“…It should also be emphasized that careful interpretation of PD‐L1 staining in cytology samples is warranted because of the presence of immune cells, particularly macrophages, which are found in large numbers in pleural effusions. If these samples are evaluated without reference to a hematoxylin and eosin slide or, in some cases, complementary stains to assess immune cells such as macrophages or lymphocytes, the PD‐L1 staining might be misinterpreted as positive tumor cell PD‐L1 expression, even though tumor cells adjacent to the macrophage infiltrate have no PD‐L1 expression, as shown in Supporting Figure 6 (see online supporting information) and previously reported . In addition, although it is necessary to have ≥100 viable tumor cells present in tissue specimen slides to determine the percentage of PD‐L1–stained cells, there is no established minimum requirement for the number of tumor cells in cytology specimens.…”