SUMMARY
Despite widespread immune profiling in cancer immunotherapy, the antigen-specific responses that drive clinical outcomes remain poorly defined. In a prospective neoadjuvant trial (NCT04013854) of a single-dose anti–PD-1 (nivolumab) in stage III melanoma, we performed antigen-specific profiling of melanoma and viral-specific CD8+ T cells across blood, tumor, and lymph node compartments. Using combinatorial tetramers, we detected melanoma-specific CD8+ T cells in 72% of HLA-A1, A2, and A3 patients. These cells displayed distinct phenotypes shaped by tissue and antigen context. Tumor-infiltrating T-bet+ intermediate exhausted CD8+ T cells were strongly associated with pathologic response, while CD39+ terminal exhausted cells marked non-response. T-bet and CD39 expression also stratified responses in uninvolved lymph nodes, suggesting early divergence of therapeutic immune trajectories. Longitudinal profiling revealed that circulating melanoma-specific CD8+ T cell dynamics were antigen-dependent and associated with clinical outcomes. Our findings highlight the clinical value of antigen-specific profiling and identify mechanistic correlates of anti-PD-1 efficacy.
INTRODUCTION
Although immune checkpoint blockade (ICB) has transformed cancer treatment, durable remission is achieved in only a subset of patients1. Understanding each patient’s unique immune response to immunotherapy, the specific mechanisms of resistance, and the effective immunotherapeutic strategies tailored to his or her needs is essential for advancing precision immune-oncology. Despite the widespread immune profiling and translational studies that are being performed, identifying clinically useful biomarkers remains challenging. Current biomarkers, including tumor PD-L12–4, tumor mutational burden5–8, and IFNγ signatures9–11 are still limited because they represent imperfect surrogates for assessing tumor-specific CD8+ T cell responses. The ability to consistently monitor tumor-specific CD8+ T cells in the clinical setting would enable greater sensitivity for biomarker and mechanistic discovery.
Identifying tumor-specific CD8+ T cells is difficult due to their low frequency, the heterogeneity of immunodominant epitopes, and the diversity of HLA alleles across patients. Melanoma, however, presents an ideal model for antigen-specific immune profiling, as it is a tumor type that is immunogenic12–15 and ICB responsive16–21. Moreover, the landscape of immunogenic T cell epitopes has been clearly defined22.
Neoadjuvant anti-PD-1 therapy (αPD-1), administered before surgery, offers a unique window to observe the early immune responses elicited by checkpoint inhibition. This approach has demonstrated superiority over adjuvant treatment in event-free survival23. The pharmacodynamic effects of PD-1 blockade occur rapidly, resulting in reinvigoration of exhausted CD8+ T cells as early as 7 days post-treatment and expansion of CD8+ T cells within the tumor24. These early immune responses are associated with pathologic responses and long-term recurrence-free survival25, providing a window for evaluating tumor-specific CD8+ T cell activity and its role in therapeutic outcomes.
In this study, we apply antigen-specific immune profiling to a prospective clinical trial of neoadjuvant PD-1 blockade to track melanoma-specific CD8+ T cells in blood, tumor, and lymph node. Our findings demonstrate that features of melanoma-specific CD8+ T cells, particularly T-bet expression, have the strongest association with pathologic response following neoadjuvant PD-1 blockade.
RESULTS
Combinatorial tetramers identify melanoma and viral-specific CD8+ cells after neoadjuvant PD-1 blockade.
We applied antigen-specific profiling to a phase 2 neoadjuvant anti-PD-1 clinical trial (NCT04013854). Patients with stage IIIA-D resectable melanoma received a pre-treatment biopsy, a single dose of nivolumab (480 mg IV), and underwent definitive surgery four weeks later allowing for the collection of paired blood and tumor specimens (Figure 1A). Patients then received either nivolumab, or nivolumab + ipilimumab as adjuvant therapy, based on their pathologic response. Out of a total of 67 trial patients, 48 had adequate research specimens collected and analyzed. In addition to peripheral blood mononuclear cells (PBMCs), we collected 22 biopsies at baseline and 48 tumor resections and 20 uninvolved lymph nodes at the post-treatment resection timepoint (Table S1). For this study we focused on the early immune responses in blood and tumor after the initial dose of neoadjuvant αPD-1 in this cohort of 48 patients. The median age was 64, and 67% (n = 32) had HLA-A1, HLA-A2, and/or HLA-A3 alleles, for which tetramer reagents were available (Figure 1A and Table S1). Pathologic responses were assessed based on the percentage of viable tumor cells following the initial nivolumab dose. In summary, 42% of patients achieved pathologic responses, including 14 complete responders (pCR, no residual tumor identified), 2 near-complete responders (pNCR, 10% or less viable tumor cells), and 4 partial responders (pPR, 50% or less viable tumor cells), consistent with previous studies24,26–28 (Figure 1B and Table S1).
Figure 1. Combinatorial tetramers identify melanoma and viral-specific CD8+ T cells after neoadjuvant PD-1 blockade.

(A) Study schema with patient demographics and HLA types in neoadjuvant anti-PD-1 (nivolumab) clinical trial.
(B) Pathologic responses separated by complete response (pCR), near-complete response (pNCR), partial response (pPR), and non-response (pNR).
(C) Number of patients with melanoma and viral-specific CD8+ T cells, out of 32 HLA-A1, A2, and A3 patients.
(D) Representative flow plots of melanoma (red) and viral (blue)-specific CD8+ T cells identified by tetramers.
(E) Percentage of melanoma-specific CD8+ T cells in the blood, lymph node, and tumor. Each dot represents an individual tetramer+ population. The middle line and error bars indicate the mean ± SEM.
(F) Percentage of viral-specific CD8+ T cells in the blood, lymph node, and tumor. Each dot represents an individual tetramer+ population. The middle line and error bars indicate the mean ± SEM.
(G) Representative flow plots comparing CD39 and TIM-3 expressions on melanoma-specific CD8+ T cells and CD8+ T cells in the lymph node and tumor.
(H) PCA plot of melanoma-specific CD8+ T cells and CD8+ T cells in the lymph node and tumor. Each dot represents an individual population.
(I) Variables contributing to principal components.
(J) Representative flow plots comparing KI-67 and TCF-1 expressions on melanoma-specific CD8+ T cells in the blood, lymph node, and tumor.
(K) Representative flow plots comparing T-bet and TOX expressions on melanoma-specific CD8+ T cells in the blood, lymph node, and tumor.
(L) Expressions of selected markers on melanoma-specific CD8+ T cells in the blood, lymph node, and tumor. Each line connects paired samples from the same patient. Thick line with dots indicates the group mean.
Statistical significance was assessed using the Wilcoxon rank-sum test for unpaired comparisons and the Wilcoxon signed-rank test for paired comparisons. P-values are reported as follows: ns (not significant), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.
See also Figures S1–S3 and Tables S1–S3.
To identify melanoma-specific CD8+ T cells, we utilized a combinatorial tetramer approach where tetramers of the same specificity were double-labeled with two different fluorophores29–31 (Figure S1A). Using a combination of five flow cytometry channels, we designed tetramer panels specific for HLA-A1, A2, or A3 epitopes that were able to detect up to ten different melanoma and viral-specific CD8+ cells simultaneously in PBMC, regional lymph node, and tumor samples. These tetramer channels were embedded into a 29-parameter spectral cytometer panel for the characterization of antigen-specific CD8+ T cells.
Among 32 patients expressing HLA-A1, A2, and/or A3 alleles, we successfully identified a total of 119 melanoma-specific CD8+ T cell populations in 23 patients (72%) targeting 12 distinct epitopes (e.g. GP100, MAGE A1, and MART-1). We identified melanoma-specific CD8+ T cells in the blood for 18/32 (56%) patients and in the tumor for 21/32 (66%) patients (Table S2). Additionally, we detected 56 viral-specific CD8+ T cell populations in 18 patients (56%). Altogether we were able to profile antigen-specific CD8+ T cells in 28 out of 32 melanoma patients (Figure 1C, Tables S2 and S3) demonstrating the feasibility of our antigen-specific immune profiling approach.
CD8+ T cell atlas of blood, lymph node, and tumor.
To explore the cell states of melanoma-specific CD8+ T cells, we first constructed a CD8+ T cell atlas using flow cytometry data from peripheral blood, lymph node, and tumor samples collected from a subset of 15 patients in the cohort. Dimension reduction of CD8+ T cells using Uniform Manifold Approximation and Projection (UMAP) highlighted tissue-related patterns of co-expression (Figures S1B and S1C). For example, CD8+ tumor-infiltrating lymphocytes (TILs) were characterized by high expressions of PD-1, CD39, and TOX. In peripheral blood, CD8+ T cells exhibited elevated expressions of effector markers like CX3CR1 and T-bet. Conversely, CD8+ T cells in lymph nodes displayed greater expressions of CCR7, and TCF-1, consistent with naïve and memory phenotypes. We performed unsupervised clustering and dimensionality reduction using flowSOM32 revealing 9 distinct phenotypes of CD8+ cells across blood, lymph node, and tumor (Figures S1D–S1F). Naïve (CD45RAhi CD27hi CCR7hi), central memory (CD45RA− CCR7+ CD127hi), transitional memory (CD45RA− CCR7− CD27+), effector memory (CD45RA− CD27−), and TEMRA (CD45RA+ CD27− CX3CR1+) CD8+ T cells were predominantly in the blood (Figure S1G). We also identified subsets of exhausted CD8+ T cells (TEX) which were largely restricted to the tumor (Figure S1G). These TEX subsets included progenitor (ProgEX), intermediate (IntEX), and terminal exhausted (TermEX) CD8+ T cells, as well as an additional TCF-1− TIM-3− CD8+ T cell subset (TCF-1− TIM-3−EX). TEX was defined by high expression of the transcription factor TOX33–37, and varying degrees of inhibitory receptors including PD-1, CD39, TIGIT, and TIM-338–44. ProgEX had high expressions of TCF-1, CXCR5, and CD28, and low expressions of TIM-3 and CD39, while TermEX were characterized by high expressions of CXCR6, TIGIT, TIM-3, CD39, and PD-145–48. IntEX closely resembled TermEX but expressed high levels of T-bet and KI-67 with decreased TOX, consistent with published data49–51. TCF-1− TIM-3−EX displayed a phenotypic profile similar to that of ProgEX, but was distinguished by decreased TCF-1 expression, increased KI-67, and enhanced T-bet expression, reflecting a transitioning state from TCF-1+ TIM-3− to TCF-1− TIM-3+ populations52 (Figure S1F).
Antigen specificity and the tumor microenvironment shape CD8+ T cell phenotypes and determine their cell states.
To decipher the phenotypes of antigen-specific CD8+ cells, we then mapped melanoma and viral-specific CD8+ T cells onto the CD8+ T cell atlas. Melanoma-specific CD8+ T cells were found predominantly in the tumor, while viral-specific CD8+ T cells were mostly found in the blood (Figure S2A). Unsupervised clustering and principal component analysis grouped melanoma-specific CD8+ T cells in the blood and tumor together, separating them from viral-specific CD8+ T cells (Figures S2B–S2D). Indeed, melanoma-specific CD8+ T cells in the tumor were largely exhausted with high expressions of PD-1, CD39, and TOX, and melanoma-specific CD8+ T cells in the blood retained many of these exhaustion features, albeit at a lower expression level (Figures S2B–S2F). In contrast, viral-specific CD8+ T cells in the tumor were memory-like with expressions of TCF-1 and CCR7 and viral-specific CD8+ T cells in the blood had effector features with expressions of T-bet and CX3CR1 (Figures S2B–S2E), demonstrating that antigenic properties play a major role in defining the cell state of CD8+ T cells.
At the same time, when we compared melanoma-specific CD8+ T cells of the same specificity in blood and tumor, those in the blood had higher TCF-1 and CCR7, and lower TOX, TIM-3, and CD39 (Figure S2G). Thus, melanoma-specific CD8+ T cells in the blood appeared to be less exhausted, consistent with recent data54. Indeed, melanoma-specific CD8+ T cells within the tumor were primarily comprised of TEX subsets, whereas those in the peripheral blood were predominantly a transitional memory or naive phenotype (Figures S2H–S2K). In contrast, virus-specific CD8+ T cells, which were mostly CMV-specific, exhibited an effector memory phenotype in circulation, whereas tumor-infiltrating viral-specific CD8+ T cells showed naïve and transitional memory characteristics (Figures S2H–S2L). These results indicate that both antigen specificity and the tumor microenvironment significantly influence the phenotypic profile of CD8+ T cells.
Melanoma-specific CD8+ T cells can be identified within regional lymph nodes exhibiting signs of activation, proliferation, and exhaustion.
Tumor-draining lymph nodes are crucial for the priming and activation of tumor-specific CD8+ T cell responses57–59. Thus, we studied uninvolved regional lymph nodes that were resected along with tumor metastasis. In contrast to the lymph node metastasis which were lymph nodes completely replaced with metastatic tumor, these uninvolved lymph nodes had no evidence of tumor metastasis upon examination by a melanoma pathologist (Figure S3A).
While melanoma-specific CD8+ T cells primarily accumulated in the tumor, they could be found in the lymph node at a greater frequency than in the blood (Figures 1D and 1E). As expected, viral-specific CD8+ T cells were predominantly in blood and lymph nodes, albeit at low frequencies (Figures 1D and 1F). Globally, melanoma-specific CD8+ cells in lymph nodes resembled those present in the tumors, marked by elevated expressions of exhaustion markers such as TIM-3 and CD39 (Figures 1G–1I). Importantly, these melanoma-specific CD8+ T cells, which represent the minority of cells in the lymph node (median = 0.03% of CD8+ T cells), were distinct in phenotype from the rest of the CD8+ T cells in the lymph node, which had naïve and memory features, highlighting the importance of antigen-specific profiling for identifying small but relevant populations (Figures 1H and 1I). There were certain features that were prominent in melanoma-specific CD8+ T cells in the lymph node; notably T-bet, TOX, and KI-67 (Figure S3B). While melanoma-specific CD8+ T cells in the lymph node had generally high levels of TOX that were comparable to those in the tumor, melanoma-specific CD8+ T cells in the lymph node in certain patients displayed high KI-67, TCF-1, and T-bet, reminiscent of ProgEX and IntEx (Figures 1J–1L). Indeed, melanoma-specific CD8+ T cells in the lymph node were composed largely of exhausted CD8+ T cell subsets, in contrast with viral-specific CD8+ T cells which were mostly transitional memory in the lymph node (Figures S3C–S3E).
T-bet+ intermediate exhausted melanoma-specific CD8+ T cells in the tumor are associated with pathologic response.
To identify immune features that were associated with clinical outcomes, we employed a random forest machine-learning algorithm to analyze 79 immune features across all patients, both antigen-specific and non-antigen-specific, in the tumor following PD-1 blockade. We classified responders as patients with <50% viable tumors, (pCR, pNCR, and pPR) and patients with pNR as non-responders. Features associated with melanoma-specific CD8+ T cells, including frequency and T-bet expression on tumor-infiltrating melanoma-specific CD8+ T cells outperformed all other features (Figure 2A).
Figure 2. T-bet+ intermediate exhausted melanoma-specific CD8+ T cells in post-treatment tumors are associated with pathologic response.

(A) Seventy-nine post-treatment tumor immune parameters classified by pathologic response using random forest analysis and ranked by importance score. Error bar denotes mean +/− s.d. for 1,000 random forest iterations.
(B) Representative flow plot of the percentage of melanoma-specific CD8+ T cells in pCR and in pNR patient.
(C) Percentage of dominant melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents dominant tetramer+ population in individual patient. The bar and error bars indicate the mean ± SEM.
(D) Percentage of melanoma-specific CD8+ T cells in the tumor before and after treatment in responders and non-responders. Each line connects paired timepoints from the same patient.
(E) Pathologic response rate stratified by CART-defined percentage of dominant melanoma-specific CD8+ T cell threshold (0.08%) in the tumor.
(F) Representative flow plots comparing T-bet and CD39 expressions on melanoma-specific CD8+ T cells in pCR and pNR patient.
(G) Percentage of T-bet+ melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(H) CD39 MFI melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(I) Pathologic response rate stratified by CART-defined T-bet threshold (11.8%) in the tumor.
(J) Representative flow plot of TCF-1, TIM-3, TOX, and CD39 expressions on T-bet+ melanoma-specific CD8+ T cells in the tumor.
(K) Cell state composition of T-bet+ melanoma-specific CD8+ T cells in the tumor.
(L) Percentage of intermediate exhausted CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
Statistical significance was assessed using the Wilcoxon rank-sum test for unpaired comparisons and the Wilcoxon signed-rank test for paired comparisons. P-values are reported as follows: ns (not significant), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.
See also Figure S4.
Quantitatively, responders had a greater frequency of melanoma-specific CD8+ cells within tumors compared to non-responders (pNR), after anti-PD-1 (Figures 2B and 2C). This was true when we measured the dominant melanoma-specific CD8+ T cell population per patient (Figure 2C), or when we took the sum-total of all melanoma-specific CD8+ populations (not shown). PD-1 blockade resulted in a mean of 4-fold expansion of melanoma-specific CD8+ T cells in the tumor, but only in responders (Figure 2D). Moreover, Classification And Regression Tree (CART) analysis revealed that patients with a high frequency (>0.08%) of melanoma-specific CD8+ T cells had a much higher chance to achieve a pathologic response, compared with those with a low frequency of melanoma-specific responses (73% versus 35%) following treatment (Figure 2E).
T-bet expression on melanoma-specific CD8+ T cells was the feature most strongly associated with pathologic responses (Figure 2A), indicating that the characteristics and quality of melanoma-specific CD8+ T cells were important indicators for effective anti-tumor immunity. The percentage of T-bet-positive intratumoral melanoma-specific CD8+ T cells was strongly correlated with response to PD-1 blockade (Figures 2F and 2G). CART analysis revealed that a T-bet expression threshold of 11.8% yielded optimal predictive performance, and patients with T-bet expression greater or equal to 11.8% demonstrated a 100% likelihood of achieving a pathologic response with an area under the curve (AUC) of 1 among patients with detectable melanoma-specific CD8+ T cells (Figures 2I, S4A, and S4B).
For enhanced sensitivity in identifying melanoma-specific CD8+ T cell features associated with clinical outcomes, we restricted our Random Forest analyses to 26 variables associated with melanoma-specific CD8+ T cells. While virtually all melanoma-specific CD8+ T cells expressed CD3960, high CD39 expression levels on melanoma-specific CD8+ T cells were strongly correlated with resistance to PD-1 blockade (Figures 2F, 2H, and S4C). CD39 and T-bet expressions were associated with clinical outcomes in regard to melanoma-specific CD8+ T cells, but not for the unselected CD8+ T cell population (Figures S4D and S4E) demonstrating that antigen-specific profiling offers enhanced insights into anti-PD-1 therapy and improves our sensitivity for biomarker detection.
Given the pivotal role of T-bet-positive melanoma-specific CD8+ T cells in individuals who respond to anti-PD-1, we sought to define the cell states of these T-bet-expressing cells. T-bet-positive melanoma-specific CD8+ T cells expressed both TOX and CD39 but had more TIM-3 expressions than TCF-1 (Figure 2J) and were largely composed of IntEX by unsupervised classification (Figure 2K). Consequently, responders showed enrichment of melanoma-specific IntEX in the tumor post-treatment (Figure 2L). A substantial population of melanoma-specific CD8+ T cells in responders expressed TCF-1 but not TIM-3, characteristic of the ProgEX phenotype, which was also overrepresented in responders (Figures S4F and S4G). In contrast, CD39hi melanoma-specific CD8+ cells, typically displaying a TermEX phenotype, were mostly observed in non-responders (Figures S4H and S4I). Thus, increased melanoma-specific progenitor and intermediate exhausted CD8+ T cells in the tumor were associated with pathologic response while terminal exhaustion was associated with resistance.
To validate our findings, we analyzed antigen-specific CD8+ T cells in an independent cohort from a phase 1b neoadjuvant pembrolizumab clinical trial (NCT02434354)24 (Figures S4J and S4K). Out of 26 patients with available post-treatment samples, 19 (73%) were HLA-A1, A2, or A3, for which we performed tetramer-based profiling and phenotypic analysis. Pathologic responders exhibited higher frequencies of T-bet+ melanoma-specific CD8+ T cells with reduced CD39 expression, which aligns with results from the neoadjuvant nivolumab clinical trial (Figures S4L–S4N).
Progenitor exhausted CD8+ T cells are enriched in the regional lymph nodes of responders.
As different cell states of intratumoral melanoma-specific CD8+ T cells determined clinical outcomes, we asked whether these changes could also be identified in the uninvolved regional lymph nodes. Melanoma-specific CD8+ T cells in responders also demonstrated higher T-bet and reduced CD39 expressions in regional lymph nodes compared to non-responders (Figures 3A–3C). Indeed, melanoma-specific CD8+ T cells that had high T-bet and low CD39 levels in the lymph node were able to maintain their T-bet+ CD39lo phenotype in the tumor (Figures 3A, 3D, and 3E). Conversely, melanoma-specific CD8+ T cells with low T-bet and high CD39 in the lymph node lost their T-bet expression and gained CD39 expression in the tumor (Figures 3A, 3D, and 3E). Similar to those in the tumor, T-bet+ melanoma-specific CD8+ T cells in lymph nodes exhibited high signal of TOX. However, these cells in the lymph node characterized by TCF-1 upregulation than TIM-3 (Figure 3F) consistent with ProgEX (Figure 3G). Responders had a greater proportion of melanoma-specific ProgEX in the lymph node, while non-responders showed evidence of terminal differentiation such as TIM-3 expression (Figures 3H and 3I). Indeed, CD39hi melanoma-specific CD8+ T cells were predominantly TermEX and accumulated in the lymph nodes of non-responders (Figures 3J and 3K). Notably, the frequency of melanoma-specific IntEX in regional lymph nodes did not correlate with pathologic responses (Figure 3L). Thus, the ability of melanoma-specific CD8+ T cells to be restrained from terminal differentiation and retain a progenitor state in the lymph node was associated with its ability to maintain a functional state in the tumor microenvironment.
Figure 3. Melanoma-specific progenitor exhausted CD8+ T cells in the lymph node are associated with pathologic response.

(A) Representative flow plots comparing T-bet and CD39 expressions on melanoma-specific CD8+ T cells in a complete responder (top), and non-responder (bottom), in the lymph node and tumor.
(B) Percentage of T-bet+ melanoma-specific CD8+ T cells stratified by response in the lymph node. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(C) CD39 MFI melanoma-specific CD8+ T cells stratified by response in the lymph node. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(D) Percentage of T-bet+ melanoma-specific CD8+ T cells in the lymph node and tumor. Each line connects paired samples from the same patient.
(E) CD39 MFI melanoma-specific CD8+ T cells in the lymph node and tumor. Each line connects paired samples from the same patient.
(F) Representative flow plot of TCF-1, TIM-3, TOX, and CD39 expressions on T-bet+ melanoma-specific CD8+ T cells in the lymph node.
(G) Cell state composition of T-bet+ melanoma-specific CD8+ T cells in the lymph node.
(H) Representative flow plots of melanoma-specific CD8+ T cells from responder (blue), and non-responder (red) in the lymph node, overlayed.
(I) Percentage of progenitor exhausted CD8+ T cells stratified by response in the lymph node. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(J) Cell state composition of CD39hi (MFI>15431, CART-defined CD39 MFI threshold) melanoma-specific CD8+ T cells in the lymph node.
(K) Percentage of terminal exhausted CD8+ T cells stratified by response in the lymph node. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(L) Percentage of intermediate exhausted CD8+ T cells stratified by response in the lymph node. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
Statistical significance was assessed using the Wilcoxon rank-sum test for unpaired comparisons and the Wilcoxon signed-rank test for paired comparisons. P-values are reported as follows: ns (not significant), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.
Longitudinal profiling reveals tumor antigen–dependent dynamics of circulating melanoma-specific CD8+ T cell responses to PD-1 blockade.
To evaluate the longitudinal behavior of circulating melanoma-specific CD8+ T cells and their association with clinical outcomes, we extended our analysis to include the adjuvant phase of this neoadjuvant nivolumab clinical trial. Patients were randomized to receive adjuvant nivolumab (Nivo) or combination nivolumab plus ipilimumab (Nivo + Ipi) based on their pathologic response to a single pre-surgical dose of nivolumab (STAR Methods). Antigen-specific profiling was performed on PBMC samples at multiple timepoints including baseline, after neoadjuvant nivolumab (resection), and during adjuvant therapy starting from cycle 2 (C2D1) through at least cycle 6 (C6D1) (Figures 4A and 4B).
Figure 4. Melanoma-specific CD8+ T cell responses to immune checkpoint blockade are dependent on tumor antigen.

(A) Study schema of neoadjuvant and adjuvant phase of neoadjuvant anti-PD-1 clinical trial.
(B) Representative flow plots on the percentage of circulating melanoma-specific CD8+ T cells in pCR (top, blue) and pNR (bottom, red).
(C) Hierarchical clustering on the percentage of circulating melanoma and viral-specific CD8+ T cell from pre-treatment to C3D1 (M: Melanoma-specific, V: Viral-specific). Color scale indicates Pearson’s correlation coefficient values (r), from −1 (strong negative correlation, blue) to +1 (strong positive correlation, red), with white denoting no correlation.
(D) Cumulative frequencies of melanoma-specific CD8+ T cells across three temporal trajectories. Each line represents a distinct melanoma-specific CD8+ T cell population, the vertical distance between adjacent lines at each time point reflects its frequency.
(E) Cumulative frequencies of melanoma-specific CD8+ T cells across three temporal trajectories stratified by response.
(F) Absolute change in the percentage of circulating melanoma-specific CD8+ T cells from Pre to Resection in Traj 3 stratified by response. Each dot represents an individual tetramer+ population. The box and whiskers indicate the median and interquartile range (IQR; Q1–Q3) with whiskers extending to 1.5× IQR.
(G) Percentage of circulating melanoma-specific CD8+ T cells across time stratified by response. Each line represents a distinct melanoma-specific CD8+ T cell population. Paired comparisons were evaluated using the Wilcoxon signed-rank test.
(H) Percentage of circulating KI-67+ melanoma-specific CD8+ T cells across time stratified by response. Each line represents a distinct KI-67+ melanoma-specific CD8+ T cell population. Paired comparisons were evaluated using the Wilcoxon signed-rank test.
Statistical significance was assessed using the Wilcoxon rank-sum test for unpaired comparisons and the Wilcoxon signed-rank test for paired comparisons. P-values are reported as follows: ns (not significant), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.
See also Figure S5.
A higher baseline frequency of tumor-specific CD8+ T cells in the blood was associated with pathologic response consistent with recent findings from neoadjuvant therapy in virus-positive Merkel cell carcinoma61 (Figures S5A and S5B). Using our recently developed Cyclone approach62, we performed unsupervised clustering of antigen-specific CD8+ T cells in the blood based on their temporal trajectories (Figures 4C and 4D). Despite the presence of multiple timepoints across neoadjuvant and adjuvant phases of treatment, there were only 3 stable temporal trajectories of antigen-specific CD8+ T cells identified in the blood. These trajectories included antigen-specific CD8+ T cells that peaked at C3D1, after 1st dose of adjuvant therapy (Traj 1), those who decreased in frequency across time (Traj 2), and those that peaked after a single dose of neoadjuvant nivolumab (Traj 3). We then stratified patients by pathologic response (R = pCR, pNCR, and pPR). Non-responders (NR = pNR) had a greater expansion of circulating melanoma-specific CD8+ T cells after neoadjuvant therapy, compared to responders (Figures 4E-4G, S5C, and S5D). Minimal change was observed in viral-specific CD8+ T cells in the blood (Figure S5E).
Moreover, in these non-responders, KI-67 in melanoma-specific CD8+ T cells were sustained or increased after neoadjuvant anti-PD-1, but rapidly decreased after tumor resection, which could be observed by C2D1 (Figures 4H and S5F). These data are consistent with the idea that the amount of tumor antigen plays an important role in modulating the magnitude of systemic antigen-specific immunity63,64. In contrast, responders had low, or even decreasing KI-67, at the time of tumor resection (Figures 4H and S5F). Notably, after tumor removal, with rare exceptions, subsequent adjuvant therapy did not elicit significant changes in the frequency of circulating melanoma-specific CD8+ T cells, regardless whether the adjuvant therapy received was nivolumab (Nivo), or Nivo + Ipi (Figures 4G and S5G), indicating that the presence of antigenic stimulation is essential for maintaining T cell responses during checkpoint inhibition.
DISCUSSION
Using a combinatorial tetramer system, we performed antigen-specific CD8+ T cell profiling that revealed insights which would have been difficult to identify using non-antigen-specific approaches.
First, there was a strong association between the frequency and the phenotype of melanoma-specific CD8+ T cells with clinical efficacy. As neoantigen-specific T cells vary across patients and are cumbersome to identify65, we focused on CD8+ T cells specific to melanoma shared antigens which have been well characterized22 and are common across the broader patient population. The fact that pathologic responses were associated with T cell responses against shared antigens is consistent with early data demonstrating the recognition of melanoma shared antigens by tumor infiltrating lymphocytes66–69, and clinical responses in melanoma patients infused with MART-1 TCR-engineered T cells70. In fact, more recent studies have shown that successful immune checkpoint blockade can induce T cell responses not only against melanoma neoantigens but promote epitope spreading towards melanoma shared antigens71. Thus, for many patients, T cell responses to melanoma-shared antigens are sufficient for tumor regression which has implications for cellular therapies and vaccination approaches.
Second, when we compared melanoma-specific CD8+ T cells with the same specificity in blood and tumor, melanoma-specific CD8+ T cells in the tumor were clearly exhausted while those in the blood had a CD45RA− CD27+ transitional memory phenotype but also expressed CD39, TIGIT and TOX, albeit at a lower expression. These results are consistent with published data that demonstrate neoantigen-specific T cell clones in the blood had a TIGIT+ CD39+ phenotype, but with an enrichment of memory genes (e.g. TCF7, SELL, and LEF1) as compared to their counterparts in the tumor54. Thus, the tumor microenvironment drives T cell exhaustion through chronic antigen exposure72, inflammation73,74, and immunosuppressive factors75,76. The presence of less terminally differentiated tumor-specific CD8+ T cells in the blood raises the possibility of their recruitment back into the tumor following treatment77–80, where they could re-engage and contribute to tumor control.
Third, expression of the effector transcription factor T-bet on melanoma-specific CD8+ T cells strongly correlated with favorable clinical responses. These T-bet-expressing CD8+ T cells have a phenotype that is consistent with an intermediate (or transitory) exhausted CD8+ T cells that has been shown to mediate the clinical efficacy of PD-1 blockade81–83. In contrast, CD39hi terminally differentiated TEX were associated with therapeutic resistance, a phenotype that was seen in neoantigen-specific CD8+ T cells from non-responders to neoadjuvant αPD-1 in lung cancer84. These findings were also observed in the uninvolved regional lymph node. Melanoma-specific CD8+ T cells in the regional lymph node exhibited an exhausted phenotype which aligns with data on the early acquisition of exhaustion85 and the presence of ProgEX within tumor draining lymph nodes86,87. However, these melanoma-specific TEX were low in frequency (median = 0.03% of CD8+ T cells) and had a distinct phenotype from the rest of CD8+ T cells in the lymph node, highlighting the importance of antigen-specific profiling. ProgEX from responders had upregulation or maintenance of T-bet expression and a commitment to IntEX transition in the tumor. In contrast, melanoma-specific CD8+ T cells in the lymph node of non-responders had minimal T-bet and high CD39. These CD8+ T cells in non-responders were susceptible to terminal exhaustion in the tumor, with complete loss of T-bet and upregulation of CD39. Therefore, it may be that the quality and quantity of tumor-specific CD8+ T cells are determined early upon priming in the draining lymph node based on the nature of the lymph node microenvironment. This is consistent with recent data showing the role of IFNγ in regulating effector CD8+ T cell differentiation in the lymph node88.
Fourth, our study identified antigen-specific biomarkers of clinical efficacy, including the presence of T-bet and CD39 on melanoma-specific CD8+ T cells in the tumor after treatment. Antigen-specific profiling could complement features such as TIL score and pathologic response (which perform quite well) to elucidate mechanisms of resistance. For example, patients with TermEX CD8+ T cells may require rejuvenation of ProgEX and IntEX through CTLA-4 blockade93 or cytokine-based therapies53,94–96.
Finally, despite robust melanoma-specific CD8+ T cell responses to neoadjuvant therapy, there was no significant expansion or activation of melanoma-specific CD8+ T cells after tumor removal, during adjuvant therapy. The absence of a melanoma-specific CD8+ T cell response following adjuvant therapy supports the notion that ongoing tumor antigen presence is essential for effective immune checkpoint blockade63,64, providing a likely explanation for the superior efficacy of neoadjuvant-adjuvant αPD-1 therapy compared to adjuvant-only αPD-123. This also raises the question about the role of adjuvant ICB, which is likely beneficial for the subset of patients with residual tumor antigen after surgery. Given the exquisite sensitivity of exhausted CD8+ T cells to tumor antigen, assessing the activity and exhaustion state of melanoma-specific CD8+ T cells in nodal tissues may provide a way to monitor the presence of residual tumor cells and help stratify patients considered for adjuvant immunotherapy.
Altogether we demonstrate the feasibility of antigen-specific T cell profiling in a prospective clinical trial, and that antigen-specific T cell profiling provides enhanced sensitivity for the biomarker discovery and mechanistic understanding in the human setting.
RESOURCE AVAILABILITY
Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Alexander C. Huang (Alexander.Huang@pennmedicine.upenn.edu).
Materials availability
This study did not generate new unique reagents.
Data and code availability
Flow cytometry data and associated clinical metadata have been deposited on Omiq and available upon request.
STAR Methods
EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
Patients/Specimen collection
This phase 2 clinical trial (NCT04013854) was a single-institution, investigator-initiated study conducted at the University of Pennsylvania, with approval from the Institutional Review Board and written informed consent obtained from all participants. After providing informed consent, all patients underwent a baseline pretreatment biopsy, with tissue collection equivalent to at least two to four core biopsies. Patients then received a single intravenous dose of nivolumab (480 mg), followed by standard definitive surgery 4 weeks later. Paired blood samples were obtained at pre-treatment and post-treatment time points, along with regional lymph node samples post-treatment. Postoperatively, patients received open-label adjuvant therapy based on their pathologic response at resection. Patients with pathologic complete response (pCR) or near-complete response (pNCR) (Arm A) were treated with adjuvant nivolumab (480 mg every 4 weeks) for up to 1 year. Those with partial response (pPR) or non-response (pNR) were randomized 1:2 to receive either adjuvant nivolumab (Arm B) or combination therapy with ipilimumab (1 mg/kg) plus nivolumab (3 mg/kg) for 4 doses, followed by nivolumab alone for up to 1 year (Arm C).
Patient blood was drawn in sodium heparin tubes, kept at room temperature, and processed within 8 h of blood drawn. Whole blood was centrifuged, and plasma samples were collected, aliquoted in the volume of 1 ml per cryotube, and stored at −80 °C. Blood samples were reconstituted by adding an equal volume of Hank’s balanced salt solution (HBSS, Corning MT21021CM) to the volume of collected plasma. Reconstituted blood samples were diluted two-fold in HBSS and Ficoll (Ficoll-Paque PLUS density gradient medium, GE Healthcare Life Sciences, 17–1440-03) was layered underneath. The volume of Ficoll-Paque PLUS used was equal to that of the undiluted blood sample. The buffy coat was collected and washed twice with HBSS. Ammonium-chloride-potassium lysis buffer (Lonza 10–548E) was used to lyse the red blood cells, if necessary.
Fresh melanoma resection specimens were documented with gross photographs and processed for tumor-infiltrating lymphocyte isolation within 4 hours of excision. Samples were cut into 2–3-mm pieces with a scalpel in an RPMI 1640 (Corning 10–040-CM)-containing petri dish. A 70-μm cell strainer (Falcon 352350) was placed on top of a 50-ml conical tube and the tissue fragments were transferred to the cell strainer with a pipette. The cells were released by gently grinding the tissue fragments using the thumb depressor of a sterile syringe plunger placed against the cell strainer. Cells were washed twice and counted in a hemocytometer. The numbers of lymphocytes and melanoma cells were distinguished by their morphology and recorded separately.
Other tumor biopsies and resection specimens were processed by the Department of Pathology and Laboratory Medicine at the University of Pennsylvania, where tissues were immediately snap-frozen or formalin-fixed and paraffin-embedded. Pathologic response was assessed 4 weeks post-treatment, with the percentage of viable tumor determined by a melanoma pathologist. pCR was defined as the absence of viable tumor based on hematoxylin and eosin (H&E) staining, pNCR as <10% viable tumor, and pPR as ≤50% viable tumor cells.
For validation, we analyzed samples from a phase 1b clinical trial (NCT02434354) enrolling patients with stage III or resectable stage IV melanoma. Participants underwent baseline tumor biopsy followed by a single 200 mg dose of pembrolizumab. Surgical resection was performed three weeks later, and paired tumor samples were collected pre- and post-treatment. After surgery, patients received adjuvant pembrolizumab every three weeks for up to one year, or until disease progression or treatment-limiting toxicity.
All relevant clinical characteristics, including age and sex, are listed in Table S1.
METHOD DETAILS
Peptide reconstitution and tetramer preparation
Peptides were reconstituted to a concentration of 10 mg/mL in 100% DMSO at room temperature and mixed thoroughly. Biotinylated monomers were obtained from either BioLegend (Flex-T™ monomer UVX, 0.2 mg/mL) or the NIH Tetramer Core Facility (2 mg/mL) and diluted to 0.1 mg/ml with phosphate-buffered saline (PBS). For UV-exchange, 49.5 μL of Monomer-PBS solution was combined with 0.9 μL of the 10 mg/mL peptide solution in a V-bottom 96-well plate, ensuring thorough mixing. The peptide-monomer-PBS mixture was exposed to UV light (366 nm) for 30 minutes in a cold room without a plate cover. Following UV exchange, the plate was incubated at 37°C with a plate cover for an additional 30 minutes. After incubation, the mixture was centrifuged at 2,000 rpm for 2 minutes to separate aggregates. From the top layer of the supernatant, 15 μL of peptide-monomer-PBS (approximately 1.5 μg of monomer) was transferred into each of two PCR tubes. To each tube, 1.5 μL of a different fluorophore-conjugated streptavidin (BioLegend) was added, and the reaction was incubated for at least 1 hour at room temperature or overnight at 4°C. To terminate the reaction, 2.4 μL of 0.4 mM free biotin was added to each tube to block excess streptavidin binding sites. Quality control testing of the tetramers from each batch was performed using HLA-matched, CMV-positive PBMCs from the same healthy donor to ensure consistency. The prepared tetramers were stored at 4°C in the dark and were utilized within four months to maintain optimal performance.
Combinatorial tetramer flow cytometry staining
Single-cell suspensions prepared from blood and tissue samples were treated with DNase I, washed with FACS staining buffer (PBS supplemented with 2% fetal bovine serum), and aliquoted into U-bottom 96-well plates for staining. Both staining samples and non-staining controls were included. Tetramer Staining: Tetramers were diluted in FACS staining buffer containing Brilliant Stain Buffer (BD Biosciences; used to reduce spectral overlap) and milk to prevent tetramer aggregation. A total of 50 μL of diluted tetramer solution was added to the staining samples, while 50 μL of tetramer-free buffer was added to the non-staining controls. The plates were incubated for 20 minutes at room temperature in the dark. Following incubation, the samples were washed with 150 μL of FACS staining buffer, centrifuged at 2,000 rpm for 3 minutes, and the supernatant was discarded. Surface Staining: Surface antibodies and live/dead viability dyes were diluted in FACS staining buffer supplemented with Brilliant Stain Buffer. A total of 50 μL of diluted surface staining solution was added to the staining samples, while 50 μL of surface stain-free buffer was added to the non-staining controls. The samples were incubated for 20 minutes at room temperature in the dark, followed by a wash with 150 μL of FACS staining buffer, centrifugation at 2,000 rpm for 3 minutes, and removal of supernatants. Fixation and Permeabilization: Fixation and permeabilization were performed using commercially available buffer set (Invitrogen™ eBioscience™ Foxp3 / Transcription Factor Staining Buffer Set) prepared according to the manufacturer's instructions. A total of 50 μL of fixation/permeabilization buffer was added to each sample, and the plates were incubated for 20 minutes at room temperature in the dark. The samples were then washed with 150 μL of permeabilization buffer, centrifuged at 2,000 rpm for 3 minutes, and supernatants were removed. Intracellular Staining: Intracellular antibodies were diluted in permeabilization staining buffer containing Brilliant Stain Buffer. A total of 50 μL of diluted intracellular staining solution was added to the staining samples, while 50 μL of intracellular stain-free buffer was added to the non-staining controls. The plates were incubated overnight at 4°C in the dark. Following incubation, the samples were washed with 150 μL of permeabilization buffer, centrifuged at 2,000 rpm for 3 minutes, and supernatants were removed. Finally, both staining samples and non-staining controls were resuspended in 200 μL of 1% paraformaldehyde solution for fixation before flow cytometric acquisition.
Flow cytometry data acquire and basic analysis
Stained samples and corresponding non-staining controls were acquired using a five-laser Cytek® Aurora spectral flow cytometer at an event rate of less than 8,000 events per second. Autofluorescence subtraction was performed by subtracting the fluorescence intensity of non-staining controls from the staining samples from the same patient. Spectral fluorescence unmixing was conducted using single-color controls according to the manufacturer's instructions. Further compensation and adjustments were applied using SpectroFlo® software. Basic data analysis, including gating and population quantification, was performed using FlowJo software (BD Biosciences).
Combinatorial tetramer flow cytometry analysis
To detect and analyze melanoma-specific CD8+ T cells, we employed combinatorial tetramer staining approaches, as previously described29–31. Briefly, peptide-MHC class I tetramers were conjugated to distinct fluorochromes and combined to uniquely label antigen-specific T cell populations. Events were considered tetramer-positive if they displayed specific staining in exactly two combinatorial channels. A threshold of ≥5 tetramer-positive events was set to define a sample as tetramer-positive, ensuring robust detection above background. For phenotypic analyses (e.g., exhaustion markers, differentiation status), a minimum of ≥10 tetramer-positive cells per population was required for inclusion in downstream characterization.
UMAP and cluster analysis of flow cytometry data
CD8+ T cells gated from flow cytometry data in FlowJo software were exported and analyzed using the OMIQ cloud-based platform, which integrates machine learning and traditional analytical pipelines. For each sample, CD8+ T cells were subsampled to a maximum of 25,000 events. Uniform Manifold Approximation and Projection (UMAP) was performed using markers of interest, excluding PD-1 (due to reduced staining intensity in post-treatment samples resulting from competition with the anti-PD-1 therapeutic antibody nivolumab) and LAG-3 (as it was not significantly expressed). UMAP settings included: Neighbors = 15, Minimum Distance = 0.4, Components = 2, and Epochs = 200. FlowSOM clustering was performed on the same set of markers used for UMAP, with the following parameters: xdim = 10, ydim = 10, number of training iterations = 10, and k values = 9 (alternative k values were also evaluated). Clustering employed the Run Consensus Metaclustering method. Clustered heatmaps were generated based on median fluorescence intensity values for each cluster, with clustering applied to both rows and columns.
PCA analysis of flow cytometry data
Antigen-specific CD8+ and all CD8+ T cells were gated from flow cytometry data using FlowJo software and analyzed for the expression of individual markers. Marker-positive cell counts were summarized and processed for PCA to explore high-dimensional data structure and identify key contributors to data variance. PCA was conducted in R using the prcomp function, with all variables standardized to ensure comparability across markers. The scree plot was used to evaluate the variance explained by each principal component and to select the most informative dimensions for further analysis. Clustering of samples on the principal component space was visualized using PC1 versus PC2. Specificity groups (e.g., tumor-specific, viral-specific, and CD8+ T cells) and sample sources (PBMC, tumor, and lymph node) were distinguished by unique colors and shapes to highlight their differences in marker expression patterns. To further dissect the contributions of individual markers, variable loadings for PC1 and PC2 were extracted and visualized using contribution plots, providing insights into their role in separating distinct cell populations.
Random forest analysis of flow cytometry data and pathologic response
Antigen-specific and all CD8+ T cells were gated from flow cytometry data using FlowJo software. The expression of individual markers on these cells was analyzed for each sample. In addition, patients' pathologic response metadata were incorporated to build predictive models. Random forest models were developed to predict the response variable using the expression of individual markers as input features for each patient. The models were implemented using the randomForest package in R, with 5000 trees and proximity estimation enabled to enhance model robustness and prediction accuracy. Variable importance scores were computed for each marker to assess its contribution to predicting the response. To ensure the reliability of the importance scores, 1000 independent random forest models were generated. The variable importance scores were extracted from each iteration and averaged, reducing the impact of model stochasticity. This iterative approach mitigated variability, providing a more stable estimate of each marker’s relevance. Markers with the highest importance scores were compared between responders and non-responders to identify potential associations with pathologic response or non-response. The statistical significance of these associations was evaluated to determine which markers were most predictive of clinical outcomes.
In vitro expansion of melanoma-specific CD8+ T cell
To generate a short-term ex vivo expanded Nras Q61K-specific T cell line, tumor-infiltrating lymphocytes (TILs) were cultured with irradiated (2,500 rads) autologous mature dendritic cells (DCs) pulsed with 40 μg of Nras Q61K peptide (ILDTAGKEEY) at a TIL:DC ratio of 20:1. Cultures were maintained in Optimizer CTS™ T Cell Expansion Media (Gibco), supplemented with 5% human AB serum and interleukin-2 (IL-2; 50 U/mL). On day 10, T cells were harvested, enumerated, and re-stimulated with irradiated (10,000 rads) Nras peptide-pulsed (500 ng/106 cells) HLA-A*01:01 / 4–1BBL artificial antigen-presenting cells (aAPCs; K562). T cells and aAPCs were co-cultured at a 1:1 ratio in media containing IL-2 (500 U/mL) for an additional 10 days. To evaluate the frequency of Nras Q61K-specific T cells in expanded cultures, cells were stained with custom Nras Q61K peptide/HLA-A*01:01 tetramers. Tetramer staining was performed on pre- and post-expansion T cells by incubating cells with tetramers for 20 minutes at 4°C.
Classification and regression tree analysis
A Classification and Regression Tree (CART) analysis was employed to determine the optimal cut point for each biomarker. CART was implemented using Minitab® and Salford Predictive Modeler® (SPM) software to stratify patients into pathologic responders and non-responders based on biomarker expression levels. The classification trees were constructed to iteratively partition the data, identifying the biomarker threshold that best differentiated the two patient groups. At each node, the algorithm evaluated all possible split points to minimize within-group variance and maximize between-group variance, ensuring an optimal division of patients.
ROC/AUC estimation
Logistic regression analysis was conducted using the Statistical Analysis System (SAS), Version 9.4, to assess the performance of the biomarkers in distinguishing pathologic responders from non-responders. The receiver operating characteristic (ROC) curve was generated, and the associated area under the curve (AUC) was calculated to evaluate the discriminatory ability of each biomarker.
Hierarchical clustering on antigen-specific CD8+ T cells
A correlation matrix of antigen-specific CD8+ T-cell features was generated to evaluate similarity across four longitudinal timepoints (Pre, Resection, C2D1, and C3D1). These timepoints were chosen because later samples (e.g., C4D1) contained missing data for some patients. Pairwise correlations were calculated using Pearson’s method, followed by hierarchical clustering to identify groups of antigen-specific CD8+ T cells exhibiting similar temporal dynamics.
QUANTIFICATION AND STATISTICAL ANALYSIS
Data are presented as mean ± standard error of the mean (SEM) unless otherwise specified. Statistical significance was assessed using the Wilcoxon rank-sum test for unpaired comparisons and the Wilcoxon signed-rank test for paired comparisons. P-values are reported as follows: ns (not significant), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. All analyses were conducted using R or GraphPad Prism.
Supplementary Material
Table S3. Tetramer antigen sequences, related to Figure 1
Table S1. Neoadjuvant anti-PD-1 (nivolumab) clinical trial patient metadata, related to Figure 1
Table S2. Antigen-specific CD8+ T cell summary, related to Figure 1
Document S1. Figures S1–S5
Figure S1. CD8+ T cell atlas of blood, lymph node, and tumor. Related to Figure 1. (A) Representative combinatorial tetramer flow plot identifying melanoma (red) and viral (blue)-specific CD8+ T cells.
(B) UMAP projection of CD8+ T cells stratified by tissue type.
(C) UMAP visualization of feature expressions.
(D) UMAP projection of CD8+ T cell states in combined blood, lymph node, and tumor defined by unsupervised clustering.
(E) UMAP projection of CD8+ T cell states in the blood and tumor.
(F) Expression of selected markers for CD8+ T cell states.
(G) Cell state composition of CD8+ T cells in the blood and tumor.
Figure S2. Antigen-specific CD8+ T cell phenotypes in the blood and tumor. Related to Figure 1. (A) Melanoma (red) and viral (blue)-specific CD8+ T cells projected on CD8+ T cell atlas in the blood and tumor. CD8+ T cells are in gray.
(B) Expression of selected markers on melanoma and viral-specific CD8+ T cells in the blood and tumor.
(C) PCA plot of melanoma and viral-specific CD8+ T cells in the blood and tumor. Each dot represents an individual antigen-specific population.
(D) Variables contributing to principal components.
(E) UMAP visualization of feature expressions.
(F) MFI of exhausted T cell features on melanoma and viral-specific CD8+ T cells in the blood and tumor. Each dot represents an individual tetramer+ population.
(G) MFI of progenitor and terminal exhausted T cell features on melanoma-specific CD8+ T cells in the blood and tumor separated by antigen. Each line connects paired samples from the same patient.
(H) Cell state composition of melanoma and viral-specific CD8+ T cells in the tumor.
(I) Cell state composition of melanoma and viral-specific CD8+ T cells in the blood.
(J) Cell state composition of melanoma and viral-specific CD8+ T cells in the tumor. Each dot represents an individual tetramer+ population.
(K) Cell state composition of melanoma and viral-specific CD8+ T cells in the blood. Each dot represents an individual tetramer+ population.
(L) MFI of memory and effector T cell features on viral-specific CD8+ T cells in the blood and tumor separated by antigen. Each line connects paired samples from the same patient.
Figure S3. Melanoma-specific CD8+ T cells in the uninvolved regional lymph nodes demonstrate features of exhaustion. Related to Figure 1. (A) Representative gross pathology (left) and H&E stain (right) of uninvolved lymph node (top) and lymph node metastasis (bottom).
(B) Expression of selected markers for CD8+ T cells in the lymph node (white), melanoma-specific CD8+ T cells in the lymph node (purple), and melanoma-specific CD8+ T cells in the tumor (red).
(C) UMAP projection of CD8+ T cell states in the lymph node.
(D) Melanoma (red) and viral (blue)-specific CD8+ T cells projected on CD8+ T cell atlas in the lymph node. CD8+ T cells in gray.
(E) Cell state composition of all CD8+, melanoma and viral-specific CD8+ T cells in the lymph node.
Figure S4. T-bet+ CD39low melanoma-specific CD8+ T cells are associated with pathologic response. Related to Figure 2. (A) AUC curves for immune parameters associated with pathologic response.
(B) Pathologic response rate stratified by CART-defined T-bet threshold (11.8%) in the patients with tetramer+ tumor.
(C) Twenty-six post-treatment melanoma-specific CD8+ T cell parameters classified by pathological response using random forest analysis and ranked by importance score. Error bar denotes mean ± s.d. for 1,000 random forest iterations.
(D) Percentage of T-bet+ and T-bet MFI of all CD8+ T cells stratified by response in the tumor. Each dot represents an individual patient.
(E) Percentage of CD39+ and CD39 MFI of all CD8+ T cells stratified by response in the tumor. Each dot represents an individual patient.
(F) Representative flow plots of melanoma-specific CD8+ T cells from responder (blue), and non-responder (red) in the tumor, overlayed.
(G) Percentage of progenitor exhausted CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population.
(H) Cell state composition of CD39hi (MFI>15431, CART-defined CD39 MFI threshold) melanoma-specific CD8+ T cells in the tumor.
(I) Percentage of terminally exhausted CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population.
(J) Study schema with patient demographics and HLA types in neoadjuvant anti-PD-1 (pembrolizumab) clinical trial.
(K) Pathologic responses separated by complete response (pCR), near-complete response (pNCR), partial response (pPR), and non-response (pNR) (pembrolizumab).
(L) Representative flow plots comparing T-bet and CD39 expressions on melanoma-specific CD8+ T cells in pCR and pNR patient (pembrolizumab).
(M) Percentage of T-bet+ melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population (pembrolizumab).
(N) CD39 MFI melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population (pembrolizumab).
Figure S5. Pharmacodynamic changes in the percentage of circulating antigen-specific CD8+ T cells. Related to Figure 4 (A) Representative flow plots of the percentage of circulating melanoma-specific CD8+ T cells in pCR and pNR patients before anti-PD-1 neoadjuvant therapy.
(B) Percentage of circulating melanoma-specific CD8+ T cells stratified by response before anti-PD-1 neoadjuvant therapy. Each dot represents an individual tetramer+ population.
(C) Percentage of circulating melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents a distinct melanoma-specific CD8+ T cell population.
(D) Percentage of total circulating melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents the total melanoma-specific CD8+ T cell frequency in an individual patient.
(E) Percentage of circulating viral-specific CD8+ T cells across time stratified by Response (R top; NR bottom). Each line represents a distinct viral-specific CD8+ T cell population.
(F) Percentage of circulating KI-67+ melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents a distinct KI-67+ melanoma-specific CD8+ T cell population.
(G) Percentage of circulating melanoma-specific CD8+ T cells across time stratified by adjuvant treatment (Nivo top; Nivo + Ipi bottom). Each line represents a distinct melanoma-specific CD8+ T cell population.
KEY RESOURCES TABLE
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Antibodies | ||
| BD Horizon™ BUV395 Mouse Anti-Human CD45RA | BD Biosciences | Cat# 568712; clone HI100 |
| LIVE/DEAD™ Fixable Blue Dead Cell Stain Kit, for UV excitation | Invitrogen | Cat# L23105; RRID: AB_2796682 |
| BD Horizon™ BUV496 Mouse Anti-Human CD8 | BD Biosciences | Cat# 612942; clone RPA-T8 |
| BD OptiBuild™ BUV563 Mouse Anti-Human CD69 | BD Biosciences | Cat# 748764; RRID: AB_2873167; clone FN50 |
| BD OptiBuild™ BUV615 Mouse Anti-Human NKG2A (CD159a) | BD Biosciences | Cat# 752302; clone 131411 |
| BD OptiBuild™ BUV661 Mouse Anti-Human CD226 | BD Biosciences | Cat# 749934; RRID: AB_2874171; clone DX11 |
| BD Horizon™ BUV737 Mouse Anti-T-bet | BD Biosciences | Cat# 568166; clone O4-46 |
| BD OptiBuild™ BUV805 Mouse Anti-Human CD28 | BD Biosciences | Cat# 742037; clone CD28.2 |
| BD Horizon™ BV480 Mouse Anti-Human CD14 | BD Biosciences | Cat# 566190; clone MφP9 |
| BD Horizon™ V500 Mouse anti-Human CD19 | BD Biosciences | Cat# 561125; RRID: AB_10562391; clone HIB19 |
| Spark Violet™ 538 anti-human CD4 Antibody | BioLegend | Cat# 344673; RRID: AB_2890774; clone SK3 |
| Brilliant Violet 570™ anti-human CD127 (IL-7Rα) Antibody | BioLegend | Cat# 351308; RRID: AB_2832685; clone A019D5 |
| Brilliant Violet 711™ anti-human Perforin Antibody | BioLegend | Cat# 308130; clone dG9 |
| Brilliant Violet 750™ anti-human CD197 (CCR7) Antibody | BioLegend | Cat# 353253; RRID: AB_2800945; clone G043H7 |
| Mouse Anti-Human IgG4 Fc-AF488 | SouthernBiotech | Cat# 9200-30; clone HP6025 |
| Spark Blue™ 550 anti-human CD3 Antibody | BioLegend | Cat# 344852; RRID: AB_2819985; clone SK7 |
| BD Horizon™ BB700 Rat Anti-Human CXCR5 (CD185) | BD Biosciences | Cat# 566469; RRID: AB_2869769; clone RF8B2 |
| TIGIT Monoclonal Antibody (MBSA43), PerCP-eFluor™ 710, eBioscience™ | Invitrogen | Cat# 46-9500-42; RRID: AB_10853679; clone MBSA43 |
| BD OptiBuild™ BB790 Rat Anti-Human CX3CR1 | BD Biosciences | Custom design; clone 2A9-1 |
| BD Horizon™ RY586 Mouse Anti-Ki-67 | BD Biosciences | Cat# 568147; clone B56 |
| PE/Dazzle™ 594 anti-human CD152 (CTLA-4) Antibody | BioLegend | Cat# 369616; RRID: AB_2632878; clone BNI3 |
| PE/Fire™ 640 anti-human CD103 (Integrin αE) Antibody | BioLegend | Cat# 350243; RRID: AB_2924550; clone Ber-ACT8 |
| PE/Cyanine5 anti-human CD366 (Tim-3) Antibody | BioLegend | Cat# 345052; RRID: AB_2819987; clone F38-2E2 |
| FOXP3 Monoclonal Antibody (PCH101), PE-Cyanine5.5, eBioscience™ | Invitrogen | Cat# 35-4776-42; RRID: AB_11218682; clone PCH101 |
| PE/Cyanine7 anti-human CD279 (PD-1) Antibody | BioLegend | Cat# 329918; clone EH12.2H7 |
| PE/Fire™ 810 anti-human CD223 (LAG-3) Antibody | BioLegend | Cat# 369223; clone 7H2C65 |
| TOX Antibody, anti-human/mouse, APC, REAfinity™ | Miltenyi Biotec | Cat# 130-118-335; clone REA473 |
| BD Horizon™ R718 Mouse Anti-TCF-7/TCF-1 | BD Biosciences | Cat# 567587; clone S33-966 |
| APC/Fire™ 750 anti-human CD39 Antibody | BioLegend | Cat# 328229; RRID: AB_2650838; clone A1 |
| APC/Fire™ 810 anti-human CD27 Recombinant Antibody | BioLegend | Cat# 393213; RRID: AB_2860961; clone QA17A18 |
| Biological samples | ||
| Patient PBMC and tissue | University of Pennsylvania | IRB |
| Chemicals, peptides, and recombinant proteins | ||
| eBioscience™ Foxp3/Transcription Factor Staining Buffer Set | Invitrogen | Cat# 00-5523-00 |
| CMV pp65 (YSE) YSEHPTFTSQY | GenScript | Custom peptide synthesis |
| FluNP (CTE) CTELKLSDY | GenScript | Custom peptide synthesis |
| CMV pp50 (VTE) VTEHDTLLY | GenScript | Custom peptide synthesis |
| Tyrosinase (SSD) SSDYVIPIGTY | GenScript | Custom peptide synthesis |
| MAGE A1 (EAD) EADPTGHSY | GenScript | Custom peptide synthesis |
| MAGE A3 (EVD) EVDPIGHLY | GenScript | Custom peptide synthesis |
| AIM2 (RSD) RSDSGQQARY | GenScript | Custom peptide synthesis |
| N-RAS Q61R (ILD) ILDTAGREEY | GenScript | Custom peptide synthesis |
| CMVpp65 (NLV) NLVPMVATV | GenScript | Custom peptide synthesis |
| MART-1 (ELA) ELAGIGILTV | GenScript | Custom peptide synthesis |
| GP100 280-9V (YLE) YLEPGPVTV | GenScript | Custom peptide synthesis |
| GP100 209-2M (IMD) IMDQVPFSV | GenScript | Custom peptide synthesis |
| GP100 (KTW) KTWGQYWQV | GenScript | Custom peptide synthesis |
| GnTV (VLP) VLPDVFIRCV | GenScript | Custom peptide synthesis |
| LAGE-1 (MLM) MLMAQEALAFL | GenScript | Custom peptide synthesis |
| SSX-2 (KAS) KASEKIFYV | GenScript | Custom peptide synthesis |
| NY-ESO-1 (SLL) SLLMWITQA | GenScript | Custom peptide synthesis |
| PRDX5 (AMA) AMAPIKVRL | GenScript | Custom peptide synthesis |
| CMVpp50 (TVR) TVRSHCVSK | GenScript | Custom peptide synthesis |
| Flu M1 (SII) SIIPSGPLK | GenScript | Custom peptide synthesis |
| EBV BRLF1 (RVR) RVRAYTYSK | GenScript | Custom peptide synthesis |
| EBV EBNA 3A (RLR) RLRAEAQVK | GenScript | Custom peptide synthesis |
| GP100 (ALL) ALLAVGATK | GenScript | Custom peptide synthesis |
| GP100 (LIY) LIYRRRLMK | GenScript | Custom peptide synthesis |
| GP100 (ALN) ALNFPGSQK | GenScript | Custom peptide synthesis |
| TAG (RLS) RLSNRLLLR | GenScript | Custom peptide synthesis |
| MAGE A1 (SLF) SLFRAVIIK | GenScript | Custom peptide synthesis |
| Flex-T™ HLA-A∗01:01 Monomer UVX | BioLegend | Cat# 280001 |
| Flex-T™ HLA-A∗02:01 Monomer UVX | BioLegend | Cat# 280003 |
| Flex-T™ HLA-A∗03:01 Monomer UVX | BioLegend | Cat# 280005 |
| Deposited data | ||
| This work (flow data) | Omiq | |
| Software and algorithms | ||
| R | R | 4.5.1 |
| RStudio | Posit | 2025.09.1+401 |
| FlowJo | BD Biosciences | Version 10.10.0 |
| randomForest | CRAN | Version 4.7-1.2 |
| stats | CRAN | Version 3.6.2 |
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grants R01 CA273018 (ACH), R01 CA258113 (XX), P50 CA261608 (ACH, GCK, XX, RKA), P30 CA016520 (ACH). In addition, ACH was supported by Damon Runyon Clinical Investigator Award, Damon Runyon Continuation Award, Doris Duke Clinical Scientist Development Award, the W.W. Smith Charitable Trust Award, and the Institute for Translational Medicine and Therapeutics of the Perelman School of Medicine, and the Pew-Stewart Stewart Scholars Program in Cancer Research. Work in the Huang Lab is funded by the Tara Miller Foundation and the Parker Institute for Cancer Immunotherapy. We thank CIC IGR-Curie 1428, Ton Schumacher and Mireille Toebes, for sharing tetramer reagents. We are grateful to the NIH Tetramer Core for providing biotinylated UV cleavable monomers. The authors thank the Human Immunology Core at the Perelman School of Medicine at the University of Pennsylvania for assistance with PBMC processing, plasma/serum isolation and tumor/tissue dissociation. The HIC is supported in part by NIH P30 AI045008 and P30 CA016520. HIC RRID: SCR_022380. DY and ACH thank the Robert Wood Johnson Distinction in Research Program, and the SAC including Dr. Cole, and Dr. Yurchenco. We thank E. John Wherry for his continual advice and support. ACH thanks Beatriz Carreno, Gerald Linette, Tara Mitchell, E. John Wherry, Ivan Maillard, and Jesus Christ for making me the investigator I am today.
DECLARATION OF INTERESTS
BMS provided funding for the clinical trial of neoadjuvant nivolumab. ACH performed consulting work for Immunai and received research funding from Bristol Myers Squibb (BMS) and Merck. TCM received honorarium for Scientific Advisory Board participation from: BMS, GigaGen, Merck, Pliant, Pfizer. GCK is on the Merck Advisory Board. RKA is a consultant for Tasca therapeutics and received clinical trial funding from Merck, BMS, and deciphera.
REFERENCES
- 1.Wolchok JD, Chiarion-Sileni V, Gonzalez R, Grob J-J, Rutkowski P, Lao CD, Cowey CL, Schadendorf D, Wagstaff J, Dummer R, et al. (2022). Long-Term Outcomes With Nivolumab Plus Ipilimumab or Nivolumab Alone Versus Ipilimumab in Patients With Advanced Melanoma. Journal of Clinical Oncology 40, 127–137. 10.1200/jco.21.02229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, Powderly JD, Carvajal RD, Sosman JA, Atkins MB, et al. (2012). Safety, Activity, and Immune Correlates of Anti–PD-1 Antibody in Cancer. New England Journal of Medicine 366, 2443–2454. 10.1056/nejmoa1200690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Garon EB, Rizvi NA, Hui R, Leighl N, Balmanoukian AS, Eder JP, Patnaik A, Aggarwal C, Gubens M, Horn L, et al. (2015). Pembrolizumab for the Treatment of Non–Small-Cell Lung Cancer. New England Journal of Medicine 372, 2018–2028. 10.1056/nejmoa1501824. [DOI] [PubMed] [Google Scholar]
- 4.Doroshow DB, Bhalla S, Beasley MB, Sholl LM, Kerr KM, Gnjatic S, Wistuba II, Rimm DL, Tsao MS, and Hirsch FR. (2021). PD-L1 as a biomarker of response to immune-checkpoint inhibitors. Nat Rev Clin Oncol 18, 345–362. 10.1038/s41571-021-00473-5. [DOI] [PubMed] [Google Scholar]
- 5.Rizvi NA, Hellmann MD, Snyder A, Kvistborg P, Makarov V, Havel JJ, Lee W, Yuan J, Wong P, Ho TS, et al. (2015). Mutational landscape determines sensitivity to PD-1 blockade in non–small cell lung cancer. Science 348, 124–128. 10.1126/science.aaa1348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hugo W, Zaretsky JM, Sun L, Song C, Moreno BH, Hu-Lieskovan S, Berent-Maoz B, Pang J, Chmielowski B, Cherry G, et al. (2016). Genomic and Transcriptomic Features of Response to Anti-PD-1 Therapy in Metastatic Melanoma. Cell 165, 35–44. 10.1016/j.cell.2016.02.065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hellmann MD, Ciuleanu T-E, Pluzanski A, Lee JS, Otterson GA, Audigier-Valette C, Minenza E, Linardou H, Burgers S, Salman P, et al. (2018). Nivolumab plus Ipilimumab in Lung Cancer with a High Tumor Mutational Burden. New England Journal of Medicine 378, 2093–2104. 10.1056/nejmoa1801946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yarchoan M, Hopkins A, and Jaffee EM. (2017). Tumor Mutational Burden and Response Rate to PD-1 Inhibition. New England Journal of Medicine 377, 2500–2501. 10.1056/nejmc1713444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ayers M, Lunceford J, Nebozhyn M, Murphy E, Loboda A, Kaufman DR, Albright A, Cheng JD, Kang SP, Shankaran V, et al. (2017). IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade. J Clin Invest 127, 2930–2940. 10.1172/JCI91190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Reijers ILM, Rao D, Versluis JM, Menzies AM, Dimitriadis P, Wouters MW, Spillane AJ, Klop WMC, Broeks A, Bosch LJW, et al. (2023). IFN-gamma signature enables selection of neoadjuvant treatment in patients with stage III melanoma. J Exp Med 220. 10.1084/jem.20221952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cristescu R, Mogg R, Ayers M, Albright A, Murphy E, Yearley J, Sher X, Liu XQ, Lu H, Nebozhyn M, et al. (2018). Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy. Science 362. 10.1126/science.aar3593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Azimi F, Scolyer RA, Rumcheva P, Moncrieff M, Murali R, McCarthy SW, Saw RP, and Thompson JF. (2012). Tumor-infiltrating lymphocyte grade is an independent predictor of sentinel lymph node status and survival in patients with cutaneous melanoma. J Clin Oncol 30, 2678–2683. 10.1200/JCO.2011.37.8539. [DOI] [PubMed] [Google Scholar]
- 13.Clark WH Jr., Elder DE, Guerry D.t., Braitman LE, Trock BJ, Schultz D, Synnestvedt M, and Halpern AC. (1989). Model predicting survival in stage I melanoma based on tumor progression. J Natl Cancer Inst 81, 1893–1904. 10.1093/jnci/81.24.1893. [DOI] [PubMed] [Google Scholar]
- 14.Clark WH Jr., From L, Bernardino EA, and Mihm MC. (1969). The histogenesis and biologic behavior of primary human malignant melanomas of the skin. Cancer Res 29, 705–727. [PubMed] [Google Scholar]
- 15.Clemente CG, Mihm MC Jr., Bufalino R, Zurrida S, Collini P, and Cascinelli N. (1996). Prognostic value of tumor infiltrating lymphocytes in the vertical growth phase of primary cutaneous melanoma. Cancer 77, 1303–1310. 10.1002/(SICI)1097-0142(19960401)77:7<1303::AID-CNCR12>3.0.CO;2-5. [DOI] [PubMed] [Google Scholar]
- 16.Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, Gonzalez R, Robert C, Schadendorf D, Hassel JC, et al. (2010). Improved Survival with Ipilimumab in Patients with Metastatic Melanoma. New England Journal of Medicine 363, 711–723. 10.1056/nejmoa1003466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Robert C, Thomas L, Bondarenko I, O'Day S, Weber J, Garbe C, Lebbe C, Baurain J-F, Testori A, Grob J-J, et al. (2011). Ipilimumab plus Dacarbazine for Previously Untreated Metastatic Melanoma. New England Journal of Medicine 364, 2517–2526. 10.1056/nejmoa1104621. [DOI] [PubMed] [Google Scholar]
- 18.Ascierto PA, Long GV, Robert C, Brady B, Dutriaux C, Di Giacomo AM, Mortier L, Hassel JC, Rutkowski P, McNeil C, et al. (2019). Survival Outcomes in Patients With Previously Untreated BRAF Wild-Type Advanced Melanoma Treated With Nivolumab Therapy. JAMA Oncology 5, 187. 10.1001/jamaoncol.2018.4514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Robert C, Long GV, Brady B, Dutriaux C, Maio M, Mortier L, Hassel JC, Rutkowski P, McNeil C, Kalinka-Warzocha E, et al. (2015). Nivolumab in Previously Untreated Melanoma withoutBRAFMutation. New England Journal of Medicine 372, 320–330. 10.1056/nejmoa1412082. [DOI] [PubMed] [Google Scholar]
- 20.Robert C, Schachter J, Long GV, Arance A, Grob JJ, Mortier L, Daud A, Carlino MS, McNeil C, Lotem M, et al. (2015). Pembrolizumab versus Ipilimumab in Advanced Melanoma. New England Journal of Medicine 372, 2521–2532. 10.1056/nejmoa1503093. [DOI] [PubMed] [Google Scholar]
- 21.Robert C, Ribas A, Schachter J, Arance A, Grob J-J, Mortier L, Daud A, Carlino MS, McNeil CM, Lotem M, et al. (2019). Pembrolizumab versus ipilimumab in advanced melanoma (KEYNOTE-006): post-hoc 5-year results from an open-label, multicentre, randomised, controlled, phase 3 study. The Lancet Oncology 20, 1239–1251. 10.1016/s1470-2045(19)30388-2. [DOI] [PubMed] [Google Scholar]
- 22.Andersen RS, Thrue CA, Junker N, Lyngaa R, Donia M, Ellebaek E, Svane IM, Schumacher TN, Thor Straten P, and Hadrup SR. (2012). Dissection of T-cell antigen specificity in human melanoma. Cancer Res 72, 1642–1650. 10.1158/0008-5472.CAN-11-2614. [DOI] [PubMed] [Google Scholar]
- 23.Patel SP, Othus M, Chen Y, Wright GP, Yost KJ, Hyngstrom JR, Hu-Lieskovan S, Lao CD, Fecher LA, Truong T-G, et al. (2023). Neoadjuvant–Adjuvant or Adjuvant-Only Pembrolizumab in Advanced Melanoma. New England Journal of Medicine 388, 813–823. 10.1056/nejmoa2211437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Huang AC, Orlowski RJ, Xu X, Mick R, George SM, Yan PK, Manne S, Kraya AA, Wubbenhorst B, Dorfman L, et al. (2019). A single dose of neoadjuvant PD-1 blockade predicts clinical outcomes in resectable melanoma. Nature Medicine 25, 454–461. 10.1038/s41591-019-0357-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Menzies AM, Amaria RN, Rozeman EA, Huang AC, Tetzlaff MT, van de Wiel BA, Lo S, Tarhini AA, Burton EM, Pennington TE, et al. (2021). Pathological response and survival with neoadjuvant therapy in melanoma: a pooled analysis from the International Neoadjuvant Melanoma Consortium (INMC). Nat Med 27, 301–309. 10.1038/s41591-020-01188-3. [DOI] [PubMed] [Google Scholar]
- 26.Amaria RN, Reddy SM, Tawbi HA, Davies MA, Ross MI, Glitza IC, Cormier JN, Lewis C, Hwu W-J, Hanna E, et al. (2018). Neoadjuvant immune checkpoint blockade in high-risk resectable melanoma. Nature Medicine 24, 1649–1654. 10.1038/s41591-018-0197-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hellmann MD, Chaft JE, William WN, Rusch V, Pisters KMW, Kalhor N, Pataer A, Travis WD, Swisher SG, and Kris MG. (2014). Pathological response after neoadjuvant chemotherapy in resectable non-small-cell lung cancers: proposal for the use of major pathological response as a surrogate endpoint. The Lancet Oncology 15, e42–e50. 10.1016/s1470-2045(13)70334-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Forde PM, Chaft JE, Smith KN, Anagnostou V, Cottrell TR, Hellmann MD, Zahurak M, Yang SC, Jones DR, Broderick S, et al. (2018). Neoadjuvant PD-1 Blockade in Resectable Lung Cancer. New England Journal of Medicine 378, 1976–1986. 10.1056/nejmoa1716078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Toebes M, Rodenko B, Ovaa H, and Schumacher TNM. (2009). Generation of Peptide MHC Class I Monomers and Multimers Through Ligand Exchange. Current Protocols in Immunology 87. 10.1002/0471142735.im1816s87. [DOI] [PubMed] [Google Scholar]
- 30.Hadrup SR, Bakker AH, Shu CJ, Andersen RS, Van Veluw J, Hombrink P, Castermans E, Thor Straten P, Blank C, Haanen JB, et al. (2009). Parallel detection of antigen-specific T-cell responses by multidimensional encoding of MHC multimers. Nature Methods 6, 520–526. 10.1038/nmeth.1345. [DOI] [PubMed] [Google Scholar]
- 31.Newell EW, Klein LO, Yu W, and Davis MM. (2009). Simultaneous detection of many T-cell specificities using combinatorial tetramer staining. Nature Methods 6, 497–499. 10.1038/nmeth.1344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Van Gassen S, Callebaut B, Van Helden MJ, Lambrecht BN, Demeester P, Dhaene T, and Saeys Y. (2015). FlowSOM: Using self-organizing maps for visualization and interpretation of cytometry data. Cytometry Part A 87, 636–645. 10.1002/cyto.a.22625. [DOI] [PubMed] [Google Scholar]
- 33.Alfei F, Kanev K, Hofmann M, Wu M, Ghoneim HE, Roelli P, Utzschneider DT, von Hoesslin M, Cullen JG, Fan Y, et al. (2019). TOX reinforces the phenotype and longevity of exhausted T cells in chronic viral infection. Nature 571, 265–269. 10.1038/s41586-019-1326-9. [DOI] [PubMed] [Google Scholar]
- 34.Khan O, Giles JR, McDonald S, Manne S, Ngiow SF, Patel KP, Werner MT, Huang AC, Alexander KA, Wu JE, et al. (2019). TOX transcriptionally and epigenetically programs CD8(+) T cell exhaustion. Nature 571, 211–218. 10.1038/s41586-019-1325-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Scott AC, Dundar F, Zumbo P, Chandran SS, Klebanoff CA, Shakiba M, Trivedi P, Menocal L, Appleby H, Camara S, et al. (2019). TOX is a critical regulator of tumour-specific T cell differentiation. Nature 571, 270–274. 10.1038/s41586-019-1324-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Seo H, Chen J, Gonzalez-Avalos E, Samaniego-Castruita D, Das A, Wang YH, Lopez-Moyado IF, Georges RO, Zhang W, Onodera A, et al. (2019). TOX and TOX2 transcription factors cooperate with NR4A transcription factors to impose CD8(+) T cell exhaustion. Proc Natl Acad Sci U S A 116, 12410–12415. 10.1073/pnas.1905675116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yao C, Sun HW, Lacey NE, Ji Y, Moseman EA, Shih HY, Heuston EF, Kirby M, Anderson S, Cheng J, et al. (2019). Single-cell RNA-seq reveals TOX as a key regulator of CD8(+) T cell persistence in chronic infection. Nat Immunol 20, 890–901. 10.1038/s41590-019-0403-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gupta PK, Godec J, Wolski D, Adland E, Yates K, Pauken KE, Cosgrove C, Ledderose C, Junger WG, Robson SC, et al. (2015). CD39 Expression Identifies Terminally Exhausted CD8+ T Cells. PLOS Pathogens 11, e1005177. 10.1371/journal.ppat.1005177. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Khan O, Giles JR, McDonald S, Manne S, Ngiow SF, Patel KP, Werner MT, Huang AC, Alexander KA, Wu JE, et al. (2019). TOX transcriptionally and epigenetically programs CD8+ T cell exhaustion. Nature 571, 211–218. 10.1038/s41586-019-1325-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Robert, Comps-Agrar L, Hackney J, Yu X, Huseni M, Yang Y, Park S, Javinal V, Chiu H, Irving B, et al. (2014). The Immunoreceptor TIGIT Regulates Antitumor and Antiviral CD8 + T Cell Effector Function. Cancer Cell 26, 923–937. 10.1016/j.ccell.2014.10.018. [DOI] [PubMed] [Google Scholar]
- 41.Barber DL, Wherry EJ, Masopust D, Zhu B, Allison JP, Sharpe AH, Freeman GJ, and Ahmed R. (2006). Restoring function in exhausted CD8 T cells during chronic viral infection. Nature 439, 682–687. 10.1038/nature04444. [DOI] [PubMed] [Google Scholar]
- 42.Sakuishi K, Apetoh L, Sullivan JM, Blazar BR, Kuchroo VK, and Anderson AC. (2010). Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restore anti-tumor immunity. Journal of Experimental Medicine 207, 2187–2194. 10.1084/jem.20100643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Jin H-T, Anderson AC, Tan WG, West EE, Ha S-J, Araki K, Freeman GJ, Kuchroo VK, and Ahmed R. (2010). Cooperation of Tim-3 and PD-1 in CD8 T-cell exhaustion during chronic viral infection. Proceedings of the National Academy of Sciences 107, 14733–14738. 10.1073/pnas.1009731107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Blackburn SD, Shin H, Haining WN, Zou T, Workman CJ, Polley A, Betts MR, Freeman GJ, Vignali DA, and Wherry EJ. (2009). Coregulation of CD8+ T cell exhaustion by multiple inhibitory receptors during chronic viral infection. Nat Immunol 10, 29–37. 10.1038/ni.1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Im SJ, Hashimoto M, Gerner MY, Lee J, Kissick HT, Burger MC, Shan Q, Hale JS, Lee J, Nasti TH, et al. (2016). Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature 537, 417–421. 10.1038/nature19330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Im SJ, Hashimoto M, Gerner MY, Lee J, Kissick HT, Burger MC, Shan Q, Hale JS, Lee J, Nasti TH, et al. (2016). Defining CD8+ T cells that provide the proliferative burst after PD-1 therapy. Nature 537, 417–421. 10.1038/nature19330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Miller BC, Sen DR, Al Abosy R, Bi K, Virkud YV, LaFleur MW, Yates KB, Lako A, Felt K, Naik GS, et al. (2019). Subsets of exhausted CD8(+) T cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol 20, 326–336. 10.1038/s41590-019-0312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Utzschneider DT, Charmoy M, Chennupati V, Pousse L, Ferreira DP, Calderon-Copete S, Danilo M, Alfei F, Hofmann M, Wieland D, et al. (2016). T Cell Factor 1-Expressing Memory-like CD8(+) T Cells Sustain the Immune Response to Chronic Viral Infections. Immunity 45, 415–427. 10.1016/j.immuni.2016.07.021. [DOI] [PubMed] [Google Scholar]
- 49.Beltra J-C, Manne S, Abdel-Hakeem MS, Kurachi M, Giles JR, Chen Z, Casella V, Ngiow SF, Khan O, Huang YJ, et al. (2020). Developmental Relationships of Four Exhausted CD8+ T Cell Subsets Reveals Underlying Transcriptional and Epigenetic Landscape Control Mechanisms. Immunity 52, 825–841.e828. 10.1016/j.immuni.2020.04.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Beltra J-C, Abdel-Hakeem MS, Manne S, Zhang Z, Huang H, Kurachi M, Su L, Picton L, Ngiow SF, Muroyama Y, et al. (2023). Stat5 opposes the transcription factor Tox and rewires exhausted CD8+ T cells toward durable effector-like states during chronic antigen exposure. Immunity 56, 2699–2718.e2611. 10.1016/j.immuni.2023.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Hudson WH, Gensheimer J, Hashimoto M, Wieland A, Valanparambil RM, Li P, Lin J-X, Konieczny BT, Im SJ, Freeman GJ, et al. (2019). Proliferating Transitory T Cells with an Effector-like Transcriptional Signature Emerge from PD-1+ Stem-like CD8+ T Cells during Chronic Infection. Immunity 51, 1043–1058.e1044. 10.1016/j.immuni.2019.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Zhou P, Shi H, Huang H, Sun X, Yuan S, Chapman NM, Connelly JP, Lim SA, Saravia J, Kc A, et al. (2023). Single-cell CRISPR screens in vivo map T cell fate regulomes in cancer. Nature 624, 154–163. 10.1038/s41586-023-06733-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Beltra JC, Abdel-Hakeem MS, Manne S, Zhang Z, Huang H, Kurachi M, Su L, Picton L, Ngiow SF, Muroyama Y, et al. (2023). Stat5 opposes the transcription factor Tox and rewires exhausted CD8(+) T cells toward durable effector-like states during chronic antigen exposure. Immunity 56, 2699–2718 e2611. 10.1016/j.immuni.2023.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Yossef R, Krishna S, Sindiri S, Lowery FJ, Copeland AR, Gartner JJ, Parkhurst MR, Parikh NB, Hitscherich KJ, Levi ST, et al. (2023). Phenotypic signatures of circulating neoantigen-reactive CD8(+) T cells in patients with metastatic cancers. Cancer Cell 41, 2154–2165 e2155. 10.1016/j.ccell.2023.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Gill AL, Wang PH, Lee J, Hudson WH, Ando S, Araki K, Hu Y, Wieland A, Im S, Gavora A, et al. (2023). PD-1 blockade increases the self-renewal of stem-like CD8 T cells to compensate for their accelerated differentiation into effectors. Science Immunology 8. 10.1126/sciimmunol.adg0539. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Miller BC, Sen DR, Al Abosy R, Bi K, Virkud YV, Lafleur MW, Yates KB, Lako A, Felt K, Naik GS, et al. (2019). Subsets of exhausted CD8+ T cells differentially mediate tumor control and respond to checkpoint blockade. Nature Immunology 20, 326–336. 10.1038/s41590-019-0312-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.MartiíN-Fontecha A, Sebastiani S, HoöPken UE, Uguccioni M, Lipp M, Lanzavecchia A, and Sallusto F. (2003). Regulation of Dendritic Cell Migration to the Draining Lymph Node. The Journal of Experimental Medicine 198, 615–621. 10.1084/jem.20030448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Bousso P, and Robey E. (2003). Dynamics of CD8+ T cell priming by dendritic cells in intact lymph nodes. Nature Immunology 4, 579–585. 10.1038/ni928. [DOI] [PubMed] [Google Scholar]
- 59.Du Bois H, Heim TA, and Lund AW. (2021). Tumor-draining lymph nodes: At the crossroads of metastasis and immunity. Science Immunology 6. 10.1126/sciimmunol.abg3551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Simoni Y, Becht E, Fehlings M, Loh CY, Koo SL, Teng KWW, Yeong JPS, Nahar R, Zhang T, Kared H, et al. (2018). Bystander CD8(+) T cells are abundant and phenotypically distinct in human tumour infiltrates. Nature 557, 575–579. 10.1038/s41586-018-0130-2. [DOI] [PubMed] [Google Scholar]
- 61.Pulliam T, Jani S, Jing L, Ryu H, Jojic A, Shasha C, Zhang J, Kulikauskas R, Church C, Garnett-Benson C, et al. (2024). Circulating cancer-specific CD8 T cell frequency is associated with response to PD-1 blockade in Merkel cell carcinoma. Cell Reports Medicine 5, 101412. 10.1016/j.xcrm.2024.101412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Wang K, Coutifaris P, Brocks D, Wang G, Azar T, Solis S, Nandi A, Anderson S, Han N, Manne S, et al. (2024). Combination anti-PD-1 and anti-CTLA-4 therapy generates waves of clonal responses that include progenitor-exhausted CD8(+) T cells. Cancer Cell 42, 1582–1597 e1510. 10.1016/j.ccell.2024.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Huang AC, Postow MA, Orlowski RJ, Mick R, Bengsch B, Manne S, Xu W, Harmon S, Giles JR, Wenz B, et al. (2017). T-cell invigoration to tumour burden ratio associated with anti-PD-1 response. Nature 545, 60–65. 10.1038/nature22079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Liu J, Blake SJ, Yong MC, Harjunpaa H, Ngiow SF, Takeda K, Young A, O'Donnell JS, Allen S, Smyth MJ, and Teng MW. (2016). Improved Efficacy of Neoadjuvant Compared to Adjuvant Immunotherapy to Eradicate Metastatic Disease. Cancer Discov 6, 1382–1399. 10.1158/2159-8290.CD-16-0577. [DOI] [PubMed] [Google Scholar]
- 65.Carreno BM, Magrini V, Becker-Hapak M, Kaabinejadian S, Hundal J, Petti AA, Ly A, Lie W-R, Hildebrand WH, Mardis ER, and Linette GP. (2015). A dendritic cell vaccine increases the breadth and diversity of melanoma neoantigen-specific T cells. Science 348, 803–808. 10.1126/science.aaa3828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66.Wolfel T, Van Pel A, Brichard V, Schneider J, Seliger B, Meyer zum Buschenfelde KH, and Boon T. (1994). Two tyrosinase nonapeptides recognized on HLA-A2 melanomas by autologous cytolytic T lymphocytes. Eur J Immunol 24, 759–764. 10.1002/eji.1830240340. [DOI] [PubMed] [Google Scholar]
- 67.Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Sakaguchi K, Appella E, Yannelli JR, Adema GJ, Miki T, and Rosenberg SA. (1994). Identification of a human melanoma antigen recognized by tumor-infiltrating lymphocytes associated with in vivo tumor rejection. Proc Natl Acad Sci U S A 91, 6458–6462. 10.1073/pnas.91.14.6458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68.Kawakami Y, Eliyahu S, Sakaguchi K, Robbins PF, Rivoltini L, Yannelli JR, Appella E, and Rosenberg SA. (1994). Identification of the immunodominant peptides of the MART-1 human melanoma antigen recognized by the majority of HLA-A2-restricted tumor infiltrating lymphocytes. J Exp Med 180, 347–352. 10.1084/jem.180.1.347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 69.Cox AL, Skipper J, Chen Y, Henderson RA, Darrow TL, Shabanowitz J, Engelhard VH, Hunt DF, and Slingluff CL Jr. (1994). Identification of a peptide recognized by five melanoma-specific human cytotoxic T cell lines. Science 264, 716–719. 10.1126/science.7513441. [DOI] [PubMed] [Google Scholar]
- 70.Morgan RA, Dudley ME, Wunderlich JR, Hughes MS, Yang JC, Sherry RM, Royal RE, Topalian SL, Kammula US, Restifo NP, et al. (2006). Cancer regression in patients after transfer of genetically engineered lymphocytes. Science 314, 126–129. 10.1126/science.1129003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 71.Lo JA, Kawakubo M, Juneja VR, Su MY, Erlich TH, Lafleur MW, Kemeny LV, Rashid M, Malehmir M, Rabi SA, et al. (2021). Epitope spreading toward wild-type melanocyte-lineage antigens rescues suboptimal immune checkpoint blockade responses. Science Translational Medicine 13, eabd8636. 10.1126/scitranslmed.abd8636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 72.Wherry EJ. (2011). T cell exhaustion. Nat Immunol 12, 492–499. 10.1038/ni.2035. [DOI] [PubMed] [Google Scholar]
- 73.Teijaro JR, Ng C, Lee AM, Sullivan BM, Sheehan KC, Welch M, Schreiber RD, de la Torre JC, and Oldstone MB. (2013). Persistent LCMV infection is controlled by blockade of type I interferon signaling. Science 340, 207–211. 10.1126/science.1235214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Wilson EB, Yamada DH, Elsaesser H, Herskovitz J, Deng J, Cheng G, Aronow BJ, Karp CL, and Brooks DG. (2013). Blockade of chronic type I interferon signaling to control persistent LCMV infection. Science 340, 202–207. 10.1126/science.1235208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 75.Saadey AA, Yousif A, Osborne N, Shahinfar R, Chen YL, Laster B, Rajeev M, Bauman P, Webb A, and Ghoneim HE. (2023). Rebalancing TGFbeta1/BMP signals in exhausted T cells unlocks responsiveness to immune checkpoint blockade therapy. Nat Immunol 24, 280–294. 10.1038/s41590-022-01384-y. [DOI] [PubMed] [Google Scholar]
- 76.Kim CG, Jang M, Kim Y, Leem G, Kim KH, Lee H, Kim TS, Choi SJ, Kim HD, Han JW, et al. (2019). VEGF-A drives TOX-dependent T cell exhaustion in anti-PD-1-resistant microsatellite stable colorectal cancers. Sci Immunol 4. 10.1126/sciimmunol.aay0555. [DOI] [PubMed] [Google Scholar]
- 77.Lucca LE, Axisa PP, Lu B, Harnett B, Jessel S, Zhang L, Raddassi K, Zhang L, Olino K, Clune J, et al. (2021). Circulating clonally expanded T cells reflect functions of tumor-infiltrating T cells. J Exp Med 218. 10.1084/jem.20200921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 78.Pauken KE, Shahid O, Lagattuta KA, Mahuron KM, Luber JM, Lowe MM, Huang L, Delaney C, Long JM, Fung ME, et al. (2021). Single-cell analyses identify circulating anti-tumor CD8 T cells and markers for their enrichment. J Exp Med 218. 10.1084/jem.20200920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 79.Gros A, Parkhurst MR, Tran E, Pasetto A, Robbins PF, Ilyas S, Prickett TD, Gartner JJ, Crystal JS, Roberts IM, et al. (2016). Prospective identification of neoantigen-specific lymphocytes in the peripheral blood of melanoma patients. Nat Med 22, 433–438. 10.1038/nm.4051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 80.Wu TD, Madireddi S, de Almeida PE, Banchereau R, Chen YJ, Chitre AS, Chiang EY, Iftikhar H, O'Gorman WE, Au-Yeung A, et al. (2020). Peripheral T cell expansion predicts tumour infiltration and clinical response. Nature 579, 274–278. 10.1038/s41586-020-2056-8. [DOI] [PubMed] [Google Scholar]
- 81.Beltra JC, Manne S, Abdel-Hakeem MS, Kurachi M, Giles JR, Chen Z, Casella V, Ngiow SF, Khan O, Huang YJ, et al. (2020). Developmental Relationships of Four Exhausted CD8(+) T Cell Subsets Reveals Underlying Transcriptional and Epigenetic Landscape Control Mechanisms. Immunity 52, 825–841 e828. 10.1016/j.immuni.2020.04.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 82.Chen Y, Zander RA, Wu X, Schauder DM, Kasmani MY, Shen J, Zheng S, Burns R, Taparowsky EJ, and Cui W. (2021). BATF regulates progenitor to cytolytic effector CD8(+) T cell transition during chronic viral infection. Nat Immunol 22, 996–1007. 10.1038/s41590-021-00965-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 83.Hudson WH, Gensheimer J, Hashimoto M, Wieland A, Valanparambil RM, Li P, Lin JX, Konieczny BT, Im SJ, Freeman GJ, et al. (2019). Proliferating Transitory T Cells with an Effector-like Transcriptional Signature Emerge from PD-1(+) Stem-like CD8(+) T Cells during Chronic Infection. Immunity 51, 1043–1058 e1044. 10.1016/j.immuni.2019.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 84.Caushi JX, Zhang J, Ji Z, Vaghasia A, Zhang B, Hsiue EH, Mog BJ, Hou W, Justesen S, Blosser R, et al. (2021). Transcriptional programs of neoantigen-specific TIL in anti-PD-1-treated lung cancers. Nature 596, 126–132. 10.1038/s41586-021-03752-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 85.Rudloff MW, Zumbo P, Favret NR, Roetman JJ, Detres Roman CR, Erwin MM, Murray KA, Jonnakuti ST, Dundar F, Betel D, and Philip M. (2023). Hallmarks of CD8(+) T cell dysfunction are established within hours of tumor antigen encounter before cell division. Nat Immunol 24, 1527–1539. 10.1038/s41590-023-01578-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 86.Rahim MK, Okholm TLH, Jones KB, McCarthy EE, Liu CC, Yee JL, Tamaki SJ, Marquez DM, Tenvooren I, Wai K, et al. (2023). Dynamic CD8+ T cell responses to cancer immunotherapy in human regional lymph nodes are disrupted in metastatic lymph nodes. Cell 186, 1127–1143.e1118. 10.1016/j.cell.2023.02.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 87.Prokhnevska N, Cardenas MA, Valanparambil RM, Sobierajska E, Barwick BG, Jansen C, Reyes Moon A, Gregorova P, Delbalzo L, Greenwald R, et al. (2023). CD8+ T cell activation in cancer comprises an initial activation phase in lymph nodes followed by effector differentiation within the tumor. Immunity 56, 107–124.e105. 10.1016/j.immuni.2022.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 88.Cardenas MA, Prokhnevska N, Sobierajska E, Gregorova P, Medina CB, Valanparambil RM, Greenwald R, DelBalzo L, Bilen MA, Joshi SS, et al. (2024). Differentiation fate of a stem-like CD4 T cell controls immunity to cancer. Nature 636, 224–232. 10.1038/s41586-024-08076-7. [DOI] [PubMed] [Google Scholar]
- 89.Pauken KE, Sammons MA, Odorizzi PM, Manne S, Godec J, Khan O, Drake AM, Chen ZY, Sen DR, Kurachi M, et al. (2016). Epigenetic stability of exhausted T cells limits durability of reinvigoration by PD-1 blockade. Science 354, 1160–1165. 10.1126/science.aaf2807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 90.Fairfax BP, Taylor CA, Watson RA, Nassiri I, Danielli S, Fang H, Mahe EA, Cooper R, Woodcock V, Traill Z, et al. (2020). Peripheral CD8(+) T cell characteristics associated with durable responses to immune checkpoint blockade in patients with metastatic melanoma. Nat Med 26, 193–199. 10.1038/s41591-019-0734-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 91.Valpione S, Galvani E, Tweedy J, Mundra PA, Banyard A, Middlehurst P, Barry J, Mills S, Salih Z, Weightman J, et al. (2020). Immune-awakening revealed by peripheral T cell dynamics after one cycle of immunotherapy. Nat Cancer 1, 210–221. 10.1038/s43018-019-0022-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 92.Yamauchi T, Hoki T, Oba T, Jain V, Chen H, Attwood K, Battaglia S, George S, Chatta G, Puzanov I, et al. (2021). T-cell CX3CR1 expression as a dynamic blood-based biomarker of response to immune checkpoint inhibitors. Nat Commun 12, 1402. 10.1038/s41467-021-21619-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 93.Wang K, C.P., Brocks David, Solis S, Han N, Manne S, Kiner Evgeny, Sachar C, George S, Yan P, Kier M, Laughlin AI, Kothari S, Giles JR. and Mathew D, Ghinnagow R, Alanio C, Flowers A, Xu W, Tenney D, Xu X, Amaravadi RK, Karakousis GC, Schuchter LM, Buggert M, Wherry EJ, Minn A, Weber J, Mitchell TC, Herati R, Huang AC. (2023). Reinvigoration of Progenitor-Exhausted CD8 T Cells by anti-CTLA-4 Contributes to the Sustained Activity of Combination Checkpoint Blockade.. Available at SSRN: https://ssrn.com/abstract=4493245or 10.2139/ssrn.4493245. [DOI] [Google Scholar]
- 94.Codarri Deak L, Nicolini V, Hashimoto M, Karagianni M, Schwalie PC, Lauener L, Varypataki EM, Richard M, Bommer E, Sam J, et al. (2022). PD-1-cis IL-2R agonism yields better effectors from stem-like CD8(+) T cells. Nature 610, 161–172. 10.1038/s41586-022-05192-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 95.Hashimoto M, Araki K, Cardenas MA, Li P, Jadhav RR, Kissick HT, Hudson WH, McGuire DJ, Obeng RC, Wieland A, et al. (2022). PD-1 combination therapy with IL-2 modifies CD8(+) T cell exhaustion program. Nature 610, 173–181. 10.1038/s41586-022-05257-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 96.West EE, Jin HT, Rasheed AU, Penaloza-Macmaster P, Ha SJ, Tan WG, Youngblood B, Freeman GJ, Smith KA, and Ahmed R. (2013). PD-L1 blockade synergizes with IL-2 therapy in reinvigorating exhausted T cells. J Clin Invest 123, 2604–2615. 10.1172/JCI67008. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S3. Tetramer antigen sequences, related to Figure 1
Table S1. Neoadjuvant anti-PD-1 (nivolumab) clinical trial patient metadata, related to Figure 1
Table S2. Antigen-specific CD8+ T cell summary, related to Figure 1
Document S1. Figures S1–S5
Figure S1. CD8+ T cell atlas of blood, lymph node, and tumor. Related to Figure 1. (A) Representative combinatorial tetramer flow plot identifying melanoma (red) and viral (blue)-specific CD8+ T cells.
(B) UMAP projection of CD8+ T cells stratified by tissue type.
(C) UMAP visualization of feature expressions.
(D) UMAP projection of CD8+ T cell states in combined blood, lymph node, and tumor defined by unsupervised clustering.
(E) UMAP projection of CD8+ T cell states in the blood and tumor.
(F) Expression of selected markers for CD8+ T cell states.
(G) Cell state composition of CD8+ T cells in the blood and tumor.
Figure S2. Antigen-specific CD8+ T cell phenotypes in the blood and tumor. Related to Figure 1. (A) Melanoma (red) and viral (blue)-specific CD8+ T cells projected on CD8+ T cell atlas in the blood and tumor. CD8+ T cells are in gray.
(B) Expression of selected markers on melanoma and viral-specific CD8+ T cells in the blood and tumor.
(C) PCA plot of melanoma and viral-specific CD8+ T cells in the blood and tumor. Each dot represents an individual antigen-specific population.
(D) Variables contributing to principal components.
(E) UMAP visualization of feature expressions.
(F) MFI of exhausted T cell features on melanoma and viral-specific CD8+ T cells in the blood and tumor. Each dot represents an individual tetramer+ population.
(G) MFI of progenitor and terminal exhausted T cell features on melanoma-specific CD8+ T cells in the blood and tumor separated by antigen. Each line connects paired samples from the same patient.
(H) Cell state composition of melanoma and viral-specific CD8+ T cells in the tumor.
(I) Cell state composition of melanoma and viral-specific CD8+ T cells in the blood.
(J) Cell state composition of melanoma and viral-specific CD8+ T cells in the tumor. Each dot represents an individual tetramer+ population.
(K) Cell state composition of melanoma and viral-specific CD8+ T cells in the blood. Each dot represents an individual tetramer+ population.
(L) MFI of memory and effector T cell features on viral-specific CD8+ T cells in the blood and tumor separated by antigen. Each line connects paired samples from the same patient.
Figure S3. Melanoma-specific CD8+ T cells in the uninvolved regional lymph nodes demonstrate features of exhaustion. Related to Figure 1. (A) Representative gross pathology (left) and H&E stain (right) of uninvolved lymph node (top) and lymph node metastasis (bottom).
(B) Expression of selected markers for CD8+ T cells in the lymph node (white), melanoma-specific CD8+ T cells in the lymph node (purple), and melanoma-specific CD8+ T cells in the tumor (red).
(C) UMAP projection of CD8+ T cell states in the lymph node.
(D) Melanoma (red) and viral (blue)-specific CD8+ T cells projected on CD8+ T cell atlas in the lymph node. CD8+ T cells in gray.
(E) Cell state composition of all CD8+, melanoma and viral-specific CD8+ T cells in the lymph node.
Figure S4. T-bet+ CD39low melanoma-specific CD8+ T cells are associated with pathologic response. Related to Figure 2. (A) AUC curves for immune parameters associated with pathologic response.
(B) Pathologic response rate stratified by CART-defined T-bet threshold (11.8%) in the patients with tetramer+ tumor.
(C) Twenty-six post-treatment melanoma-specific CD8+ T cell parameters classified by pathological response using random forest analysis and ranked by importance score. Error bar denotes mean ± s.d. for 1,000 random forest iterations.
(D) Percentage of T-bet+ and T-bet MFI of all CD8+ T cells stratified by response in the tumor. Each dot represents an individual patient.
(E) Percentage of CD39+ and CD39 MFI of all CD8+ T cells stratified by response in the tumor. Each dot represents an individual patient.
(F) Representative flow plots of melanoma-specific CD8+ T cells from responder (blue), and non-responder (red) in the tumor, overlayed.
(G) Percentage of progenitor exhausted CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population.
(H) Cell state composition of CD39hi (MFI>15431, CART-defined CD39 MFI threshold) melanoma-specific CD8+ T cells in the tumor.
(I) Percentage of terminally exhausted CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population.
(J) Study schema with patient demographics and HLA types in neoadjuvant anti-PD-1 (pembrolizumab) clinical trial.
(K) Pathologic responses separated by complete response (pCR), near-complete response (pNCR), partial response (pPR), and non-response (pNR) (pembrolizumab).
(L) Representative flow plots comparing T-bet and CD39 expressions on melanoma-specific CD8+ T cells in pCR and pNR patient (pembrolizumab).
(M) Percentage of T-bet+ melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population (pembrolizumab).
(N) CD39 MFI melanoma-specific CD8+ T cells stratified by response in the tumor. Each dot represents an individual tetramer+ population (pembrolizumab).
Figure S5. Pharmacodynamic changes in the percentage of circulating antigen-specific CD8+ T cells. Related to Figure 4 (A) Representative flow plots of the percentage of circulating melanoma-specific CD8+ T cells in pCR and pNR patients before anti-PD-1 neoadjuvant therapy.
(B) Percentage of circulating melanoma-specific CD8+ T cells stratified by response before anti-PD-1 neoadjuvant therapy. Each dot represents an individual tetramer+ population.
(C) Percentage of circulating melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents a distinct melanoma-specific CD8+ T cell population.
(D) Percentage of total circulating melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents the total melanoma-specific CD8+ T cell frequency in an individual patient.
(E) Percentage of circulating viral-specific CD8+ T cells across time stratified by Response (R top; NR bottom). Each line represents a distinct viral-specific CD8+ T cell population.
(F) Percentage of circulating KI-67+ melanoma-specific CD8+ T cells across time stratified by response (R top; NR bottom) and patient. Each line represents a distinct KI-67+ melanoma-specific CD8+ T cell population.
(G) Percentage of circulating melanoma-specific CD8+ T cells across time stratified by adjuvant treatment (Nivo top; Nivo + Ipi bottom). Each line represents a distinct melanoma-specific CD8+ T cell population.
Data Availability Statement
Flow cytometry data and associated clinical metadata have been deposited on Omiq and available upon request.
