KEY POINTS
- This PRISMA systematic review and meta-analysis included 52 studies evaluating surface-guided or optical-surface-monitoring-assisted DIBH during left-sided breast or chest-wall RT, with or without regional nodal irradiation.
- Compared with free breathing, SGRT/OSMS-assisted DIBH reduced mean heart dose by 1.21 Gy (95% CI 0.93–1.48 Gy) and mean LAD dose by 7.05 Gy (95% CI 4.71–9.39 Gy). Heterogeneity was substantial for both outcomes (I² 84.5% and 93.8%).
- Pulmonary exposure was also lower: ipsilateral lung mean dose decreased by 0.88 Gy (95% CI 0.42–1.34) and lung V20 by 2.30 percentage points (95% CI 1.32–3.27). However, the V20 effect became close to null under a common-effect sensitivity model, making that endpoint less robust.
- Secondary cardiac metrics moved in the same direction. Heart Dmax decreased by 14.86 Gy, heart V5 by 5.74 percentage points, LAD Dmax by 20.11 Gy, and ipsilateral lung V5 by 2.67 percentage points.
- Comparative analyses found no statistically significant pooled differences in image-verified residual setup errors between SGRT-DIBH and comparator workflows. Importantly, these analyses were not designed to establish equivalence or non-inferiority.
- In single-arm SGRT-DIBH studies, pooled image-verified residual errors were 0.88 mm laterally, 2.02 mm superior–inferiorly, and 1.13 mm anterior–posteriorly. The pooled 3D-vector displacement was 6.21 mm, but this estimate came from only two highly heterogeneous studies.
- SGRT systems, ROI definitions, gating thresholds, reference surfaces, image-verification strategies and QA procedures varied substantially. The authors therefore recommend maintaining appropriate internal image verification, particularly for complex targets and regional nodal treatment.
CLINICAL TAKEAWAY
SGRT-supported DIBH has a well-established dosimetric rationale for left-sided breast RT, with particularly meaningful reductions in heart and LAD exposure. Surface monitoring can support positioning and breath-hold reproducibility, but it remains a surrogate for internal anatomy and should not automatically replace appropriately selected IGRT.