KEY POINTS
- Investigators retrospectively replanned 20 patients with middle-distal esophageal or esophagogastric junction cancer previously treated with neoadjuvant chemoradiotherapy according to CROSS, delivering 41.4 Gy in 23 fractions.
- Each patient was replanned with both IMRT and VMAT using explicit heart-base objectives while maintaining PTV V95 >95% and conventional lung, spinal cord, bowel, and kidney constraints.
- Mean heart dose fell from 19.7 Gy to 15.9 Gy with IMRT and 16.1 Gy with VMAT. Heart V20 decreased from 43.6% to 28.2–28.3%, while V30 fell from 17.8% to 14.0–14.3%.
- Some of the largest reductions occurred in individual conduction and valvular structures: atrioventricular node mean dose fell from 28.7 to 18.2–18.3 Gy, aortic valve root from 20.2 to 9.5–9.9 Gy, and sinoatrial node from 13.8 to 6.8–7.6 Gy.
- The cardiac avoidance area mean dose decreased from 19.5 Gy to 13.3 Gy with IMRT and 13.0 Gy with VMAT. LAD mean dose fell from 13.7 to ~8.6–8.7 Gy, while LAD V15 was reduced from approximately 42% to <10%.
- The proportion meeting a sinoatrial-node Dmean <10 Gy increased from 40% to 95% with either replanning method; the aortic valve root target was achieved in 85–90% vs 20% originally. In contrast, stringent CAA and AV-node goals remained achievable in only about 5%, reflecting direct anatomical proximity to the target.
- Additional cardiac sparing increased planning complexity: monitor units rose by approximately 43% with IMRT and 26% with VMAT. Mean lung dose nevertheless remained approximately 9.0–9.2 Gy, with VMAT showing somewhat more favorable V20/V30 behavior.
CLINICAL TAKEAWAY
Heart-base structures can be deliberately spared during neoadjuvant distal esophageal RT without sacrificing target coverage or conventional thoracic OAR constraints. Whether reducing these substructure doses actually prevents arrhythmias, MACE, or mortality remains unknown, so the proposed objectives should not yet be treated as validated clinical constraints.