Accuracy of PD-L1 tumour staining of cytological and histological samples of lung adenocarcinoma

Abstract

Background: Immune checkpoint blockade with novel biological agents are effective in treatment of advanced NSCLC for patients with programmed death-ligand 1 (PD-L1) expression. Assessment of PD-L1 expression has previously been limited to histology specimens. We examined the accuracy of PD-L1 assessment of cytology specimens acquired via endobronchial ultrasound (EBUS). Our aim was to review whether there was concordance of PD-L1 tumour staining between cytology samples from EBUS and histology samples from surgically resected tumours.

Method: Retrospective chart review identified 27 patients who had EBUS sampling and subsequently underwent a surgical resection of adenocarcinoma. EBUS specimens processed into cell blocks and histological samples of resected tumours were each stained for PD-L1 expression.

Results: 27 matched cytology & histology specimens underwent PD-L1 assessment. Specimens were acquired by fine needle aspiration (n=1), bronchial brushings (n=23) and transbronchial needle aspiration (n=4). No false positive results were observed. There was concordance in 4 samples confirming PD-L1 expression in tumour cells from cell blocks and histology specimens. Three samples had a negative cell block with a positive biopsy however had insufficient cells in the cell block for analysis.

Conclusion: Our findings indicate accurate PD-L1 assessment of cytology specimens may be obtained by EBUS when cell block specimens have adequate cellular content for assessment. False negative results may be seen in inadequate specimens (<50 cells).