KEY POINTS
- This retrospective study included 194 consecutive patients with esophageal cancer treated with IMPT from 2019–2025. Patients received either 23 × 1.8 Gy(RBE) preoperative or 28 × 1.8 Gy(RBE) definitive radiotherapy using two or three posterior proton beams, with weekly repeat 4D CT evaluation.
- Mean heart dose increased by 0.4 Gy per week (p<0.001) and heart V40Gy by 4.1 cc per week (p<0.001). Across treatment, mean heart dose was on average 0.7 Gy higher than at planning and heart V40Gy was 7.3 cc higher.
- Anatomical changes occurred systematically during treatment: heart volume decreased 1.2% per week, anterior-posterior and left-right ITV dimensions each decreased 0.4 mm per week, and the heart shifted anterior-posteriorly by approximately 0.7 mm per week. Diaphragm baseline changes were also strongly associated with heart-dose variation.
- An Elastic Net model predicted mean heart-dose changes with R²=0.49 and mean absolute error 0.9±0.1 Gy. The strongest predictors were initial heart dose, diaphragm baseline shift, heart-volume change, and anterior-posterior ITV change; temporal validation in 2024 patients achieved R²=0.50 and MAE 0.8 Gy.
- A mean heart-dose increase of >1.5 Gy occurred at least once in 92/194 patients (47%). Lowering the trigger to 1.0 or 0.5 Gy would theoretically have increased adaptation requirements to 63% and 76% of patients, respectively.
- After the heart-dose adaptation protocol was introduced, 46/111 patients exceeded the 1.5-Gy threshold and 25 underwent replanning specifically because of heart-dose increase. In these adapted patients, the average excess mean heart dose fell from 1.5 to 0.8 Gy (p<0.01).
- Diaphragm shifts below approximately −1 cm and heart-volume reductions greater than 50 cc were particularly associated with large mean heart-dose increases. The study did not test whether adaptive heart-dose reduction translated into fewer cardiac events.
CLINICAL TAKEAWAY
The cardiac advantage of proton therapy for esophageal cancer cannot be assumed to remain constant after planning. Weekly anatomical assessment, with particular attention to diaphragm position and heart-volume change, can identify patients losing that advantage and enable selective replanning rather than routine adaptation of everyone.