PD-L1 expression on circulating CD45(-) cells is an independent prognostic factor for overall survival (OS) in patients (Pts) across all stages of treatment-naïve lung cancer in a prospective, multicenter study

Background

Biopsies are currently utilized for profiling (histologic, molecular) of lung cancer prior to treatment. Profiling tumors from circulating markers could avoid invasive procedures and enable more frequent, kinetic disease monitoring. PD-L1 expression on lung cancer biopsy correlates with efficacy of immune checkpoint inhibitors. Using the Epic Sciences PD-L1 assay, we sought to correlate patient outcome with PD-L1(+) circulating CD45(-) cells, prior to its evaluation as a predictive biomarker for immune checkpoint inhibitors.

Methods

Peripheral blood was collected from 124 pts with diverse staging and histology at three clinical sites prior to diagnostic biopsy, with some having follow-up draws. All nucleated cells were plated onto glass slides and circulating cells of interest identified by immunofluorescence and morphological features. Prognostic capacity of PD-L1(+) cells (CK + /-, CD45-, PD-L1 + , malignant nuclear morphology) were assessed with Kaplan-Meier and Cox Proportional Hazard (PH) models.

Results

When detected, PD-L1 subcellular localization was primarily membranous. Pts with >1 PD-L1(+) cell/mL in baseline samples had worse overall survival (OS) (n = 22/124, mOS: 17.2 months vs. not reached, HR = 3.22, p = 0.0015) and follow-up (n = 5/22 mOS = 2.1 months vs. not reached, HR = 4.25, p = 0.0302). High baseline PD-L1(+) cell burden was additive and independent to AJCC staging in Cox PH models (HR = 2.32, p = 0.0447). Single-cell sequencing is underway.

Conclusions

In a multicenter cohort, circulating PD-L1(+)/CD45(-) cell burden predicted worse OS in pre-biopsy and follow-up lung cancer blood samples. This warrants prospective investigation as a predictive biomarker to PD-1 axis immune checkpoint inhibitors.

Clinical trial identification

HHSN261201200049C

Legal entity responsible for the study

National Cancer Institute, Epic Sciences

Funding

National Cancer Institute, Epic Sciences

Disclosure

R. Graf, D. Lu, R. Krupa, J. Louw, L. Dugan, A. Jendrisak, S. Orr, K. Bethel, Y. Wang, M. Landers, R. Dittamore: Epic Sciences Employee. All other authors have declared no conflicts of interest.