Immune Correlates in a Real-World EGFR-mutated Non-Small Cell Lung Cancer Cohort Before and After First-Line Osimertinib

IASLC WCLC 2026

Aug 19, 2026
Oncology
Presentation

T. Azenkot, K. Lu, E.B. Jaeger, U. Jariwala, S. Fragkogianni, J. Mercer, J.D. Patel, S.P. Patel

Introduction
Treatment with osimertinib has been shown to induce tumor microenvironment changes in EGFR-mutated non-small cell lung cancer (NSCLC). Optimal systemic therapy after osimertinib resistance is not well established. A molecular understanding of resistance mechanisms could inform drug development and patient selection for subsequent therapy. Here we describe immune biomarkers and cell infiltrates in tumor samples before and after osimertinib treatment in a retrospective, real-world cohort.

 

Methods
Patients with EGFR exon 19 deletion or L858R-altered NSCLC (N=525) treated with first-line (1L) osimertinib monotherapy who underwent DNA (xT) and RNA (xR) sequencing were identified in the Tempus database. Tumor samples were obtained prior to 1L therapy (pre, n=378) or 1-24 months after 1L therapy initiation (post, n=179). RNA expression was normalized as transcripts per million (TPM) and reported as log2(TPM+1). Immune infiltration was estimated using quanTIseq. Tumor mutational burden (TMB) and PD-L1 were evaluated via xT and immunohistochemistry. Groups were compared using Chi-squared/Fisher’s and Wilcoxon rank sum tests for categorical and continuous variables, respectively.

 

Results
Tissue samples (n=557) from 525 patients (32 paired) were evaluated. Patients were a median (range) of 67 (27-88) years old; 69% were female and at least 54% were never smokers. Median (interquartile range) TMB was higher in post- than pre-treatment samples (4.74 [2.63, 6.84] vs. 3.68 [2.11, 5.79]; p<0.001). The frequency of high TMB (>10 mut/mb) and PD-L1 expression (>50%) was also greater in post-treatment samples (11.1% vs. 3.7%, p=0.001; 31.4% vs. 14.7%, p<0.001). In pre- and post-treatment samples, the most common alterations occurred in TP53 (65% vs. 72%), CDKN2A (22% vs. 32%), CDKN2B (20% vs. 30%), MTAP (16% vs. 27%), and RB1 (8% vs. 16%); MET (0.3% vs. 16%) and EGFR C797S (0.5% vs. 7.8%) alterations were also observed. Among 10 immune cell subtypes assessed, infiltration of M1 macrophages, neutrophils, regulatory T, B, natural killer, and CD8+ T cells were significantly lower in post-treatment samples (Table).

 

Conclusions
Though TMB and PD-L1 expression were higher in post-osimertinib treatment tissue samples, overall immune cell infiltration was paradoxically lower in post-treatment samples. This may reflect immunosuppressive remodeling from acquired osimertinib resistance and may provide context for limited clinical activity of anti-PD-(L)1 therapies in second-line settings. Further intra-patient longitudinal characterization is warranted in EGFR-mutated NSCLC after osimertinib treatment.