The estimated dose of radiation to immune cells predicts lymphopenia and reveals technique-Dependent immune sparing in postoperative radiotherapy for early esophageal cancer.

BACKGROUND

Radiation-induced lymphopenia (RIL) affects patient survival, but previous studies have mostly focused on locally advanced patients receiving chemoradiotherapy, which involves many confounding factors. This study aimed to investigate whether the estimated dose of radiation to immune cells (EDRIC) correlates with hematologic toxicity (especially lymphopenia) in early-stage esophageal squamous cell carcinoma (ESCC) patients treated with postoperative radiotherapy alone, and to compare EDRIC differences among four modern radiotherapy techniques.

MATERIALS AND

METHODS

We retrospectively analyzed 30 patients with early-stage ESCC who received postoperative radiotherapy (5040 cGy in 28 fractions) between 2020 and 2025. None of the patients received chemotherapy or immunotherapy, ensuring treatment homogeneity. EDRIC was calculated using the Ladbury formula. Four treatment plans-intensity-modulated radiotherapy (IMRT), volumetric-modulated arc therapy (VMAT), intensity-modulated proton therapy (IMPT), and spot scanning proton arc therapy (SPArc)-were generated for each patient in the RayStation system. EDRIC values were compared using the Friedman test followed by Bonferroni-corrected Wilcoxon signed-rank tests.

RESULTS

Complete blood count data were available for 14 of the 30 patients. In this exploratory analysis, only the absolute lymphocyte count (ALC) showed a significant correlation with EDRIC (adjusted R2 = 0.727, 95% CI: 0.35-0.91, P < 0.001). Moreover, after adjusting for baseline ALC, multivariate regression indicated that EDRIC remained associated with the nadir ALC (P = 0.004). EDRIC outperformed mean lung dose (adjusted R2 = 0.428), mean heart dose (adjusted R2 = 0.482), and mean body dose (adjusted R2 = 0.588) in predicting ALC decline. The EDRIC values of proton techniques (IMPT: 1.802 ± 0.602; SPArc: 1.404 ± 0.448) were significantly lower than those of photon techniques (IMRT: 3.193 ± 1.165; VMAT: 3.407 ± 1.340) (all P < 0.001). SPArc produced the lowest EDRIC, which was significantly better than IMPT (P < 0.001), while no significant difference was observed between IMRT and VMAT. SPArc also showed the best cardiopulmonary protection, with the lowest normal tissue complication probabilities for radiation pneumonitis (0.96%) and pericardial effusion (10.57%).

CONCLUSION

In this exploratory study of early-stage ESCC patients, EDRIC was associated with RIL, explaining approximately 73% of the variation in ALC. EDRIC is highly dependent on radiotherapy technique, with proton therapy-especially SPArc-offering significant immune-sparing advantages. These findings warrant validation in larger prospective cohorts.