Context.—

Programmed death ligand-1 (PD-L1) expression in non–small cell lung carcinoma (NSCLC) is heterogeneous and known to be underestimated on small biopsies. Correlation of PD-L1 expression with clinicopathologic features may provide additional useful information. To our knowledge, the clinicopathologic features of NSCLC have not been reported for subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells.

Objective.—

To investigate the clinicopathologic characteristics of NSCLC subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells.

Design.—

PD-L1 immunohistochemistry with the SP142 clone was performed on whole-tissue sections and given semiquantitative scores (0/1/2/3) according to percent of PD-L1+ tumor cells (TCs) and percent tumor area with PD-L1+ tumor-infiltrating immune cells (ICs).

Results.—

Adenocarcinoma cases that were scored either TC 1/2/3 or IC 1/2/3 included most (22 of 34; 65%) high–histologic grade cases and most (25 of 36; 69%) solid subtype cases. Compared with the adenocarcinoma TC 0 and IC 0 subset, the TC 1/2/3 or IC 1/2/3 subset correlated with higher histologic grade (P = .005, χ2 test for trend) and solid subtype (P < .001, Fisher exact test). Compared with the adenocarcinoma TC 0/1 or IC 0/1 subset, the TC 2/3 or IC 2/3 subset correlated with higher histologic grade (P = .002, χ2 test for trend), solid subtype (P < .001, Fisher exact test), and higher smoking pack-years (P = .01, Mann-Whitney test).

Conclusions.—

Lung adenocarcinoma subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells correlated with high histologic grade, solid subtype, and high smoking pack-years.

Several immune checkpoint inhibitors that target the interaction between programmed death receptor-1 (PD-1) and programmed death ligand-1 (PD-L1) have been successful in clinical trials for the treatment of non–small cell lung carcinoma (NSCLC). Nivolumab and pembrolizumab target PD-1, thereby inhibiting its interaction with PD-L1 expressed by tumor cells and tumor-infiltrating immune cells.1,2  Atezolizumab targets PD-L1, thereby inhibiting its interaction with PD-1 and B7.1 on T cells.3  PD-L1 expression by immunohistochemistry is currently the only available predictive biomarker for NSCLC response to immune checkpoint blockade.

When selecting immune checkpoint blockade as first-line therapy for metastatic NSCLC, as recently approved by the US Food and Drug Administration (FDA) for pembrolizumab,4  evidence of likely response, including PD-L1 status, assumes greater importance than for second-line therapy.5  One potential problem is that PD-L1 expression tends to be underestimated on small biopsies compared with surgical resection specimens.6  Clinicopathologic features that correlate with PD-L1 expression may be useful to mitigate this problem. In patients with a small biopsy that is PD-L1 but with clinicopathologic features that strongly correlate with PD-L1 expression, rebiopsy might be a consideration so that patients who are candidates for first-line immune checkpoint blockade are not missed.

Recently, the FDA approved atezolizumab for second-line treatment of metastatic NSCLC regardless of PD-L1 status.7  Evidence for this decision comes from the POPLAR and OAK clinical trials in which atezolizumab improved survival compared with docetaxel in patients with previously treated advanced or metastatic NSCLC. In the POPLAR clinical trial, stratification by PD-L1 expression was predictive of response to atezolizumab.3  To establish this, the POPLAR study analyzed the atezolizumab treatment arm as subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells. To our knowledge, the clinicopathologic features of NSCLC have not been reported for subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells. Given the potential usefulness of clinicopathologic correlations with PD-L1 expression, the aim of the present study is to investigate the clinicopathologic features of NSCLC subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells.

Study Samples

Institutional Review Board approval of research protocols for this project was obtained through Houston Methodist Hospital Research Institute (Houston, Texas). Surgical pathology blocks were obtained from a series of NSCLC from attempted curative surgical resections of lung cancers between 1975 and 1991 at Houston Methodist Hospital. Retrospective chart review was performed to obtain pathology reports, including original cell type diagnosis, and clinical data, including patient age, sex, smoking history, and survival status. Routine hematoxylin-eosin sections were obtained from these blocks and reviewed for histologic diagnosis.

Tumors were reclassified according to 2015 World Health Organization (WHO) criteria.8  Tumors lacking both adenocarcinoma and squamous differentiation were reclassified as follows. Immunohistochemistry results for thyroid transcription factor-1 (TTF-1), napsin A, p40, and DG-3 + cytokeratin 5 (CK5) cocktail from a previous study were reviewed.9  Tumors positive for adenocarcinoma markers (TTF-1 or napsin A) were reclassified as adenocarcinoma, solid subtype, whereas tumors positive for squamous cell carcinoma markers (p40 or CK5 + DG3) were reclassified as nonkeratinizing squamous cell carcinoma. Tumors lacking both adenocarcinoma and squamous differentiation were excluded from the study if immunohistochemistry results were unavailable or if the immunophenotype was ambiguous. Tumors with sarcomatoid or neuroendocrine morphology were excluded from the study. Subtypes were recorded for adenocarcinoma cases according to 2015 WHO criteria.

Histologic Grading

Well, moderately, and poorly differentiated adenocarcinomas were considered grades 1, 2, and 3, respectively. In the event of mixed histologic grades, only the highest histologic grade was recorded.

PD-L1 Immunohistochemistry

Immunohistochemistry on whole-tissue sections was carried out with an automated stainer (Leica Bond III, Leica Biosystems, Buffalo Grove, Illinois) using anti–PD-L1 (clone SP142) obtained from Spring Bioscience (Pleasanton, California). Incubation time was 30 minutes and dilution was 1:1000. Results were evaluated against a negative control.

PD-L1 expression was scored semiquantitatively according to percent of PD-L1+ tumor cells (TC 0 for <1%, TC 1 for 1%–4%, TC 2 for 5%–49%, and TC 3 for ≥50%) and percent tumor area with PD-L1+ tumor-infiltrating immune cells (IC 0 for <1%, IC 1 for 1%–4%, IC 2 for 5%–9%, and IC 3 for ≥10%), as in the POPLAR study.3  Closely associated peritumoral alveolar macrophages expressing PD-L1 were counted in the IC score, but alveolar macrophages distant from the tumor were not.

Statistics

Statistical calculations were performed using GraphPad Prism version 7 software (GraphPad Software Inc, La Jolla, California). P values less than .05 were considered statistically significant.

Following histologic review of recut sections, 90 of 227 cases (40%) lacked both adenocarcinoma and squamous differentiation. Of these 90 cases, 30 cases (33%) had immunohistochemistry results available for reclassifying by 2015 WHO criteria. There were 22 of 30 cases (73%) reclassified as adenocarcinoma, solid subtype; 7 of 30 cases (23%) reclassified as nonkeratinizing squamous cell carcinoma; and 1 case was excluded because of ambiguous immunophenotype. Three cases were excluded because of neuroendocrine morphology, and zero cases had sarcomatoid morphology. A total of 125 adenocarcinomas and 38 squamous cell carcinomas were included in the study.

The full range of PD-L1 expression semiquantitative scores was observed in both adenocarcinoma and squamous cell carcinoma groups (Table 1). PD-L1 expression was usually focal and involved aggregates of cells at the periphery of the tumor, as others have observed.10  Tumor-infiltrating immune cells with PD-L1 expression typically had abundant cytoplasm and may be macrophages or dendritic cells. Figures 1 and 2 depict TC 3 and IC 3 staining, respectively.

Table 1. 

Programmed Death Ligand-1 (PD-L1) Expression in Lung Adenocarcinomas and Squamous Cell Carcinomas

Programmed Death Ligand-1 (PD-L1) Expression in Lung Adenocarcinomas and Squamous Cell Carcinomas
Programmed Death Ligand-1 (PD-L1) Expression in Lung Adenocarcinomas and Squamous Cell Carcinomas
Figure 1

Representative image of solid adenocarcinoma with programmed death ligand-1 (PD-L1) expression in tumor cells (TCs) that was scored as TC3 (original magnification ×100).

Figure 2. Representative image of solid adenocarcinoma with programmed death ligand-1 (PD-L1) expression in tumor-infiltrating immune cells (ICs) that was scored as IC3 (original magnification ×100).

Figure 1

Representative image of solid adenocarcinoma with programmed death ligand-1 (PD-L1) expression in tumor cells (TCs) that was scored as TC3 (original magnification ×100).

Figure 2. Representative image of solid adenocarcinoma with programmed death ligand-1 (PD-L1) expression in tumor-infiltrating immune cells (ICs) that was scored as IC3 (original magnification ×100).

Close modal

The clinicopathologic variables of adenocarcinomas and squamous cell carcinomas were analyzed separately. No clinicopathologic variable correlated with the squamous cell carcinoma TC 1/2/3 or IC 1/2/3 subset (data not shown). The most common histologic grade of the adenocarcinoma TC 1/2/3 or IC 1/2/3 subset was high grade (22 of 56; 39%), whereas the most common histologic grade of the adenocarcinoma TC 0 or IC 0 subset was low grade (31 of 69; 45%) (Table 2). Because histologic grade is considered a progressive spectrum, we used a χ2 test for trend to examine statistical differences between subsets. The adenocarcinoma TC 1/2/3 or IC 1/2/3 subset correlated with high histologic grade (P = .005, χ2 test for trend). Age, sex, smoking pack-years, and stage did not correlate with the adenocarcinoma TC 1/2/3 or IC 1/2/3 subset.

Table 2. 

Clinicopathologic Characteristics of Lung Adenocarcinoma Subset Defined by 1% Programmed Death Ligand-1 (PD-L1) Expression in Either Tumor Cells (TCs) or Tumor-Infiltrating Immune Cells (ICs)

Clinicopathologic Characteristics of Lung Adenocarcinoma Subset Defined by 1% Programmed Death Ligand-1 (PD-L1) Expression in Either Tumor Cells (TCs) or Tumor-Infiltrating Immune Cells (ICs)
Clinicopathologic Characteristics of Lung Adenocarcinoma Subset Defined by 1% Programmed Death Ligand-1 (PD-L1) Expression in Either Tumor Cells (TCs) or Tumor-Infiltrating Immune Cells (ICs)

The adenocarcinoma TC 1/2/3 or IC 1/2/3 subset correlated with solid subtype compared with other subtypes (P < .001, Fisher exact test; Table 2). A detailed analysis of adenocarcinoma subtypes shows all adenocarcinoma subtype groups had cases with PD-L1 expression (see Supplemental digital content, containing 3 tables, at www.archivesofpathology.org in the November 2017 table of contents). Subtype groups with the least frequent PD-L1 expression included acinar (6 of 29; 21%), mucinous (3 of 11; 27%), and papillary (4 of 11; 36%) subtypes. Subtype groups with the most frequent PD-L1 expression included solid (25 of 36; 69%), micropapillary (7 of 14; 50%), and lepidic (11 of 23; 48%).

The clinicopathologic variables of the adenocarcinoma TC 2/3 or IC 2/3 subset were analyzed (Supplemental Table 2). As was observed in the TC 1/2/3 or IC 1/2/3 subset, the TC 2/3 or IC 2/3 subset correlated with histologic grade (P = .002, χ2 test for trend) and solid subtype (P < .001, Fisher exact test). The adenocarcinoma TC 2/3 or IC 2/3 subset also correlated with higher smoking pack-years (P = .01, Mann-Whitney test), with a range from 25 to 165 pack-years.

Kaplan-Meier survival analyses of various combinations of subsets did not show significant differences in disease-specific survival (Supplemental Table 3). For NSCLC, the TC 2/3 or IC 2/3 subset trended toward an improvement in disease-specific survival (P = .08, log rank test).

Although our literature search did not produce prior studies of the clinicopathologic features of NSCLC subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells, there are numerous studies of the clinicopathologic features of NSCLC subsets defined solely by PD-L1 expression in tumor cells. Several studies report correlation of PD-L1 expression in tumor cells with poor differentiation and solid histology.1116  In our study the adenocarcinoma TC 1/2/3 or IC 1/2/3 subset and TC 2/3 or IC 2/3 subset correlated with high histologic grade and solid subtype.

The correlation of PD-L1 expression with histologic grade and solid subtype may be clinically useful. Currently, frontline use of pembrolizumab in advanced lung cancer is limited to patients with high expression of PD-L1.4  In patients with a small biopsy that is negative for PD-L1 expression but with clinicopathologic features that would suggest a high probability of PD-L1 expression, rebiopsy might be considered to make sure that patients who are candidates for frontline immune checkpoint inhibitors are not missed.

In our study, medium-to-high PD-L1 expression was associated with higher smoking pack-years. This result is in good agreement with prior studies showing an association between PD-L1 expression and higher smoking pack-years in KRAS-mutated NSCLC17  and lung adenocarcinomas.13 

Investigations of the relationship between PD-L1 expression and survival in NSCLC have not yielded consistent results. A 2015 meta-analysis found PD-L1 expression correlated with poor overall survival.18  In our study PD-L1 expression did not correlate with disease-specific survival; however, the NSCLC TC 2/3 or IC 2/3 subset trended toward improved survival. It should be noted that our cohort is weighted toward lower stage. Cooper et al19  found PD-L1 expression in early-stage NSCLC was associated with improved survival, which may help explain our trending result.

A limitation of this study is the lack of an established grading system for lung adenocarcinomas.8,20  However, lung adenocarcinoma subtypes have been defined by the 2011 International Association for the Study of Lung Cancer (IASLC) classification, and these criteria were adopted in the 2015 WHO classification. A study of the reproducibility of lung adenocarcinoma histologic subtyping by 2011 IASLC criteria achieved a κ score of .77 for typical patterns, indicating good reproducibility.21  Strengths of our study include the use of whole-tissue sections for PD-L1 immunohistochemistry and the use of 2015 WHO criteria for tumors lacking adenocarcinoma and squamous differentiation that would have been diagnosed as large cell carcinoma by older classification systems.

In conclusion, we have examined the clinicopathologic features of NSCLC subsets defined by PD-L1 expression in either tumor cells or tumor-infiltrating immune cells because of the clinical relevance of such subsets. The lung adenocarcinoma TC 1/2/3 or IC 1/2/3 subset and TC 2/3 or IC 2/3 subset correlated with high histologic grade and solid subtype, and the latter subset also correlated with higher smoking pack-years.

1
Borghaei
H,
Paz-Ares
L,
Horn
L,
et al.
Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer
.
N Engl J Med
.
2015
;
373
(
17
):
1627
1639
.
2
Herbst
RS,
Baas
P,
Kim
DW,
et al.
Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomised controlled trial
.
Lancet
.
2016
;
387
(
10027
):
1540
1550
.
3
Fehrenbacher
L,
Spira
A,
Ballinger
M,
et al.
Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial
.
Lancet
.
2016
;
387
(
10030
):
1837
1846
.
4
US Food and Drug Administration
.
Pembrolizumab (KEYTRUDA) checkpoint inhibitor. United States Food and Drug Administration Web site
. ,
2017
.
5
Miller
RA,
Cagle
PT,
Bernicker
EH.
First-line immune therapy-implications for pathologists
.
Arch Pathol Lab Med
.
2016
;
140
(
8
):
739
740
.
6
Ilie
M,
Long-Mira
E,
Bence
C,
et al.
Comparative study of the PD-L1 status between surgically resected specimens and matched biopsies of NSCLC patients reveal major discordances: a potential issue for anti-PD-L1 therapeutic strategies
.
Ann Oncol
.
2016
;
27
(
1
):
147
153
.
7
FDA approves Roche's cancer immunotherapy TECENTRIQ (atezolizumab) for people with a specific type of metastatic lung cancer [press release]
.
Roche Web site
.
January
7
,
2017
.
8
Travis
WD,
Brambilla
E,
Burke
AP,
Marx
A,
Nicholson
AG,
eds
.
WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart. 4th ed
.
Lyon, France
:
IARC Press;
2015
.
WHO Classification of Tumours; vol 7
.
9
Brown
AF,
Sirohi
D,
Fukuoka
J,
et al.
Tissue-preserving antibody cocktails to differentiate primary squamous cell carcinoma, adenocarcinoma, and small cell carcinoma of lung
.
Arch Pathol Lab Med
.
2013
;
137
(
9
):
1274
1281
.
10
Herbst
RS,
Soria
JC,
Kowanetz
M,
et al.
Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients
.
Nature
.
2014
;
515
(
7528
):
563
567
.
11
Yang
CY,
Lin
MW,
Chang
YL,
Wu
CT,
Yang
PC.
Programmed cell death-ligand 1 expression in surgically resected stage I pulmonary adenocarcinoma and its correlation with driver mutations and clinical outcomes
.
Eur J Cancer
.
2014
;
50
(
7
):
1361
1369
.
12
Zhang
Y,
Wang
L,
Li
Y,
et al.
Protein expression of programmed death 1 ligand 1 and ligand 2 independently predict poor prognosis in surgically resected lung adenocarcinoma
.
Onco Targets Ther
.
2014
;
7
:
567
573
.
13
Cha
YJ,
Kim
HR,
Lee
CY,
Cho
BC,
Shim
HS.
Clinicopathological and prognostic significance of programmed cell death ligand-1 expression in lung adenocarcinoma and its relationship with p53 status
.
Lung Cancer
.
2016
;
97
:
73
80
.
14
Koh
J,
Go
H,
Keam
B,
et al.
Clinicopathologic analysis of programmed cell death-1 and programmed cell death-ligand 1 and 2 expressions in pulmonary adenocarcinoma: comparison with histology and driver oncogenic alteration status
.
Mod Pathol
.
2015
;
28
(
9
):
1154
1166
.
15
Saruwatari
K,
Ikemura
S,
Sekihara
K,
et al.
Aggressive tumor microenvironment of solid predominant lung adenocarcinoma subtype harboring with epidermal growth factor receptor mutations
.
Lung Cancer
.
2016
;
91
:
7
14
.
16
Takada
K,
Okamoto
T,
Shoji
F,
et al.
Clinical significance of PD-L1 protein expression in surgically resected primary lung adenocarcinoma
.
J Thorac Oncol
.
2016
;
11
(
11
):
1879
1890
.
17
Calles
A1
,
Liao
X,
Sholl
LM,
et al
.
Expression of PD-1 and its ligands, PD-L1 and PD-L2, in smokers and never smokers with KRAS-mutant lung cancer
.
J Thorac Oncol
.
2015
;
10
(
12
):
1726
1735
.
18
Wang
A,
Wang
HY,
Liu
Y,
et al.
The prognostic value of PD-L1 expression for non-small cell lung cancer patients: a meta-analysis
.
Eur J Surg Oncol
.
2015
;
41
(
4
):
450
456
.
19
Cooper
WA,
Tran
T,
Vilain
RE,
et al.
PD-L1 expression is a favorable prognostic factor in early stage non-small cell carcinoma
.
Lung Cancer
.
2015
;
89
(
2
):
181
188
.
20
Travis
WD,
Brambilla
E,
Noguchi
M,
et al.
International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma
.
J Thorac Oncol
.
2011
;
6
(
2
):
244
285
.
21
Thunnissen
E,
Beasley
MB,
Borczuk
AC,
et al.
Reproducibility of histopathological subtypes and invasion in pulmonary adenocarcinoma: an international interobserver study
.
Mod Pathol
.
2012
;
25
(
12
):
1574
1583
.

Author notes

The authors have no relevant financial interest in the products or companies described in this article.

Supplemental digital content is available for this article at www.archivesofpathology.org in the November 2017 table of contents.

Supplementary data