記住我
Research ArticleImmunologyOncology
Open Access | 
10.1172/JCI177082
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Hansen, U. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Church, C. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Carnaz Simões, A. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Frej, M. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Bentzen, A. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Tvingsholm, S. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Becker, J. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Fling, S. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Ramchurren, N. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Topalian, S. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Nghiem, P. in: JCI | PubMed | Google Scholar
1Section of Experimental and Translational Immunology, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark.
2PokeAcell Aps, BioInnovation Institute, Copenhagen, Denmark.
3Department of Dermatology, Department of Medicine, University of Washington, Seattle, Washington, USA.
4Department of Translational Skin Cancer Research, University Hospital Essen and German Cancer Consortium (DKTK), Essen, Germany.
5German Cancer Research Center (DKFZ), Heidelberg, Germany.
6Department of Dermatology, University Hospital Essen, Essen, Germany.
7Fred Hutchinson Cancer Center, Seattle, Washington, USA.
8Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.
9Bloomberg-Kimmel Institute for Cancer Immunotherapy, Sidney Kimmel Comprehensive Cancer Center, Baltimore, Maryland, USA.
Address correspondence to: Sine Reker Hadrup, Kemitorvet Bld. 204, 2800 Kongens Lyngby, Denmark. Phone: 45.27.12.52.21; Email: sirha@dtu.dk.
Find articles by Hadrup, S. in: JCI | PubMed | Google Scholar
Published January 30, 2024 - More info
Published in Volume 134, Issue 8 on April 15, 2024
Related article:
Abstract
Merkel cell carcinoma (MCC) is an aggressive, fast-growing, highly metastatic neuroendocrine skin cancer. The Merkel cell polyomavirus (MCPyV) is an oncogenic driver in the majority of MCC tumors. In this issue of the JCI, Hansen and authors report on their tracking of CD8+ T cells reactive to MCPyV T antigen (T-Ag) in the peripheral blood of 26 patients with MCC who were undergoing frontline anti–programmed cell death protein-1 (anti–PD-1) immunotherapy. They discovered unique T cell epitopes and used the power of bar-coded tetramers to portray immune checkpoint inhibitor–induced immunogenicity as a predictor of clinical response. These findings provide the foundation for therapeutic possibilities for MCC, including vaccines and adoptive T cell– and T cell receptor–driven (TCR-driven) treatments.
Authors
Michael K. Wong, Cassian Yee
× AbstractMerkel cell carcinoma (MCC) is a highly immunogenic skin cancer primarily induced by Merkel cell polyomavirus, which is driven by the expression of the oncogenic T antigens (T-Ags). Blockade of the programmed cell death protein-1 (PD-1) pathway has shown remarkable response rates, but evidence for therapy-associated T-Ag–specific immune response and therapeutic strategies for the nonresponding fraction are both limited. We tracked T-Ag–reactive CD8+ T cells in peripheral blood of 26 MCC patients under anti-PD1 therapy, using DNA-barcoded pMHC multimers, displaying all peptides from the predicted HLA ligandome of the oncoproteins, covering 33 class I haplotypes. We observed a broad T cell recognition of T-Ags, including identification of 20 T-Ag–derived epitopes we believe to be novel. Broadening of the T-Ag recognition profile and increased T cell frequencies during therapy were strongly associated with clinical response and prolonged progression-free survival. T-Ag–specific T cells could be further boosted and expanded directly from peripheral blood using artificial antigen-presenting scaffolds, even in patients with no detectable T-Ag–specific T cells. These T cells provided strong tumor-rejection capacity while retaining a favorable phenotype for adoptive cell transfer. These findings demonstrate that T-Ag–specific T cells are associated with the clinical outcome to PD-1 blockade and that Ag-presenting scaffolds can be used to boost such responses.
Graphical Abstract
Introduction
Immunotherapy based on blockade of the programmed cell death protein-1/programmed cell death ligand 1 (PD-1/PD-L1) pathway has demonstrated pronounced efficacy in restoring antitumor activity in some patients with solid tumors (1). The rare skin cancer Merkel cell carcinoma (MCC) has demonstrated profound responsiveness to PD-1/PD-L1 blockade, with some of the highest response rates (56%–62%) obtained among all solid cancers (2–4). Of MCC cases, 80% are caused by Merkel cell polyomavirus (MCPyV), the oncogenic potential of which requires integration into the host genome and truncation of the large T antigen (LTA), leading to the expression of the viral oncogenic T antigens (T-Ags; LTA and small T-Ag [STA]) (5). Virus-positive MCC has a low tumor mutation burden (TMB) (6), which for other tumor types is associated with lower response rates to immune checkpoint inhibition (ICI) (1). Instead, it is hypothesized that the viral T-Ags are targets for the immune recognition responsible for the high ICI response rate in this tumor type. A persistent expression of these T-Ags is essential for tumorigenesis and maintenance (7, 8), making them ideal targets for adaptive immune control. In fact, immune surveillance has already proven critical for tumor control, through positive associations between survival and intratumoral levels of both CD3+ and CD8+ lymphocytes (9, 10) and CD8+ T cells reactive toward KLLEIAPNC, an epitope embedded in the overlapping sequence of LTA and STA (common T-Ag [CT]) (11). Serum levels of anti–T-Ag antibodies have additionally served as an indicator of disease burden (12), and T-Ag–reactive CD8+ T cells are exclusively detected in MCC patients with virus-positive tumors compared with healthy donors and MCC patients with virus-negative tumors (13, 14). Furthermore, clinical evidence suggests that strong adaptive immune recognition is important for tumor control. This includes the rare events of complete spontaneous regression observed (15) together with increased incidence rates in individuals suffering from systemic immune suppression, such as those with HIV and solid organ transplant recipients (16, 17). Thus, we hypothesize that MCPyV-reactive CD8+ T cells play a role in the antitumor response generated by ICI in MCC patients.
Understanding the immune response to MCC during ICI therapy is important for improving therapeutic efficacy further. Across solid tumors, others have suggested that blockade of the PD-1/PD-L1 pathway leads not only to reinvigoration of preexisting dysfunctional T cells in the tumor microenvironment, but also infiltration of new tumor-reactive T cell clones (18, 19). This is supported by evidence of rapid and robust T cell proliferation in the periphery following ICI therapy (20–22). Such peripheral immune induction would be easily measurable with a minimally invasive blood-based source, which would allow fast evaluation and prediction of therapy response. In the current study, we utilized high-throughput screening technology with DNA barcode-labeled multimers (23). This allowed us to study a comprehensive panel of potential T-Ag–derived T cell epitopes, covering the full HLA ligandome and restricted to a broad range of HLA haplotypes, thereby capturing the majority of T-Ag–reactive CD8+ T cells present in the circulation of ICI-treated MCC patients to evaluate their response to therapy.
Despite the high response rate to PD-1/PD-L1 blockade, in the end, around half of MCC patients do not derive durable benefit from therapy, and no strong alternative therapeutic strategy for this cohort currently exists. Given the tumor-exclusive and required expression of the T-Ags, these are ideal targets for a precision-targeted T cell therapy approach. Furthermore, the T-Ags are shared across all patients and would not require personalized antigen (Ag) selection. To date, adoptive cell therapy (ACT) of single epitope expanded T cells has been clinically tested in MCC, but with limited efficacy due to acquired immune escape by HLA class I allelic loss (24, 25). Currently, high-affinity TCR-transgenic T cells targeting the CT-derived epitope, KLLEIAPNC (ClinicalTrials.gov NCT03747484) are being clinically evaluated. Thus, to leverage the knowledge of a broad repertoire of CD8+ T cell epitopes derived from T-Ag presented herein and in literature (13, 14, 26), we examined the capacity to expand such T-Ag–specific T cells from the peripheral blood through coordinated peptide-HLA and cytokine stimulation by artificial Ag-presenting scaffolds (27). This approach maintains a broad T cell recognition and clonality profile, while generating T cell products with advantageous phenotype and potent tumor-rejection capacity. Such a therapeutic strategy could be used to further boost the T cell compartment in patients with partial response to ICI or facilitate antitumor responses in patients with primary ICI resistance.
ResultsBroad recognition of MCPyV-derived epitopes across a wide range of HLA haplotypes. To perform a comprehensive evaluation of circulating MCPyV-specific CD8+ T cells in MCC patients undergoing ICI, we first generated an extensive library of potential CD8+ T cell epitopes from the T-Ag proteins (truncated LTA and STA) and viral capsid protein 1 (VP1). We included the full predicted ligandome for 33 HLA class I haplotypes to ensure broad patient coverage through in silico binding prediction of 9- and 10-mer peptides to the 33 HLA class I haplotypes using netMHCpan 4.0 (28) with a predicted eluted ligand percentile rank score cutoff of 2. This resulted in 1,490 unique peptide-MHC (pMHC) complexes used for evaluation of T cell recognition, of which 714 presented T-Ag–derived peptides, and these were distributed with 7–38 peptides presented per HLA haplotype (Figure 1, A and B, and Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/JCI177082DS1). Additional control epitopes from common nononcogenic viruses, including CMV, EBV, and influenza (FLU), were available for 10 of the HLA haplotypes and served as technical validation for the T cell detection process (Figure 1B). These will be referred to as CEF peptides.
Figure 1Screening with DNA barcode–labeled pMHC multimers. (A) Schematic overview of the in silico peptide prediction for selecting the library of MCPyV-derived peptides. Created with BioRender. (B) The distribution of the peptides across the 33 HLA haplotypes, colored based on the protein of origin. (C) Experimental workflow for the detection of multimer-reactive CD8+ T cells. Created with BioRender. (D) Flow plots for the 2 multimer design strategies with either single multimer color for all peptides (PE) (left) or 2 multimer colors separating T-Ag peptides (PE) and VP1+CEF peptides (APC) (right). (E) Representative examples of screening results for patient 4 and 2 non–ICI-treated healthy donors screened in parallel. T cell recognition of a given epitope is defined by significant enrichment of the pMHC-assigned DNA barcode with log-fold change > 2 and FDR < 0.001, indicated by the dotted line. T cell epitopes are colored based on the protein of origin. (F) Combined screening results for all patients divided based on HLA haplotype. Number of patients screened with a given haplotype is indicated above the graph.
Peripheral blood samples from 26 patients enrolled in the Cancer Immunotherapy Trials Network CITN-09 clinical trial (ClinicalTrials.gov NCT02267603) were included with 1–4 PBMC samples obtained before and/or on anti-PD1 therapy (Supplemental Table 2). Patient blood samples were screened with HLA-matched DNA-barcoded pMHC multimers carrying the above-selected peptide library as schematically depicted in Figure 1C. Utilizing this screening technology allowed us to identify T cell recognition against a large number of pMHC specificities simultaneously while maintaining the pMHC-specificity knowledge, since every peptide specificity is identifiable by its DNA barcode tag (23). The patients’ samples were screened with 45–302 pMHC multimers covering on average 74% of their HLA class I haplotypes. The HLA haplotype C*07:02 was later excluded due to technical concerns, and 4 haplotypes (B*37:01, B*40:01/02, and C*02:02) were not represented in our patient material (Supplemental Figure 1A). For half of the patients, all multimers had a common phycoerythrin (PE) label for sorting multimer-binding CD8+ T cells. For the other half, 2 fluorescent labels, PE and allophycocyanin (APC), were associated with either T-Ag or the controls VP1 and CEF, respectively (Figure 1D, full gating strategy in Supplemental Figure 1B). This allowed us to include a 12-parameter T cell phenotype panel during the staining step in order to compare the phenotypes of T cells recognizing oncogenic versus nononcogenic viral elements. The associated DNA barcodes from the sorted multimer-binding cells, irrespective of their fluorescence label, were amplified and sequenced to reveal DNA barcodes enriched in the sorted T cell fraction compared with baseline levels with a false discovery rate (FDR) of less than 0.001, defining significant T cell recognition of the corresponding peptide.
The enrichment of pMHC-binding T cells (log fold change) of all pMHC multimers for patient 4 is shown in Figure 1E, with the dotted line representing the threshold of significantly enriched pMHC binding from T cells and the vertical lines separating the 4 blood samples screened. No T-Ag–specific cells were detected prior to therapy, but following ICI treatment, recognition of a CT- and LTA-derived epitope was detected, in particular the A*01:01-restricted epitope, AAFKRSCLK, which was recognized by T cells in PBMCs at all time points after treatment initiation. Additional VP1 and CEF epitopes were recognized by T cells throughout and are the only epitope types recognized in the 2 healthy donors screened in parallel as technical controls for the screening process (Figure 1E).
In total, 172 multimer-reactive CD8+ T cell populations were detected across all samples and protein types with restriction to 20 of the 28 included HLA haplotypes (Figure 1F). Large variations can be observed between the HLA haplotypes in terms of recognized epitopes, which may potentially be affected by the low representation of patients with certain HLA types.
T cell reactivity detected to 32 T-Ag–derived epitopes exclusively in MCC patients. Of all the multimer-reactive CD8+ T cell populations, 46 were T-Ag–specific, and hence tumor relevant, with 1–8 populations detected in 14 patients with MCPyV-positive tumors at summed estimated frequencies among CD8+ T cells ranging from 0.1% to 1% (Figure 2A). Since multiple blood samples were screened from individual patients, the number of unique T-Ag epitopes recognized per patient ranged from 1 to 6. The 2 patients with MCPyV-negative tumors had no detectable T-Ag recognition, in line with the lack of tumor expression of these oncogenes. VP1 and CEF epitopes were detectable in a large proportion of the patients as well (Supplemental Figure 1, C and D). Forty healthy donors were screened in parallel with the patient cohort, and here, no T cell recognition was observed against T-Ags, only against VP1- and CEF-derived epitopes, thus validating the tumor-specific characteristics of T-Ag expression and T cell recognition (Figure 2B), in agreement with our previous observations (13, 14).
Figure 2Characterization of the 32 recognized T-Ag epitopes. (A) The numbers of unique (bar) and total (dot) T-Ag epitopes recognized by T cells across all time points in 14 out of 26 MCC patients with tumor MCPyV status are indicated. The size of the circles varies with the summed frequency of T-Ag–specific T cells, across all time points. (B) T cell recognition of the 3 proteins within the healthy donor cohort (n = 40). ****P < 0.0001, Kruskal-Wallis test with Dunn’s correction. (C) Prevalence of the 32 recognized T-Ag epitopes out of screened patients with aa, HLA haplotype, and number of screened patients provided. Unreported epitopes are highlighted in bold. (D–G) Epitopes divided into their proteins of origin (LTA, orange; STA, blue; CT, green) and displayed as either total T cell populations detected (D), unique CD8+ T cell epitopes (E), prevalence in cohort (F), or immunogenic peptides out of total peptides screened within each protein (G). (H) Bars show the number of peptides screened within each HLA haplotype, with the blue fraction indicating those recognized by T cells (left y axis) and diamonds marking the percentage of immunogenic peptides within each HLA (right y axis).
Overall, we detected T cell recognition toward 32 T-Ag epitopes with a prevalence between 11% and 100% of the screened patients for a given HLA haplotype (Figure 2C). However, for 5 epitopes, only a single patient was screened; therefore, excluding these could give a more correct prevalence of 11% to 42%. Of the detected epitopes, 20 were previously unreported (13, 14, 26). The CT region appeared more immunogenic compared with the nonoverlapping sequences of LTA and STA, both in terms of the total numbers of T cell populations detected (Figure 2D) and the number of unique T cell epitopes derived from this region (Figure 2E). In addition, a higher proportion of patients had T cells recognizing a minimum of 1 epitope derived from the CT region (Figure 2F), and the fraction of the predicted peptides recognized by T cells (i.e., defined as immunogenic peptides) was higher for CT (Figure 2G). These observations can potentially be explained by the higher copy number of this region, since it is expressed with both LTA and STA. The T-Ag epitopes were restricted to 16 different HLA haplotypes, with B*07:02 and B*51:01 showing the highest percentages of immunogenic epitopes out of total screened peptides, (27% and 38%, respectively; Figure 2H). The immunogenic T-Ag epitopes could be distinguished from the nonimmunogenic peptides by an improved MHC-binding affinity (Supplemental Figure 2, A and B), but no difference was observed in their MHC-binding stability (Supplemental Figure 2, C and D).
T-Ag–specific T cell populations are associated with clinical response to ICI. To test the hypothesis that T-Ag–specific T cells contribute to tumor recognition and elimination following ICI treatment, we evaluated the association of T-Ag–restricted T cell recognition with clinical outcomes. The patients were grouped according to RECIST criteria (29), as either responders (complete response [CR] or partial response [PR]) or nonresponders (stable disease [SD] or progressive disease [PD]). The overall kinetics of the T-Ag–specific T cells during ICI therapy was different between the 2 patient groups, with a substantial increase in T-Ag–specific T cells observed only in the clinical response group (Figure 3A). Since the posttherapy blood samples were not available at all potential time points (3, 12, and 18 weeks) from several patients, the following analyses were performed with a pooled posttherapy measurement, either the 3-week sample, the 12-week sample, or an average of these 2 time points when both were available. Data from the individual time points are plotted in Supplemental Figure 3. We detected a significantly higher number of T-Ag–specific T cell populations in the clinical responders after treatment initiation (Figure 3B) compared with nonresponders. These T cell populations were enriched both in terms of their breadth of response, i.e., the number of different T-Ag–derived epitopes recognized (Figure 3B and Supplemental Figure 3A), and the magnitude of such T cell populations in the peripheral blood, i.e., the sum of estimated frequencies of T-Ag–specific T cells out of CD8+ T cells at each given time point (Figure 3C and Supplemental Figure 3B). Interestingly, the induction of T-Ag–specific T cells was observed early during ICI therapy for patients with a CR outcome (3–12 weeks), but seemed slightly delayed for the patients with a PR (12 weeks, Supplemental Figure 3, C and D), although it is plausible that this trend in PR patients may be related to limited patient sample availability for the 3-week time point.
Figure 3T-Ag–reactive T cells are associated with clinical benefit of ICI. (A) Number of T-Ag–reactive T cell populations detected during ICI therapy. The patients are divided based on their RECIST criteria into responders (CR and PR, n = 17) and nonresponders (SD and PD, n = 7) and colored accordingly with size of circles indicating the summed estimated (est.) frequency of T-Ag–reactive T cells out of CD8+. *P < 0.05, 2-way ANOVA. (B) Number of T-Ag–specific T cells detected for the 2 patient groups before and after therapy initiation. The pooled posttherapy number was based on either 3-week or 12-week time points or an average of both. *P < 0.05, Kruskal-Wallis test with Dunn’s correction. (C) The sum of estimated frequency of T-Ag–reactive T cells before and after therapy for patient groups. *P < 0.05, Kruskal-Wallis test with Dunn’s correction. (D) Change in the sum of estimated frequency before and after therapy. *P < 0.05, Mann-Whitney U test. (E) Number of VP1- and CEF-specific T cells detected for the 2 patient groups before and after therapy initiation. B–E are presented with box plots displaying the interquartile range. (F) Progression-free survival curves split based on detectable (n = 13) or nondetectable (n = 7) T-Ag–reactive T cells at any time point after ICI therapy initiation. Significance levels and hazard ratios are denoted; log-rank (Mantel-Cox) test. (G) Progression-free survival curves for detectable (Detect.) T-Ag–reactive T cells split by median baseline tumor burden (diameter = 42 mm) and for nondetectable (Nondetect.) T-Ag–reactive T cells split by median baseline tumor burden (diameter = 15 mm). Significance levels are denoted, log-rank (Mantel-Cox) test. TB, tumor burden.
In the pretreatment blood samples, we found that T-Ag was exclusively recognized in patients who later responded to treatment, yet in the majority of responding patients, such T cell populations were undetectable prior to treatment initiation. Thus, while few patients had detectable T-Ag–specific T cell populations before ICI treatment, the capacity to mount or enhance the T cell response in association with ICI therapy was significantly stronger in the responder group, as measured by the increase in T-Ag–specific T cell frequencies from before to after therapy (Figure 3D). In contrast, no differences were observed in T cell recognition of the control peptides, VP1 and CEF, that were screened in parallel (Figure 3E).
To further investigate the importance of T-Ag–reactive T cells in mediating ICI response, we evaluated progression-free survival end points. The patients were divided according to the presence or absence of detectable T-Ag–specific T cell populations after therapy initiation (Figure 3F), and each of these groups was further split based on tumor burden (Figure 3G). Those having detectable T-Ag–specific T cells showed a trend toward enhanced progression-free survival, with significant differences when including tumor burden. However, given the small-size cohort, additional adjustments for confounding factors were not feasible and may influence this trend.
T-Ag–specific CD8+ T cells are marked by distinct CD39 and Ki67 expression. For 14 of the patients’ samples, a 12-parameter flow cytometric antibody-staining panel was employed to study T cell phenotypes associated with T-Ag–specific T cells compared with VP1- and CEF-specific T cells or the remaining bulk CD8+ T cells both prior to and during ICI. For the nonresponder group, no T-Ag–reactive cells were detected prior to therapy, and therefore the phenotype is provided only for bulk CD8+ T cells with undetermined specificities. In the responders prior to ICI therapy, T-Ag–specific T cells demonstrated significantly increased levels of CD39 (Figure 4A), indicating Ag recognition and possible exhaustion. This population also displayed a high level of HLA-DR, indicating Ag-mediated activation. The remaining phenotype characteristics were similar to the VP1- and CEF-specific and bulk CD8+ T cell compartments (Figure 4A). In on-treatment specimens, CD39 expression remained increased on T-Ag–specific cells compared with VP1- and CEF-specific and bulk CD8+ T cells, both in the responding and nonresponding patients (Figure 4B), demonstrating CD39 as a signature of T-Ag recognition in MCC patients. Interestingly, an increased expression of the proliferation marker Ki67 was observed after therapy initiation, with a significant increase in Ki67+ T-Ag–specific T cells compared with bulk CD8+ T cells following ICI therapy (Figure 4B). Dividing the samples into individual time points, irrespective of response to ICI, we found that both CD39 and Ki67 expression peaked 3 weeks after treatment initiation, followed by a decline (Figure 4, C and D). The coexpression of CD39 and Ki67 appeared specifically associated with T-Ag–specific T cells at all time points, with borderline significance at week 12 (Figure 4E).
Figure 4Individual marker expression before and after ICI therapy initiation. (A and B) Expression of phenotypic markers before (A) and after ICI therapy initiation (B) for responders (blue) CD8+ T cells with T-Ag recognition (blue filled), with VP1+CEF recognition (gray filled) or unspecific bulk cells (open circle, blue outline); and nonresponders (red) CD8+ T cells with T-Ag–recognition (red filled), VP1+CEF recognition (gray filled), or unspecific bulk cells (open circle, red outline). *P < 0.05, Kruskal-Wallis test with Dunn’s correction. CM, central memory. (C) CD39 expression on T-Ag– or VP1+CEF-specific cells divided into the 3 time points tested: prior to ICI (0 weeks) and 3 weeks and 12 weeks after ICI initiation. Kruskal-Wallis test with Dunn’s correction. (D) Ki67 expression on T-Ag– or VP1+CEF-specific cells divided into the 3 time points tested. Kruskal-Wallis test with Dunn’s correction. (E) Percentage of double-positive for CD39 and Ki67 of T-Ag–specific, VP1+CEF-specific, or unspecific CD8+ T cells divided into the 3 time points tested. Kruskal-Wallis test with Dunn’s correction. All bars display the median and upper quartile. Ctrl, control.
T-Ag–specific T cells can be expanded using artificial Ag-presenting scaffolds. Given that T-Ags are strong tumor targets associated with MCC clearance after ICI, they may serve as ideal targets for adoptive T cell therapies. We therefore explored a recently described strategy to expand multiple T-Ag–specific T cell populations from peripheral blood using artificial Ag-presenting scaffolds (27). The Ag-scaffolds consisted of a dextran backbone carrying the pMHC of interest to allow pMHC-directed binding to specific CD8+ T cells as well as IL-2 and IL-21 to allow cytokine-mediated stimulation exclusively to pMHC-binding T cells in the PBMC pool (Figure 5A). The approach allowed multiple different epitope-specific T cell populations to be expanded simultaneously. We therefore selected 6 prevalent HLA-I haplotypes, A*01:01, A*02:01, A*03:01, A*24:02, B*07:02, and B*08:01, loaded with 4 to 8 T-Ag–derived epitopes each (Figure 5A). The selected epitopes were either detected in the above screening study or previously described (13, 14, 26). The in vitro cell expansion was a 2-week process with Ag-scaffolds supplemented to the cell culture on days 0, 3, 6, and 9 before the T cells were harvested and evaluated on day 14.
留言 (0)