Abstract / Summary
Abstract Purpose To evaluate the effects of arc number and collimator angle on target dosimetry, organ-at-risk (OAR) sparing, and plan complexity in volumetric modulated arc therapy (VMAT) for thyroid-associated ophthalmopathy (TAO), and to identify a preferred planning configuration. Methods and Materials: Following institutional review board approval, 24 patients with TAO were retrospectively included. For each patient, 16 VMAT plans were generated using two arc configurations (2Arc and 4Arc) and eight collimator angle combinations (C0-0H to C90-270H), resulting in 384 plans. All plans prescribed 20 Gy in 10 fractions and were normalized so that at least 95% of CTV1 was covered by the prescription dose. Differences among the 8 collimator angles were evaluated separately within the 2Arc and 4Arc configurations using the Friedman test with Nemenyi post hoc comparisons, with Kendall's W as the overall effect size. For the 2Arc-vs-4Arc comparison, Wilcoxon signed-rank tests were performed at each collimator angle level (8 angle groups × 20 metrics = 160 tests), with Benjamini-Hochberg correction applied within each angle family (m = 20 per family), and effect size r was computed. Results Across collimator angle configurations, 4Arc plans showed better target dosimetric performance than 2Arc plans, with an 11.1% reduction in homogeneity index, a 2.2% increase in conformity index, a 0.3% increase in CTV1 D98, and a 1.0% decrease in CTV1 Dmax. The gradient index increased slightly by 1.1%. Meanwhile, 4Arc plans increased total leaf travel by 94% and MUs by 2%, without significant changes in MCSv. Among OARs, aggregate maximum doses to the brainstem (− 2.2%), left (− 0.9%) and right (− 1.1%) eyes, bilateral optic nerves (− 0.9% each), optic chiasm (− 1.2%), pituitary (− 1.9%), and lenses (left − 2.9%, right − 3.6%) all decreased; the right hippocampus was the only parameter showing no significant difference at any angle. However, under the study's own per-angle Benjamini-Hochberg framework, these OAR differences reached statistical significance mainly in the small-angle range (0°–45°; optic chiasm, pituitary, and lenses were each significant only at 0°, and the eyes at 3/8 angles) and thus represent descriptive aggregate differences rather than per-angle inferential conclusions. Kendall's W was generally higher for 4Arc than for 2Arc, indicating clearer dose differentiation across collimator angles: from 0° to 60°, the homogeneity index improved from 0.218 to 0.055 and the conformity index from 0.589 to 0.898; from 0° to 75°, lens Dmax decreased by up to 59%. Conclusions Four-arc VMAT showed improved target dosimetric performance and lower doses to most OARs compared with 2Arc VMAT in TAO radiotherapy planning. Considering both arc number and collimator angle configuration, the 4Arc_C60-300H configuration provided the best balance between target dosimetry and OAR sparing. When OAR sparing was prioritized, the 4Arc_C75-285H configuration showed the lowest composite OAR-sparing score, with similar scores at 68° and 90°. The 0° and 15° collimator angles should be avoided on target-quality and OAR-Dmax grounds, although their smaller low-dose bath may be weighed in specific clinical situations.