Article Text
PDF
Abstract
Background Immunotherapy is gaining momentum, but current treatments have limitations in terms of beneficiaries. Clear cell renal cell carcinoma (ccRCC) harbors the highest expression of human leukocyte antigen E (HLA-E), ligand of NKG2A, among all solid tumors. In this study, we aim to investigate the role of NKG2A+CD8+ T cells in tumor microenvironment and its potential as a novel target in ccRCC.
Methods This study included four independent cohorts, including 234 patients from Zhongshan cohort (ZSHC) who underwent partial or radical nephrectomy at Zhongshan Hospital, and 117 metastatic patients from metastatic Zhongshan cohort (ZSHC-metastatic renal cell carcinoma) who were treated with immune checkpoint inhibitor or tyrosine kinase inhibitor alone. We also incorporated a cohort of 530 patients diagnosed with ccRCC from The Cancer Genome Atlas (referred to as TCGA-kidney renal clear cell carcinoma) and 311 patients from CheckMate cohort for bioinformatics exploration and hypothesis validation. Fresh surgical specimens from 15 patients who underwent ccRCC surgery at Zhongshan Hospital were collected for flow cytometry analysis. Another 10 fresh surgical specimens were used to investigate the therapeutic potential of NKG2A blockade after in vitro intervention. The infiltration of NKG2A+CD8+ T cells was assessed using immunohistochemical staining, flow cytometry, and immunofluorescence staining in ZSHC cohort.
Results Patients with higher infiltration of NKG2A+CD8+ T cells in ccRCC exhibited shorter overall survival and resistance to immunotherapy. NKG2A+CD8+ T cells expressed upregulated checkpoint molecules and displayed impaired effector functions, along with tissue-residency characteristics. Combination of programmed cell death protein-1 (PD-1) blockade and NKG2A blockade demonstrated an enhanced capability in reactivating CD8+ T cells effector functions.
Conclusion Intense infiltration of NKG2A+CD8+ T cells were associated with poorer prognosis and response to immunotherapy. NKG2A blockade combined with current immunotherapy exhibited a robust ability to reactivate CD8+ T cells effector functions.
- renal cell carcinoma
- immunotherapy
- tumor microenvironment
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Immunotherapy is gaining momentum, but current treatments have limitations in terms of beneficiaries. Treatment options after progression are limited. Therefore, exploration of new treatment approaches for patients with metastatic renal cell carcinoma are urgently needed.
A pronounced infiltration of CD8+ T cells is correlated with less favorable prognosis in clear cell renal cell carcinoma (ccRCC).
ccRCC harbors the highest expression of human leukocyte antigen-E (HLA-E), ligand of NKG2A, among all solid tumors.
WHAT THIS STUDY ADDS
In ccRCC, NKG2A+CD8+ T cells were associated with unfavorable prognosis and response to immunotherapy.
NKG2A+CD8+ T cells display an exhausted phenotype and were associated with an inflamed but immunosuppressive microenvironment with enriched regulatory CD4+ T cells (Tregs) infiltration.
The synergistic interplay of NKG2A blockade and current immunotherapy exhibited a formidable capacity to reactivate CD8+ T cells effector functions in ccRCC.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Combined therapy of anti-programmed cell death protein-1 (PD-1) and anti-NKG2A results in enhanced CD8+ T cells effector function, which may become an effective approach for overcoming resistance to current immunotherapy in ccRCC.
Introduction
Renal cell carcinoma (RCC) is one of the three major malignant tumors of the urinary system.1 In 2020, there were over 400,000 new cases of RCC globally,2 among which clear cell renal cell carcinoma (ccRCC) accounts for about 75% and serves as the most common pathological type.3 Compared with high 5-year survival probability (over 80%) of patients with non-metastatic localized ccRCC,4 the other 20%–50% of patients with metastatic renal cell carcinoma (mRCC) would suffer from greatly reduced survival, mostly occurring merely within 3 years after surgery.5
Immunotherapy has revolutionized the treatment of mRCC over the past decade.6 Compared with tyrosine kinase inhibitors (TKI) alone, latest results from phase III trials demonstrated that combination therapy based on immune checkpoint blockade (ICB) plus TKI significantly improved the outcomes of patients with mRCC.7 However, durable responses only achieved in a subset of patients with mRCC.8 Treatment options after progression are limited and usually far from efficacious.9 Therefore, exploration of new treatment approaches for patients with mRCC are urgently needed.
Interestingly, unlike other tumors, intense infiltration of CD8+ T cells correlated with poorer prognosis in ccRCC, which is likely due to functional subdivision of distinct CD8+ T cells subtypes.10 11 Our group previously also found that in ccRCC, TNFRSF9+CD8+ T cells possessed both exhaustion and effector phenotype and CXCL13+CD8+ T cells possessed exhausted phenotype with impaired antitumor function.12 13
NKG2A is a C-type lectin family member that can form dimers with CD94 on the cell surface.14 Unlike the programmed cell death protein-1 /programmed cell death ligand 1 (PD-1/PD-L1) pathway, NKG2A acts independently of the T-cell receptor (TCR).15 The non-classical major histocompatibility complex-I molecule human leukocyte antigen-E (HLA-E) is the main ligand for NKG2A-CD94.15 And among solid tumors, ccRCC represents the highest level of HLA-E expression.16 NKG2A is reported to be mainly expressed in CD8+αβ T cells, CD56hi natural killer (NK) cells and natural killer T (NKT) cells.15 NKG2A-CD94 binding to HLA-E can inhibit the activation of T cells and NK cells.15 17 Previous studies suggested that NKG2A antagonism primarily functions through NKG2A+CD8+ T cells rather than NKG2A+ NK cells in tumor vaccine therapy resistant individuals.18 However, the role of NKG2A+CD8+ T cells in tumor microenvironment (TME) and the therapeutic effect of targeting NKG2A in ccRCC remained unknown.
In the present study, we sought to investigate the specific phenotype of NKG2A+CD8+ T cells in ccRCC and tried to uncover its relationship with immune contexture. We further explored its predictive value for prognosis and responses to ICB. Ultimately, we sought to evaluate the potential of combining NKG2A antagonism with existing immunotherapy as a new treatment modality for patients with ccRCC.
Materials and methods
Study cohorts
Two independent in-house cohorts were enrolled in this study, namely Zhongshan Hospital cohort (ZSHC, n=234), and Zhongshan Hospital mRCC cohort (ZSHC-mRCC, n=117).
Zhongshan Hospital cohort (ZSHC) included 290 patients with ccRCC who underwent radical nephrectomy or partial nephrectomy from the Department of Urology, Zhongshan Hospital, Fudan University (Shanghai, China) between February 2005 and June 2007. The inclusion criteria for this study were as follows: (1) signed informed consent; (2) diagnosed with ccRCC pathologically; (3) age ≥18 years; (4) never received any systematic antitumor therapy before the surgery; and (5) with available clinicopathological data. 21 patients were excluded due to tissue microarray point detachment. 35 patients were excluded due to unassessable NKG2A or CD8 staining. Finally, 234 patients were recruited for further study.
Zhongshan Hospital mRCC cohort (ZSHC-mRCC) enrolled 128 patients with RCC (62 patients treated with ICB and 66 patients treated with TKI alone) from the Department of Urology, Zhongshan Hospital, Fudan University (Shanghai, China) between April 2012 and July 2023. The inclusion criteria for this cohort were as follows: (1) informed consent, (2) patients with RCC suffering metastasis, (3) available for formalin fixed paraffin embedded specimens. 11 patients were excluded due to their pathology type not being ccRCC. Finally, 117 patients were recruited for further study (58 patients treated with ICB and 59 patients treated with TKI alone). Therapeutic regimens of patients in ZSHC-mRCC (ICB) were shown in online supplemental table 1. Response to ICB and TKI alone were assessed using Response Evaluation Criteria in Solid Tumors V.1.1.19
Supplemental material
Supplemental material
Public data processing
The Cancer Genome Atlas kidney renal clear cell carcinoma cohort (TCGA-KIRC, n=530) and the CheckMate cohort (CheckMate, n=311) were included for external validation.
TCGA-KIRC cohort included 530 patients, with both clinicopathological and RNA-sequencing data available. Data was downloaded from UCSC Xena (https://xenabrowser.net/datapages/).
CheckMate cohort including three cohorts (CheckMate-025, CheckMate-009, CheckMate-010), are randomized phase III trials to compare nivolumab with everolimus in patients with advanced RCC.8 20 21 311 patients with both clinical information and RNA-sequencing data were included for further investigation. Data was acquired from previous study.8
Immunohistochemistry and immunofluorescence
This study conducted immunohistochemical (IHC) staining on NKG2A, NKG2A+CD8+ T cells and 13 types of tumor-infiltrating immune cells in the ZSHC and ZSHC-mRCC cohort. The antibody information was shown in online supplemental table 2. The specific experimental procedures were consistent with our previous study.22 The infiltration level of NKG2A+ cells and immune cells were evaluated under microscopy by two independent pathologists with relevant pathological reading abilities without knowledge of patient clinical information.
Supplemental material
Immunofluorescence staining was used to evaluate the infiltration level of NKG2A+CD8+ T cells in the ZSHC cohort. The specific experimental procedures followed previous research.22 Primary antibodies included NKG2A (ab260035, Abcam) and CD8 (ab17147, Abcam), then using Bond Polymer Refine Detection (DS9800, Leica). The infiltration level of NKG2A+CD8+ T cells was assessed by two independent pathologists in the 3DHISTECHVIEWER software without knowledge of patients’ clinical information. The average of the results from these two pathologists was calculated as the final value.
In our study, two pathologists (Dr Yunyi Kong and Dr Lingli Chen) who were blinded to the clinicopathological data and scored all samples separately. The mean count derived from their respective evaluations was used as the basis for analysis. Instances of enumeration discrepancies exceeding five cells were subject to individual re-evaluation by both pathologists, with the objective of reaching a consensus.
Construction of immune cells signature
We also used a single-cell sequencing data of ccRCC to construct signatures of NKG2A+CD8+ T cells and NKG2A+ NK cells.23 The expression matrix obtained from the single-cell sequencing was further clustered to define CD8+T-cell subgroups. Subsequently, NKG2A+ cells (KLRC1: counts >0) were further clustered within the CD8+ T-cell group. FindMarkers from “Seurat” package24 was then employed to define the marker gene set for the NKG2A+CD8+ T cells subpopulation. Ultimately, 14 genes with significant differences were identified as the signature of NKG2A+CD8+ T cells. Similarly, NKG2A+ NK cells were also classified using the FindMarkers function to obtain the signature of NKG2A+ NK cells. Gene list of NKG2A+CD8+ T cells signature and NKG2A+ NK cells signature were shown in online supplemental table 3.
Supplemental material
Flow cytometry
Twenty-five patients who received radical nephrectomy or partial nephrectomy from March 2022 to January 2023 were included in this study at Zhongshan Hospital, Fudan University. After complete removal of the tumor specimens, fresh tumor tissues were selected by cutting along the longest axis of the tumor at its base, while avoiding necrotic tissue. None of the patients received systemic antitumor treatment before surgery. All patients signed informed consent forms for the donation of biological samples and the use of related information. Among them, four specimens were excluded due to contamination, and six cases were excluded because the follow-up pathology reports indicated non-ccRCC. The final cohort for flow cytometry (FCM) experiments on ccRCC at Zhongshan Hospital consisted of 15 remaining specimens.
Specimens were digested with collagenase IV (0.05 g/40 mL prepared with penicillin–streptomycin antibodies) and Golgi blockade agent (1 µL/1 mL). The digestion was performed at 37°C for about 2 hours to obtain a single-cell suspension. Erythrocytes were removed by adding a red blood cell lysis buffer. Then Fc-receptor-blocking antibody (BD Bioscience) was used. Fixable Viability Stain 510 cell viability staining reagent was added and incubated for 20 minutes, followed by cell membrane antibody staining. If intracellular antibody staining or transcription factors were required simultaneously, a Fixation/Permeabilization Solution Kit or Transcription Factor Fixation/Permeabilization Buffer was used. FCM analysis was performed using BD FACSCelesta, and the results were analyzed using FlowJo V.10.8.1. online supplemental table 4 provides a list of FCM antibodies.
Supplemental material
In vitro intervention assay
Tumor tissues from 10 patients with pathologically diagnosed ccRCC were included for in vitro intervention. Tumor single-cell suspensions were cultured with anti-NKG2A antibody (IM2750, BECKMAN COULTER, 3 µg/mL), or anti-PD-1 antibody (nivolumab, A2002, Selleck, 5 µg/mL), or IgG4B κ isotype control antibodies (A1101, Biovision, 10 µg/mL) for 12 hours in Roswell Park Memorial Institute medium 1640 (RPMI-1640 medium) with a concentration of around 10% fetal bovine serum. After overnight culture, the suspensions were performed FCM analysis as described above.
Statistical analysis
For each sample, the average expression of all genes in a given signature was computed. The involved genes for signature calculation were listed in online supplemental table 5. The results will be presented as mean±SD or as proportions, as needed. For the comparison between grouped continuous variables, Student’s t-test and Mann-Whitney U test would be used; Wilcoxon signed-rank test was used for paired samples comparison and Spearman’s test would be employed for correlation analysis of continuous variables. The comparison of rates or proportions between groups was used the χ2 test, and if T<1 or n<40, Fisher’s exact test would be used. Kaplan-Meier curves and log-rank test were used to analyze the survival differences between subgroups. Cox regression analysis was employed to assess whether the infiltration level of NKG2A+CD8+ T cells could independently impact prognosis. All hypothesis tests considered p<0.05 as statistically significant, and all tests would be two-tailed.
Supplemental material
Results
NKG2A+CD8+ T cells are associated with poor prognosis in ccRCC
First, we sought to compare the expression of NKG2A between tumor and adjacent normal tissues. In ZSHC, IHC staining confirmed overexpression of NKG2A in tumor tissues (figure 1A, figure 1B and online supplemental figure 1A; p<0.001). Furthermore, in TCGA-KIRC, NKG2A messenger RNA (mRNA) expression was also significantly upregulated in tumor tissues compared with adjacent normal tissues (figure 1C; p<0.001).
Supplemental material
NKG2A+CD8+ T cells are associated with poor prognosis in ccRCC. (A) Representative immunohistochemical images of NKG2A+ cells in ccRCC tissues (scale bar, 20 µm). (B–C) Boxplots comparing the expression of NKG2A in ccRCC tissues and adjacent normal tissues in ZSHC cohort (B) and TCGA-KIRC (C) cohort. Data were analyzed by Mann-Whitney U test. (D–E) NKG2A+CD8+ T cells signature predict prognosis of patients with ccRCC (D) while NKG2A+ NK cells signature could not predict prognosis of patients with ccRCC (E). Kaplan-Meier analysis and log-rank test. (F) Immunofluorescence staining of NKG2A (red), CD8 (green) and DAPI (blue) in TMA of ZSHC cohort, double-positive cells are highlighted by the yellow arrow. The green and red arrows identify CD8 and NKG2A single-positive cells, respectively (scale bar, 20 µm). (G–H) Expression quantities of NKG2A+CD8+ T cells under high magnification were compared among patients in the SSIGN scoring system (G) as well as UISS staging system (H). For the SSIGN, patients with scores of 0–1 were classified as low-grade, those with scores of 2–6 as intermediate-grade, and those with scores ≥7 as high-grade. Kruskal-Wallis test was applied. (I–J) Figures demonstrating the overall survival (I) and recurrence-free survival (J) of patients in the ZSHC cohort, stratified by infiltration of NKG2A+CD8+ T cells and median value was used as the cut-off. Kaplan-Meier analysis and log-rank test. *p<0.05, **p<0.01, ***p<0.001, ns refers to no significance. ccRCC, clear cell renal cell carcinoma; DAPI, 4’,6-diamidino-2-phenylindole; HPF, high power filed; KIRC, kidney renal clear cell carcinoma cohort; mRNA, messenger RNA; NK, natural killer; SSIGN, Mayo Clinic Stage, Size, Grade, Necrosis Score; TCGA, The Cancer Genome Atlas; UISS, the University of California, Los Angeles, Integrated Staging System.
Previous studies indicated that NKG2A is predominantly expressed both in CD8+ T cells and NK cells.15 Therefore, we employed FCM to determine its expression in ccRCC. In compliance with the previous studies, we identified the presence of both NKG2A+CD8+ T cells and NKG2A+ NK cells in ccRCC (online supplemental figure 1B). And the average expression of NKG2A on CD8+ T cells is 12.8%, while NK cells exhibit an average NKG2A expression of 29.7% (online supplemental figure 1C).
Next, we attempted to explore the prognostic value of NKG2A+CD8+ T cells and NKG2A+ NK cells in ccRCC. Using single-cell sequencing data,23 we developed NKG2A+CD8+ T cells signatures and NKG2A+ NK cells signatures. In TCGA-KIRC, patients with higher expression of NKG2A+CD8+ T cells signature had worse prognosis (figure 1D; p=0.004), while the NKG2A+ NK cells signature could not indicate patient prognosis (figure 1E; p=0.144). Thus, we focused our research on the NKG2A+CD8+ T cells subset.
Then, we confirmed the substantial infiltration of NKG2A+CD8+ T cells in ccRCC by immunofluorescence in ZSHC (figure 1F). Clinical characteristics of patients in ZSHC were shown in online supplemental table 6. Higher NKG2A+CD8+ T cells infiltration correlated with higher Mayo Clinic Stage, Size, Grade, Necrosis Score (SSIGN) and the University of California, Los Angeles, Integrated Staging System (UISS) stages (figure 1G and figure 1H; p=0.047 and p=0.049, respectively). Furthermore, we found that higher infiltration of NKG2A+CD8+ T cells was associated with reduced overall survival (OS) and recurrence-free survival (RFS) in ccRCC (figure 1I and figure 1J; p=0.004 and p=0.002, respectively). Moreover, multivariate analysis incorporating clinical and pathological factors indicated that NKG2A+CD8+ T cells infiltration was an independent predictor for OS (online supplemental figure 2). In summary, NKG2A expression was upregulated in ccRCC, and higher infiltration of NKG2A+CD8+ T cells correlated with poorer prognosis.
Supplemental material
Supplemental material
NKG2A+CD8+ T cells display an exhausted phenotype
To assess the functional phenotype of NKG2A+CD8+ T cells, we then collected 15 fresh ccRCC tissue samples for FCM analysis. Compared with NKG2A–CD8+ T cells, NKG2A+CD8+ T cells significantly upregulated the expression of canonical immune checkpoints, including PD-1, lymphocyte activation gene 3 (LAG-3), T-lymphocyte-associated antigen 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), and CD39 (figure 2A–E; p=0.002, p=0.005, p=0.012, p=0.023, and p=0.020, respectively). Furthermore, NKG2A+CD8+ T cells expressed significantly lower interferon (IFN)-γ (figure 2F; p=0.039), while other effector markers like tumor necrosis facor-alpha (TNF-α) and granzyme B (GZMB) showed comparable expression between two groups (figure 2G–H; p=0.112 and p=0.268, respectively). Interestingly, we compared the degranulation function (CD107a) of the two groups and found NKG2A+CD8+ T cells exhibited stronger degranulation function than NKG2A−CD8+ T cells (figure 2I; p=0.011). We further evaluated the expression of tissue-resident marker CD103 and found NKG2A+CD8+ T cells had higher CD103 expression (figure 2J; p=0.003), suggesting a tissue-resident characteristic of NKG2A+CD8+ T cells. Moreover, no significant difference of proliferation marker Kiel-67 protein (Ki-67) and stem-like marker transcription factor 7 (TCF-7) were observed between NKG2A+CD8+ T cells and NKG2A–CD8+ T cells (figure 2K,L; p=0.069 and p=0.088, respectively). Collectively, these results demonstrated that NKG2A+CD8+ T cells exhibited an exhausted phenotype featured with elevated expression of immune checkpoints and decreased effector function.
NKG2A+CD8+ T cells display an exhausted phenotype. (A–E) The expression of immune checkpoints (PD-1+, LAG-3+, CTLA-4+, TIM-3+, CD39+) were significantly higher in NKG2A+CD8+ T cells, compared with NKG2A–CD8+ T cells (gated on CD3+CD8+ T cells). (F–H) The expression of IFN-γ was decreased in NKG2A+CD8+ T cells, while the expression of TNF-α and GZMB remained unaffected (gated on CD3+CD8+ T cells). (I) The expression of CD107a was increased in NKG2A+CD8+ T cells, compared with NKG2A–CD8+ T cells (gated on CD3+CD8+ T cells). (J) Tissue-resident marker (CD103) was upregulated in NKG2A+CD8+ T cells (Gated on CD3+CD8+ T cells). (K–L) No significant difference of Ki-67 (K) and TCF-7 (L) were observed between NKG2A+CD8+ T cells and NKG2A–CD8+ T cells. Above: representative flow cytometry plots; below: quantitative data. Mann-Whitney U test. *P<0.05, **p<0.01, ***p<0.001, ns refers to no significance. APC, Allophycocyanin; CTLA4, T-lymphocyte-associated antigen 4; GZMB, granzyme B; HAVCR2, hepatitis A virus cellular receptor 2; IFN, interferon; Ki-67, Kiel-67 protein; LAG-3, lymphocyte activation gene 3; MFI, median fluorescence intensity; PD-1, programmed cell death protein-1; TIM-3, T-cell immunoglobulin and mucin domain-containing protein 3; TCF-7, transcription factor 7; TNF, tumor necrosis facor.
NKG2A+CD8+ T cells infiltration is associated with immunosuppressive tumor microenvironment in ccRCC
In the intricate milieu of the tumor TME, tumor-infiltrating immune cells play diverse roles.25 The functionality of CD8+ T cells, in particular, can be modulated by the presence and interactions with other immune cells.26 To understand the mechanisms underlying the exhausted phenotype of NKG2A+CD8+ T cells, we then scrutinized the relationship between NKG2A+CD8+T cells and other immune cells infiltration. First, in the TCGA-KIRC and CheckMate cohort, the differences of immune cells infiltration were assessed between two groups with different expressions of NKG2A+CD8+ T-cell signature. The results revealed that patients with higher infiltration of NKG2A+CD8+ T cells had higher infiltration of CD8+ T cells, regulatory CD4+ T (Treg) and T-follicular helper (Tfh) cells compared with patients with lower expression of NKG2A+CD8+ T cells signature (figure 3A; p<0.001, p<0.001 and p<0.001, respectively). On the other hand, patients with higher infiltration of NKG2A+CD8+ T cells had lower infiltration of CD4+ T cells, Mast cells, monocyte and M2 macrophages (figure 3A; p<0.001, p<0.001, p<0.001 and p<0.001, respectively). Additionally, the total infiltration of macrophages was higher in patients with lower infiltration of the NKG2A+CD8+ T cells group (figure 3A; p<0.001). Analysis data of CheckMate cohort confirmed that patients with intense infiltration of NKG2A+CD8+ T cells had high infiltration of T cells such as CD8+ T cells, whereas myeloid cells such as macrophages showed lower infiltration (figure 3B).
NKG2A+CD8+ T cells infiltration is associated with immunosuppressive tumor microenvironment in clear cell renal cell carcinoma. (A–B) Heatmap representative of diverse immune cell populations, checkpoint molecule expression, and overall immune status (immune checkpoints, immune score, suppression score) among NKG2A+CD8+ T cells infiltration subgroups in TCGA-KIRC cohort (A) and CheckMate cohort (B). (C) Distribution of various immune cells in different infiltration groups of NKG2A+CD8+ T cells in ZSHC cohort; Green dots represent patients with low infiltration of NKG2A+CD8+ T cells, while red dots represent patients with high infiltration. (D) Boxplots illustrating the abundance of TCR-related scores across two subgroups within the TCGA-KIRC cohort. Mann-Whitney U test. *p<0.05, **p<0.01, ***p<0.001, ns refers to no significance. CTLA4, T-lymphocyte-associated antigen 4; HAVCR2, hepatitis A virus cellular receptor 2; IMDC, international metastatic renal cell carcinoma database consortium; KIRC, kidney renal clear cell carcinoma; NK, natural killer; TCGA, The Cancer Genome Atlas; TNM, Tumor-node-metastasis staging; TMB, tumor mutation burden; Treg, regulatory CD4+ T; Tfh, T-follicular helper; TCR, T-cell receptor.
We also evaluated the expression of immune checkpoints between two groups. Patients with higher infiltration of the NKG2A+CD8+ T cells showed higher expression of immune checkpoints such as PD-1, CTLA-4 and LAG-3 at the mRNA level (figure 3A,B; p<0.001, p<0.001 and p<0.001, respectively).
In order to evaluate the overall immune microenvironment status, we employed a commonly used immune score.27 The results demonstrated that patients with higher infiltration of the NKG2A+CD8+ T cells had significantly higher immune scores (figure 3A,B; p<0.001 and p<0.001, respectively), indicating a stronger immune response of these patients, characterized by increased immune cells infiltration and elevated cytokine levels. However, simultaneously, patients with higher infiltration of the NKG2A+CD8+ T cells also exhibited significantly higher suppression scores (figure 3A,B; p<0.001 and p<0.001, respectively),28 and immune checkpoint score,29 suggesting these patients harbored an immunosuppressed microenvironment (figure 3A,B; p<0.001 and p<0.001, respectively).
Subsequently, we also conducted IHC in the ZSHC, the infiltration levels of 13 types of immune cells were identified and evaluated. The results showed that patients with higher infiltration of NKG2A+CD8+ T cells had more CD8+ T cells and Treg cells infiltration (figure 3C and online supplemental figure 3A; p=0.011 and p=0.006, respectively), while lower infiltration of M2 macrophages and NK cells (figure 3C and online supplemental figure 3A; p=0.018 and p=0.046, respectively). These findings further validated the results from TCGA-KIRC and CheckMate cohort (figure 3A,B). Moreover, the intense infiltration of CD8+ T cells was accompanied by a higher abundance infiltration of Treg cells (figure 3C). TCR Shannon score reflecting TCR diversity and TCR Richness score were higher in NKG2A+CD8+T cells highly infiltrated group (figure 3D; p<0.001 and p<0.001, respectively). However, the TCR evenness score was found to be lower in NKG2A+CD8+ T cells higher infiltration group (figure 3D; p<0.001). Molecular features of ccRCC have been reported to be associated with immune status, and we have explored the relationship between common mutations and NKG2A+CD8+ T cells infiltration. Our investigation has found that SETD2 mutation patients have more NKG2A+CD8+ T cells infiltration (online supplemental figure 3B). Collectively, our data showed that patients with high infiltration of NKG2A+CD8+T cells had an infiltrated but suppressive immune microenvironment.
Supplemental material
NKG2A+CD8+ T cells infiltration predicts poor immunotherapy responses in ccRCC
Given that NKG2A+CD8+ T cells were in an exhausted state and their infiltration was linked to a suppressive immune milieu, we assessed whether the infiltration level of NKG2A+CD8+ T cells could impact the prognosis of ICB therapy.
In the ZSHC-mRCC cohort, we evaluated the relationship of NKG2A+CD8+ T cells and corresponding clinical annotations (figure 4A and online supplemental table 7). The clinicopathological characteristics of the patients from ZSHC-mRCC cohort are summarized in online supplemental table 7. In terms of OS and progression-free survival (PFS), patients with higher infiltration of NKG2A+CD8+ T cells showed poorer sensitivity to ICB-based therapy (figure 4B,C; p=0.010 and p=0.079, respectively), but not to TKI alone (figure 4D; p=0.687).
Supplemental material
NKG2A+CD8+ T cells infiltration predicts poor immunotherapy responses in ccRCC. (A) Heatmap showed clinical features of the ZSHC-mRCC (ICB) cohort. (B–D) Kaplan-Meier analysis of patients with RCC in ZSHC-mRCC (ICB) (B and C) cohort and ZSHC-mRCC (TKI) cohort (D). (E–H) Combination of CD8+ T cells infiltration and NKG2A expression could predict response to nivolumab in CheckMate cohort. (I–J) NKG2A+CD8+ T cells infiltration could predict prognosis of patients with ccRCC treated with nivolumab (I), but not everolimus (J). Kaplan-Meier analysis and log-rank test. ccRCC, clear cell renal cell carcinoma; CR, complete response; ISUP, international society of urological pathology; ICB, immune checkpoint blockade; mRCC, metastatic renal cell carcinoma; NA, not available; ORR, objective response rate; PR, partial response; PD, progressive disase; SD, stable disease; TKI, tyrosine kinase inhibitors.
We further tried to validate the above findings in the CheckMate cohort. First, we assessed the impact of CD8+ T cells and NKG2A mRNA on the OS of the nivolumab treated arm. We found that neither the CD8+ T cells infiltration level nor the NKG2A mRNA level alone could significantly predict the prognosis in both nivolumab and everolimus treatment arm (online supplemental figure 4A–D; p=0.732, p=0.251, p=0.799 and p=0.332, respectively). Furthermore, when combining CD8+ T cells infiltration with NKG2A mRNA level, we observed that patients with higher CD8+ T cells infiltration had worse OS only when their NKG2A mRNA expression was at a high level (figure 4E,F; p=0.594 and p=0.041, respectively). However, this phenomenon was not observed in the everolimus arm (figure 4G,H; p=0.975 and p=0.248, respectively).
Supplemental material
By further simulating the infiltration levels of NKG2A+CD8+ T cells using the NKG2A+CD8+ T cells signature, we found that patients with higher NKG2A+CD8+ T cells signature had worse OS with nivolumab treatment (figure 4I; p=0.037), but not with everolimus treatment (figure 4J; p=0.818). Interestingly, within the cohort characterized by high infiltration of NKG2A+ CD8+ T cells, there was an upregulation in the expression of T-effector, myeloid inflammation and IFN-γ-related features, concomitant with a discernible downregulation in the expression of angiogenesis feature (online supplemental figure 5; p<0.001, p<0.001, p<0.001 and p=0.021, respectively).30 31 This suggested that high NKG2A+ CD8+ T cells infiltration could lead to resistance to immunotherapy, and thus new treatment approaches are urgently needed.
Supplemental material
NKG2A blockade synergizes with anti-PD-1 therapy in reinvigorating CD8+ T cells cytotoxicity
Since higher infiltration of NKG2A+ CD8+ T cells was associated with poorer response to immunotherapy, we further explored potential treatments for these patients by using an in vitro intervention model with fresh ccRCC tissues. Intervention was divided into four groups, namely, control group, PD-1 monoclonal antibody (nivolumab) treated group, NKG2A monoclonal antibody (mAb) treated group, and combined therapy group (figure 5A). After in vitro intervention with nivolumab, patients with lower NKG2A expression on CD8+ T cells exhibited significant upregulated expression of IFN-γ and GZMB in CD8+ T cells (figure 5B–D), confirming the result that patients with higher infiltration of NKG2A+ CD8+ T cells had worse responses to immunotherapy. Additionally, after nivolumab intervention, CD8+ T cells showed an increased expression of NKG2A (figure 5E; p=0.047), suggesting that combination therapy based on anti-NKG2A and anti-PD-1 might add potential value to PD-1 blockade monotherapy.
NKG2A blockade synergizes with anti-PD-1 therapy in reinvigorating CD8+ T cells cytotoxicity. (A) An in vitro intervention model was established using fresh ccRCC tissues from 10 patients to assess the ability of PD-1 monoclonal antibody and NKG2A monoclonal antibody to reactivate CD8+ T cells. (B–D) 10 ccRCC tissue samples were collected and divided into two subgroups based on the expression of NKG2A on CD8+ cells. (B) In tumors with NKG2Alow CD8+ T cells infiltration group (C) the application of the PD-1 inhibitor nivolumab significantly increased the expression of IFN-γ and GZMB in CD8+ T cells; however, the efficacy of nivolumab diminished in tumors with NKG2AhighCD8+ T cells infiltration group (D). Statistical significance was determined using paired t-test or Wilcoxon signed-rank test. Each point represents an individual patient. (E) After nivolumab treatment, NKG2A expression on CD8+ T cells was upregulated. (F–H) The NKG2A and PD-1 dual intervention group exhibits a stronger ability to reactivate CD8+ T cells, with a certain advantage in improving both the degranulation (F) and cytotoxicity (G–H) functions of CD8+ T cells. (I) No significant difference of TCF-7 was observed among four groups. Significance values were determined using RM one-way analysis of variance and Tukey’s multiple comparison test. *p<0.05, **p<0.01, ***p<0.001, ns refers to no significance. ccRCC, clear cell renal cell carcinoma; GZMB, granzyme B; IFN, interferon; PD-1, programmed cell death protein-1; RM, repeated measures; TCF-7, transcription factor 7.
Subsequently, we conducted a comprehensive evaluation of the dynamic functional alterations within CD8+ T cells under varying intervention regimens. Remarkably, both the group receiving anti-NKG2A treatment and co-inhibition of NKG2A and PD-1 treatment exhibited profound enhancements in the degranulation function (CD107a) of CD8+ T cells (figure 5F; p=0.047 and p=0.013, respectively), while the group treated with nivolumab failed to manifest such improvement (figure 5F; p=0.241). Combination of anti-NKG2A and anti-PD-1 treatment boosted the cytotoxicity of CD8+ T cells with the observation of a significant augmentation of IFN-γ secretion was discernible solely in the combinatorial therapy group (figure 5G; p=0.017), and outperformed the anti-NKG2A monotherapy group (figure 5G; p=0.011). Furthermore, it is noteworthy that the combinatorial therapy could significantly elevate the expression of GZMB within CD8+ T cells (figure 5H; p=0.007). But there was no difference in the expression of TCF-7 (figure 5I). Overall, these findings indicated that the combined therapy of anti-PD-1 and anti-NKG2A results in enhanced CD8+ T cells effector function, which may potentially benefit patients resistant to anti-PD-1/PD-L1-based immunotherapy.
Discussion
TME assumes a crucial role in orchestrating the dynamics of tumor proliferation, metastasis as well as response to therapeutic modalities.32 Within this complex milieu, adaptive immune responses triggered by adaptive immune cells holds paramount significance in the context of immunotherapy for ccRCC, exerting an influence on responsiveness and prognosis of patients.33 Among these immune orchestrators, CD8+ T cells are regarded as the most powerful effectors in the antitumor immune response, stand as the very backbone of cancer immunotherapy.34 Interestingly, in contrast to other tumors, a pronounced infiltration of CD8+ T cells is correlated with less favorable prognosis in ccRCC, which is likely due to functional subdivision of distinct CD8+ T cells subtypes.10–13 Thus, further clustering of CD8+ T cells in ccRCC holds significant implications. In light of these intricate immune dynamics within the ccRCC microenvironment, our investigation delves further into the specific phenotype of NKG2A+CD8+ T cells and its potential role in treatment resistance.
Single agent immunotherapy could induce treatment resistance in a portion of patients.35 In the current study, we have discovered that this might be associated with the specific phenotype of NKG2A+CD8+ T cells. Using FCM, we discerned that NKG2A+CD8+ T cells express higher levels of various immune checkpoints compared with NKG2A−CD8+ T cells, including PD-1, LAG-3, TIM-3, and CD39. Besides, we observed an impaired effector function of NKG2A+CD8+ T cells in ccRCC with reduced secretion of IFN-γ. Previous researches indicated that NKG2A+CD8+ T cells exhibit an activated phenotype and represent a subset of transiently exhausted CD8+ T cells.36 An elevated infiltration of NKG2A+CD8+ T cells might potentially correlate with enhanced immunotherapy response.36 Our research delineated that patients with higher infiltration of NKG2A+CD8+ T cells manifest suboptimal responses to anti-PD-1 therapy. This could be explained by the exhausted phenotype of NKG2A+CD8+ T cells in ccRCC presented in this study. An additional rationale may be attributed to the functional status of NKG2A+CD8+ T cells being contingent on the expression of its corresponding ligand (HLA-E). Considering the elevated expression levels of HLA-E observed in ccRCC relative to other solid tumors,16 it is plausible to hypothesize that HLA-E might be another contributor to the discrepant effector function state of NKG2A+CD8+ T cells.
We have also revealed that individuals with higher NKG2A+CD8+ T cells infiltration exhibited increased infiltration of Treg cells. Additionally, there is an overall higher expression of immune checkpoints in microenvironment enriched with NKG2A+CD8+ T cells infiltration. Immune score analysis indicates that individuals with higher infiltration of NKG2A+CD8+ T cells display stronger immune response. Interestingly, meanwhile, this occurrence is accompanied by a generalized augmentation in the expression of immune checkpoints, contributing to the formation of a relatively suppressive TME in these patients. TCR Shannon score reflects the diversity of TCR repertoire and has been related to positive immunotherapeutic responses in various cancers.37 38 Indeed, previous study of the RCC TME reported that the TCR diversity was lower than in normal tissue.39 Intriguingly, in our study, we found TCR diversity, reflected by TCR Shannon score, was enriched in patients with high NKG2A+CD8+ T cells infiltration, and ICB resistant subgroup lack of specific TCR amplification for tumor antigens. Consequently, individuals with higher infiltration of NKG2A+CD8+ T cells would tend to experience relatively unfavorable prognosis and resistance to anti-PD-1/PD-L1-based immunotherapy.
ICBs have advanced the treatment landscape of mRCC,40 However, there exist individuals who remain unresponsive to contemporary immunotherapy.20 Furthermore, most patients may inevitably acquire resistance after a period of treatment where treatment options remain limited following immunotherapy resistance.41 Patients with higher infiltration of NKG2A+CD8+ T cells were observed to suffer diminished overall prognosis when treated with anti-PD-1 therapy. Notably, existing guidelines endorse limited alternative treatment options for this group of patients.42 Recently, a phase II clinical trial of NKG2A inhibitors reported combination of durvalumab and monalizumab (NKG2A monoclonal antibodies) outperformed durvalumab monotherapy (objective response rate(ORR): 35.5% vs 17.9%) in lung cancer, underscoring the significant promise of combined immunotherapy.43 We speculate that an amalgamation of NKG2A and PD-1 immune checkpoint inhibitors may proffer prospective benefits in ccRCC. Through an in vitro intervention model, we have validated for the first time in human tissue-derived samples that the combination of PD-1 and NKG2A blockade exhibited enhanced capacity to reactivate CD8+ T cells, outperforming the efficacy of single ICB therapy in ccRCC.
Our study held retrospective nature with limited samples included. For example, in vitro intervention experiments did not obtain metastatic sample tissues for validation. More in-depth research and clinical trials are needed in the future to confirm the potential of NKG2A as a new target in ccRCC.
In conclusion, we demonstrated NKG2A+CD8+ T cells were associated with unfavorable prognosis and response to immunotherapy. The synergistic interplay of NKG2A blockade and current immunotherapy exhibited a formidable capacity to reactivate CD8+ T cells effector functions. These findings herald fresh perspectives for the advancement of ccRCC immunotherapy.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The study followed the Declaration of Helsinki and was approved by the Clinical Research Ethics Committee of Zhongshan Hospital, Fudan University (No. B2022-114R). Signed informed consent was obtained from each patient.
Acknowledgments
We thank Dr Lingli Chen (Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, China) and Dr Yunyi Kong (Department of Pathology, Fudan University Shanghai Cancer Center, Shanghai, China) for their excellent pathological technology help.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
YQ, LL, WJ and ZX contributed equally.
Contributors YQ, LL, WJ and ZX for acquisition of data, analysis and interpretation of data, statistical analysis and drafting of the manuscript. JW and SD for technical and material support. JG and JX for study concept and design, analysis and interpretation of data, drafting of the manuscript, obtained funding and study supervision. JX as the guarantor for this study. All authors read and approved the final manuscript.
Funding This study was funded by grants from National Natural Science Foundation of China (81974393, 82272776, 82272786) and Shanghai Municipal Health Bureau Project (2020CXJQ03, 202240291). All these study sponsors have no roles in the study design, in the collection, analysis and interpretation of data.
Competing interests No, there are no competing interests.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.