Skull Dosimetry for Mitigating Acute Hematologic Toxicity in Patients With Non-small Cell Lung Cancer With Brain Metastases Undergoing Cranial Radiation Therapy and Immunotherapy.

[PURPOSE] Brain radiation therapy (RT) combined with immunotherapy is standard for patients with non-small cell lung cancer (NSCLC) with brain metastases. However, RT-induced hematologic toxicities (HTs) can cause treatment interruption and lymphopenia, potentially impairing immunotherapy efficacy. Because the skull bone marrow has recently been identified as a key hematopoietic site, this study evaluated the impact of skull RT dose on HTs in patients with NSCLC undergoing cranial RT and immunotherapy.

[METHODS AND MATERIALS] This retrospective study consecutively included 191 patients with NSCLC with brain metastases treated between June 2018 and April 2023. Stereotactic RT or whole-brain RT was selected based on lesion characteristics. All doses were converted to equivalent 2-Gy fractions (EQD2). HTs were graded using the National Cancer Institute Common Toxicity Criteria for Adverse Events version 5.0. Skull dosimetric parameters (Dmean, V5 - V50) were analyzed for associations with grade ≥ 3 HTs (HT3+) using logistic regression. Normal tissue complication probability models and optimal cutoff thresholds were derived using logistic and receiver operating characteristic analyses. Based on these cutoffs, post-RT absolute lymphocyte count (ALC) and immunotherapy response rates were assessed in stratified groups.

[RESULTS] HT3+ occurred in 18.8% (36/191) of patients, with neutropenia (18.9%), leukopenia (15.2%), and lymphopenia (15.7%) being most frequent. Multivariate analysis, adjusting for age and other covariates, revealed significant associations between HT3+ and Dmean as well as V5 -V20 (all P < .05). The normal tissue complication probability models predicted a 50% probability of HT3+ at the following thresholds: Dmean = 41.3 Gy, V5 = 94.6%, V10 = 86.0%, and V20 = 77.0%. Receiver operating characteristic analysis identified optimal HT3+ cutoffs of Dmean≥ 32.2 Gy, V5≥ 82.0%, V10≥ 64.7%, and V20≥ 53.4%. Patients above these thresholds had over twice the HT3+ risk, significantly lower post-RT ALC, and reduced immunotherapy response rates (all P < .05), with higher post-RT ALC predicted better response.

[CONCLUSIONS] Higher skull radiation doses significantly increase severe HTs in patients with NSCLC receiving cranial RT and immunotherapy. Incorporating skull dose constraints may mitigate HTs, preserve lymphocytes, and enhance immunotherapy efficacy, warranting validation in prospective studies.