Abstract
Background/Aim: PD-L1 inhibitors have been approved for cisplatin-ineligible urothelial cancer patients relapsing after radical cystectomy. A prerequisite for therapy is a positive PD-L1 expression in the tumor tissue, whereas no options are available for patients with negative PD-L1 status. However, studies revealed that many PD-L1-negative patients also responded to PD-L1 therapy. This study investigated the feasibility of PD-L1 mRNA complementary RNA in situ hybridization (RNAish) analysis to detect PD-L1-responders independent of PD-L1 protein status. Materials and Methods: Immunohistochemistry and RNA in situ hybridization were used to assess PD-L1 protein and mRNA in radical cystectomy tissue from patients with advanced and metastasized urothelial cancer. Results: In this study, PD-L1 protein and mRNA were detected in ≥90% of the examined tissue. Positive PD-L1 mRNA expression (≥1%) on TC and IC could be evaluated in 77% and 31% of the cases, respectively. Moreover, scatterplot analysis revealed a PD-L1 mRNA positive and PD-L1 protein negative subpopulation. According to the CPS score, positive PD-L1 protein expression could be evaluated in 88% and positive PD-L1 mRNA expression in 71% of the cases. Scatterplot analysis of the CPS scores revealed a CPS protein negative but CPS mRNA positive small subpopulation. Conclusion: The feasibility of RNAish on formalin-fixed tissue could be proven. Moreover, complementary PD-L1 RNAish identified a sub-population of PD-L1 protein-negative and PD-L1 mRNA-positive patients, which may benefit from PD-L1 therapy.
Immune checkpoint inhibitors (ICI) have been clinically investigated in recent years. As a result, FDA approved the first immune checkpoint inhibitors for first- or second-line treatment of patients with malignant melanoma in 2011 (1). ICI explicitly inhibits the binding of proteins like the cytotoxic T lymphocyte-associated protein-4 (CTLA-4), the programmed death (PD)-1 T cell receptor, and its ligand PD-L1 with their partner activating the immune system to attack cancer cells (2). Subsequently, inhibition of these proteins by ICI activates immune cells, such as the cytotoxic CD8+ T cells, thereby targeting the tumor cells. Due to its beneficial influence on cancer patients, ICIs have been approved for multiple cancers, including non-small cell lung cancer, head and neck squamous cell carcinoma, renal cancer, and urothelial cancer (3, 4).
Urothelial cancer (UC) is one of the most aggressive forms of cancer, with approximately 573,278 new cases and 212,536 yearly deaths worldwide in 2020 (5). UC is classified as either non-muscle invasive (NMI) or muscle-invasive (MI) disease with a higher incidence of metastases in MI disease. Most UC diagnoses are NMI (~75%), and the probability of progression at 5 years varies between 0.93% to 40% according to the risk group (6, 7). In contrast, MI presents as ~20% of all cases with a 5-year survival of ~50%, with survival rates varying from 32% in patients with lymph node involvement to 75% in those without (8). The current gold-standard treatment for MI UC is the surgical removal of the bladder (radical cystectomy). However, ~50% of patients will still relapse and proceed to develop metastasis and 5% of patients present with primary metastatic disease. Metastatic patients are treated first-line with platin-based combination chemotherapy (9). Nevertheless, many patients with metastatic UC are ineligible for cisplatin-based chemotherapy and face recurrence after the first-line therapy. Therefore, there is a significant medical need to identify novel therapeutic options.
Since 2016 ICI has been approved as first line in PD-L1 highly positive patients (cisplatin-ineligible and PD-L1 highly positive, Atezolizumab, Pembrolizumab) and as a second line treatment of palliative UC patients (Atezolizumab, Nivolumab, and Pembrolizumab) (9-13). However, due to the FDA and EMA restrictions, first-line treatment with Atezolizumab and Pembrolizumab in cisplatin-ineligible patients, requires PD-L1 positivity assessed by the PD-L1 score (IHC). In contrast, there is no treatment option for PD-L1-negative patients (9, 14). In this context, different evaluation algorithms have been developed for the various available and approved drugs, targeting either immune cells (IC), tumor cells (TC), or even the combined view of the two cell types (combined positive score, CPS). However, the currently valid approvals and cut-offs are subject to dynamics (4, 15, 16).
However, these ICI cannot induce a response in all PD-L1-positive patients (12, 17). On the other hand, responses to ICI have been reported despite a lack of PD-L1 (17). These observations and experiences suggest that there might be differences in the biology of the tumors, which are only partially revealed by the various available diagnostic antibodies. Therefore, this pilot study investigated the feasibility of PD-L1 mRNA complementary RNA in situ hybridization (RNAish) analysis to detect PD-L1-responders independent of PD-L1 protein status.
Materials and Methods
Patients and study design. This retrospective pilot study includes 21 patients with radical cystectomy for advanced and lymphatic metastasized urothelial cancer between 2005 and 2015. The study was approved by the local institutional review board of the Faculty of Medicine of the Technische Universität Dresden (ethics vote EK43022017). Furthermore, it was conducted per the Declaration of Helsinki, according to the ICH Harmonized Tripartite Guideline for Good Clinical Practice (18).
Immunohistochemistry. Immunohistochemistry (IHC) was performed as published previously (19). Formalin-fixed, paraffin-embedded (FFPE) tissue sections were immunostained with the VENTANA PD-L1 (SP142) assay (Roche, Basel, Switzerland) optimized for the detection of immune cells and PD-L1 (Clone E1L3N) antibody (Cell Signaling, Danvers, MA, United States), followed by counterstaining with hematoxylin, dehydration, and mounting of the slides according to the manufacturer’s protocol (Benchmark; Ventana Medical Systems, Tucson, AZ, USA). PD-L1-stained slides were scored for PD-L1 protein expression on immune cell (IC)-positivity (% of tumor area covered by stained IC) and tumor cell (TC)-positivity (% of positive PD-L1 TC in the tumor area). PD-L1 IC-positivity was defined as staining in granulocytes, lymphocytes, macrophages, and dendritic cells of any intensity. PD-L1 TC-positivity was defined as membranous PD-L1 staining of any intensity. Moreover, the combined positive score (CPS) was calculated as the total number of PD-L1 positive TC and IC in the tumor area divided by the total number of viable TC multiplied by 100% (20). PD-L1 expression values ≥1% and a CPS ≥1 were defined as positive.
RNA in situ hybridization. RNA in situ hybridization (RNAish) was performed to detect PD-L1 mRNA using the ViewRNA ISH Tissue 2-Plex Assay (Affymetrix, Thermo Fisher, Waltham, MA, USA) using only Fast Red chromogen according to the manufacturer’s protocol. FFPE tissue sections (4 μm thick) were briefly baked for 1 h at 60°C. The slides were deparaffinized with xylene three times for 10 min and allowed to air dry after two rinses with 100% ethanol for 5 min. The slide’s heat pretreatment was performed for 5 min, followed by protease digestion for 10 min. The target-specific PD-L1 probe set (CD274, VA1-14391-01, Affymetrix, Thermo Fisher) 1:40 diluted in Diluent QT was hybridized overnight at 40°C in a hybridization oven. The signal amplification was performed the next day by sequential hybridization steps with PreAmplifier Mix for 25 min, Amplifier Mix for 20 min, and Label Probe 1-alkaline phosphatase (AP) Mix (1:1,000) for 15 min each at 40°C. Finally, the slides were incubated with Fast Red Substrate for 30 min at 40°C, followed by two 5 min washes in PBS. The counterstaining was carried out with Gill’s Hematoxylin I, and the slides were mounted with Mowiol 4-88/DABCO (Carl Roth GmbH, Karlsruhe, Germany) for visualization.
Image quantification was performed for each case by counting tumor and immune cells in three tumor areas (hot spots) whereas both cell types produce 100% of the cell count. In addition, the percentage of PD-L1 positive tumor cells (TC) and PD-L1 positive immune cells (IC) was determined by counting the “dot-like” red color hybridization signals in the tumor cell nuclei and cytoplasm. For each PD-L1 staining, a control section without a target-specific PD-L1-probe set was prepared and evaluated in the same way to exclude unspecific reactions of endogenous alkaline phosphatases in the tumor tissue.
Statistical analysis. Prism 9.5 (GraphPad Software, San Diego, CA, USA) was used for all statistical analyses. The Pearson correlation coefficient (r) was calculated and interpreted, as suggested by Schober et al., for correlation analysis (21).
Results
PD-L1 protein and PD-L1 mRNA are expressed in the majority of the examined tissues. The tissues of 21 patients with radical cystectomy for advanced and lymphatic metastasized urothelial cancer between 2005 and 2015 were stained for PD-L1 protein and PD-L1 mRNA. Representative immunohistochemical images for PD-L1 protein and PD-L1 mRNA are displayed in Figure 1. PD-L1 protein expression in tumor and immune cells could be evaluated in 20 out of 21 cases (95%). Likewise, PD-L1 mRNA could be evaluated in 19 cases (90%). Only one case was utterly negative.
Expression of PD-L1 mRNA and protein in tumor and immune cells. Due to constantly changing cut-off values in the different lines of therapy and also the fact that patients with low PD-L1 positivity can respond to appropriate therapy, PD-L1 expression on more than 1% of tumor cells or immune cells was defined as positive (9-13). Therefore, positive PD-L1 protein expression (≥1%) on TC and IC could be evaluated in 38% (Figure 2A, Figure 3A) and 69% (Figure 2B, Figure 3B) of the 21 cases, respectively. In line with the guidelines for PD-L1 positivity, the PD-L1 mRNA was assessed. Positive PD-L1 mRNA expression (≥1%) on TC and IC could be evaluated in 77% (Figure 2A, Figure 3A) and 31% (Figure 2B, Figure 3B) of the 21 cases, respectively.
Correlation analysis reveals a PD-L1 mRNA positive and PD-L1 protein negative subpopulation. To analyze if there is a correlation between the PD-L1 mRNA positive and protein expression, the Pearson correlation coefficient (r) was calculated. In TC and IC, the mRNA expression only showed a significant moderate correlation with the PD-L1 protein expression. However, the analysis reveals that 59% of the TC cells and 6% of IC are PD-L1 protein negative but PD-L1 mRNA positive (Figure 3).
Evaluation of the complementary analysis of PD-L1 mRNA for the combined positive score. The combined positive score has been developed as a clinically relevant and highly reproducible scoring method for PD-L1 (16). To assess if RNAish may have a benefit as a complementary analysis for the CPS, the CPS was calculated for the PD-L1 mRNA and protein expression. According to the CPS score, positive PD-L1 protein expression (≥1%) could be evaluated in 88% (Figure 4B) of the 21 cases. In line with this, positive PD-L1 mRNA expression (≥1%) could be evaluated in 71% (Figure 4B) of the 21 cases. Correlation analysis revealed a strong correlation between the calculated protein and mRNA CPS. In contrast to the PD-L1 analysis in TC or IC alone, only 2% of the CPS negative but CPS mRNA positive could be detected.
Discussion
Since their approval by FDA and EMA, ICI therapies with anti-PD-(L)1 antibodies have become the most critical medical therapies in urogenital cancers, including kidney and UC (22). During PD-L1 evaluation, the expression of PD-L1 on IC and TC is assessed, as both cell types have been shown to have prognostic value (23). Especially UC patients ineligible for cisplatin have benefited from the approved PD-L1 therapies (9, 12, 14). The inclusion criterion for the first-line therapy for these patients is a positive PD-L1 protein detection in the tissue (24). However, many patients still do not benefit from the PD-L1 therapy despite proven PD-L1 positivity (12, 17, 25). Moreover, Rosenberg and colleagues reported that 8% PD-L1 negative patients responded to PD-L1 therapy (17). Therefore, complementary biomarkers are necessary to identify PD-L1 positive non-responders and PD-L1 negative non-responders. Several studies have reported on complementary diagnostic and prognostic biomarkers in melanoma, non-small cell lung cancer, and renal cancer (26, 27). Oh and colleagues reported high soluble PD-L1 (sPD-L1) levels in serum from melanoma and non-small cell lung cancer associated with low ICI response rates (27). In renal carcinoma, serum sPD-L1 has been reported to predict treatment response to nivolumab in patients with mRCC (26). However, by the publication date of this paper, sPD-L1 had not yet been established as a prognostic marker. In UC, a biomarker study investigating Ki-67, MRE11 and PD-L1, failed to have any predictive value (28). The present pilot study investigated the feasibility of complementary PD-L1 mRNA detection by RNAish analysis on TC and IC. In a similar approach, Duncan and colleagues reported the feasibility of detecting PD-L1 mRNA in TC by RNAish (29). In addition, the group reported the possibility of PD-L1 mRNA and protein detection in TC on formalin-fixed tissue. Moreover, Erber et al. suggested quantitative PD-L1 mRNA determination as an alternative to IHC as there is no interobserver variability in RNA results (30).
Moreover, the PD-L1 mRNA extent in TC cells was associated with PD-L1 protein status determined by IHC. In contrast, in the present study, no association could be found between PD-L1 mRNA and PD-L1 protein either on TC or on IC. Moreover, scatter plot analysis revealed a high number of PD-L1 mRNA positive but PD-L1 protein negative TC. PD-L1 expression on tumor cells has been reported to be downregulated during cell cycle traversal, and PD-L1 status changes due to cell plasticity (31). This result may explain the temporary negative PD-L1 protein status due to the tumor cells’ proliferative nature and UC’s high cell plasticity (32-34).
The CPS scoring method is a reliable prognostic marker in multiple tumor entities (4, 15, 16). Due to the involvement of the PD-L1 expression of IC, 88% of the cases were scored as PD-L1 protein positive. Moreover, this correlated strongly with the CPS calculated from the PD-L1 mRNA. Still, a small population of PD-L1 protein hostile and PD-L1 mRNA positive subpopulations could only be identified by complementary PD-L1 RNAish analysis.
Conclusion
In this pilot study, the feasibility of RNAish on 8–18-year-old formalin-fixed tissue could be proven. Moreover, complementary PD-L1 RNAish identifies a sub-population of PD-L1 protein-negative and PD-L1 mRNA-positive patients, which may benefit from PD-L1 therapy. In further studies, tissues from UC patients treated with ICI will be analyzed for PD-L1 protein and mRNA to compare the data with treatment response to assess the predictive value of the complementary PD-L1 RNAish.
Acknowledgements
We thank all the urology surgeons (Department of Urology, University Hospital Carl Gustav Carus, Dresden, Germany). In addition, we would like to thank the patients who kindly provided samples.
Footnotes
Authors’ Contributions
Conceptualization: US, PH, HE, GB. Methodology: US, MG, DA. Formal analysis: US, MG KB. Investigation: US, MG, DA, KB, CT. Writing-Original Draft Preparation: US, HE. Writing-Review & Editing: HE, MG, DA, KJ, KB, CT, PH, HE, GB. Supervision: PH, HE, GB.
Conflicts of Interest
The Authors declare that they have no competing interests.
- Received June 28, 2023.
- Revision received July 26, 2023.
- Accepted August 7, 2023.
- 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).