Melanoma is a common and aggressive cancer, with rising incidence in most developed countries. Major discoveries in melanoma biology have been rapidly translated, allowing cures for patients in late-stage disease. Despite these advances, many tumors remain refractory, in part due to an incomplete understanding of the genes and pathways gained or lost during melanoma tumorigenesis. To address this gap and provide a broadly useful resource for the scientific community, we established melPDomiX, a multi-omics cohort of melanoma-patient-derived xenografts. By linking mutations with transcriptomic and proteomic features, melPDomiX enables systematic characterization of gain- and loss-of-function alterations in treatment-refractory melanoma. Using multi-omics integration and structural-context representation, we demonstrate how this resource distinguishes gain- from loss-of-function variants and uncovers new candidate melanoma drivers and therapeutic targets. Together, melPDomiX provides a comprehensive, deeply profiled set of tumor models that supports mechanistic discovery and facilitates the development of improved treatments for this devastating heterogeneous malignancy.

Uveal melanoma is a rare but aggressive intraocular malignancy with limited therapeutic options and a high risk of metastatic relapse. Although its mutational landscape is well defined, how genetic factors, uveal location, and disease progression shape cell states and the tumor ecosystem remains poorly understood. We integrate single-cell RNA sequencing and spatial transcriptomics across a cohort spanning healthy uveal tissue, primary tumors, and metastases. We uncover a continuous shift in melanoma cell states during disease progression, from differentiated, pigmentation-associated programs toward stress- and epithelial-to-mesenchymal transition-associated states. This trajectory is modulated by uveal tract location, with iris tumors exhibiting a more differentiated phenotype and choroidal tumors enriched for stemness-associated features that localize to melanoma-rich regions correlating with genomic instability. In parallel, genetic background defines distinct melanoma cell states and tumor–immune interactions. Together, these findings define clinically relevant cellular states and provide a framework for understanding metastatic progression and therapeutic vulnerability.

BACKGROUND:Randomized trials demonstrated superior outcomes with neoadjuvant immune checkpoint inhibitors (ICI) compared with adjuvant ICI alone for patients with resectable macroscopic melanoma. However, the optimal perioperative strategy remains undetermined, and biomarkers are lacking. Real-world data reporting the feasibility of limited resection and the utility of circulating tumor DNA (ctDNA) in predicting melanoma recurrence could help inform optimal practice. METHODS:Retrospective data regarding treatment, outcomes, and safety were collected for 76 patients who had melanoma treated with neoadjuvant ICI from 2020 to 2024. Signatera ctDNA results were available for 22 patients. RESULTS:Of the 76 patients, 42 (55%) received ipilimumab plus nivolumab (Ipi/Nivo), 31 (41%) received anti-PD1 monotherapy, and 3 (4%) received nivolumab plus relatlimab. The patients included 64 (84%) who underwent planned surgical excision, 25 who underwent upfront total lymph node dissection (TLND), and 32 who underwent index node excision (INE), with reflex TLND performed for 7 patients. The overall major pathologic response (MPR) rate was 55% (49% with Ipi/Nivo and 65% with monotherapy). Six (21%) of 28 non-MPR patients did not receive adjuvant therapy. After a median follow-up period of 15.8 months, recurrence rates were comparable between the INE (9%) and TLND (12%) cohorts. At the time of surgery, ctDNA was undetectable in the majority of the MPR patients (9/11) and the minority of the non-MPR patients (4/11). Postoperatively, ctDNA was detectable in 1 of 20 patients. CONCLUSIONS:A personalized surgical approach to neoadjuvant ICI was feasible, with comparable recurrence rates between the patients who underwent INE and those who underwent TLND. For the patients without MPR, subsequent adjuvant therapy improved recurrence-free survival (p = 0.046). The ctDNA results correlated with the clinical outcomes, suggesting that ctDNA may complement pathologic response in guiding management.

e21503 Background: Immune checkpoint inhibitor (ICI) therapy targeting PD-1 and CTLA-4 (e.g., ipilimumab-nivolumab) has revolutionized treatment of advanced melanoma. While there are few reliable predictors of response, ICI-associated vitiligo is a common immune-related adverse event in melanoma and is associated with favorable clinical outcomes. Understanding the immunological correlates of ICI-associated vitiligo may therefore provide crucial insights into the mechanisms underlying durable response to melanoma immunotherapy. Methods: We performed single-cell multiomic profiling (gene expression, CITE-seq, T cell receptor [TCR] sequencing) of matched primary tumor, lymph node metastasis, and vitiligo skin from a metastatic melanoma patient who both achieved durable response to ipilimumab-nivolumab and developed vitiligo. To contextualize these findings, we integrated ~57,000 T cells from publicly available single-cell datasets encompassing ICI-treated melanoma, autoimmune vitiligo, and healthy skin. Results: Across tissues and datasets, we identified six CD4⁺ and eight CD8⁺ T cell states conserved across conditions, though exhibiting tissue-specific abundance shifts. CD8⁺ T cell populations revealed the most pronounced differences, most notably demonstrating enrichment within vitiligo skin of a distinct MHC II-expressing CD8⁺ effector population with high expression of GZMA , GZMK , and GZMH , along with activation-associated programs suggesting recent TCR engagement and a non-terminal inflammatory effector state. Additional cytotoxic CD8⁺ subsets with checkpoint-associated and tissue-resident memory-like features were seen across tumor and skin, whereas CD4⁺ T cell states demonstrated more modest tissue-specific variation. In contrast, transcriptionally related CD8⁺ T cells within tumor exhibited characteristics suggesting functional limitation within the suppressive tumor microenvironment. Through leveraging paired single-cell TCR sequencing from the same patient, ongoing analyses are evaluating the clonal relationships between vitiligo- and tumor-infiltrating T cells to determine whether shared TCRs adopt divergent transcriptional programs across tissues. Conclusions: Our data support a framework whereby ICI-associated vitiligo reflects tissue-specific effector-state programming of melanoma-reactive CD8⁺ T cells, providing mechanistic insight into the link between treatment-induced autoimmunity and effective anti-tumor immunity.

9574 Background: Neoadjuvant ICI renders MPR in 50-60% of resectable Stage III/IV melanoma. Early data illustrated the ability to de-escalate surgery in cases of MPR in the index lymph node (ILN). However complete lymph node dissection (CLND), a procedure with considerably morbidity, remains the standard of care. Real-world data demonstrating the ability to de-escalate surgery while maintaining disease control, can inform optimal practice. Methods: In our retrospective study performed at two academic institutions (MGB/DFCI), patients with resectable, macroscopic Stage III/IV cutaneous melanoma treated with neoadjuvant ICI were identified. Pre-treatment fiducial placement was routinely performed in cases of macroscopic nodal disease to facilitate index node excision (INE). Data pertaining to treatment course, anti-tumor outcomes, and safety were collected. Signatera ctDNA was evaluated as a non-invasive biomarker of early response. Results: From 2020-2025, 126 patients received neoadjuvant ICI: 60 anti-PD-1 monotherapy, 62 ipilimumab/nivolumab (I/N) and 4 nivolumab/relatlimab (N/R). The majority had macroscopic stage III disease (95%) and completed the full neoadjuvant course (86%). Ten patients (8%) did not undergo surgery (clinical CR, n = 3; toxicity, n = 1; progression, n = 6). Of the 116 (92%) who underwent surgery; 43 (37%) had an upfront TLND; 52 (45%) had an INE and 21 (18%) excision of N1c or M1a disease. After INE (n = 52), reflex CLND was performed in 7 (13%) all due to non-MPR. The MPR of the total cohort was 54% (n = 63) with a pCR rate of 51% (n = 59). The MPR/pCR rate was similar between ICI regimens. In the MPR cohort, 35 underwent an INE only whereas 17 underwent CLND; the remainder had excision of N1c/M1a disease. Post-operatively, 15 required drain placement, of whom 2 had an INE. Adjuvant ICI was utilized in 23 patients, the majority receiving peri-operative pembrolizumab (n = 20). After a median follow-up of 15.5mths (3.2-68.7mths), in those who underwent surgery, 20% (23/116) have recurred; 3 with a MPR (INE, n = 1; excision, n = 2). Of these, 2 were distant metastases with no nodal recurrence. Notably, 33 patients had pre-operative ctDNA results (Signatera). At the time of surgery, ctDNA was undetectable in the majority of MPR patients (19/20), while the majority of non-MPR patients remained detectable (8/13). Pre-surgery ctDNA assessment prognosticated RFS for the 20 patients who were detectable at baseline (p = 0.035) with ctDNA clearance associated with both MPR and improved RFS. Updated clinical and ctDNA data will be presented at the meeting. Conclusions: Similar to published data, patients with a MPR had low rates of recurrence, with no difference between those who underwent INE compared to upfront CLND. Pre-operative ctDNA levels, particularly ctDNA clearance, correlate with clinical outcomes and may inform de-escalation strategies.

Melanoma patients treated with anti-PD1 or anti-PD1/anti-CTLA4 present intrinsic or acquired resistance to these immune checkpoint blockade (ICB) treatments. However, the interactions between tumor-stromal-immune cells in the microenvironment (TME) underlying these resistances remain unclear. In this study, we used a cohort of melanoma patients (n = 61) treated with ICB, pre-treatment (n = 33) and post-progression (n = 28), for whole-exome sequencing, single-nucleus RNA sequencing and high-resolution spatial imaging. We developed a framework to analyze response and resistance (intrinsic and acquired) via tumor-intrinsic programs, immune features, and tumor-stromal-immune interactions. Patients with low immune infiltration exhibited enrichment of macrophages associated with hypoxia and angiogenesis phenotypes. Samples with high immune infiltrates displayed an enrichment in lymphocytes, particularly TCF7+ CD8+ T cells. Pre-treatment samples from patients experiencing durable clinical benefit are encriched in memory B cells and plasmablast-like B cells.The presence of Bcells and of follicular dendritic cells in non-lymph node biopsies support the presence of tertiary lymphoid structures within the TME. Lastly, the presence of a TIMD4-expressing macrophage subset correlates with lymphocyte infiltration, indicating its potential role in promoting anti-tumor immunity. Our work defines TME immune profile interaction with tumor-stromal cells associated with clinical outcomes in melanoma. NCI HTAN Tumor Immunology: Checkpoints, Prevention, and Treatment (TIPT)

ABSTRACT Purpose A significant proportion of patients with locoregional (stage III) cutaneous melanoma recur despite adjuvant systemic therapy. Staging criteria and surgical nodal management have changed since the trials were completed. Data assessing the effect of systemic therapy compared to surveillance are limited, and factors associated with recurrence are unclear. We assessed the efficacy of adjuvant systemic therapy in real‐world patients and assessed whether baseline genomic characteristics could prognosticate or predict benefit from therapy. Methods We collected demographic, histopathologic, clinical, and genomic data for patients diagnosed with stage III cutaneous melanoma. Outcomes of interest were recurrence‐free survival (RFS) and distant‐metastasis‐free survival (DMFS). Survival analysis was performed using the Kaplan–Meier method with log‐rank analysis. Univariate and multivariate analyses were performed using a Cox regression analysis. Results Two hundred and fifteen patients were included, of which 65 and 76 were treated with BRAF/MEK inhibitors (BRAFi/MEKi) and anti‐PD1 adjuvant systemic therapy respectively. Seventy four underwent active surveillance. Both adjuvant therapies reduced the hazard of recurrence when compared to patients undergoing active surveillance: anti‐PD1 HR: 0.32 (p < 0.01) and BRAFi/MEKi HR: 0.39 (p = 0.03). Anti‐PD1‐treated patients with a BRAF V600 mutation had a shorter RFS than patients with BRAF WT melanoma (p < 0.01); this was validated in external data where the presence of a BRAF V600 mutation was associated with an increased hazard recurrence (HR: 2.1, p = 0.025). Conclusion Adjuvant systemic therapy improved RFS in our cohort. We found that BRAF V600 mutation was associated with a worse RFS for adjuvant anti‐PD1 monotherapy. The effect of BRAF mutation on the response to anti‐PD1 therefore may be considered when choosing between adjuvant anti‐PD1 and BRAFi/MEKi for patients with BRAF V600 mutant melanoma.

BACKGROUND:Tebentafusp has significantly improved overall survival in HLA-A*02:01+ metastatic uveal melanoma (mUM) patients even in those with a best objective response of progressive disease. Thus, strategies to maintain tebentafusp therapy are critical. Here, we examine the efficacy and safety of adding concurrent local therapy (CLT) to tebentafusp upon radiological progression with tebentafusp alone. PATIENTS AND METHODS:This multicenter retrospective study included mUM patients treated with tebentafusp and CLT, consisting of extrahepatic soft tissue irradiation and liver-directed therapies (LDTs). Efficacy of target and nontarget sites were assessed per RECIST version 1.1. PFS with tebentafusp alone (PFS1) was compared to that after adding CLTs to tebentafusp upon progression (PFS1+PFS2). ctDNA responses were explored. RESULTS:Of the 30 eligible patients, 21 (70%) received concurrent LDT, 7 (23%) had extrahepatic irradiation, and 2 (7%) had both. The objective response rate (ORR) was 12% (95% CI, 3-32) for tebentafusp alone and 28% (95% CI, 14-47) after adding CLTs. The disease-control rate with tebentafusp alone was 44% (95% CI, 25-65) vs 63% (95% CI, 44-78) after CLT. Median PFS1 was 5.8 months (95% CI, 2.8-13.4), while median PFS1+PFS2 was 14.8 months (95% CI, 9.2-NA). CLT thereby allowed treatment beyond progression with tebentafusp for approximately 9 months. Two patients (66%) had decreased ctDNA with tebentafusp alone, while 4 (100%) had decreased ctDNA after CLT. There were no treatment discontinuations due to toxicities from tebentafusp with CLT. CONCLUSIONS:CLT with tebentafusp was well-tolerated, extending the duration of tebentafusp benefit in a highly selected mUM population. This merits further studies to assess clinical utility.

9528 Background: Adjuvant PD1 treatment improves clinical outcomes in high-risk resected melanoma. We have shown that adjuvant PD1 can lead to irAEs that become chronic in up to 46% of treated patients (pts). We performed longer follow-up (f/u) to further characterize chronic irAEs from adjuvant PD1 treatment and assessed risk factors to determine predictors for their development. Methods: We retrospectively analyzed pts treated with adjuvant PD1 for resected stage III-IV melanoma from 2015-2024 from 6 institutions. All pts had at least 12 months of f/u after PD1 initiation. We collected demographics, treatment details, and outcomes. We characterized type, grade, management, duration, and resolution of acute (onset during PD1) and chronic (persisting at least 3 months after PD1 cessation) irAEs. We performed Olink 96-protein inflammation assay in plasma from pts with and without chronic non-endocrine irAEs at 12 months after PD1 initiation. Results: We included 304 pts; 184 (61%) were male, and median age at PD1 initiation was 64 years. Among all pts, 221 (73%) developed acute irAEs, and 147 (48%) developed chronic irAEs; 59 pts had chronic endocrine irAEs, 99 had chronic non-endocrine irAEs, and 11 had both. At last f/u (median 61.4 months), 104 (34%) pts had ongoing irAEs. The most common chronic irAEs were hypothyroidism/thyroiditis (n=45, 15%), arthritis (n=25, 8%), dermatitis (n=17, 6%), hypophysitis/adrenal insufficiency (n=16, 5%), and xerostomia (n=10, 3%). Twenty (7%) pts experienced chronic toxicities outside of classical irAEs, most often fatigue (n=14, 5%), orthostasis (n=2, 1%), and headache (n=2, 1%). We then assessed risk factors for chronic irAEs compared with acute, resolving irAEs (excluding endocrine irAEs since nearly all become chronic). We found that peak steroid dose was similar in patients with and without chronic irAEs (median 50 mg for both groups, p=0.33). Time to irAE onset was similar in patients with and without chronic irAEs (median 91 vs. 114 days, p=0.78). Time to steroids from symptom onset trended longer for those with chronic irAEs (median 7 vs. 4 days, p=0.18) but was not statistically significant. In proteomic analysis, 24/96 cytokines had higher expression (0 with lower expression) in pts with chronic irAE (n=17) compared with controls (n=10), including IL-8 (p=0.02), IL-17 (p=0.049), TNF (p=0.02), VEGFA (p=0.005), and soluble PD-L1 (p=0.03). Conclusions: Among this large cohort of pts with melanoma treated with adjuvant PD1, chronic irAEs were common, persistent, and associated with elevated circulating cytokines, which could suggest possible therapeutics. No obvious predictors of chronic irAEs were identified outside of organ affected; analyses are ongoing. Given the long-term survival of pts treated with adjuvant PD1, monitoring and managing chronic irAEs is crucial.

Melanoma is a common and aggressive cancer, with a rising incidence in most developed countries. Major discoveries in melanoma biology have been rapidly translated, allowing cures for some late-stage patients. Despite these advances, incomplete knowledge of genes and pathways that are gained or lost during melanogenesis prohibits cures for many patients. To identify gain-of-function and loss-of-function drivers of melanoma, we established a multi-omics cohort of melanoma patient-derived xenografts. By linking mutations with gene and protein expression, we characterize gain or loss-of-function of specific melanoma drivers in treatment refractory tumor models. We use multi-omics integration and structural-context representation to distinguish gain from loss of function variants, revealing new candidate melanoma genes and targets. This study provides a comprehensive resource of tumor models with the genetic, molecular and structural features of gain and loss of function melanoma drivers, allowing future development of better therapeutics for this devastating and heterogenous malignancy.

Immune checkpoint inhibitors (ICI) can achieve durable responses in patients with advanced melanoma, and results from clinical trials suggest cure may be possible for a subset of patients. Despite clinical trial data, little is known about the risk, character, and clinical outcome of late recurrences after ICI. This study aimed to explore the disease outcomes and survival in a cohort of patients with long-term responses to ICI.We retrospectively identified patients treated with ICI for advanced melanoma with long-term disease control, defined as not requiring a subsequent line of systemic therapy within 3 years of ICI commencement. We analysed disease characteristics, treatment, toxicity, recurrence patterns, management, and outcomes.A total of 567 patients were identified with a median follow-up of 7.1 years: 504 (89%) without disease progression within 3 years (cohort 1) and 63 (11.1%) with disease progression within 3 years managed without a change in systemic therapy (cohort 2). Subsequent progression after 3 years occurred for 39 (7.7%) patients in cohort 1, compared to 14 (22%) in cohort 2. Predictors for late progression after 3 years were a non-complete radiological response (CR) best response and prior progression within 3 years. Thirty-two patients (5.6%) died during follow-up, 8 (1.4%) from melanoma, 6 (1.2%) from cohort 1 and 2 (3.2%) from cohort 2.In this population of patients with advanced melanoma with long-term disease control from ICI, the risk of subsequent disease progression and death was low. This suggests that a significant proportion of long-term ICI responders are likely cured and may inform the frequency and duration of follow-up.

T cell exhaustion is a major driver of immune checkpoint blockade (ICB) resistance and clinically effective strategies to prevent or reverse T cell exhaustion to restore ICB sensitivity are lacking. CD38, an ecto-enzyme involved in NAD+ catabolism, is highly expressed in exhausted CD8+ T cells in human melanoma, yet its role in T cell exhaustion remains to be elucidated. Here we show that CD38+CD8+ T cells are enriched during tumor progression and following unsuccessful ICB treatment and are strongly associated with ICB resistance in melanoma. Chronic TCR activation and type I interferon stimulation upregulate CD38 in human T cells, leading to mitochondrial dysfunction and reduced anti-tumor activity. Disrupting CD38 restored cellular NAD+ pools and T cell bioenergetics, leading to improved T cell proliferation, enhanced effector function and elevation in T cells “stemness” markers that are associated with increased response to ICB in human melanoma. Importantly, targeting CD38 in a cohort of patient-derived organotypic tumor spheroid (PDOTS), human living tumor explants, with intact immune compartment, demonstrates that CD38-directed therapies can overcome ICB resistance in clinically resistant human melanoma, an effect that is furthered enhanced by supplementation with NAD+. These results emphasize the need for further preclinical and clinical evaluation of CD38 directed therapies in melanoma and underscore the importance of NAD+ as a vital metabolite to enhance those therapies. Citation Format: Or-Yam Revach, Angelina Cicerchia, Ofir Shorer, Boryana Petrova, Seth Anderson, Joshua Park, Chen Lee, Arnav Mehta, Samuel Wright, Niamh McNamee, Aya Tal-Mason, Giulia Cattaneo, Payal Tiwari, Hongyan Xie, Johanna Sweere, Li-Chun Cheng, Natalia Sigal, Elizabeth Enrico, Marisa Miljkovic, Shane Evans, Ngan Nguyen, Mark Whidden, Ramji Srinivasan, Matthew Spitzer, Yi Sun, Tatyana Sharova, Aleigha Lawless, William Michaud, Martin Rasmussen, Jacy Fang, Claire Palin, Feng Chen, Xinhui Wang, Cristina Ferrone, Donald Lawrence, Ryan Sullivan, David Liu, Uma Sachdeva, Debattama Sen, Keith Flaherty, Robert Manguso, Lloyd Bod, Manolis Kellis, Genevieve Boland, Keren Yizhak, Jiekun Yang, Naama Kanarek, Moshe Sade-Feldman, Nir Hacohen, Russell Jenkins. Overcoming resistance to immunotherapy by targeting CD38 in human tumor explants [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Functional and Genomic Precision Medicine in Cancer: Different Perspectives, Common Goals; 2025 Mar 11-13; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85(5 Suppl):Abstract nr B034.

BACKGROUND:Uveal melanoma (UM) is the most common primary ocular malignancy with a high rate of metastases. While immune checkpoint inhibitors (ICIs), including ipilimumab and nivolumab (ipi + nivo), have shown efficacy in metastatic cutaneous melanoma, their success in metastatic UM (MUM) remains limited. This study evaluates toxicity and outcomes of ipi + nivo in the largest, multicenter MUM cohort. METHODS:We analyzed 131 MUM patients treated with ipi + nivo from 2016 to 2024 across 5 international centers. Rates of toxicity, response, and survival outcomes were assessed. RESULTS:Among 131 patients, 37.4% of patients received 4 cycles of ipi + nivo. The most common reason for ipi + nivo discontinuation (31.3%) was toxicity. Of all treated patients, 80.2% experienced immune-related adverse events (irAEs). The overall response rate (ORR) was 16.4%, and the disease control rate (DCR) was 43.4%. Progression-free survival (PFS) was three months, and the median overall survival (OS) was 18 months. Patients receiving ipi + nivo as second-line therapy had lower ORR compared to patients who received ipi + nivo as first-line therapy (P = .04). Patients with exclusively extrahepatic metastases had a better ORR and OS compared to those with hepatic or mixed metastases (P = .02, P = .02, respectively). 20.6% of patients developed eosinophilia during treatment, which was associated with improved median OS (24 months vs 15 months, P = .02). CONCLUSIONS:Ipi + nivo shows moderate efficacy and clinically relevant toxicities in patients with MUM. Eosinophilia is a potential prognostic biomarker, that merits further investigation.

Patients with melanoma brain metastases (MBM) have a poor prognosis despite improvements in treatments. The combination of ipilimumab and nivolumab remains the standard of care for this patient population and there is limited data on the combination of nivolumab and relatlimab (nivo-rela) in MBM. We performed a retrospective cohort study of patients with MBM treated with nivo-rela from May 2022 to December 2023 at two academic institutions in Boston, MA. Information pertaining to patient demographics, survival, timing of therapy (radiation, nivo-rela), response rates and adverse events were collected. Progression-free survival and overall survival were calculated. A total of 39 patients were identified with a median follow-up of 26 months. Most patients were male (25 of 39 [64.1%]), with a median age of 69 years (range 31-87 years). Estimated survival rates were 79% (95% CI: 63% to 89%) at one year and 69% (95% CI: 50% to 82%) at two years following MBM diagnosis. We saw intracranial clinical benefit (response or stable disease) in 19 of 39 (48.7%) patients. Extracranial response was largely concordant with intracranial benefit with durable clinical benefit, defined as CR, PR or SD for 6 months, in 14 of 39 (35.9%) patients. A total of 30 patients (77%) received radiation therapy (SRT or both WBRT and SRT). Among them, 46.7% (14 of 30) showed either a decrease or stable intracranial disease after starting nivo-rela, compared to 55.6% (5 of 9) in the non-radiation therapy group. The combination of nivo-rela showed intracranial activity in patients with MBM, particularly in the front-line setting. Prospective trials are needed to validate these findings and provide data on the optimal use of systemic and local therapies in this challenging patient population.

Long-standing goals of cancer immunotherapy are to activate cytotoxic antitumor T cells across a range of affinities for tumor antigens while suppressing regulatory T cells. Computational protein design has enabled the precise tailoring of proteins to meet specific needs. Here, we report a de novo designed IL-21 mimic, 21h10, with high stability and signaling potency in humans and mice. In murine and ex vivo human organotypic tumor models, 21h10 showed robust antitumor activity, with more prolonged signaling and stronger antitumor activity than native IL-21. 21h10 induced pancreatitis that could be mitigated by TNF blockade without compromising antitumor efficacy. Although antidrug antibodies to 21h10 formed, they were not neutralizing. 21h10 induced highly cytotoxic T cells with a range of affinities, robustly expanding intratumoral low-affinity cytotoxic T cells and driving high expression of IFN-γ and granzyme B compared with native IL-21, while increasing the frequency of IFN-γ + T helper 1 cells and reducing regulatory T cells. The full human-mouse cross-reactivity, high stability and potency, and low-affinity antitumor responses support the translational potential of 21h10.

Checkpoint inhibitors have revolutionized cancer treatment, but a significant proportion of patients do not respond to these therapies, underscoring the need for alternative strategies. Although gene therapy has made substantial strides, its application in solid tumors remains underexplored, with limited treatments approved. Here, we further investigated a gene therapy approach with non-replicating adenoviral vectors encoding the alternate reading frame (ARF) and interferon beta (IFNb) and tested it in a clinically relevant setting. We previously showed that this combined gene therapy induces immunogenic cell death in melanoma models, and now, we utilize patient-derived organotypic tumor spheroids (PDOTS), a model that closely recapitulates the immune environment of tumors, to test its effects using patient tumors. Our results demonstrate, for the first time, the effectiveness of using PDOTS to evaluate viral-vector-based gene therapies. While the addition of anti-PD-1 did not enhance therapeutic outcomes, the gene therapy alone suppressed tumor growth and triggered antitumor immune responses across different cancer models, notably those with low immunogenicity and specific genome profiles. These findings suggest that this gene therapy could serve as a valuable alternative for patients not responsive to checkpoint inhibitors and who have solid tumors with limited treatment and impaired p53-ARF-MDM2 pathways, such as liposarcomas.

Clinical decisions for immune checkpoint blockade (ICB) treatment currently rely on tumor-based biomarkers such as mutational burden, microsatellite instability, and PD-L1 expression. While these markers serve as valuable proxies for tumor antigenicity and the immune status of the tumor microenvironment, they fail to account for the fitness of a patient’s T cells, a critical factor in anti-tumor immunity. Direct assessment of T cell functional capacity could complement existing biomarkers to improve patient stratification and efficacy prediction for ICB therapies. Our previous work showed that single-cell mass measurements serve as a label-free means of assessing immune cell activation and ICB response. T cells from patients with various malignancies displayed significant variability in both their activation capacity and ICB response ex vivo, indicating that mass measurements capture key functional differences in T cells across patients.Here we present a next-generation platform that measures cell mass alongside linked measurements of volume, density, and morphological features. This platform incorporates fluorescence exclusion measurements to determine single-cell volume which, when combined with cell mass, enables precise determination of single-cell density. Additionally, we utilize inline brightfield imaging to capture images from individual cells and extract morphological features using an autoencoder-based reconstruction model.Our results show that multiparametric measurements significantly enhance the ability to characterize T cell functional states. In activation experiments using healthy donor T cells, these combined signatures reveal notable phenotypic changes as early as six hours after activation—changes that remain undetectable with mass measurements alone.In a cohort of peripheral blood samples from patients with advanced melanoma, we observed substantial heterogeneity in T cell biophysical signatures at baseline and following stimulation, with and without immune checkpoint blockade. These signatures were distinct from those observed in T cells derived from healthy donor PBMCs.These findings underscore the potential of a blood-based functional precision medicine assay to assess patients' baseline immune fitness. When integrated with existing clinical biomarkers, these unique functional signatures offer orthogonal insights that could improve patient stratification for immune checkpoint inhibitor therapies. Moreover, the multiparametric nature of these measurements aligns well with emerging multimodal AI approaches, paving the way for more effective identification of patients most likely to benefit from these treatments. Robert Kimmerling, Selim Olcum, Mark Stevens, Rachel LaBella, Madeleine Vacha, Katelin Katsis, Reginald Aikins, Steven Wasserman, Tatyana Sharova, Aleigha Lawless, Sonia Cohen, Genevieve Boland. Beyond tumor biomarkers: Multiparametric biophysical assessment of T cells from peripheral blood to inform checkpoint blockade therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4621.

Tumor-infiltrating lymphocyte (TIL) therapy is FDA-approved for patients with treatment-resistant advanced melanoma, but the TIL subpopulations critical for tumor eradication remains incompletely understood. Using patient-derived TIL-melanoma co-cultures, we identified and characterized a novel subset of CD8+ TIL, capable of class I HLA-independent cancer cell lysis. The lymphotoxin β receptor (LTβR) and interferon (IFN) sensing pathways were nominated as key determinants of TIL-mediated cancer cell killing from a whole-genome, loss-of-function CRISPR screen. Validation studies confirmed that dual LTβR and IFN sensing is necessary and sufficient for cancer cell lysis, and that expanded CD8+ TIL express high lymphotoxin β (LTB) and upregulate lymphotoxin α (LTA) upon coculture with cancer cells. Leveraging paired scRNA-seq and scTCR-seq data, we confirmed that enrichment of LTB + CD8 + T cells is associated with clinical response to TIL, and that LTB + CD8 + TIL are expanded from putative neoantigen-reactive, LTB lo CD8+ T cells in resected tumors.

e14001 Background: Patients with melanoma brain metastases (MBM) have a poor prognosis. Immune checkpoint inhibitors (ICI) have improved the prognosis of patients with MBM. The combination of anti-PD1 (nivolumab, nivo) and anti-cytotoxic T-lymphocyte antigen-4 (ipilimumab, ipi) has demonstrated the highest intracranial and extracranial response rates and durability. Combination ipilimumab / nivolumab (ipi/nivo) is the standard first line (1L) therapy for patients with MBM. Data pertaining to the efficacy of the alternative, more tolerable, front-line regimen of nivolumab/relatlimab (nivo/rela) is lacking. Methods: We performed a retrospective cohort study of patients with MBM treated with nivo-rela from May 2022 to December 2023. Information pertaining to patient demographics, timing of therapy (radiation, nivo-rela), response rates and adverse effects were collected. The intracranial response was categorized as an increase, stable, or decrease in brain lesion size. Extracranial response was defined as durable clinical benefit (stable disease (SD), partial response (PR), or complete response (CR)) sustained for 6 months. Progression free survival and overall survival was calculated. Results: A total of 44 patients were identified with a median follow-up of 25 months. The majority of patients were male (28 of 44 [63.6%]), with a median age of 69 years (range 31-87 years). Most patients were BRAFV600 mutated (29 of 44 [65.9%]). The median time from primary melanoma diagnosis to the development of MBM was 34 months (range 0 - 644 months). The median interval between MBM diagnosis and the initiation of nivo-rela was 8.6 months (range -21-94 months) with 6 (13.6%) treated in the 1L setting, 16 (36.4%) treated in 2L, and 22 (50%) in the ≥3L and beyond. Thirty-nine patients had brain metastasis at the time of nivo-rela initiation. Most patients (30 of 44 [68.2%]) had 1-3 MBM, while (14 of 44 [31.8%]) had more than 3 MBM. Two patients had leptomeningeal disease. (33 of 44 [75.0%]) received local intracranial stereotactic radiotherapy (SRT). The estimated survival rates were 77% (95% CI: 62% to 87%) at one year and 67% (95% CI: 50% to 80%) at two years following MBM diagnosis. We saw an intracranial clinical benefit (decreasing and stable) in (20 of 44 [45.5%]) patients. For extracranial response, we saw durable clinical benefit in (17 of 44 [38.6%]) patients. Conclusions: The combination of nivo-rela showed intracranial activity in patients with MBM, particularly in the 1L. Prospective trials are needed to confirm these findings and provide data on the optimal use of systemic and local therapies in this challenging patient population. Patient demographics. N Mean S.D. Min Median Max % Months from Prim. Dx to BM Dx 44 62.2 101.6 0.0 34.1 643.7 Months after BM to start Nivo-Rela 44 15.0 20.9 -20.7 8.6 94.3 Age at BM diagnosis 44 66.9 11.2 31.0 69.0 87.0 Female 16 36.4 Male 28 63.6 1-3 MBM 30 68.2 3+ MBM 14 31.8

Immune checkpoint inhibitor (ICI) therapies have markedly improved the prognosis for patients with stage III & IV metastatic melanoma by prolonging progression-free and overall survival rates. However, the variability in immune evasion and resistance mechanisms presents significant challenges to the clinical efficacy of ICIs. This project aims to define drivers of immunotherapy response and resistance by employing advanced genomic, single-cell mRNA analyses, and spatial profiling techniques on tissue biopsies from metastatic melanoma patients. In this study, we developed a framework to analyze response and resistance, both intrinsic and acquired, via immune features in the tumor microenvironment in a standardized, uniformly processed, and deeply clinically annotated cohort of metastatic melanoma patients (n=61) treated with ICB as part of the human tumor atlas network (HTAN) initiative. From the tumor samples, we conducted single-nucleus RNA sequencing, and for a subset of the samples high-resolution spatial imaging (including protein mIHC, CODEX, and transcriptomics MERFISH). Standardized processing and data pipelines allowed for integrating genomic, transcriptomic, and spatial features to elucidate characteristics and mechanisms in tumor microenvironment and their relationships with resistance. Studies of the pretreatment samples demonstrated that CD4 and CD8 T cells, particularly TCF7+ CD8 T cells, are significantly more prevalent in responders to immunotherapy. Conversely, macrophages, especially Angio-TAMs, show higher levels of enrichment in non-responders. Moreover, the presence of B cells and follicular dendritic cells in non-lymph node biopsies supports the presence of tertiary lymphoid structures within the TME. Three levels of immune enrichment were identified through spatial analysis, and samples with more immune enrichment tend to have better responses. We also identified 10 recurrent cellular neighborhoods (RCNs) and found that RCN2,4, and 7 with high lymphocyte infiltration are significantly more enriched in responders than non-responders. In addition, RCN4 is associated with immune infiltration while RCN7 is TLS-like. Our findings indicate that patients with low immune infiltration exhibited enrichment of macrophages associated with the hypoxia and angiogenesis phenotypes, while patients with high immune infiltrates displayed enrichment in lymphocytes, particularly TCF7+ CD8+ T cells, confirming previous findings and indicating a robust T cell-mediated immune response. This project integrates genomic, transcriptomic, and spatial features to elucidate shared tumor and microenvironmental states and their relationships with resistance, and guide more personalized and effective treatment strategies for metastatic melanoma. Xinyu Cui, Giuseppe Tarantino, Yiwen He, Priyanka Solanky, Kathleen Pfaff, Aaron R. Thorner, Tyler J. Aprati, Boyang Zhang, Timothy Blosser, Emily Robitschek, Jiajia Chen, Junko Tsuji, Elliot Boblitt, Allison Frangieh, Hannah M. Faulkner, Marta Holovatska, Aleigha Lawless, Michael Manos, Karla Helvie, Tatyana Sharova, Dennie Frederick, James L. Fahey, Diego Villamarin, Sachi Krishna, Chanell Mangum, Ajit J. Nirmal, Domenic Abbondanza, Cai McCann, Bruce Johnson, Alex K. Shalek, Eliezer Van Allen, Xiaowei Zhuang, Ryan Sullivan, Barbara E. Engelhardt, Samouil L. Farhi, Scott Rodig, F. Stephen Hodi, Genevieve M. Boland, David Liu. Dissecting tumor-immune microenvironment in response and resistance to immune checkpoint blockade in metastatic melanoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4536.