Introduction

Oncology Reports

1021-335X 1791-2431

D.A. Spandidos

10.3892/or.2022.8284

OR-47-04-08284

Articles

Differential expression of PD-L1 and PD-L2 is associated with the tumor microenvironment of TILs and M2 TAMs and tumor differentiation in non-small cell lung cancer

Sumitomo

Ryota

1 2 3 Huang

Cheng-Long

1 2 Fujita

Masaaki

2 Cho

Hiroyuki

1 Date

Hiroshi

1Department of Thoracic Surgery, Tazuke Kofukai Medical Research Institute, Kitano Hospital, Kita-ku, Osaka 530-8480, Japan 2Department of Oncology, Tazuke Kofukai Medical Research Institute, Kitano Hospital, Kita-ku, Osaka 530-8480, Japan 3Department of Thoracic Surgery, Faculty of Medicine, Kyoto University, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan

Correspondence to: Dr Cheng-Long Huang, Department of Thoracic Surgery, Tazuke Kofukai Medical Research Institute, Kitano Hospital, 2-4-20 Ohgimachi, Kita-ku, Osaka 530-8480, Japan, E-mail: chuang@kitano-hp.or.jp

04 2022

15 02 2022

47 4

06102021 01022022

2022

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

To improve the treatment strategy of immune-checkpoint inhibitors for non-small cell lung cancer (NSCLC), a comprehensive analysis of programmed death-ligand (PD-L)1 and PD-L2 expression is clinically important. The expression of PD-L1 and PD-L2 on both tumor cells (TCs) and tumor-infiltrating immune cells (ICs) was investigated, with respect to tumor-infiltrating lymphocytes (TILs) and M2 tumor-associated macrophages (TAMs), which are key components of the tumor microenvironment, in 175 patients with resected NSCLC. The TIL and M2 TAM densities were associated with the expression of PD-L1 on the two TCs (both P<0.0001) and ICs (both P<0.0001). The TIL and M2 TAM densities were also associated with the expression of PD-L2 on both TCs (P=0.0494 and P=0.0452, respectively) and ICs (P=0.0048 and P=0.0125, respectively). However, there was no correlation between the percentage of PD-L1-positive TCs and the percentage of PD-L2-positive TCs (r=0.019; P=0.8049). Meanwhile, tumor differentiation was significantly associated with the PD-L1 expression on TCs and ICs (P=0.0002 and P<0.0001, respectively). By contrast, tumor differentiation was inversely associated with the PD-L2 expression on both TCs and ICs (P=0.0260 and P=0.0326, respectively). In conclusion, the combined evaluation of PD-L1 and PD-L2 expression could be clinically important in the treatment strategy of immune-checkpoint inhibitors in patients with NSCLC. In particular, the evaluation of PD-L2 expression may be necessary for patients with PD-L1-negative NSCLC.

programmed death-ligand 1/2 tumor cell immune cell macrophage lymphocyte tumor microenvironment differentiation lung cancer

Funding: No funding was received.

Introduction

Non-small cell lung cancer (NSCLC), accounting for ~85% of all cases of lung cancer, remains to be the leading cause of cancer-related mortality worldwide, despite the availability of advanced cytotoxic chemotherapies and molecular-targeted therapies, such as EGFR-tyrosine kinase inhibitors (1,2). However, recently, agents that target the programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) axis, such as immune-checkpoint inhibitors, have been widely used as a standard treatment for patients with metastatic NSCLC (3–5). Pembrolizumab, an anti-PD-1 antibody, has been approved as monotherapy in patients with tumors that have highly upregulated expression of PD-L1 on tumor cells (TCs) (4). This finding made PD-L1 testing a mandatory diagnostic test during treatment planning in patients with NSCLC. Furthermore, effective clinical response to atezolizumab, an anti-PD-L1 antibody, is observed not only in patients with tumors with high PD-L1 expression on TCs, but also in patients with tumors that expressed high levels of PD-L1 on tumor-infiltrating immune cells (ICs) (6). These observations suggest that the PD-L1 expression not only on TCs but also ICs serves an important role in regulating the anti-tumor T cell response. In addition, the PD-L1 expression on TCs and ICs is reported to be affected by microenvironment stimuli, including tumor-infiltrating lymphocytes (TILs) and M2 tumor-associated macrophages (TAMs) (7,8).

On the other hand, recent clinical studies report that PD-L2, another PD-1 ligand, is also widely expressed in numerous types of cancer, including NSCLC (9–14). Several studies reveal that PD-L2 is also expressed by both TCs and various ICs, depending on the microenvironment stimuli (15–17). In addition, experimental studies report that PD-L2-expressing TCs are resistant to treatment with anti-PD-L1 antibody alone and that this resistance is overcome by an anti-PD-1 antibody or in combination with an anti-PD-L2 antibody (18,19). A clinical study reports that clinical response to pembrolizumab in patients with head and neck squamous cell carcinoma may be related partly to blockade of PD-1/PD-L2 interactions (20). However, the clinical significance of the PD-L2 expression in NSCLC is still controversial (7,9–11).

Taken together, to improve the treatment strategy of immune-checkpoint inhibitors for patients with NSCLC, a comprehensive analysis of the biological mechanisms and clinical significance of PD-L1 and PD-L2 expression was considered to be clinically important. Therefore, a the present study was performed to evaluate the expression of PD-L1 and PD-L2 on both TCs and ICs in patients with NSCLC. In addition, the association between TILs and M2 TAMs, which are key components of the tumor microenvironment (TME), on the expression of PD-L1 and PD-L2 was also analyzed.

Materials and methods <sec> <title>Patients

Consecutive 175 patients with NSCLC, who underwent surgery at the Department of Thoracic Surgery, Kitano Hospital (Osaka, Japan) between November 2011 and December 2014, were included. The present study was approved by the Ethics Committee (approval no. P181200300) and written informed consent was provided from each patient. Pathological staging was determined using the 8th Tumor Node Metastasis (TNM) classification system (21). The histological type and the grade of differentiation of the tumors were determined according to the classification system developed by the World Health Organization (22). The medical records and histopathological diagnosis from the patients were fully documented.

Immunohistochemistry

Immunohistochemical studies were performed to evaluate the TIL distribution by CD3 staining, the M2 TAM distribution by CD163 staining (8,23), PD-L1 expression on TCs and ICs by the Ventana SP263 assay and PD-L2 expression on TCs and ICs, using the Ventana BenchMark GX system (Ventana Medical Systems; Roche Diagnostics), according to the recommended protocol. The following antibodies were used: Rabbit monoclonal anti-human CD3 (clone 2GV6; prediluted; Ventana Medical Systems; Roche Diagnostics), PD-L1 (clone SP263; prediluted; Ventana Medical Systems; Roche Diagnostics) (24) and PD-L2 (cat. no. 18251-1-AP; 1:200; ProteinTech Group, Inc.) and mouse monoclonal anti-human CD163 (clone 760-4437; prediluted; Ventana Medical Systems; Roche Diagnostics). The tissues were fixed in 10% neutral-buffered formalin for 24 h at room temperature. After dehydration in graded ethanol series followed by xylene at room temperature, the tissues were embedded in paraffin at 60°C. Formalin-fixed paraffin-embedded tissue was cut into 4-µm sections and mounted on poly-L-lysine-coated slides. The sections were deparaffinized and rehydrated using EZ Prep (Ventana Medical Systems; Roche Diagnostics) at 75°C. Antigen retrieval was performed using Cell Conditioner 1 (Ventana Medical Systems; Roche Diagnostics) for 64 min at 100°C against CD3, PD-L1 and PD-L2 and 32 min at 100°C against CD163. The sections were then incubated with the specific primary antibody for 16 min at 37°C against CD3, CD163 and PD-L1 and 2 h at 37°C for PD-L2. Subsequently, the sections were treated with the OptiView HQ Linker (Ventana Medical Systems; Roche Diagnostics) for 8 min at 37°C and the OptiView HRP Multimer (Ventana Medical Systems; Roche Diagnostics) for 8 min at 37°C. Finally, counterstaining was performed with Mayer's hematoxylin and Scott's tap water bluing reagent at 37°C.

The evaluation of the stained tissue sections was performed by two investigators (RS and CLH) blinded to the study. The cases with discrepancies were jointly re-evaluated until a consensus was reached. For CD3 and CD163 staining, the five most representative high-power fields (magnification, ×400; 0.0625 mm²) of the tumor stroma were selected. Tumor stroma was defined as the area where tumor stromal cells accounted for >70% of the total cells (25). The number of CD3-positive cells and CD163-positive cells in each area was counted and the average number of fields in each area was calculated. Finally, the CD3-positive cell density in the tumor stroma (TIL density) and the CD163-positive macrophage density in the tumor stroma (M2 TAM density) were defined as the cell number per mm². PD-L1 and PD-L2 expression was calculated as the percentage of membrane staining on TCs or ICs, respectively, in the overall area of the tumor, regardless of intensity.

Statistical analysis

The statistical significances regarding continuous variables were assessed using either a t-test, ANOVA with Bonferroni/Dunn post hoc test or Pearson's correlation coefficient. Categorical variables were compared using a χ² test. Statistical analyses were performed using SPSS v23.0 for Windows (IBM Corp.). All P-values were based on the two-sided statistical analysis and P<0.05 was considered to indicate a statistically significant difference.

Results <sec> <title>Distribution and clinical significance of TILs among resected NSCLCs

Immunohistochemistry for CD3 exhibited a membranous and cytoplasmic staining pattern (Fig. 1A, C and E). The TIL density varied among the 175 tumor tissues (mean ± standard deviation, 948.1±890.6; Table I). As the TIL density cut-off (524) demonstrated the highest significance with respect to the percentage of PD-L1-positive TCs and PD-L2-positive TCs, the sample was classified as TIL-high when the TIL density was >524. A total of 71 tumors (40.6%) were classified as TIL-low and 104 tumors (59.4%) were classified as TIL-high. With respect to tumor histology, the TIL density was significantly higher in squamous cell carcinoma compared with that in adenocarcinoma (P=0.0206). In addition, with respect to tumor differentiation, the TIL density was significantly higher in moderately and poorly differentiated tumors compared with that in well-differentiated tumors (P=0.0130).

Distribution and clinical significance of M2 TAMs among resected NSCLCs

Immunohistochemistry for CD163 exhibited a membranous and cytoplasmic staining pattern (Fig. 1G). The M2 TAM density also varied among the 175 tumor tissues (mean ± standard deviation, 382.5±381.9; Table I). There was a weak correlation between TIL and M2 TAM densities (r=0.262; P=0.0004; Fig. 2). The sample was classified as M2 TAM-high when the M2 TAM density was >380 due to the highest significance in the level of C-reactive protein, a marker of the inflammatory response, as previously reported (23). A total of 108 tumors (61.7%) were classified as M2 TAM-low and 67 tumors (38.3%) were classified as M2 TAM-high. The M2 TAM density was also significantly higher in squamous cell carcinoma compared with that in adenocarcinoma (P=0.0036). The M2 TAM density was also significantly higher in poorly differentiated tumors compared with that in well- and moderately differentiated tumors (P=0.0015). Furthermore, the M2 TAM density was significantly higher in node-positive tumors and advanced stage (P=0.0165 and P=0.0388, respectively).

Expression of PD-L1 on TCs and ICs with respect to TILs and M2 TAMs

The percentage of PD-L1-positive TCs varied among the 175 tumor tissues (mean ± standard deviation; 15.6±27.0%; Fig. 1B). PD-L1 expression on TCs was significantly higher in squamous cell carcinoma compared with that in adenocarcinoma (P=0.0001).

The percentage of PD-L1-positive ICs also varied (mean ± standard deviation, 9.4±10.9%; Fig. 1D). PD-L1 expression on ICs was also significantly higher in squamous cell carcinoma compared with that in adenocarcinoma (P=0.0173). Furthermore, PD-L1 expression on ICs was significantly associated with tumor status, nodal status and pathological stage (P=0.0104, P=0.0166 and P=0.0027, respectively).

With respect to TILs, the TIL density was significantly correlated with the percentage of PD-L1-positive TCs (r=0.365; P<0.001). The percentage of PD-L1-positive TCs was significantly higher in the TIL-high group compared with that in the TIL-low group (22.2±30.8% vs. 6.1±16.0%; P<0.0001; Fig. 3A). Furthermore, the TIL density was also significantly correlated with the percentage of PD-L1-positive ICs (r=0.751; P<0.001). The percentage of PD-L1-positive ICs was significantly higher in the TIL-high group compared with that in the TIL-low group (14.4±11.2% vs. 1.9±3.5%; P<0.0001; Fig. 3B).

As previously reported (8), with respect to M2 TAMs, the M2 TAM density was significantly correlated with the percentage of PD-L1-positive TCs (r=0.389; P<0.001). The percentage of PD-L1-positive TCs was significantly higher in the M2 TAM-high group compared with that in the M2 TAM-low group (26.5±32.3% vs. 8.9±20.5%; P<0.0001; Fig. 3C). Furthermore, the M2 TAM density was significantly correlated with the percentage of PD-L1-positive ICs (r=0.375; P<0.001). The percentage of PD-L1-positive ICs was significantly higher in the M2 TAM-high group compared with that in the M2 TAM-low group (14.2±12.0% vs. 6.4±8.8%; P<0.0001; Fig. 3D).

With respect to the combined evaluation of TILs and M2 TAMs, the percentage of PD-L1-positive TCs was significantly the highest in both the TIL-high and M2 TAM-high tumors (Fig. 3E). The percentage of PD-L1-positive ICs was also significantly the highest in both the TIL-high and M2 TAM-high tumors (Fig. 3F).

Expression of PD-L2 on TCs and ICs with respect to TILs and M2 TAMs

The percentage of PD-L2-positive TCs varied among the 175 tumor tissues (mean ± standard deviation, 14.6±22.9%; Fig. 1F) and there were <1% in 70 (40.0%) tumors, 1–49% in 84 (48.0%) tumors and ≥50% in 21 (12.0%) tumors (Table II). The percentage of PD-L2-positive ICs also varied (mean ± standard deviation, 12.5±18.4%; Fig. 1H) and there were <1% in 50 (28.6%) tumors, 1–9% in 69 (39.4%) tumors and ≥10% in 56 (32.0%) tumors (Table II).

With respect to TILs, the percentage of PD-L2-positive TCs was significantly higher in the TIL-high group compared with that in the TIL-low group (17.4±25.9% vs. 10.5±17.0%; P=0.0494; Fig. 3G). In addition, the TIL density was significantly correlated with the percentage of PD-L2-positive ICs (r=0.226; P=0.003). The percentage of PD-L2-positive ICs was significantly higher in the TIL-high group compared with that in the TIL-low group (15.7±19.7% vs. 7.8±15.3%; P=0.0048; Fig. 3H).

With respect to M2 TAMs, the percentage of PD-L2-positive TCs was significantly higher in the M2 TAM-high group compared with that in the M2 TAM-low group (19.0±27.9% vs. 11.9±18.9%; P=0.0452; Fig. 3I). Furthermore, the percentage of PD-L2-positive ICs was also significantly higher in the M2 TAM-high group compared with that in the M2 TAM-low group (16.9±19.8% vs. 9.8±17.0%; P=0.0125; Fig. 3J).

With respect to the combined evaluation of TILs and M2 TAMs, the percentage of PD-L2-positive TCs was significantly the lowest in both the TIL-low and M2 TAM-low tumors (Fig. 3K). The percentage of PD-L2-positive ICs was significantly lower in both the TIL-low and M2 TAM-low tumors compared with that in both the TIL-high and M2 TAM-high tumors (P=0.0011; Fig. 3L).

Correlations between the expression of PD-L1 and PD-L2 on the TCs and ICs among resected NSCLC

There was no correlation between the percentage of PD-L1-positive TCs and the percentage of PD-L2-positive TCs (r=0.019; P=0.8049; Fig. 4A). On the other hand, the percentage of PD-L1-positive TCs was significantly correlated with the percentage of PD-L1-positive ICs (r=0.396; P<0.0001; Fig. 4B). In addition, the percentage of PD-L2-positive TCs also was significantly correlated with the percentage of PD-L2-positive ICs (r=0.488; P<0.0001; Fig. 4C).

Expression of PD-L1 and PD-L2 with respect to tumor differentiation

PD-L1 expression on TCs was significantly associated with tumor differentiation (P=0.0002; Table II), as previously reported (8). The percentage of PD-L1-positive TCs was 6.0±17.0% in well-differentiated tumors, 13.1±24.2% in moderately differentiated tumors and 32.8±34.9% in poorly differentiated tumors. The percentage of PD-L1-positive TCs was significantly higher in poorly differentiated tumors compared with that in well- and moderately differentiated tumors (P<0.0001 and P=0.0001, respectively; Fig. 5A).

Furthermore, PD-L1 expression on ICs was also significantly associated with tumor differentiation (P<0.0001; Table II), as previously reported (8). The percentage of PD-L1-positive ICs was 5.7±8.6% in well-differentiated tumors, 8.3±10.0% in moderately differentiated tumors and 16.2±12.6% in poorly differentiated tumors. The percentage of PD-L1-positive ICs was significantly higher in poorly differentiated tumors compared with that in well- and moderately differentiated tumors (P<0.0001 and P=0.0001, respectively; Fig. 5B).

On the other hand, PD-L2 expression on TCs was inversely associated with tumor differentiation (P=0.0260; Table II). The percentage of PD-L2-positive TCs was 23.5±25.9% in well-differentiated tumors, 13.4±22.4% in moderately differentiated tumors and 9.2±19.1% in poorly differentiated tumors. The percentage of PD-L2-positive TCs was significantly higher in well-differentiated tumors compared with that in poorly and moderately differentiated tumors (P=0.0088 and P=0.0234, respectively; Fig. 5C).

The PD-L2 expression on ICs was also inversely associated with tumor differentiation (P=0.0326; Table II). The percentage of PD-L2-positive ICs was 19.3±22.5% in well-differentiated tumors, 11.4±17.4% in moderately differentiated tumors and 9.1±15.3% in poorly differentiated tumors. The percentage of PD-L2-positive ICs was significantly higher in well-differentiated tumors compared with that in poorly and moderately differentiated tumors (P=0.0196 and P=0.0269, respectively; Fig. 5D).

Discussion

A comprehensive study on PD-L1 and PD-L2 expression on both TCs and ICs in NSCLC was performed. A recent study reports that PD-L1 has predominant roles in Th1-type immunity whereas PD-L2 is involved in Th2-type immunity (26). In addition, to elucidate the biological mechanisms of their regulation, TILs and M2 TAMs, which are key components of the TME and associated with tumor progression, were investigated. The evaluation of PD-L1 and PD-L2 on both TCs and ICs is clinically important and immunohistochemistry is an appropriate method for the design of the present study. A previous study reports that the Ventana PD-L1 (SP-263) assay is clinically useful for PD-L1 staining on both TCs and ICs (24). In addition, the PD-L2 expression using the Ventana system also exhibited a clear staining on both TCs and ICs in the present study.

Consequently, the present study revealed that the TIL density was strongly associated with the PD-L1 expression on both TCs and ICs. On the other hand, PD-L2 was widely expressed not only on TCs, but also on ICs in NSCLC. In addition, the TIL density was also associated with PD-L2 expression on both TCs and ICs. Initially, CD8⁺ or CD4⁺ T cells and NK cells are known to induce PD-L1 expression by producing interferon (IFN)-γ (27,28). TILs have been reported to be an important cause of PD-L1 expression on ICs, such as lymphatic endothelial cells, macrophages and monocytes (29–31). Numerous clinical studies have also revealed that TILs are associated with PD-L1 expression in human cancer, including NSCLC (7,32,33). In addition, previous studies report that TILs are also associated with PD-L2 expression in human cancer (33,34).

Based on the physiological or pathological situation, macrophages can be polarized into various phenotypes with different biological properties, such as tumor-inhibiting M1 macrophages and tumor-promoting M2 macrophages (35,36). During tumor progression, Th2-derived cytokines originating from TCs and stromal cells can induce the production of M2 TAMs in the TME, which can promote tumor cell proliferation (37). In fact, the M2 TAM density was associated with nodal status and pathological stage in the present study. Thus, M2 TAM-high tumors have more aggressive potential in NSCLC (23).

On the other hand, our previous study found that the M2 TAM density was strongly associated with PD-L1 expression on both TCs and ICs (8). In addition, the present study demonstrated that the M2 TAM density was also associated with the PD-L2 expression on both TCs and ICs. Experimental studies report that TCs can induce M2 TAMs with increased expression of PD-L1 (38,39). It is also known that PD-L1, induced by IFN-γ from TAMs, promoted the progression of lung cancer (40). Recent studies show that other signals derived from macrophages, such as TNF-α, VEGF and CXCL8, can induce PD-L1 expression (41–43). In addition, previous studies report that macrophages can induce not only PD-L1 expression, but also PD-L2 expression (44,45).

From these findings, the TIL and M2 TAM densities were associated with the expression of PD-L1 and PD-L2 on TCs and ICs. In the present study, the TIL density was significantly associated with the preoperative serum albumin level (r=0.269; P<0.001; Fig. S1A) and the preoperative peripheral blood lymphocyte count (r=0.209; P=0.006; Fig. S1B). Therefore, TILs are considered to be a host-related factor. By contrast, M2 TAMs are considered to be a tumor-related factor (23). Thus, such complex crosstalk in the TME, including TILs and M2 TAMs, could affect the expression of PD-L1 and PD-L2 on TCs and ICs in NSCLC (46).

However, the present study demonstrated the additional finding of no correlation between PD-L1 expression on TCs and PD-L2 expression on TCs, despite the possible same regulations by TILs and M2 TAMs. Several studies also report a high frequency of discordance between PD-L1 and PD-L2 expression in human cancer (47,48). By contrast, there were correlations between PD-L1 expression on TCs and PD-L1 expression on ICs and between PD-L2 expression on TCs and PD-L2 expression on ICs in the present study.

The present study revealed that tumor differentiation was strongly associated with PD-L1 expression on TCs and ICs. The percentages of PD-L1-positive TCs and PD-L1-positive ICs were higher in poorly differentiated tumors compared with that in well- and moderately differentiated tumors. A meta-analysis on PD-L1 expression in lung cancer also reports the same results (49). In addition, an experimental study reveals that PD-L1 could upregulate the β-catenin signaling pathway to induce epithelial-mesenchymal transition (50), which is associated with tumor differentiation in lung cancer (51,52). By contrast, tumor differentiation was inversely associated with PD-L2 expression on TCs and ICs in the present study. The percentages of PD-L2-positive TCs and PD-L2-positive ICs were higher in well-differentiated tumors compared with that in poorly and moderately differentiated tumors.

Therefore, the combined evaluation of PD-L1 and PD-L2 expression could be considered clinically important in the treatment strategy of immune-checkpoint inhibitors in patients with NSCLC. In particular, the evaluation of PD-L2 expression may be necessary for patients with PD-L1-negative NSCLC. Patients with PD-L2-positive NSCLC could be treated with anti-PD-1 antibodies, such as Pembrolizumab, and combined treatment with anti-PD-L2 antibodies in the future (18–20). In fact, in the present study, immune-checkpoint inhibitors were only used in 7 cases of PD-L1-positive tumors at the time of disease recurrence, whereas 56 cases had recurrence following surgery. Further clinical studies are required for patients with PD-L2-positive NSCLC. In addition, the present study was performed using a relatively small number of patients at one institution. Therefore, a further study using more cases is required to elucidate the clinical significance of PD-L2 expression, especially with respect to the treatment strategy of immune-checkpoint inhibitors. Furthermore, the present study was evaluated only by immunohistochemistry and a further study to investigate their gene copy numbers may be needed (53).

In conclusion, PD-L1 and PD-L2 expression on TCs and ICs was associated with TILs and M2 TAMs in NSCLC. However, there was no correlation between PD-L1 and PD-L2 expression on TCs. Meanwhile, PD-L1 expression on TCs and ICs was associated with tumor differentiation, while PD-L2 expression on TCs and ICs was inversely associated with tumor differentiation. The combined evaluation of PD-L1 and PD-L2 expression could be considered clinically important in the treatment strategy of immune-checkpoint inhibitors in patients with NSCLC. In particular, the evaluation of PD-L2 expression may be necessary for patients with PD-L1-negative NSCLC.

Supplementary Material

Supporting Data

Acknowledgements

Not applicable.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

RS, CLH and HD designed the study. RS, CLH and MF designed and performed the experiments. RS, CLH and HC collected the data. RS and CLH analyzed and interpreted the data and wrote the manuscript. RS and CLH confirm the authenticity of all the raw data. All authors have read and approved the final version of the manuscript for publication.

Ethics approval and consent to participate

The current study was approved by the Institutional Ethics Committee of the Kitano Hospital (approval no. P181200300) and written informed consent was provided from each patient. The research was conducted in compliance with the principles outlined in the Declaration of Helsinki.

Patient consent for publication

Written informed consent for publication of patient data/accompanying images was obtained.

Competing interests

The authors declare that they have no competing interests.

References 1

Ettinger

Akerley

Bepler

Blum

Chang

Cheney

Chirieac

D'Amico

Demmy

Ganti

Non-small cell lung cancer

J Natl Compr Canc Netw87408012010

10.6004/jnccn.2010.0056

20679538

Hsu

Yang

Mok

Loong

Overview of current systemic management of EGFR-mutant NSCLC

Ann Oncol29(Suppl_1)i3i92018

10.1093/annonc/mdx702

29462253

Borghaei

Paz-Ares

Horn

Spigel

Steins

Ready

Chow

Vokes

Felip

Holgado

Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer

N Engl J Med373162716392015

10.1056/NEJMoa1507643

26412456

Herbst

Baas

Kim

Felip

Perez-Gracia

Han

Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): A randomized controlled trial

Lancet387154015502016

10.1016/S0140-6736(15)01281-7

26712084

Rittmeyer

Barlesi

Waterkamp

Park

Ciardiello

von Pawel

Gadgeel

Hida

Kowalski

Atezolizumab versus docetaxel in patients with previously treated non-small-cell lung cancer (OAK): A phase 3, open-label, multicenter randomized controlled trial

Lancet3892552652017

10.1016/S0140-6736(16)32517-X

27979383

Kowanetz

Zou

Gettinger

Koeppen

Kockx

Schmid

Kadel

Wistuba

Chaft

Rizvi

Differential regulation of PD-L1 expression by immune and tumor cells in NSCLC and the response to treatment with atezolizumab (anti-PD-L1)

Proc Natl Acad Sci USA115E10119E101262018

10.1073/pnas.1802166115

30297397

Kim

Koh

Kim

Jeon

Chung

Clinicopathological analysis of PD-L1 and PD-L2 expression in pulmonary squamous cell carcinoma: Comparison with tumor-infiltrating T cells and the status of oncogenic drivers

Lung Cancer8824332015

10.1016/S0169-5002(15)50059-7

25662388

Sumitomo

Hirai

Fujita

Murakami

Otake

Huang

PD-L1 expression on tumor-infiltrating immune cells is highly associated with M2 TAM and aggressive malignant potential in patients with resected non-small cell lung cancer

Lung Cancer1361361442019

10.1016/j.lungcan.2019.08.023

31499335

Shinchi

Komohara

Yonemitsu

Sato

Ohnishi

Saito

Fujiwara

Mori

Shiraishi

Ikeda

Suzuki

Accurate expression of PD-L1/L2 in lung adenocarcinoma cells: A retrospective study by double immunohistochemistry

Cancer Sci110271127212019

10.1111/cas.14128

31294893

Matsubara

Takada

Azuma

Takamori

Toyokawa

Haro

Osoegawa

Tagawa

Kawahara

Akiba

A clinicopathological and prognostic analysis of PD-L2 expression in surgically resected primary lung squamous cell carcinoma

Ann Surg Oncol26192519332019

10.1245/s10434-019-07684-2

30815803

Takamori

Takada

Azuma

Jogo

Shimokawa

Toyokawa

Hirai

Tagawa

Kawahara

Akiba

Prognostic impact of programmed death-ligand 2 expression in primary lung adenocarcinoma patients

Ann Surg Oncol26191619242019

10.1245/s10434-019-07231-z

30815801

Baptista

Sarian

Derchain

SFM

Pinto

Vassallo

Prognostic significance of PD-L1 and PD-L2 in breast cancer

Hum Pathol4778842016

10.1016/j.humpath.2015.09.006

26541326

Zhao

Lehrer

Chang

Das

Erho

Liu

Sjostrom

Den

Freedland

Klein

The immune landscape of prostate cancer and nomination of PD-L2 as a potential therapeutic target

J Natl Cancer Inst1113013102019

10.1093/jnci/djy141

30321406

Okadome

Baba

Nomoto

Yagi

Kalikawe

Harada

Hiyoshi

Nagai

Ishimoto

Iwatsuki

Prognostic and clinical impact of PD-L2 and PD-L1 expression in a cohort of 437 oesophageal cancers

Br J Cancer122153515432020

10.1038/s41416-020-0811-0

32210369

Okazaki

Honjo

PD-1 and PD-1 ligands: From discovery to clinical application

Int Immunol198138242007

10.1093/intimm/dxm057

17606980

Rozali

Hato

Robinson

Lake

Lesterhuis

Programmed death ligand 2 in cancer-induced immune suppression

Clin Dev Immunol20216563402012

22611421

Zhong

Tumang

Gao

Bai

Rothstein

PD-L2 expression extends beyond dendritic cells/macrophages to B1 cells enriched for V(H)11/V(H)12 and phosphatidylcholine binding

Eur J Immunol37240524102007

10.1002/eji.200737461

17683117

Tanegashima

Togashi

Azuma

Kawahara

Ikeguchi

Sugiyama

Kinoshita

Akiba

Kashiwagi

Takeuchi

Immune suppression by PD-L2 against spontaneous and treatment-related antitumor immunity

Clin Cancer Res25480848192019

10.1158/1078-0432.CCR-18-3991

31076547

Umezu

Okada

Sakoda

Adachi

Ojima

Yamaue

Eto

Tamada

Inhibitory functions of PD-L1 and PD-L2 in the regulation of anti-tumor immunity in murine tumor microenvironment

Cancer Immunol Immunother682012112019

10.1007/s00262-018-2263-4

30357491

Yearley

Gibson

Moon

Murphy

Juco

Lunceford

Cheng

Chow

LQM

Seiwert

PD-L2 Expression in Human Tumors: Relevance to Anti-PD-1 Therapy in Cancer

Clin Cancer Res23315831672017

10.1158/1078-0432.CCR-16-1761

28619999

Amin

Edge

Greene

AJCC Cancer Staging Manual

8th edition

Springer

New York

2017

10.1007/978-3-319-40618-3

Travis

Brambilla

Burke

Marx

Nicholson

WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart

4th edition

International Agency for Research on Cancer

Lyon, France

2015

Sumitomo

Hirai

Fujita

Murakami

Otake

Huang

M2 tumor-associated macrophages promote tumor progression in non-small-cell lung cancer

Exp Ther Med18449044982019

31777551

Tsao

Kerr

Kockx

Beasley

Borczuk

Botling

Budendorf

Chirieac

Chen

Chou

PD-L1 immunohistochemistry comparability study in real-life clinical samples: Results of Blueprint print phase 2 project

J Thorac Oncol13130213112018

10.1016/j.jtho.2018.05.013

29800747

Maeda

Yoshida

Umakoshi

Nanjo

Shiraishi

Saito

Kohno

Konno

Saito

The intratumoral distribution influences the prognostic impact of CD68- and CD204-positive macrophages in non-small cell lung cancer

Lung Cancer1231271352018

10.1016/j.lungcan.2018.07.015

30089583

Tanaka

Ichimura

Kubota

Saito

Nakamura

Ishitsuka

Watanabe

Fujisawa

Mizuno

Takahashi

Differential involvement of programmed cell death ligands in skin immune responses

J Invest Dermatol142145154.e82022

10.1016/j.jid.2021.06.026

34310947

Sanmamed

Chen

Inducible expression of B7-H1 (PD-L1) and its selective role in tumor site immune modulation

Cancer J202562612014

10.1097/PPO.0000000000000061

25098285

Chen

Feng

Wang

Dai

Zhang

Interferon-γ-induced PD-L1 surface expression on human oral squamous carcinoma via PKD2 signal pathway

Immunobiology2173853932012

10.1016/j.imbio.2011.10.016

22204817

Lane

Femel

Breazeale

Loo

Thibault

Kaempf

Mori

Tsujikawa

Chang

Lund

IFNγ-activated dermal lymphatic vessels inhibit cytotoxic T cells in melanoma and inflamed skin

J Exp Med215305730742018

10.1084/jem.20180654

30381467

Qian

Wang

Yang

Wang

Luo

Zhao

Liu

Chu

The IFN-γ/PD-L1 axis between T cells and tumor microenvironment: Hints for glioma anti-PD-1/PD-L1 therapy

J Neuroinflammation152902018

10.1186/s12974-018-1330-2

30333036

Chen

Crabill

Pritchard

McMiller

Wei

Pardoll

Pan

Topalian

Mechanisms regulating PD-L1 expression on tumor and immune cells

J Immunother Cancer73052019

10.1186/s40425-019-0770-2

31730010

Arrieta

Montes-Servin

Hernandez-Martinez

Cardona

Cases-Ruiz

Crispin

Motola

Flores-Estrada

Barrera

Expression of PD-1/PD-L1 and PD-L2 in peripheral T-cells from non-small cell lung cancer patients

Oncotarget81019941020052017

10.18632/oncotarget.22025

29254220

Kitsou

Ayiomamitis

Zaravinos

High expression of immune checkpoints is associated with the TIL load, mutation rate and patient survival in colorectal cancer

Int J Oncol572372482020

10.3892/ijo.2020.5062

32468013

Zhang

Hua

Liu

Liang

Meng

Wei

Zhang

A PD-L2-based immune marker signature helps to predict survival in resected pancreatic ductal adenocarcinoma

J Immunotherapy Cancer72332019

10.1186/s40425-019-0703-0

31464648

Mei

Xiao

Guo

Liu

Lin

Liao

You

Liu

Prognostic impact of tumor-associated macrophage infiltration in non-small cell lung cancer: A systemic review and meta-analysis

Oncotarget734217342282016

10.18632/oncotarget.9079

27144518

Jackute

Zemaitis

Pranys

Sitkauskiene

Miliauskas

Vaitkiene

Sakalaukas

Distribution of M1 and M2 macrophages in tumor islets and stroma in relation to prognosis of non-small cell lung cancer

BMC Immunol1932018

10.1186/s12865-018-0241-4

29361917

Mantovani

Sozzani

Locati

Allavena

Sica

Macrophage polarization: Tumor-associated macrophages as a paradigm for polarized M2 mononuclear phagocytes

Trends Immunol235495552002

10.1016/S1471-4906(02)02302-5

12401408

Gabrusiewicz

Wei

Hashimoto

Marisetty

Ott

Wang

Hawke

Healy

Glioblastoma stem cell-derived exosomes induce M2 macrophages and PD-L1 expression on human monocytes

Oncoimmunology7e14129092018

10.1080/2162402X.2017.1412909

29632728

Wen

Liu

Gao

Zhou

Zhang

Zhao

Chen

Zhang

Huang

Tumor cell-released autophagosomes (TRAPs) promote immunosuppression through induction of M2-like macrophages with increased expression of PD-L1

J Immunother Cancer61512018

10.1186/s40425-018-0452-5

30563569

Zhang

Zeng

Zhu

Liu

Ning

Zeng

Zhang

Chen

Huang

PD-L1 induced by IFN-γ from tumor-associated macrophages via the JAK/STAT3 and PI3K/AKT signalling pathways promoted progression of lung cancer

Int J Clin Oncol22102610332017

10.1007/s10147-017-1161-7

28748356

Tsukamoto

Imai

Ishimoto

Komohara

Yamashita

Nakagawa

Umezaki

Yamao

Kitano

Miyata

PD-L1 expression enhancement by infiltrating macrophage-derived tumor necrosis factor-α leads to poor pancreatic cancer prognosis

Cancer Sci1103103202019

30426611

Lai

Wahyuningtyas

Aui

Chang

Autocrine VEGF signalling on M2 macrophages regulates PD-L1 expression for immunomodulation of T cells

J Cell Mol Med23125712672019

30456891

Lin

Liu

Chen

Jiang

Chen

Zhang

Tumor-associated macrophage-derived CXCL8 determines immune evasion through autonomous PD-L1 expression in gastric cancer

Gut68176417732019

10.1136/gutjnl-2018-316324

30661053

Horlad

Yano

Pan

Ohnishi

Fujiwara

Endo

Kikukawa

Okuno

Matsuoka

An IL-27/Stat3 axis induces expression of programmed death ligands (PD-L1/2) on infiltrating macrophages in lymphoma

Cancer Sci107169617042016

10.1111/cas.13065

27564404

Cai

Yuan

Zhu

Yang

Tang

Cong

Glioma-associated stromal cells stimulate glioma malignancy by regulating the tumor immune microenvironment

Front Oncol116729282021

10.3389/fonc.2021.672928

33996602

Wang

Cang

Guo

Crosstalk between cancer and immune cells: Role of tumor-associated macrophages in the tumor microenvironment

Cancer Med8470947212019

10.1002/cam4.2327

31222971

Menguy

Prochazkova-Carlotti

Beylot-Barry

Saltel

Vergier

Merlio

Pham-Ledard

PD-L1 and PD-L2 are differentially expressed by macrophages or tumor cells in primary cutaneous diffuse large B-cell lymphoma, Leg type

Am J Surg Pathol423263342018

10.1097/PAS.0000000000000983

29112015

Pinato

Vallipuram

Evans

Wong

Zhang

Brown

Dina

Trivedi

Akarca

Marafioti

Programmed cell death ligand expression drives immune tolerogenesis across the diverse subtypes of neuroendocrine tumors

Neuroendocrinology1114654742021

10.1159/000506745

32097935

Wan

Song

Zhan

Zhang

Liu

The clinicopathological and prognostic significance of PD-L1 expression assessed by immunohistochemistry in lung cancer: A meta-analysis of 50 studies with 11,383 patients

Transl Ling Cancer Res84294492019

10.21037/tlcr.2019.08.04

31555517

Hua

Qiu

Hao

Zou

Guo

Chen

Sui

PD-L1 promotes tumor growth and progression by activating WIP and β-catenin signaling pathways and predicts poor prognosis in lung cancer

Cell Death Dis115062020

10.1038/s41419-020-2701-z

32632098

Sato

Shames

Hasegawa

Emerging evidence of epithelial-to-mesenchymal transition in lung carcinogenesis

Respirology17104810592012

10.1111/j.1440-1843.2012.02173.x

22452538

Brabletz

Schuhwerk

Brabletz

Stemmler

Dynamic EMT: A multi-tool for tumor progression

EMBO J40e1086472021

10.15252/embj.2021108647

34459003

Inoue

Yoshimura

Nishimoto

Inui

Karayama

Yasui

Hozumi

Suzuki

Furuhashi

Fujisawa

Evaluation of programmed death ligand 1 (PD-L1) gene amplification and response to nivolumab monotherapy in non-small cell lung cancer

JAMA Netw Open3e20118182020

10.1001/jamanetworkopen.2020.11818

32955570

Figure 1.

Immunostaining of lung cancer. (A) A squamous cell carcinoma with a high density of TIL and (B) positive expression of PD-L1 on TCs. (C) A squamous cell carcinoma with a high density of TIL and (D) positive expression of PD-L1 on ICs. (E) An adenocarcinoma with a high density of TIL and (F) positive expression of PD-L2 on TCs. (G) A squamous cell carcinoma with a high density of M2 TAM and (H) positive expression of PD-L2 on ICs. Scale bar=100 µm. TIL, tumor-infiltrating lymphocyte; PD-L, programmed death-ligand; TCs, tumor cells; ICs, tumor-infiltrating immune cells; TAM, tumor-associated macrophage.

Figure 2.

Association between TIL and M2 TAM densities. TIL, tumor-infiltrating lymphocyte; TAM, tumor-associated macrophage. TIL, tumor-infiltrating lymphocyte; TAM, tumor-associated macrophage.

Figure 3.

PD-L1 expression on (A) TCs and (B) ICs with respect to TIL density. PD-L1 expression on (C) TCs and (D) ICs with respect to M2 TAM density. PD-L1 expression on (E) TCs and (F) ICs with respect to TIL and M2 TAM densities. PD-L2 expression on (G) TCs and (H) ICs with respect to TIL density. PD-L2 expression on (I) TCs and (J) ICs with respect to M2 TAM density. PD-L2 expression on (K) TCs and (L) ICs with respect to TIL and M2 TAM densities. PD-L, programmed death-ligand; TCs, tumor cells; ICs, tumor-infiltrating immune cells; TAM, tumor-associated macrophage; TIL, tumor-infiltrating lymphocyte.

Figure 4.

(A) Correlation between PD-L1 expression on TCs and PD-L2 expression on TCs. (B) Correlation between PD-L1 expression on TCs and PD-L1 expression on ICs. (C) Correlation between PD-L2 expression on TCs and PD-L2 expression on ICs. PD-L, programmed death-ligand; TCs, tumor cells; ICs, tumor-infiltrating immune cells.

Figure 5.

PD-L1 expression on (A) TCs and (B) ICs with respect to tumor differentiation. PD-L2 expression on (C) TCs and (D) ICs with respect to tumor differentiation. TCs, tumor cells; ICs, tumor-infiltrating immune cells. PD-L, programmed death-ligand; TCs, tumor cells; ICs, tumor-infiltrating immune cells.

Table I.

Distributions of TIL and M2 TAM density among NSCLC patients according to clinicopathological characteristics.

			TIL				M2 TAM

		n	low	high	Mean ± standard deviation	P-value	low	high	Mean ± standard deviation	P-value
Smoking
Non-smoker		90	37	53	981.2±931.3	0.6139^a	59	31	357.7±362.0	0.3782^a
Smoker		85	34	51	913.0±849.6		49	36	408.7±402.4
Tumor status
T0		9	6	3	712.7±854.2	0.3270^b	8	1	205.3±308.8	0.1526^b
T1		84	40	44	874.8±847.4		57	27	353.7±378.5
T2, T3, T4		82	25	57	1048.9±934.9		43	39	431.4±387.6
Nodal status
N0		137	61	76	890.2±884.1	0.1032^a	88	49	346.2±337.1	0.0165^a,c
N1, N2, N3		38	10	28	1156.5±894.3		20	18	513.4±495.7
Pathological stage
0		8	6	2	641.6±884.3	0.0632^b	8	0	107.5±102.7	0.0388^b,c
I		111	51	60	842.8±816.2		71	40	358.2±352.2
II		26	7	19	1264.0±1190.4		12	14	517.9±427.0
III		30	7	23	1145.4±790.0		17	13	428.2±450.5
Differentiation
Well		35	18	17	765.4±717.5	0.0130^b,c	27	8	252.5±270.9	0.0015^b,c
Moderately		105	46	59	881.0±864.8		68	37	363.7±361.7
Poorly		35	7	28	1331.8±1028.3		13	22	568.8±467.5
Histology
Ad		141	61	80	857.5±801.0	0.0206^b,c	98	43	336.1±356.2	0.0036^b,c
Sq		27	9	18	1356.1±1244.2		10	17	594.2±474.7
La		7	1	6	1198.6±556.9		0	7	499.9±154.0
Total number of patients	175	71	104	948.1±890.6		108	67	382.5±381.9

P-value determined using a t-test.

P-value determined using ANOVA followed by a Bonferroni/Dunn test.

P<0.05. TIL, tumor-infiltrating lymphocyte; TAM, tumor-associated macrophage; NSCLC, non-small-cell lung cancer; Ad, adenocarcinoma; Sq, squamous cell carcinoma; La, large cell carcinoma.

Table II.

Distributions of PD-L1 and PD-L2 expressions among NSCLC patients according to clinicopathological characteristics.

			PD-L1 expression on TCs				PD-L1 expression on ICs				PD-L2 expression on TCs				PD-L2 expression on ICs

		n	<1%	1-49%	≥50%	P-value	<1%	1-9%	≥10%	P-value	<1%	1-49%	≥50%	P-value	<1%	1-9%	≥10%	P-value
Smoking
Non-smoker		90	50	25	15	0.5946^a	27	32	31	0.9917^a	30	47	13	0.1596^a	24	35	31	0.7427^a
Smoker		85	51	18	16		25	31	29		40	37	8		26	34	25
Tumor status
T0		9	6	2	1	0.1414^a	5	3	1	0.0104^a,b	2	7	0	0.0218^a,b	2	1	6	0.1569^a
T1		84	56	15	13		25	38	21		26	44	14		22	34	28
T2, T3, T4		82	39	26	17		22	22	38		42	33	7		26	34	22
Nodal status
N0		137	84	33	20	0.0848^a	46	51	40	0.0166^a,b	56	64	17	0.8082^a	44	49	44	0.0852^a
N1, N2, N3		38	17	10	11		6	12	20		14	20	4		6	20	12
Pathological stage
0		8	6	2	0	0.3067^a	5	3	0	0.0027^a,b	2	6	0	0.6336^a	2	1	5	0.2355^a
I		111	69	25	17		38	43	30		46	50	15		36	41	34
II		26	13	8	5		4	10	12		12	12	2		6	14	6
III		30	13	8	9		5	7	18		10	16	4		6	13	11
Differentiation
Well		35	25	9	1	0.0002^a,b	12	17	6	<0.0001^a,b	7	22	6	0.0260^a,b	8	10	17	0.0326^a,b
Moderately		105	63	27	15		36	39	30		47	46	12		34	39	32
Poorly		35	13	7	35		4	7	24		16	16	3		8	20	7
Histology
Ad		141	84	40	17	0.0001^a,b	45	56	40	0.0173^a,b	53	70	18	0.6431^a	42	59	40	0.2096^a
Sq		27	15	3	9		6	6	15		13	11	3		6	7	14
La		7	2	0	5		1	1	5		4	3	0		2	3	2
Total number	175	101	43	31		52	63	60		70	84	21		50	69	56

P-value determined using a χ² test.

statistical significance <0.05. NSCLC, non-small-cell lung cancer; TCs, tumor cells; ICs, tumor-infiltrating immune cells; Ad, adenocarcinoma; Sq, squamous cell carcinoma; La, large cell carcinoma.