Various Subtypes of EGFR Mutations in Patients With NSCLC Define Genetic, Immunologic Diversity and Possess Different Prognostic Biomarkers

Abstract

Based on data analysis of 9649 Chinese primary NSCLC patients, we calculated the exact proportion of EGFR subtypes in NSCLC and evaluated the TMB level, PD-L1 expression level and tumor immune microenvironment among different EGFR mutation subtypes. Postoperative follow-up data for 98 patients were collected and analyzed. The results showed that several uncommon EGFR mutation subtypes have a higher proportion of TMB-high or strong positive PD-L1 expression than the total EGFR mutation group. In addition, different subtypes have different characteristics related to the immune microenvironment, such as G719 mutations being associated with more CD8⁺ T cell infiltration into tumors; except for EGFR 19del, CD8⁺ T cell infiltration into tumors of other EGFR mutation subtypes were similar to that of wildtype EGFR. Moreover, follow-up results revealed that components of the immune microenvironment have prognostic value for NSCLC patients, with different prognostic biomarkers for NSCLC patients with and without EGFR mutations. These results suggest that patients with different EGFR mutations need to be treated differently. The prognosis of NSCLC patients may be assessed through components of tumor immune microenvironment, and ICIs treatment may be considered for those with some uncommon EGFR mutation subtypes.

Keywords: EGFR - epidermal growth factor receptor; NSCLC; biomarkers; immune microenviroment; prognosis.

Figure 1

The frequencies of different subtypes of EGFR mutations in Chinese primary NSCLC. (A) The proportion of different histologic types of NSCLC, others including adenosquamous carcinoma, sarcomatoid carcinoma, typical carcinoid, atypical carcinoid, lyphoepithelionma-like carcinoma, etc. (B) The ratio of NSCLC patients with EGFR mutation and without EGFR mutation. (C) The ratio of different EGFR mutation subtypes in NSCLC patients. (D) The ratio of patients with EGFR mutation and without EGFR mutation in different histologic types of NSCLC. (E) Comparison of selected gene alteration frequencies in this research and MSKCC cohort. (F) Flow charts for data analysis in this research. AC, adenocarcinoma; SCC, squamous cell carcinoma; LCC, large cell carcinoma.

Figure 2

TMB level evaluated both through WES and panel sequencing. (A) WES calculated TMB level in NSCLC patients with and without EGFR mutation. (B) Panel sequencing calculated TMB level in NSCLC patients with and without EGFR mutation. (C) Panel sequencing calculated TMB level in NSCLC patients of different EGFR mutation subtypes. (D) Ratio of TMB high or low evaluated by WES and panel sequencing. (E) Ratio of TMB high or low evaluated by panel sequencing in different subtypes of EGFR mutation. *means p < 0.05, **means p < 0.01, ***means p < 0.001, ****means p < 0.0001.

Figure 3

Comparation of tumor PD-L1 expression and ratio among different EGFR mutation subtypes and difference of TIME among different TPS groups. PD-L1 expression was quantified as negative (TPS<1%), intermediate positive (1%≤TPS<49%), and strong positive (TPS≥50%) for available cases and tabulated across different EGFR mutation subtypes. (A) Representative staining of PD-L1 expression through IHC (200X). (B) Comparation of PD-L1 expression between tumors with and without EGFR mutation, and among different EGFR mutation subtypes. (C) Ratio of patients with MSI-H in Chinese NSCLC. A total of 3876 were available for count the ratio of MSI-H. A total of 141 NSCLC patients were available for analysis of both PD-L1 expression and TIME. (D) Representative staining for CD8⁺T cells, CD56^bright and CD56^dim NK cells, CD68⁺HLA-DR⁺ M1, CD68⁺HLA-DR^- M2 TAM. CD8 (white), CD56 (purple), CD68 (green), HLA-DR (red), panCK/S100(cyan), DAPI (blue). (E, F) different immune cell counts in intra-tumoral region, stroma, and sum of both. *means p < 0.05, **means p < 0.01, ***means p < 0.001.

Figure 4

Immune cell counts both in tumor (intra-tumoral region) and stroma of 154 NSCLC patients. Immune cell counts both in tumor (intra-tumoral region) and stroma of 154 NSCLC patients of EGFR wild type and with different EGFR mutation subgroups. (A) Numbers of different immune cells in tumor were compared between without and with EGFR mutation groups. (B) Numbers of different immune cells in stroma were compared between without and with EGFR mutation groups. (C) Numbers of different immune cells in total (stroma and tumor) were compared between without and with EGFR mutation groups. (D–H) Numbers of CD8⁺ T cells, M1, M2 TAM, CD56bright NK cells, CD56dim NK cells in tumor, stroma, in total (tumor and stroma) were compared among patients of different EGFR mutation subgroups. (in the group termed as “other uncommon mutations”, contained EGFR amplification, EGFR p.E709A, EGFR p.E709K, EGFR p.E709V, EGFR p.F254L, EGFR p.H773dup, EGFR p.H870R, EGFR p.I740_K745dup, EGFR p.L747P, EGFR p.N468K, EGFR p.R451H, EGFR p.R776C, EGFR p.R776L and EGFR p.S768_D770dup). *means p < 0.05, **means p < 0.01.

Figure 5

Analysis of association with patients’ outcome. Kaplan Meier estimates for progression-free survival (PFS); patients were stratified according to with (EGFR+) and without EGFR mutation (EGFR-) (A) different subtypes of EGFR mutation (B) TMB (C) CD8+T (D) M1 (E) M2 macrophages (F), CD56bright NK (G) and CD56dim NK cells (H).

Figure 6

The prognostic value of different immune cells was evaluated in NSCLC patients with and without EGFR mutation. Kaplan Meier estimates for PFS; Patients without EGFR mutation (EGFR-) were stratified according to CD8+T (A), M1 macrophages (B) M2 (C) CD56bright NK (D) and CD56dim NK (E) Patients with EGFR mutation (EGFR+) were stratified according to CD8+T (F) M1 macrophages (G) M2 (H) CD56bright NK cells (I) and CD56dim NK cells (J).