onc brain

The longer read

The practical question this planning study speaks to is not whether VMAT beats DCAT in general, but whether the convenience of a single coplanar VMAT arc gives up anything for a mid-sized brain metastasis. On that point the data are fairly clear. Coplanar-only VMAT produced the largest normal-brain volumes at every level from 15 Gy down to 5 Gy, with V12Gy of 15.1 cm3 against 11.6 cm3 for DCAT and 9.4 cm3 for non-coplanar VMAT. The low-dose bath was the most striking divergence, with V5Gy of 96.8 cm3 for the coplanar plan against 49.2 and 42.9 cm3 for the two non-coplanar techniques. In other words, the arc geometry mattered more than the modulation: modulation layered onto non-coplanar arcs bought a further improvement, but modulation on a single coplanar arc was a step backward from conventional conformal arcs.

The population was chosen deliberately. Lesions of 2 to 3 cm sit in the zone where Vogelbaum and colleagues reported higher rates of local recurrence and radionecrosis than for lesions of 2 cm or less, so this is where a reduction in V12Gy has the most plausible clinical leverage. The authors lean on V12Gy as a radionecrosis predictor, which is a reasonable working assumption but remains an inference here: no patient in this comparison was treated three ways, and no toxicity or control endpoint was measured. The magnitude of the V12Gy difference between NcVMAT and DCAT is also modest relative to the spread within each arm, as the standard deviations show, so a reader should expect the benefit to vary considerably across individual lesions and to matter most for the larger and less spherical ones.

The design has one real strength over the prior comparison by Molinier and colleagues: all three techniques were planned in the same TPS with the same dose algorithm, normalization and target heterogeneity, so differences reflect technique rather than software. The flip side is that this is a single-center Eclipse experience, with inverse plans shaped by a ring contour and normal-tissue objective, and DCAT plans shaped by manual MLC adjustment. Inverse-optimized plan quality is notoriously dependent on planner and optimizer, so the size of the NcVMAT advantage may shrink or grow elsewhere. The study also does not address single-isocenter multi-target planning, which is where many centers now use VMAT most, or fractionated stereotactic RT, which is commonly chosen for lesions in this size range.

The trade-off a clinic must weigh is time. Both VMAT techniques had longer beam-on times than DCAT, and non-coplanar delivery adds couch rotations. In frameless mask-based SRS, longer treatment raises intrafraction motion risk against a 1 mm PTV margin, which could erode some of the dosimetric gain in practice. For a center already delivering non-coplanar arcs, the findings support moving those arcs to VMAT optimization when normal-brain dose is binding, such as a larger lesion or a pt likely to need repeat SRS. For a center delivering single coplanar VMAT arcs to solitary lesions, the data argue that the plan should at least be checked against a non-coplanar alternative.