KEY POINTS
- The quality-assurance study included 39 consecutive patients receiving left-sided locoregional breast RT with internal mammary-node irradiation in DIBH, matched in size to a previous 39-patient 2021 cohort.
- Patients received 40 Gy in 15 fractions using tangential 3D-CRT. Chest-wall position was measured from cine MV imaging throughout treatment; the departmental target was to minimize intrafraction displacement around the steep IMN dose gradient.
- Four changes were introduced together: relocation of the respiratory marker from the upper abdomen to the sternum, a 5-second delay before setup imaging after entering breath hold, IMN-prioritized image matching and systematic early cine-MV review.
- Errors >3 mm fell from 21.2% in the historical cohort to 4.7% (55/1,176 fields; p<0.0001). Errors >5 mm fell from 3.0% to 0.6%. Mean, systematic and random errors also improved from 0.7/1.4/1.7 mm to 0.1/0.7/1.3 mm, respectively.
- Early cine-MV review was performed in 38/39 patients and led to treatment-delivery interventions in 6 patients (15.4%), including additional coaching or modification of the gating window.
- Moving the physical gating block onto the sternum introduced a new problem. Among 18 evaluable patients in a prospective toxicity cohort, 11/18 (61%) reported a skin reaction beneath the block: nine grade 1, one grade 2 with blistering and one grade 3 with ulceration.
- Because all four workflow interventions were introduced together, their individual contributions cannot be separated. The comparison was also historical rather than randomized, and delivered IMN dose was not recalculated in the new cohort.
CLINICAL TAKEAWAY
The study shows how apparently small workflow details can materially change DIBH geometric accuracy, especially when IMNs sit near a steep tangential-field gradient. Better chest-wall tracking came with a clear physical-marker bolus effect, so centres using marker-block gating need to optimize both geometry and skin dose.