Association of human leukocyte antigen (HLA) status with immune checkpoint inhibitor (ICI) pneumonitis in patients with non-small cell lung cancer (NSCLC).

Abstract

8620 Background: Pneumonitis is a potentially serious complication occurring in 5-20% of lung cancer patients receiving ICI. The genetic risk factors for ICI pneumonitis are unclear. Given the association between HLA genetic variants and autoimmune diseases, we hypothesized that certain HLA genetic variants are associated with risk of ICI pneumonitis. Methods: Patients with advanced EGFR/ ALK wild type, PD-L1 tumor proportion score ≥50% NSCLC who received ICI were enrolled from year 2019 to 2022. Germline HLA typing was performed on buccal swab specimens using the Alltype Fastplex assay. Clinicopathological characteristics, treatment courses and survival outcomes were retrospectively collated. The primary endpoint was the correlation of common HLA variants (present in >30% of subjects) to the risk of ICI pneumonitis. Results: 42 patients (all Chinese, 36 male, 33 current/former smokers) were enrolled. Median age was 71 years old (range 51-91). 34, 6, 1 and 1 patients had adenocarcinoma, NSCLC not otherwise specified, squamous cell carcinoma and large cell neuroendocrine carcinoma respectively. Majority (33/42, 79%) received pembrolizumab monotherapy while 9 (21%) patients received pembrolizumab plus chemotherapy. Median follow up time was 27.8 months. At data cutoff, 28 (67%) patients received ICI for more than 1 year and 10 (24%) patients stopped ICI after completion of 2-year course of treatment. Seven (17%) patients developed pneumonitis, of which 3 were grade 2 and 4 were grade 3 in severity. Median time to development of pneumonitis was 6.9 months (range 0.3-25.2). Demographic characteristics were similar between patients who did or did not experience pneumonitis. In the entire cohort, twenty-two (52%) patients harbored the class II DPB1*02 allele. This allele was significantly associated with development of ICI pneumonitis. All 7 patients who had ICI pneumonitis harbored the DPB1*02 allele. The incidence of pneumonitis in patients harboring the DPB1*02 allele was 32% compared to 0% in those without this allele (p=0.009). Patients who experienced pneumonitis had worse survival rates than those without (p=0.03), but expression of DPB1*02 allele was not associated with inferior survival outcomes. Conclusions: Our results suggested that expression of DPB1*02 allele may be associated with development of ICI pneumonitis. A larger case-control study is ongoing to validate this association. [Table: see text]