Abstract
Background/Aim: Programmed cell death ligand 1 (PD-L1) is an immune checkpoint protein involved in immune evasion of malignant tumors. Confirmation of PD-L1 expression in non-small cell lung cancer (NSCLC) is necessary for the determination of immunotherapy using immune checkpoint inhibitors (ICIs). PDL-1 expression is currently analyzed by immunohistochemistry and is the only available biomarker that can guide the treatment of NSCLC using ICIs. The present study was conducted to compare the expression of three different commercial clones of PD-L1 in order for immunohistochemistry (IHC) for these clones to become more reliable for surgical pathologists. Materials and Methods: This study examined the expression of PD-L1 in 76 cases of resected lung cancer using IHC. Three clones were examined: SP263, SP142, and 22C3PharmDx, which are commercially approved for quantifying PD-L1 expression in lung cancer. Results: Of the 76 patients whose samples were evaluated for PD-L1 using the IHC 22C3pharmDx assay, 19 (25.0%) had a tumor proportion score (TPS) of ≥50% and 41 (53.9%) had a PD-L1 TPS of ≥1%. Furthermore, using the SP263, 48.7% had a TPS of ≥1% and 18.4% of >50%. The SP142 assay was used to evaluate tumor cells (TCs) and immune cells (ICs). Twenty (26.3%) cases were positive for TCs and 25 (32.9%) were reactive for ICs. Conclusion: These three commercial PD-L1 clones are comparable for detecting primary targets for anti-tumor immunotherapies. Careful evaluation by a pathologist is necessary to minimize misinterpretation errors.
Lung cancer is the most common type of cancer worldwide. In 2018, there were 1.8 million new cases, estimated to be responsible for 1.6 million deaths, worldwide. It remains the leading cause of cancer-related deaths (1). Immunotherapy has emerged as an important and effective therapeutic modality for lung cancer, especially non-small cell lung cancer (NSCLC), in the past few years. The latest immunotherapy uses monoclonal antibodies (mAbs) against immune checkpoint molecules, such as receptor programmed cell death 1 (PD-1) or its ligand (PD-L1). Immune checkpoint inhibitors (ICIs), anti-PD-1 and anti-PD-L1, have been approved for the treatment of malignant tumors, including NSCLC. PD-L1, also known as CD274 and B7-H1, is a major immune checkpoint protein that promotes anti-tumor immunosuppression (2). The importance of PD-L1-based immunosuppression is underscored by the advent of anti-PD-1/PD-L1 immunotherapy, which has moved to the front-most line of cancer treatment (3, 4). A major inducer of PD-L1 expression in vivo is interferon-gamma (IFN-γ) released by CD8+ T cells (5). PD-L1 expression is observed on the surface of various cells, including macrophages, antigen-presenting cells (APC), B and T lymphocytes, and epithelial, muscle, and endothelial cells (6), whereas the PD-1 receptor is mainly expressed by activated cytotoxic T cells. PD-L1 ligand binds to the PD-1 receptor on activated T cells and this linkage suppresses the immune system (6, 7). The link between PD-1 and PD-L1 prevents autoimmune responses in peripheral tissues during inflammation (8). This binding also inhibits interleukin-2 production and T-cell activation through the reduced phosphorylation of chain-associated protein kinase 70 and protein kinase C-θ (9).
ICIs have recently been developed to restore T cell cytotoxicity (10), primarily by targeting the PD-1/PD-L1 axis, and include PD-1 inhibitors such as nivolumab (Opdivo®, Bristol-Myers Squibb Co., New York, NY, USA) and pembrolizumab (Ketruda®, Merck & Co., New York, NY, USA) and PD-L1 inhibitors such as atezolizumab (Tecentriq®, Genentech, San Francisco, CA, USA), durvalumab (Imfinzi®, Astra-Zeneca, Cambridge, UK), and avelumab (Bavencio®, EMD Serono, Rockland, MA, USA).
In routine practice, immunohistochemistry (IHC) analysis of PD-L1 expression is essential to determine a patient’s eligibility for PD-1/PD-L1 immunotherapy, and quantitative detection of its expression may be useful for monitoring the therapeutic response. Additionally, preliminary studies using IHC in formalin-fixed and paraffin-embedded (FFPE) tissue samples showed that PD-L1 expression in human cancers could predict clinical responses to PD-1/PD-L1 immunotherapy (2, 11-13). Therefore, it is essential to establish a stable IHC system that can reliably detect real PD-L1 (+) cases in pathology laboratories.
In the present study, we aimed to compare three different clones of PD-L1 for the positive rates in NSCLC and identify which clones can be reliably used by the surgical pathologist to evaluate IHC PD-L1 expression on FFPE specimens. Of the three assays, SP263 and SP142 were run on the Ventana platform and 22C3 was run on the Dako platform. In addition, we analyzed the relationship between PD-L1 expression and clinicopathological factors to determine its prognostic and predictive value.
The study was approved by the Ethics Committee of Chosun University Hospital (Institutional Review Board of Chosun University Hospital, Gwangju, Republic of Korea), which waived the requirement for written informed consent because of the nature of the study (IRB no.: 2020-11-032).
Materials and Methods
Case selection. Among the patients who underwent lobectomy or segmentectomy under the diagnosis of lung cancer at Chosun University Hospital from February 2018 to May 2022, 76 with well-preserved medical records were selected discontinuously.
Histopathological analysis.
Microscopic examination. Patients’ clinical records and tissue slides were retrospectively analyzed. For observation, paraffin-embedded tissues fixed in 10% neutral buffered formalin were cut to a thickness of 4-5 μm, and H&E staining was performed to prepare slides. A slide for IHC was prepared by selecting a block containing a representative site corresponding to the purpose of the study. We reviewed each H&E slide and evaluated the TNM stage, lymphovascular invasion, pleural invasion, and histological classification of the tumor.
Immunohistochemistry. PD-L1 expression in three different clones (SP263, SP142, and 22C3 pharmDx) was evaluated using IHC. Representative images of IHC of PD-L1 expression are shown in Figure 1. Information on the three commercially available Abs utilized in this study is summarized in Table I. Two assays, SP263 and SP142, were performed on the Ventana platform. IHC was performed using a benchmark ULTRA (Ventana Medical Systems, Roche group, Tucson, AZ, USA). The expression and amplification of each clone of PD-L1 were performed using the OptiView DAB IHC detection kit and OptiView Amplification Kit. Immunolocalization of PD-L1 was performed using a secondary antibody and multimeric anti-hapten-HRP conjugate. Specific Ab-enzyme complexes were visualized using an enzymatic reaction product. Each antibody was stained with positive and negative controls, normal tonsil and placental tissue were used as a positive and negative control of SP142 and SP263 clone of PD-L1 staining, respectively. The 22C3PharmDx assay was developed for use on the Dako platform, which was not set in our laboratory. We used the results of PD-L1 IHC 22C3pharmDx by Seegen Medical, which is a qualitative IHC assay using monoclonal mouse anti-PD-L1, clone 22C3, intended for use in the detection of PD-L1 protein in FFPE NSCLC tissues using the EnVision FLEX visualization system on Autostainer Link 48 (Agilent Technologies, Santa Clara, CA, USA).
Determination of immunohistochemical staining. The staining results were evaluated by a pathologist who was blinded to the clinical course of the patient. SP142 PD-L1 was read separately for immune cells (ICs) and tumor cells (TCs). ICs were read as positive for staining at ≥5%, with a dark brown punctate staining pattern in intratumoral and peritumoral ICs (lymphocytes, macrophages, dendritic cells, and granulocytes). TCs were stained with a dark brown cell membrane pattern and read as positive for staining at ≥5%. SP263 PD-L1 expression was scored as the extent of membranous staining (negative, 0; <5%, 1; 5-50%, 2; ≥50%, 3). 22C3 PD-L1 protein expression was determined using the tumor proportion score (TPS), which is the percentage of viable tumor cells showing partial or complete membrane staining at any intensity. The specimen was considered to have low PD-L1 expression if the TPS ≥1% and high PD-L1 expression if the TPS ≥50%.
Statistical analysis. Statistical analyses were performed using the STAT View software package (Abacus Concepts, Berkeley, CA, USA). The expression of three types of PD-L1 clones was examined: SP-142 in TC and IC, SP-263, and 22C3. The correlation between the different clones and that between various clinicopathological factors and PD-L1 clone expression were analyzed using the χ2-test. p<0.05 was considered to indicate a statistically significant difference.
Results
Clinical and pathological parameters. The clinicopathological characteristics of the patients are summarized in Table II. The mean age was 67 years (range=40-84 years), and the male-to-female ratio was 54:22, with a male predominance. Seventy-six cases of NSCLC comprised adenocarcinoma (n=55), squamous cell carcinoma (n=16), and other non-small cell carcinomas (n=5). The tumor stage was pT1 in 45 patients (59.2%), T2 in 19 (25.0%), pT3 in 10 (13.2%), and pT4 in 2 (2.6%). Lymph node metastasis was observed in 13 (17.3%) patients. EGFR mutations were examined in 71 of 76 cases; 22 cases showed EGFR mutations, such as L858R, exon19 deletion, and L861Q.
Expression of PD-L1 assays. Figure 2 shows the representative staining pattern for each antibody. For the SP263 clone of PD-L1, 37 cases showed immunoreactivity, 13 cases (17.1%) had a score of 1 (<5%), 10 cases (13.2%) had a score of 2 (5-50%), and 14 cases (18.4%) had a score of 3 (≥50%). SP142 PD-L1 was scored in TCs and ICs in 20 cases (26.3 %) and 25 cases (32.9%), respectively. We reclassified cases showing expression in TCs or ICs as positive (32 cases, 42.1%). For 22C3 pharmDx PD-L1, 41 cases (53.9%) showed reactivity, 22 cases (28.9%) showed low expression (≥1%), and 19 cases had high expression (≥25.0%). Clinically, high expression of 22C3 pharmDx is utilized as a drug cutoff value, and we classified the data separately according to modified 22C3 scores, such as no and low expression (scored as negative, 57 cases, 75.0%) and high expression (as positive, 19 cases, 25.0%). We also reclassified cases showing expression in TCs or ICs as positive (Table III).
Relationships between different clones of PD-L1 expression and clinicopathologic parameters. We examined the expression patterns of PD-L1 using three mAbs (SP263, SP142, and 22C3) and the relationship between PD-L1 expression and various clinicopathologic factors (Table IV). Of several clinicopathologic factors, histologic type and EGFR mutation were related to the high expression levels of SP142 PD-L1 in ICs (p=0.02, 0.02, respectively).
Comparison of PD-L1 expression as detected by three mAbs (SP263, SP142, and 22C3). The expression levels of PD-L1 in all assays were significantly correlated (Table V, Table VI, and Table VII). The expression of the SP263 clones was significantly associated with that of the SP142 clone in TCs and ICs (p<0.001). The expression in the SSC3 assay also showed a significant correlation with that of the SP142 clone in TCs (p<0.001) and ICs (p=0.003). The correlation between SP263 PD-L1 expression and that in the 22C3 assay (p<0.001) and between SP142 PD-L1 expression in TCs and ICs (p<0.001) were also statistically significant.
Discussion
Cancer immune surveillance is thought to be an important host-protective process that prevents cancer development and maintains cellular homeostasis (14). In the body’s immune system, T cells are activated by recognizing the MHC-antigen complex through T cell receptors and regulating T cell activation through the regulation of co-stimulatory and co-inhibitory signals (15). However, in cancer cells, the immune evasion mechanism is activated, which defends against the immune attack of T cells by expressing a protein that prevents the destruction of cancer cells (14). Historically, lung cancer was considered non-immunogenic due to loss of tumor antigens, loss of sensitivity to complement, and lysis of T cells and natural killer cells (16). However, recent studies have shown that lung cancers can develop mechanisms that evade the attention of the immune system for development and progression (17), such as the loss of MHC antigen expression, secretion of immunosuppressive cytokines, induction of immunosuppressive T cells, and failure of immune competence due to the disease itself or treatment in the later stages of the disease (18). Finally, a complex set of immune checkpoint regulators appears to be important for fine-tuning the immune response, a process that tumors may adopt to protect against immune attacks (18).
One of the first inhibitory immune checkpoints targeted was cytotoxic T lymphocyte-associated protein 4 (CTLA-4), an immune checkpoint receptor expressed on the surface of T cells interacting with B7 molecules on the APCs as ligands (18). However, recently, much attention has been paid to how this interaction is blocked in lung cancers such as NSCLC by immunomodulatory therapies, such as PD-1 and its ligand PD-L1, and therapeutic monoclonal antibodies specific for PD-1 or PD-L1 (19).
Cancer cells express PD-L1 as a cell surface protein, which binds to PD-1 expressed on T cells and suppresses the antitumor immunity of T cells (15). Immunotherapy induces immune cells to selectively attack cancer cells by injecting artificial immune proteins, including ICIs (CTLA-4, PD-1, and PD-L1 inhibitors), immune cell therapy agents, and immune virus therapeutic agents, into the body to stimulate the immune system. Immune anticancer drugs improve the side effects of first-generation anticancer drugs and overcome the resistance to second-generation targeted anticancer drugs (15, 20). Owing to these advantages, interest in immunotherapy is increasing, and after the autologous tumor vaccine Provenge® (Sipuleucel-T) was approved as a treatment for prostate adenocarcinoma in 2010, several drugs that suppress immune checkpoints have been approved (15).
Recently, to target the PD-1/PD-L1 signaling pathway, PD-1 and PD-L1 inhibitors have been developed and approved to induce T-cell apoptosis against cancer cells (21). Keytruda® (pembrolizumab), the first PD-1 inhibitor approved by the US Food and Drug Administration (FDA) in 2014, was approved for the treatment of metastatic melanoma and NSCLC in 2015 (22). In Korea, it has been approved for the treatment of metastatic melanoma, NSCLC, lymphoma, squamous cell carcinoma (SqCC) of the head and neck, and urothelial cancer (UC) (23). Opdivo® (nivolumab) was approved by the US FDA in 2015 as an inhibitor of PD-1 (22). In Korea, it has since been approved for the treatment of metastatic melanoma, NSCLC, lymphoma, SqCC of the head and neck, and UC (23).
Following the development of PD-1 inhibitors, inhibitors of PD-L1 expression in cancer cells have also been developed. The first PD-L1 inhibitor, Tecentriq® (atezolizumab), was approved by the US FDA in 2016 (22). In 2017, it was approved for the treatment of locally advanced or metastatic NSCLC and UC in Korea (23). In the same year, a second PD-L1 inhibitor, Imfinzi® (durvalumab), was approved by the FDA for the treatment of severe bladder cancer that progresses after surgery or chemotherapy (22). These immunotherapeutic agents that inhibit the interaction between PD-1 of immune cells and PD-L1 in cancer cells have obtained good clinical results for the treatment of early melanoma, and the indications are expanding to various cancer treatments, such as lymphoma (20).
The assessment of PD-L1 expression is now routinely used in clinical practice, but it is not a simple task as PD-L1 exhibits intratumoral heterogeneity (24). IHC for detecting protein activity can be influenced by the selection of various factors, including primary antibody clones, detection systems, and platforms involving complex biochemistry (24). It is important to select the correct cutoff level to define the PD-L1 (+) and PD-L1 (−) groups (24).
Pathologists must understand the unique nature of the PD-L1 test to correctly interpret PD-L1 IHC results and communicate with clinicians to recommend the most effective treatment options (24). Currently, three commercially available antibodies are available at our institution for measuring PD-L1 protein expression in FFPE lung tissue specimens (SP263, SP142, and 22C3 assays). Each assay uses different automated staining systems, detection systems, and assessment means and thresholds to determine positive PD-L1 protein expression. The 22C3 pharmDx assay has been developed for use on the Dako platform, whereas SP263 and SP142 use the Ventana BenchMark platforms, which are more common in pathology laboratories. Regarding the interpretation of PD-L1 staining, in the 22C3PharmDx assay, complete circumferential or partial linear membrane staining of cancer cells, regardless of staining intensity, is considered positive. In the SP263 and SP142 assays, any membranous and/or cytoplasmic expression in cancer cells is considered positive (24). In the SP142 analysis, expression in immune and cancer cells is included as a positive criterion.
There are many issues related to the interpretation of PD-L1, which are important for pathologists, such as distinguishing the immunoreactivity of tumor cells from that of inflammatory cells, the localization of immunostaining (membrane or cytoplasm), and the scoring of expression percentages, particularly around the thresholds of clinical significance, regardless of the assay type (24). Additionally, each assay has a specific cutoff value for positive tumor cells, and even the same drug may have different percentages depending on whether it is a first-line, second-line, or add-on treatment. In addition, when an assay is applied for two-drug prescriptions, such as SP263, different cutoffs may be applied depending on the drug (durvalumab and nivolumab) (24). Pathologists should also pay close attention to insurance policies related to PD-L1 testing, as these inconsistent cutoffs have not been proven in clinical trials and are a special circumstance related to Korea’s national insurance (24).
Several international projects have been initiated to standardize various PD-L1 assays, and among them, the result of the Blueprint project led by “the International Association for the Study of Lung Cancer”is attracting attention.
In this study, clones SP263, 22C3PharmDx, and 28-8 showed high agreement in the staining ratio of tumor cells regardless of the staining intensity. Conversely, when the SP142 clone was used, low expression of PD-L1 was observed in tumor cells (25). In the present study, we analyzed PD-L1 expression and clinical characteristics of Korean patients with NSCLC. Of the 76 patients whose samples were evaluated for PD-L1 using the IHC 22C3 pharmDx assay, 19 (25.0%) had a PD-L1 TPS of ≥50%, and 41 (53.9%) had a PD-L1 TPS of ≥1%. Using SP263, 48.7% had a TPS of ≥1% and 18.4% had a TPS of >50%. The SP142 assay was evaluated for TC and IC. Twenty (26.3%) cases were positive for TC and 25 (32.9%) were reactive for IC.
The interpretation and reporting format for PD-L1 IHC analysis in lung cancer are different from those of conventional IHC assays, in that each assay should be interpreted according to the relevant criteria for each drug. Therefore, specialized training programs for pathologists are required to maintain consistency and quality of interpretation (24). Moreover, closer collaboration between oncologists and pathologists is needed to ensure that each patient receives the most appropriate treatment.
Acknowledgements
This study was supported by research funds from Chosun University, Republic of Korea, 2020.
Footnotes
Authors’ Contributions
Conceptualization: RH, SGP. Pathologic methodology: RH. Data analysis: SBL, HJL. Figures: RH, SBL, HJL. Funding acquisition: RH. Supervision: RH, SGP. Writing (original and review and editing): RH, SGP.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
- Received October 14, 2022.
- Revision received October 26, 2022.
- Accepted November 7, 2022.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).