Alterations in PD-L1 expression associated with acquisition of resistance to ALK inhibitors in Anaplastic lymphoma kinase-rearranged lung cancer

Abstract

Introduction: Anaplastic lymphoma kinase (ALK) inhibitor is a standard therapy for patients with ALK-rearranged non-small cell lung cancer (NSCLC). However, most patients who respond initially develop resistance over time. Programmed cell death–ligand 1 (PD-L1) expressed on tumor cells induces immune escape and promotes tumor progression. The relationships between the resistance of ALK-positive NSCLC tumors to ALK inhibitors and the programmed cell death-1 (PD-1)/PD-L1 pathway have not been well-defined. Thus, we evaluated alterations in PD-L1 following acquisition of resistance to ALK inhibitors in ALK-positive lung cancer tissues and cell lines. Materials and Methods: Tumors were analyzed from 26 ALK-positive metastatic NSCLC patients (11 ALK inhibitor-naïve and 15 ALK inhibitor-resistant patients). The expression of PD-L1 and lymphocyte markers were assessed by immunohistochemistry and compared between the tumor specimens before and after treatment with ALK inhibitors. The PD-L1 H-score was calculated as a product of the intensity score (0, no staining

1, weak staining

2, moderate staining

and 3, strong staining) and proportion, and PD-L1 positivity was determined by the intensity and proportion using a cutoff of 10%. We also established ALK inhibitor-resistant cell lines (H3122CR1, LR1, and CH1) by exposing the parental H3122 ALK-translocated lung cancer cell line to ALK inhibitors. Then, the double-resistant cell lines H3122CR1LR1 and CR1CH1 were developed by exposing the crizotinib-resistant cell line H3122CR1 to other ALK inhibitors. We compared the alterations in PD-L1 expression levels using western blotting, flow cytometry, and quantitative polymerase chain reaction. We also investigated gene expression in H3122, CR1, LR1, CH1, CR1LR1, and CR1CH1 cell lines using RNA sequencing. We examined these properties in single- and double-resistant cell lines, each compared with H3122 parental cell lines. We then examined the associated biological processes and pathways using the differentially expressed gene data. Results: The mean value of the PD-L1 H-score was 6.5 pre-treatment (ALK inhibitor-naïve cells) and 35.0 post-treatment (ALK inhibitor-resistant cells), and the fold difference was 5.42 (p=0.163). PD-L1 positivity is more evident in post-treatment samples than in pre-treatment samples (3 [20.0%] vs. 0 [0.0%] patients, p=0.175). The mean value of CD68/mm2 was 181.6±115.4 and 90.8±48.9 in pre-treatment and post-treatment tumor samples, respectively (p=0.030). In the in vitro experiments, PD-L1 was expressed at higher levels in ALK inhibitor-resistant cell lines than in the ALK inhibitor-naïve parental cell line H3122. Furthermore, PD-L1 expression in the double-resistant cell lines was much higher than that in the single resistant cell lines. This trend was consistent at the total protein, surface protein, and mRNA levels. RNA sequencing demonstrated that expression of immune-related genes, including PGLYRRP4, CCL20, DEFB4A, LTB, and CDH6, was largely different in the single- and double-resistant cells compared with the parental H3122 cells. The more the cell lines acquired resistance to ALK inhibitors, the higher PD-L1expression was observed. In the analysis of biologic processes and pathways, immune-related pathways, including cytokine pathways, were found to be significantly involved in ALK inhibitor resistance. Conclusions: PD-L1 expression increased following acquisition of ALK inhibitor resistance in ALK-positive NSCLC patient tumors and cell lines. Interestingly, this trend became more evident as the cells acquired additional resistance. The role of PD/1PD-L1 pathway in ALK inhibitor resistance therefore merits further investigation.