P2.09-23 PD-L1 Expression of Paired Primary Resected Non Small Cell Carcinoma and Metastatic Lymph Node Fine Needle Aspirates

Abstract

Background: Small biopsy or cytology samples may present with a different level of PD-L1 expression compared to resected samples (which usually entail scoring of a much greater number of cells) resulting in relatively increased or reduced tumour proportion scores (TPSs). Many PD-L1 results are based on cytology fine needle lymph node aspirates (FNLNAs), encompassing analysis of metastatic disease and the substitution of cytology for histology samples. We compared the PD-L1 TPS of metastatic FNLNAs with that of resected non small cell carcinoma. Method: The pathology archive at Wythenshawe Hospital was searched for cases with adequate material over a period spanning 2010-2016. The Ventana SP263 PD-L1 clone was used to stain blocks select from 50 resected NSCCs and matched FNLNA cell blocks from each individual, along with a fresh H&E and negative PD-L1 control section. Result: Four of the cell block sections were inadequate for TPS assessment. The remaining 46 cases comprised 21 adenocarcinomas, 3 large cell carcinomas, 1 large cell neuroendocrine carcinoma, 1 atypical carcinoid tumour, and 20 squamous carcinomas. 34 cell block PD-L1 TPSs (68%) were in broad agreement with the corresponding resection block TPS, based on cutoff levels of 1% and 50%. Of the 12 (32%) cases in which differences occurred, 6 (50%) reflected an increase in TPS from resection to FNA, while 6 reflected a decrease causing a change in therapeutic cutoff. Nine of the FNLNAs were sampled after the resection, favouring the presence of recurrent disease. Conclusion: The majority of metastatic TPS FNLNAs are in broad agreement with a primary resected carcinoma TPS. FNLNAs tended to score less in the 1-49% category, possibly due to limits of cellularity. In addition to heterogeneity of expression, sampling of recurrent rather than residual disease may contribute to discrepancies.