Tumor Microenvironment Suggests Potential Benefit of PD-1 Inhibitors in Pretreated EGFR+ NSCLC

March 6, 2020
Lisa Astor

Changes in the tumor microenvironment following progression on EGFR tyrosine kinase inhibitor therapy suggest a potential for benefit from PD-1/PD-L1 inhibitors in patients with EGFR-positive non–small cell lung cancer, according to the results of a retrospective study published in Clinical Cancer Research.

Changes in the tumor microenvironment (TME) following progression on EGFR tyrosine kinase inhibitor (TKI) therapy suggest a potential for benefit from PD-1/PD-L1 inhibitors in patients withEGFR-positive non—small cell lung cancer (NSCLC), according to the results of a retrospective study published inClinical Cancer Research.

“[W]e have shown that there were significant changes in the TME, including PD-L1 expression in tumor cells and [tumor-infiltrating lymphocyte (TIL)] densities, between baseline and the time of development of EGFR TKI resistance,” the study authors, led by Kohsuke Isomoto, MD, of Kindai University Faculty of Medicine in Osaka-Sayama, Japan, wrote in the published report. “The observed increase in PD-L1 expression suggests that EGFR TKI therapy might confer a promising TME for subsequent [immune checkpoint inhibition (ICI)] treatment.”

The study examined the medical records of patients withEGFR-mutant advanced or recurrent NSCLC who were treated at one of 4 medical centers in Japan. A total of 138 patients were identified who had undergone a biopsy after disease progression on an EGFR TKI. Immunohistochemistry and next-generation sequencing testing were performed to analyze the TME and tumor mutational burden (TMB) of each biopsy.

The median age of all the patients was 68 years (range, 42-86) and 59% were female. A majority of the patients were nonsmokers (63%), had an ECOG performance status of 0 or 1 (75%), and had stage IIIB/IV NSCLC at the start of EGFR TKI therapy. More than half of the patients (55%) received gefitinib (Iressa) as their first EGFR inhibitor. Forty-four percent received osimertinib (Tagrisso) at some point during the course of treatment, and 20% received treatment with a PD-1 inhibitor following progression on an EGFR TKI.

When rebiopsied, a greater number of patients had both central nervous system (from 31% to 51%) and extra thoracic metastasis (from 58% to 75%) than at baseline. Additionally, 11 patients were identified as harboring T790M.

PD-L1 expression levels changed significantly from baseline (pre-EGFR TKI treatment) to the time of drug resistance (P= .0007). At baseline, 14% of patients had high levels (≥50%) of PD-L1 expression, and this amount increased to 28% after EGFR TKI treatment (P= .0010). Patients who did not acquire T790M (n = 70) also showed similar increases in high PD-L1 levels at progression (P= .0125). Isomoto et al noted that this was consistent with previous findings of an association between T790M negative status and high PD-L1 expression levels.

Both CD8-positive and FOXP3-positive TIL densities dropped after EGFR TKI therapy compared with baseline levels (P= .0274 andP<.0001, respectively). The density of CD8-positive TILs was higher in patients with high levels of PD-L1 expression after EGFR inhibition than in those with low or no PD-L1 expression (P<.0001). Patients with high PD-L1 expression levels had higher FOXP3-positive densities both before and after EGFR TKI therapy.

Neither CD73-positive TIL density nor CD73 expression in tumor cells underwent a notable change from baseline to the time of progression (P= .0653). CD73 expression in tumor cells at the time of progression was significantly higher in patients with ≥50% PD-L1 expression compared with those with lower PD-L1 expression (P= .0004). Expression of CD73 in tumor cells was also higher after EGFR TKI therapy from baseline in patients with high levels of PD-L1 expression (P= .0155).

PD-L1 expression increased after treatment with osimertinib in 8 evaluable patients, but no significant changes were seen in densities of CD8-positive, FOXP3-positive, and CD73-positive TILs with osimertinib therapy. Five of 7 patients evaluated for TMB showed a decrease in mutational burden following treatment.

Progression-free survival (PFS) was evaluated in 27 of the 28 patients treated with PD-1/PD-L1 ICI according to tumor cell PD-L1 expression levels. The median PFS was longer in patients with higher post—EGFR TKI treatment levels of PD-L1 expression than in those with expression levels below 50%. Two patients whose expression levels increased to ≥50% at progression each had a PFS of more than 6 months. Patients who were negative for T790M but had high levels of PD-L1 expression demonstrated the most favorable survival outcomes.

After EGFR TKI treatment, TMB was higher than at baseline in the 15 evaluable patients with available paired samples (P= .0508). Those with high PD-L1 expression levels that did not acquire T790M by the time of progression had longer PFS on ICI therapy than the patients who were T790M positive. This was not considered to be due to differences in TMB according to T790M status though, either before or after EGFR TKI therapy (P= .4430 andP= .3866, respectively).

The study authors noted that survival outcomes in patients withEGFR-mutant NSCLC were still modest with ICI treatment, even in patients with high PD-L1 expression levels, compared with patients who areEGFRwild type. They suggested that this could be due to increases in regulatory T-cell density and CD73 expression in tumor cells, but these immunosuppressive factors could potentially be overcome with the use of combination therapies.

The retrospective nature of the study was one limitation of the study as well as the variation of EGFR TKI treatments and cytotoxic chemotherapy received by some patients before the subsequent biopsy. The authors recommended prospective studies in larger patient populations to confirm the results.


  1. Isomoto K, Haratani K, Hayashi H, et al. Impact of EGFR-TKI Treatment on the Tumor Immune Microenvironment in EGFR Mutation—Positive Non&ndash;Small Cell Lung Cancer [published online January 14, 2020].Clin Cancer Res.doi: 10.1158/1078-0432.CCR-19-2027.