Abstract
Background/Aim: Pulmonary pleomorphic carcinoma (PPC) is a rare and aggressive tumor that is resistant to treatment. The expression and prognostic value of programmed cell death-ligand 1 (PD-L1) and its association with epithelial–mesenchymal transition (EMT) in PPC remains unclear. Patients and Methods: The expression of PD-L1 and EMT markers, such as E-cadherin, vimentin, zinc finger E-box-binding homeobox 1 (ZEB-1), and cellular mesenchymal–epithelial transition (c-Met) was evaluated by immuno - histochemistry in 16 patients with PPC who underwent surgical resection. Results: The expression of PD-L1 varied between carcinomatous and sarcomatous areas. Positive correlations between PD-L1 and vimentin expression in carcinomatous areas (r=0.668, p=0.005) and PD-L1 and ZEB-1 expression in sarcomatous areas (r=0.562, p=0.023) were found. High PD-L1 and ZEB-1 expression in sarcomatous areas predicted poor survival (p=0.045 and p=0.012, respectively). Conclusion: PD-L1 expression associated with ZEB1 expression in the sarcomatoid component of patients with PPC may be useful for predicting patient prognosis.
Pulmonary pleomorphic carcinoma (PPC) is a type of poorly differentiated non-small cell lung carcinoma (NSCLC), including squamous cell carcinoma, adenocarcinoma, or undifferentiated NSCLC. PPC contains ≥10% sarcomatous components, such as spindle and giant cells, and is categorized as a major subtype of pulmonary sarcomatoid carcinoma (PSC) (1). PSC is a rare neoplasm accounting for approximately <1% of all lung cancers. It is associated with poorer prognosis than other subtypes of NSCLC without any sarcomatous component due to its great aggressiveness and frequent development of chemoresistance (2, 3).
Immune-checkpoint inhibitors, including anti-programmed cell death-1 (anti-PD-1) and anti-programmed cell death-ligand 1 (anti-PD-L1), as monotherapies or in combination with chemotherapies are currently a standard of care for patients with NSCLC (4-7). PD-L1 expression is used as a predictive marker for the efficacy of treatments, despite its incompleteness (4, 8-10). Previous studies using immunohistochemistry (IHC) showed dramatic responses to anti-PD-1 or anti-PD-L1 therapies in patients with PPC, whose tumors highly expressed PD-L1 (11-14). Indeed, PPC tends to show high levels of PD-L1 expression compared with the average NSCLC tumors (15-17). Based on the evidence obtained from these studies, high PD-L1 expression is recognized as an important predictive biomarker for response to anti-PD-1/PD-L1 therapy in patients with PPC. However, potential differences in PD-L1 expression between the carcinomatous and sarcomatous components in PPC have not been determined. In addition, the prognostic value of PD-L1 expression in PPC remains controversial.
Epithelial–mesenchymal transition (EMT) is a biological process through which epithelial cells acquire the properties of mesenchymal cells. In cancer, EMT is involved in tumor initiation, invasion, metastasis, and resistance to therapy, thus leading to poor prognosis (18). It plays a pivotal role in tumor immunosuppression and immune evasion, resulting in cancer progression (18, 19). Recently, an association between the EMT status and PD-L1 expression was identified in various types of cancer, including lung adenocarcinoma (19, 20). In sarcomatoid carcinoma, it is suggested that the mesenchymal sarcomatous and epithelial carcinomatous components are derived from a common ancestor, and the former originates from the latter through EMT (21, 22). However, a correlation between the EMT status and PD-L1 expression in PPC remains elusive.
Cellular mesenchymal–epithelial transition (c-Met) factor belongs to the family of receptor tyrosine kinases that is encoded by the MET proto-oncogene. MET abnormalities contribute to an immunosuppressive tumor microenvironment (23). Positive correlation between MET anomalies and PD-L1 expression in several types of cancer, including lung cancer, has been reported (24-26). Recently, MET exon 14 skipping mutation, an essential therapeutic target of MET inhibitors, is more frequently detected in PSC than in the average NSCLC tumors (27). However, a correlation between c-Met and PD-L1 expression in PPC has not been shown. Moreover, the prognostic value of c-Met protein expression in PPC has not been clarified yet.
In this study, we evaluated the expression levels of PD-L1, EMT-related markers, and c-Met in both carcinomatous and sarcomatous areas of patients with PPC. The objective of this study was to investigate the role of PD-L1 as a prognostic marker.
Patients and Methods
Patients and tumor samples. Consecutive patients with PPC who underwent surgical resection at Nippon Medical School Hospital (Tokyo, Japan) between January 2003 and December 2017 were included in this study. All the clinical data were retrieved from the medical records of patients. The pathological stage was classified according to the 8th edition of the TNM classification developed by the International Union Against Cancer. PPC was diagnosed according to the 2015 World Health Organization guidelines (1). The surgical specimens were fixed with formalin and embedded in paraffin. Specimens were stained using hematoxylin and eosin. Representative slides containing the tumors with both carcinomatous and sarcomatous components on maximum cut surface in each case were examined. Carcinomatous areas were defined as those in which tumors consisting of adenocarcinoma, squamous cell carcinoma, or undifferentiated NSCLC cells existed. Sarcomatous areas were defined as those in which the tumors consisting of spindle and/or giant cells existed. All the histological materials included in this series were reviewed by two observers (K.H. and S.K.). Written informed consent was provided by all the patients, and the specimens were collected according to the tenets of the Declaration of Helsinki established in 2013. The study was approved by the Ethics Committee Review Board at Nippon Medical School Hospital (approval no. B-2020-286).
Antibodies and IHC. IHC staining was performed as previously described (28). The following primary antibodies were used for IHC staining: PD-L1 (clone 28-8, dilution 1:50; Abcam, Cambridge, UK), E-cadherin (clone G-10, dilution 1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), vimentin (clone V9, dilution 1:100; DakoCytomation, Glostrup, Denmark), zinc finger E-box-binding homeobox 1 (ZEB-1) (polyclonal, dilution 1:1,000; Sigma–Aldrich, St. Louis, MO, USA), and c-Met (clone SP59, dilution 1:500; Abcam). IHC evaluation was independently performed both in carcinomatous and sarcomatous areas. IHC scores were defined as the percentage values of tumor cells that showed positive staining, with scores ranging from 0 to 100. The percentage values were evaluated every 5%, and positive expression was defined as ≥5% staining. High expression of PD-L1 was defined as ≥50% staining. Staining score analyses were performed by two independent observers (K.H. and S.K.) who were blinded to the clinical data of patients.
Statistical analysis. Overall survival (OS) was defined as the time from the date of surgical resection until the date of death from any cause, or the last follow-up. The survival curves were constructed through the Kaplan–Meier method and compared using the log-rank test. Correlations of each IHC score were analyzed using Spearman’s rank correlation coefficient (r). Two-sided p-values <0.05 denoted statistically significant differences. All statistical analyses were conducted using the JMP statistical software package for Windows, version 11 (SAS Institute, Cary, NC, USA).
Results
Patient characteristics and immunohistochemical scores. Of the 18 patients with PPC eligible for inclusion in this study, two patients were excluded due to inappropriate quality of pathological evaluation. Finally, 16 patients with PPC were included in this study; Table I shows the baseline characteristics of these patients [median age: 68 years; male: 11 (69%); smokers: 12 (75%)]. The most common histological subtype among the carcinomatous components was adenocarcinoma (six patients; 38%). Eleven patients (69%) had sarcomatous components consisting of only spindle cells. All patients, except for one patient with pathological stage IV disease, underwent complete resection. Representative images of tumors in carcinomatous and sarcomatous areas are shown in Figure 1A. The IHC scores of all patients are summarized in Table II. Expression of PD-L1 was detected in the carcinomatous areas of 14 patients (88%). Nine patients (56%) showed high expression of PD-L1 in carcinomatous areas. The frequency of positive staining and high expression for PD-L1 in sarcomatous areas was consistent with those recorded in carcinomatous areas. Representative images of PD-L1-negative and -positive tumors in each area are shown in Figure 1B. Expression of E-cadherin was observed in the carcinomatous areas of 12 patients (75%). In contrast, five patients (31%) showed positive staining for E-cadherin in sarcomatous areas. Fourteen patients (88%) showed positive staining for vimentin in sarcomatous areas; however, expression of vimentin in carcinomatous areas was also identified in approximately half of the patients. The positive rate of ZEB-1 expression was 50% (eight patients) and 25% (four patients) in sarcomatous areas and carcinomatous areas, respectively. The rate of positive staining for c-Met was 38% (six patients) and 25% (four patients) in carcinomatous and sarcomatous areas, respectively.
Patient characteristics (n=16).
Representative images of pulmonary pleomorphic carcinoma specimens stained using (A) hematoxylin and eosin, and (B) PD-L1. PD-L1: Programmed cell death-ligand 1.
Immunohistochemical staining scores of all patients (n=16).
Correlations between PD-L1 and EMT status/c-Met. Using IHC scores, we evaluated the relationship between PD-L1 and EMT-related markers and between PD-L1 and c-Met in each area (Table III). PD-L1 had a positive correlation with vimentin (r=0.668, p=0.005) and ZEB-1 (r=0.562, p=0.023) in carcinomatous and sarcomatous areas, respectively. In addition, no correlation was observed between PD-L1 and E-cadherin or c-Met expression.
Correlations of immunohistochemical staining scores between antibodies in carcinomatous and sarcomatous areas.
Differential expression of PD-L1 between carcinomatous and sarcomatous areas. Next, using IHC scores, we evaluated the relationship between the expression levels of each marker in carcinomatous and sarcomatous areas (Table IV). ZEB-1 and c-Met had a positive correlation in carcinomatous and sarcomatous areas (r=0.704, p=0.002; and r=0.775, p<0.001, respectively). Vimentin expression between carcinomatous and sarcomatous areas was slightly correlated. However, there was no correlation of PD-L1 expression between carcinomatous and sarcomatous areas. In summary, the expression levels of PD-L1 in each patient varied between carcinomatous and sarcomatous areas.
Correlations of immunohistochemical staining scores for each antibody between carcinomatous and sarcomatous areas.
High PD-L1 expression in sarcomatous areas predicts poor survival in patients with PPC. Finally, we evaluated the prognostic value of the expression of PD-L1, EMT-related markers, and c-Met in carcinomatous and sarcomatous areas (Table V). The median follow-up time was 48.7 months (range=5.3-137 months). The median OS was 67.6 months (95% confidence interval=2.6 months–not reached). High expression of PD-L1 in carcinomatous areas did not predict OS. In contrast, high expression of PD-L1 in sarcomatous areas was significantly associated with poor prognosis (p=0.045) (Table V; Figure 2A and B). In addition, the expression of ZEB-1 in sarcomatous areas was correlated with shorter OS (p=0.012) (Table V; Figure 2C). Also, the expression of c-Met in sarcomatous areas was linked to shorter OS (p=0.058) (Table V; Figure 2D).
Overall survival according to the immunohistochemical staining scores in carcinomatous and sarcomatous areas.
Kaplan–Meier curves for overall survival in patients with pulmonary pleomorphic carcinoma according to (A) high PD-L1 expression in carcinomatous areas, (B) high PD-L1 expression in sarcomatous areas, (C) ZEB-1 expression in sarcomatous areas, and (D) c-Met expression in sarcomatous areas. c-Met: Cellular mesenchymal–epithelial transition factor; PD-L1: programmed cell death-ligand 1; ZEB-1: zinc finger E-box-binding homeobox 1.
Discussion
In this study, we identified different PD-L1 expression patterns between carcinomatous and sarcomatous areas in patients with PPC. In addition, high PD-L1 expression associated with ZEB-1 expression in sarcomatous areas predicted poor prognosis in PPC.
A previous research study demonstrated that PD-L1 expression was higher in sarcomatous areas than carcinomatous areas in patients with PPC (15). However, the correlation of PD-L1 expression levels between carcinomatous and sarcomatous areas had not been analyzed. The present study revealed that the expression levels of PD-L1 differed between carcinomatous and sarcomatous areas. Furthermore, we showed that high PD-L1 expression in sarcomatous areas was associated with worse prognosis in patients with PPC. Previous studies have reported that PD-L1 expression may indicate poor prognosis in PPC (16, 29). In contrast, another study demonstrated that PD-L1 expression was associated with favorable prognosis in PPC (30). However, the cutoff value of PD-L1 expression for survival analysis varied between these studies. Furthermore, the prognostic power of PD-L1 expression in carcinomatous and sarcomatous areas was not evaluated in those studies. In the present study, high PD-L1 expression in carcinomatous areas was not predictive of survival. According to these findings, the biological role of PD-L1 in PPC may vary depending on tumor morphology. PD-L1 expression is affected by various factors, such as histology, stage of disease, oncogenic drivers, prior chemotherapies, and prior radiotherapies (10, 31). PD-L1 expression in sarcomatous areas may be an essential predictor of survival in PPC. Among the 16 patients with PPC included in this study, two received anti-PD-1 therapy with pembrolizumab after disease recurrence. Both patients showed partial response to pembrolizumab. One patient had tumors with high PD-L1 expression in both carcinomatous and sarcomatous areas. The other patient had tumors with high and negative PD-L1 expression in sarcomatous and carcinomatous areas, respectively. This result supports the differential biological role of PD-L1 between the carcinomatous and sarcomatous components in PPC. PD-L1 expression in sarcomatous areas may be a significant predictive biomarker for response to anti-PD-1/PD-L1 therapy in patients with PPC.
We also showed that ZEB-1 was positively correlated with PD-L1 expression in sarcomatous areas, but not in carcinomatous areas. ZEB-1 is a major transcriptional factor activating EMT. An association between ZEB-1 and PD-L1 expression in cancer has been gradually demonstrated (19, 32). The ZEB-1/microRNA-200 axis regulates the function of PD-L1, leading to immunosuppression of cytotoxic tumor-infiltrating lymphocytes and metastasis in lung cancer (33). In vitro, knockdown of ZEB-1 suppressed PD-L1 expression in mesenchymal lung tumor cell lines, while constitutive ZEB-1 expression upregulated PD-L1 in non-metastatic epithelial cells (33). In contrast, through IHC analysis in lung adenocarcinoma, it has been reported that ZEB-1 was not associated with PD-L1 expression (20). The relationship between PD-L1 and ZEB1 may be stronger in mesenchymal tumor cells than epithelial tumor cells. This conclusion is consistent with the results of our investigation in PPC. Furthermore, we observed that ZEB-1 expression in sarcomatous areas was a worse prognostic factor in PPC. ZEB-1 plays pleiotropic roles ranging from cancer initiation to metastasis. This is also reflected by its correlation with unfavorable clinical outcomes in numerous types of cancer (18, 32). A previous report indicated that high ZEB-1 expression in sarcomatous areas of patients with PPC is associated with poor prognosis (34). ZEB-1 expression in sarcomatous components may reflect more aggressive features of PPC. A differential immunosuppressive tumor microenvironment may be involved in the EMT between carcinomatous and sarcomatous components in PPC. Currently, clinical trials of PD-1/PD-L1 inhibitors in combination with targeting EMT-related molecules in patients with several malignancies are being conducted (19). A recent preclinical research study demonstrated that silencing of ZEB-1 could enhance the effect of anti-PD-1 antibody in lung cancer (35). Targeting ZEB-1 in combination with PD-1/PD-L1 inhibitors may be a promising therapeutic strategy against PPC.
In this study, there was no correlation observed between c-Met and PD-L1 expression. In lung adenocarcinoma, PD-L1 expression is significantly increased in c-Met-positive tumors compared with c-Met-negative tumors (26). Conversely, a recent report indicated that low PD-L1 expression (<50%) is associated with high MET copy number (≥2.3) in patients with PPC (36). Hence, the association between c-Met and PD-L1 in PPC may be controversial. Moreover, we showed that c-Met expression in sarcomatous areas was linked to shorter survival times. The prognostic value of c-Met expression in patients with PSC has not been reported yet. c-Met activation is involved in the communication between mesenchymal and epithelial cells (23, 37). c-Met in mesenchymal tumor cells may affect the progression of PPC. However, c-Met in sarcomatous components may contribute to tumor aggressiveness in PPC through a different action mechanism, not PD-L1, because of their weaker association.
Several limitations of this study should be considered. Firstly, this was a retrospective analysis performed at a single institution. Secondly, the sample size was relatively small; this is a common limitation of studies examining PPC due to the rarity of this disease. A multicenter trial involving a sufficient number of patients with PPC is warranted to confirm the findings of this study.
In conclusion, we revealed that the expression levels of PD-L1 differed between carcinomatous and sarcomatous areas. High PD-L1 expression associated with ZEB-1 expression in sarcomatous components was an unfavorable prognostic factor in patients with PPC. Analysis of PD-L1 expression in sarcomatous components through IHC may be useful in predicting survival in patients with PPC.
Acknowledgements
This study was supported in part by a grant-in-aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (grant 16K09592 to M. Seike), and the Clinical Rebiopsy Bank Project for Comprehensive Cancer Therapy Development from the Ministry of Education, Culture, Sports, Science and Technology Supported Program for the Strategic Research Foundation at Private Universities (grant S1311022 to A. Gemma and M. Seike).
Footnotes
Authors’ Contributions
Conception and design: K. Hisakane, M. Seike, T. Sugano, S. Kunugi; Acquisition of data: K. Hisakane, K. Matsuda, S. Kunugi, S. Nakamichi, M. Matsumoto, A. Miyanaga, R. Noro, Y. Minegishi; Analysis and interpretation of data: All Authors; Drafting the manuscript or revising it critically for important intellectual content: K. Hisakane, M. Seike, T. Sugano, K. Kubota, A. Gemma; Final approval of the version to be published: All Authors; Agreement to be accountable for all aspects of the work: All Authors.
Conflicts of Interest
The Authors have no conflicts of interest to disclose in relation to this study.
- Received March 26, 2021.
- Revision received April 15, 2021.
- Accepted April 16, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.