PD-L1/pS6 in Circulating Tumor Cells (CTCs) during Osimertinib Treatment in Patients with Non-Small Cell Lung Cancer (NSCLC)

Abstract

The PD-1/PD-L1 axis provides CTCs an escape route from the immune system. Phosphorylation of the ribosomal protein S6 is implicated in the same pathway, following mTOR activation. The aim of the study was to investigate the expression of PD-L1 and pS6 in CTCs from NSCLC patients under Osimertinib treatment at a single cell level. CTCs were isolated using ISET from NSCLC patients’ blood [37 at baseline, 25 after the 1st cycle, and 23 at the end of treatment (EOT)]. Staining was performed using immunofluorescence. Cytokeratin-positive (CK+) CTCs were detected in 62% of patients. CK+PD-L1+CD45− and CK+pS6+ phenotypes were detected in 38% and 41% of the patients at baseline, in 28% and 32% after 1st cycle, and in 30% and 35% at EOT, respectively. Spearman’s analysis revealed statistically significant correlations between PD-L1 and pS6 phenotypes at all time points. Survival analysis revealed that CK+pS6+ (p = 0.003) and CKlowpS6+ (p = 0.021) phenotypes after 1st cycle were related to significantly decreased one-year progression-free survival (PFS12m) and PFS, respectively. CK+PD-L1+CD45−phenotype at baseline and after 1st cycle showed a trend for decreased PFS12m. Increased expression of PD-L1/pS6 in CTCs of Osimertinib-treated NSCLC patients implies the activation of the corresponding pathway, which is potentially associated with poor clinical outcomes.

Keywords: Osimertinib; PD-L1; circulating tumor cells; non-small cell lung cancer; pS6.

Figure 1

PD-L1 and pS6 expression in NSCLC patients’ CTCs. (A) CTCs stained with CK (red), PD-L1 (green) and CD45 (grey). Nuclei (blue) were stained with DAPI. (B) CTCs stained with CK (red), pS6 (green) and nuclei (blue). The overlay of all images is also presented. Scale bars = 10 μm.

Figure 2

Phenotypes of PD-L1 and pS6 identified in NSCLC patients and their CTCs. (A) Percentage of total patients with the identified phenotypes of PD-L1 (p < 0.001 among the phenotypes at baseline) and (B) mean percentage of CTCs with the identified phenotypes of PD-L1 (p < 0.001 and p = 0.003 among the phenotypes at baseline and EOT, respectively). (C) Percentage of total patients with the identified phenotypes of pS6 (p = 0.021 and p = 0.033 among the phenotypes at baseline and EOT, respectively) and (D) mean percentage of CTCs with the identified phenotypes of pS6 (p = 0.003 and p = 0.001 among the phenotypes at baseline and EOT, respectively). Results [in (B,D)] are expressed as mean ± SEM. B (white bars), baseline; P 1st (purple bars), post 1st cycle; EOT (green bars), end of treatment.

Figure 3

Phenotypes of PD-L1 and pS6 identified in thirteen NSCLC patients for which samples for all time points are available. Percentage of total patients with the identified phenotypes of (A) PD-L1 and (B) pS6. B (white bars), baseline; P 1st (purple bars), post 1st cycle; EOT (green bars), end of treatment. (C) Kaplan–Meier analysis of 1-year OS (OS_12m) for pS6 for the thirteen patients with samples at all time points. OS_12m according to the presence or absence of the phenotype CK^lowpS6⁺ (B, p = 0.014). B, baseline.

Figure 4

Kaplan–Meier analysis of 1-year PFS (PFS_12m) and PFS for PD-L1 and pS6. PFS_12m according to the presence or absence of the identified PD-L1 and pS6 phenotypes; (A) CK⁺PD-L1⁺CD45⁻ (B, p = 0.073), (B) CK⁺PD-L1⁺CD45⁻ (P 1st, p = 0.094) and (C) CK⁺pS6⁺ (P 1st, p = 0.003). (D) PFS for the presence or absence of CK^lowpS6⁺ (P 1st, p = 0.021). B, baseline; P 1st, post 1st cycle.