Understanding Toxicity of Thoracic Radiation Therapy : The Influence of Immune Checkpoint Inhibitors and Alpha-2-Macroglobulin

Abstract

Radiation therapy (RT) is essential in the treatment of thoracic malignancies. New challenges arise from the combination of radiation therapy (RT) and immune checkpoint inhibitors (ICI) as well as the identification of radioprotectors and predictive factors for adverse events (AE). In a retrospective study, data on common AEs was collected for 79 patients having received both thoracic RT and ICI therapy. Grade ≥ 2 pneumonitis, esophagitis, and dermatitis rates were 6.3, 7.6., and 10.1%, respectively. No differences in AE rates were seen whether the treatment timing was concurrent, closely timed (≤ four weeks interval), or sequential (> four weeks to six months interval). For the alpha-2-macroglobulin (α2M) study, 258 patients that received thoracic radiation for any kind of malignancy had pre-treatment serum α2M levels measured. A2M, which has shown radioprotective effects in preclinical studies, as well as a range of clinical and dosimetric factors were included in predictive models to analyze their association with the development of radiation pneumonitis and esophagitis. Grade ≥ 2 radiation pneumonitis and esophagitis rates of 14 and 23.6% were reported. The mean α2M level was 217.3 mg/dl in current smokers compared to 207.3 and 185.4 mg/dl in former and never smokers. Final predictive models included D65 in lung, max dose in heart, and treatment days for pneumonitis as well as D25, D40, and treatment days for esophagitis. Our studies showed no significant increase of common AEs after thoracic RT and ICI therapy compared to RT or ICI alone and no difference of AE rates with different treatment timing intervals. Although pre-RT levels of α2M did not improve the predictive power of our models for pneumonitis, esophagitis, and dermatitis, we found a univariate association between α2M levels and radiation esophagitis as well as the smoking status of the patients. Furthermore, we were able to validate several clinical and dosimetric factors with our predictive models.