EBUS-TBNA as a Promising Method for the Evaluation of Tumor PD-L1 Expression in Lung Cancer

PD‐L1 expression in non‐small cell lung carcinoma: Comparison among cytology, small biopsy, and surgical resection specimens

“…Finally, in an analysis of a subset of patients for whom both cytologic EBUS‐FNA specimens and histologic biopsy or surgical resection samples were available, Sakakibara et al observed poor correlation of PD‐L1 expression (correlation coefficient, 0.19; P = .49) in 15 cases with 100 to 1000 tumor cells but good correlation (correlation coefficient, 0.68; P = .0019) in 18 cases with > 2000 tumor cells. Unfortunately, the number of those cases that were NSCLC was not reported, and the antibody used (EPR1161; Abcam, Cambridge, Massachusetts) was not among those used in conjunction with therapies that currently are approved by the FDA or in clinical trials, as outlined in the Blueprint comparison . Overall, these reports corroborate the feasibility and concordance of PD‐L1 expression between cytologic and histologic specimens as measured by IHC in the current study.…”

“…Furthermore, FNA specimens had a significantly greater number of tumor cells and less crush artifact compared with the tissue biopsies. However, that study included 27 neuroendocrine carcinoma specimens (28%), the small cell variant of which may demonstrate high cellularity and extensive crush artifact . In the current study, approximately 90% of cytologic specimens, all of which were NSCLC, provided sufficient cellularity for the quantification of PD‐L1 expression.…”

“…To the best of our knowledge, previously reported data regarding the feasibility of quantification of PD‐L1 expression are limited. In their study comparing PD‐L1 expression between 97 cytologic EBUS‐FNA specimens of primary lung malignancy, at least 70 of which were NSCLC and a subset for which concomitant histologic transbronchial biopsy specimens were available, Sakakibara et al noted that only 1 cytologic cell block contained <100 tumor cells . Furthermore, FNA specimens had a significantly greater number of tumor cells and less crush artifact compared with the tissue biopsies.…”

Clinical performance of endobronchial ultrasound‐guided transbronchial needle aspiration for assessing programmed death ligand‐1 expression in nonsmall cell lung cancer

“…Although there are no major studies to answer this question, two studies looked at PD‐L1 expression in histology and cytology samples from the same patients and found 80% concordance with slightly higher rates of concordance for squamous cell cancer than adenocarcinoma. These studies also demonstrated that PD‐L1 positivity of EBUS‐TBNA showed a good concordance with the corresponding primary tumor as well as with lymph node metastasis at 75% and 93% respectively . Some of the discordance could be due to tumor heterogeneity and not necessarily related to the type of tissue tested.…”

Programmed death‐ligand 1 testing of lung cancer cytology specimens obtained with bronchoscopy

“…The current study was limited by its retrospective methodology and a relatively small sample size. In addition, we did not a priori develop a metric for objectively quantifying the number of cells (eg, < 100, 100‐1000, 1000‐2000, and >2000 cells) as has been used by some investigators . Future prospective studies should apply such metrics to cell blocks, especially because a previous study has suggested that the correlation of PD‐ L1 expression between EBUS‐TBNA and biopsy specimens increases as the number of tumor cells increases .…”