 1098) and HA-mNTO (9315 ± 421) and higher gamma pass rates (99.8% ± 0.39) than HA-sNTO (98.9% ± 0.68) and HA-mNTO (99.0% ± 0.51).

Optimized mNTO (25% end-dose and a 0.5 mm^-1 dose fall-off)-based noncoplanar RA plans outperformed HA in SRS treatment of single brain metastasis.