(P37) Radiation-Induced Immunosuppression in Locally Advanced Non-Small Cell Lung Cancer: Another Cardiac Toxicity?

Abstract

Recent reports have associated increased radiation therapy (RT) dose to the heart with higher rates of cardiac toxicity. Other reports have correlated RT-related immunosuppression with worsened survival in patients with solid malignancies. Given the large vascularized volumes exposed to RT during locally advanced non-small cell lung cancer (LA-NSCLC) treatment, we hypothesized that there is an association between increased heart dose and immunosuppression. A total of 400 LA-NSCLC patients with follow up laboratory values treated at a single institution with definitive intent RT +/- chemotherapy between 2001-2016 were identified. Organs at risk were reviewed and re-contoured according to the RTOG 0617 cardiac secondary analysis atlas. Absolute lymphocyte counts (ALC), absolute neutrophil counts (ANC), and neutrophil to lymphocyte ratio (NLR = ANC/ALC) were analyzed pre-RT, during RT, and at 2 and 4 months post-RT start. Patient, tumor, treatment, and dosimetric factors were correlated with hematologic toxicity and treatment outcomes. An upper tertile threshold to increase specificity of NLR was chosen to dichotomize continuous hematologic variables and identify cutpoints for hematologic values as a predictor of mortality. The median follow up was 17 (range 0.2-174) months in all patients and 46 (range 0.2-161) months in survivors. Patients received a median dose of 66 Gy (range 50-75.25 Gy) in median 2 Gy (range 1.8-2.5 Gy) fractions delivered with IMRT in 41% of patients. The estimated median overall survival (OS) was 17.9 months and progression free survival (PFS) 12.8 months. The estimated 5 year freedom from local recurrence (FFLR) and freedom from distant metastasis (FFDM) were 49% and 34%, respectively. There were 331/363 (91%) cases of ≥ grade 3 lymphopenia and 129/362 (36%) ≥ grade 4 lymphopenia at the nadir. At 4 months after RT start (290 patients), the median ALC, ANC, and NLR were 600 cells/mm3 (range: 100-4600), 4700 cells/mm3 (range: 300-113,600), and 7.2 (range: 0.60-259), respectively. On multivariable analysis, male gender (Hazard Ration [HR] 1.66; 95% CI 1.20-2.29; p=0.02), RT alone (HR 2.08; 95% CI 1.9-3.96; p=0.026), heart V50 (HR 1.02; 95% CI 1.01-1.03; p<0.001), and higher NLR at 4 months (HR 1.015, 95% CI 1.005-1.025; p=0.002) were significantly associated with worse OS. An upper tertile cutpoint of NLR > 10.5 at 4 months post RT, was associated with decreased OS (p<0.001) and decreased FFDM (p=0.04). The median OS stratified by NLR > 10.5 and NLR ≤ 10.5 was 11 months and 24 months, respectively (p < 0.001). NLR > 10.5 was associated with worse FFDM (p=0.04) and was borderline for PFS (p=0.08). On multivariable analysis, CRT with adjuvant chemotherapy (HR 2.55; 95% CI 1.30-4.98; p=0.006), RT alone (HR 3.56; 95% CI 1.07-11.88; P=0.039), and heart V50 > 25% (HR 1.98; 95% CI 1.11-3.54; p=0.021) were associated with an NLR>10.5 at 4 months post-RT. Higher RT dose to the heart is associated with increased immunosuppression and worse tumor control, and may represent an additional actionable cause of increased mortality beyond cardiac toxicity alone.