Higher concordance of PD-L1 expression between biopsies and effusions in epithelioid than in nonepithelioid pleural mesothelioma

Abstract

Background: Malignant mesothelioma (MM) is a therapy-resistant tumor, often causing an effusion. Drugs targeting the programmed cell death 1 (PD-1)/programmed cell death ligand 1 (PD-L1) pathway have shown promising results, but assessment of PD-L1 expression to select patients for therapy has mainly been performed on histologic tissue samples. In a previous study, we showed that MM effusions are suitable for PD-L1 assessment with results comparable to those reported in histologic studies, but no studies have compared PD-L1 expression in histologic and cytologic samples.

Methods: PD-L1 expression was determined immunohistochemically (clone 28-8) in 61 paired samples of effusions and biopsies from patients with pleural MM, obtained at the time of diagnosis. Only cases with >100 tumor cells were included. Membranous staining in tumor cells was considered positive at ≥1%, >5%, >10%, and >50% cutoff levels.

Results: Of 61 histologic samples, PD-L1 expression was found in 28 and 7 samples at ≥1% and >50% cutoffs, respectively; the corresponding figures for cytology were 21 and 5, respectively. The overall percentage agreement between histology and cytology was 69% and 84%, with a kappa (κ) of 0.36 and 0.08 at ≥1% and >50% cutoffs, respectively. The concordance between cytology and histology tended to be higher for epithelioid MM versus nonepithelioid MM at a ≥1% cutoff. PD-L1 positivity in biopsies, but not in effusions, correlated with the histologic subtype at a ≥1% cutoff.

Conclusions: A moderate concordance of PD-L1 expression between biopsies and effusions from pleural MM, especially for the epithelioid subtype, indicates biological differences between the 2 types of specimens. Cytology and histology may be complementary.

Keywords: cell block; clone 28-8; cytology; histology; immunohistochemistry.

Figure 1

PD‐L1 reactivity in paired histologic and cytologic specimens from an epithelioid mesothelioma (original magnification ×400). (A) Histologic specimen, H&E staining. (B) Histologic specimen, PD‐L1 immunostaining (>50% PD‐L1–positive malignant cells). (C) Cytologic specimen, H&E staining. (D) Cytologic specimen, PD‐L1 immunostaining (>50% PD‐L1–positive malignant cells).

Figure 2

PD‐L1 reactivity in paired histologic and cytologic specimens from a biphasic mesothelioma. The histologic specimen is shown with the sarcomatoid component. The epithelial component was PD‐L1–negative (original magnification ×400). (A) Histologic specimen, H&E staining. (B) Histologic specimen, PD‐L1 immunostaining (>50% PD‐L1 positive malignant cells). (C) Cytologic specimen, H&E staining. (D) Cytologic specimen, PD‐L1 immunostaining (negative).