Endocrine therapies initially enforce lineage identity in hormone-driven cancers, yet sustained hormonal suppression often triggers a coordinated systems-level rewiring through lineage plasticity. Previously, lineage plasticity has been characterized as a discrete bypass mechanism, which lacks fluidic identity changes. Here, we describe the process as a progressive, adaptive trajectory of endocrine escape. We describe how the disruption of hormone receptor-anchored identity programs creates permissive conditions for transitions into HR-indifferent or neuroendocrine- or basal-like states. Next, we summarize historical methodologies for investigating lineage plasticity, such as patient-derived organoids, genetically engineered mouse models (GEMMS), lineage tracing, and single-cell multi-omics, and discuss novel systems biology approaches to enable the early detection and modeling of these unstable intermediate states. Finally, we present promising and potential uses for artificial intelligence and machine learning for dissecting therapy-induced identity shifts and resistance mechanisms. Intercepting these trajectories before lineage commitment offers a critical window for precision oncology interventions.

e17081 Background: Bipolar androgen therapy (BAT) is an investigational therapy comprising monthly supraphysiologic testosterone injections and continuous androgen deprivation therapy (ADT). BAT demonstrated clinical activity in patients with metastatic castration resistant prostate cancer (mCPRC) progressing on abiraterone acetate (Denmeade S et al JCO 2021). Preclinical studies showed that dihydrotestosterone increases the cytotoxicity of macrophages against prostate cancer cell lines in a concentration-dependent manner (Lee GT et al Endocrinology 2019). Sipuleucel-T, an autologous cellular immunotherapy manufactured from peripheral blood mononuclear cells cultured with PA2024, Prostatic Acid Phosphatase linked to GM-CSF, improved overall survival (OS) in patients with mCRPC. Interferon-gamma (IFN-g) enzyme-linked immunosorbent spot (ELISPOT) response to PA2024 was an immune response parameter that had the strongest correlation with OS benefit (Sheikh N et al. CII. 2012). The ELISPOT response was, however, detected only in 50% of patients treated with sipuleucel-T. We hypothesized that treatment with BAT prior to sipuleucel-T could enhance ELISPOT response rate. Methods: Patients with progressive mCRPC and with clinical indication for sipuleucel-T were enrolled. The key eligibilities included PSA < 20ng/mL, no opioid use for cancer-related pain, no hepatic metastasis, no prior chemotherapy for mCRPC, and no active autoimmune disease. Patients continued with ongoing ADT and received testosterone cypionate 400mg intramuscularly every 4 weeks until they are no longer deriving clinical benefit. Sipuleucel-T started after two 4-weekly testosterone injections and administered every 2 weeks for 3 cycles. Research blood was collected before each testosterone injection and each sipuleucel-T infusion. The primary endpoint was ELISPOT response, defined as > 10 IFN-g response spots to the PA2024 antigen. Results: Compared to the historical control of 50%, PA2024 ELISPOT response rate, eleven of 11 (100%) evaluable patients achieved positive ELISPOT responses ( > 10 spots) to PA2024 antigen. The mean number of spots was 232 (95% Confidence Interval (CI) 147, 317). 100% of patients achieved T cell proliferation response to PA2024 antigen ( > 12 Stimulation Index (SI)). Mean SI (95% CI) was 23 (17,30). One of two RECIST-evaluable patients achieved confirmed PR. Two of 11 evaluable patients had PSA50 responses. No grade 3 or higher adverse events or serious adverse events were observed. Conclusions: BAT followed by sipuleucel-T is safe and effective in potentiating immune response to Sipuleucel-T. Encouraging clinical activity has been observed. The study is currently enrolling to the second stage of the Simon’s design. Clinical trial information: NCT06100705 .

Higher exercise capacity and regular exercise training improve cancer prognosis at all stages of disease. However, the metabolic adaptations to aerobic exercise training that mediate tumor-host interactions are poorly understood. Here, we demonstrate that voluntary wheel running slows tumor growth and repartitions glucose uptake and oxidation to skeletal and cardiac muscle and away from breast and melanoma tumors in mice. Further, prehabilitation induces repartitioning of glucose metabolism in obese mice: Uptake and oxidation of glucose are enhanced in skeletal and cardiac muscle, and reduced in tumors. These increases in muscle glucose metabolism and reductions in tumor glucose metabolism, correlated with slower tumor progression. Using [U-13C6] glucose infusion, we show that exercise increases the fractional contribution of glucose to oxidative metabolism in muscle while reducing it in tumors, suggesting that aerobic exercise shifts systemic glucose metabolism away from the tumor microenvironment and toward metabolically active tissues. Transcriptional analysis revealed downregulation of mTOR signaling in tumors from exercised mice. Collectively, our findings suggest that voluntary exercise may suppress tumor progression by enhancing host tissue glucose oxidation and limiting tumor glucose availability, supporting a model in which exercise-induced metabolic competition constrains tumor energetics.

Background/Objectives: Immune checkpoint inhibitors (ICIs) are frontline treatment for advanced Merkel Cell Carcinoma (MCC), regardless of viral status. Frontline ICIs provide durable benefit to only half of patients, highlighting a need for alternative therapies. In this study, the objective is to leverage whole exome sequencing (WES) and transcriptome sequencing (WTS) to distinguish genomic alterations associated with ICI response. Investigate differential genomic alterations between virus-positive (VP) and virus-negative (VN)-MCC to identify novel therapeutic targets. Methods: A total of 95 MCC cases underwent WES and WTS. Utilizing computational pipelines applied to WES, we identified viral status and tumor mutational burden (TMB). RNA-seq data was used to characterize the immune microenvironment. Results: Of 95 MCC cases, 57 (60%) were VP-MCC and 38 (40%) were VN-MCC. Median TMB was higher in VN-MCC (27.5 vs. 1 Muts/Mb). Mutations in TP53, RB1, NOTCH1, KMTD2, KMT2C, and PIK3CA were primarily found in VN-MCC. MAPK Pathway Activity Score, NK cell infiltration, and the immune checkpoint gene CD276 in VN-MCC tumors were upregulated. No overall survival (OS) difference was identified between VP and VN-MCC, even after ICIs. Conclusions: MCC oncogenesis and treatment response transcend viral status. While mutational analysis confirms previous findings, assessment of the transcriptome and tumor microenvironment suggests alternate therapeutic targets.

2520 Background: Despite the transformational impact of immune checkpoint blockade, many cancer patients do not experience long-term survival. T cells with innate immune signatures can secrete inflammatory cytokines/chemokines and deliver potent cytotoxic signals potentially ideal for tumor immunity. The novel double-stranded RNA sensor RIG-I agonist SLR14 improved the control of murine melanoma. We tested the hypothesis that SLR14 transforms T cells to a cytotoxic state in immunologically “cold” human tumor specimens. Methods: We developed an approach, called PERCEPT, to directly test the response of patient tumor and immune samples to novel and established therapies ex vivo using perturbational single-cell RNA sequencing (Table). We obtained 9 surgical resections from primary or metastatic melanoma and Merkel Cell Carcinoma (MCC) tumors and lymph node metastases and made suspension replicates of tumor and infiltrating immune cell co-cultures. We stimulated for 42-48 hours (Table). We Flourescently Activated Cell Sorted live cells and then barcoded for multiplexed single-cell sequencing using 10x scRNAseq. We used CINEMA-OT to identify factors associated with response and resistance to the perturbations tested. We developed and validated CRISPR-KO MCC and small cell lung cancer (SCLC) cell lines, and co-cultured with CD14+ monocytes or monocyte-derived DCs. Results: Stimulation with RIG-I agonist SLR14 induced expression beyond canonical IFN-stimulated genes in tumor cells, NK cells, and T cells. SLR14 stimulates tumor-infiltrating T cells into antiviral states in tumor-immune co-cultures and primes in vitro T-cell production of IFNγ. However, MCC immune infiltrate responsiveness to IFN or SLR14 was notably decreased compared to the melanoma samples, and perturbational computational analyses with CINEMA-OT identified the cytokine midkine (MDK) associated with nonresponse in MCC. Knockout of MDK restored response to IFN and SLR14 by MCC and SCLC tumor cell lines, as well as co-cultured CD14+ monocytes or monocyte-derived DCs. Conclusions: Our approach revealed that midkine, a multifunctional cytokine, suppresses innate immune sensing of IFN and SLR14 in both tumor and immune cells, disrupting the tumor immunity cycle at multiple points. We show that this effect, while comparatively infrequent in melanoma, is pronounced in MCC and SCLC. Our study thus uses a direct assessment of patient tumor and immune samples to identify a novel resistance mechanism enriched in neuroendocrine tumors MCC and SCLC. Therapeutic class Stimulation Target Clinical development Immune checkpoint inhibitor αPD-1 PD-1 Standard of care Cytokine IFNγIFNβ IFNGRIFNAR Early-phase clinical trials orPre-clinical Innate immune agonist Poly(I:C)-NTPoly(I:C)-TADU-S100SLR14 TLR3MDA5/TIG-I/TLR3STINGRIG-I Combination αPD-1 + IFNβ PD-1/IFNAR

Despite promise in preclinical models, most immuno-oncology drug candidates fail in clinical trials. These failures reflect limitations in our ability to directly model the response of human tumor and immune cells to immunotherapies. To address this gap and test the effect of innate immune agonists, we developed PERCEPT, an approach that uses ex vivo perturbational single-cell RNA sequencing to compare the response of immunomodulatory treatments with unstimulated controls directly in patient samples. Using PERCEPT, we tested cytokines and innate immune agonists in melanoma and Merkel cell carcinoma (MCC) and identified the dsRNA mimetic, RIG-I agonist, Stem Loop RNA (SLR) 14 as a powerful inducer of anti-viral states and enhancer of T cell activation. We compared transcriptional responder and non-responder patient samples and identified midkine (MDK), a multifunctional cytokine, as a potent repressor of IFN signaling in both tumor and immune cells. MDK expression dampened MHC-I presentation in human tumor cells and reduced activation of antigen-presenting cells, disrupting tumor immunity at multiple levels. In contrast to prior studies, we identified MDK as specifically enriched in neuroendocrine cancers such as MCC and small cell lung cancer compared with melanoma, suggesting the importance of lineage- and context-specific targeting. Our results demonstrate the utility of high-dimensional controlled perturbation of patient samples to identify mechanisms of innate immune response and resistance and demonstrate an actionable path towards clinical development of MDK-inhibiting therapies including FDA-approved ALK inhibitors in neuroendocrine cancers. ### Competing Interest Statement A.I. co-founded RIGImmune, Xanadu Bio and PanV, and is a member of the Board of Directors of Roche Holding and Genentech. A.I. is a co-inventor on patents related to Stem-Loop RNA (SLR) compounds. NIH Common Fund, https://ror.org/001d55x84, 1R37CA279834, 5P50CA121974, 3P50DE030707, T32CA233414, 5K12CA215110, T32AI07019 NIH Common Fund, https://ror.org/001d55x84, T32GM136651 AstraZeneca (Switzerland), https://ror.org/034rhks82 Conquer Cancer Foundation, https://ror.org/027327e78, Endowed Young Investigator Award, in Memory of John R. Durant, MD

e23526 Background: Kaposi sarcoma (KS) is a tumor caused by Kaposi sarcoma herpesvirus (KSHV), also known as human herpesvirus 8 (HHV-8). Unlike other virus-associated tumors, where disease progression often continues even after viral eradication, KS requires the persistent presence of KSHV for disease progression. This unique dependency underscores the potential of strategies targeting the virus to also target the tumor. This study aims to characterize the expression of KSHV viral proteins and identify factors associated with differences in viral gene expression, ultimately formulating treatment targets. Methods: A total of 43 unique samples were collected at a single academic institution between 2001 and 2024. RNA was extracted from formalin-fixed paraffin-embedded (FFPE) tissue blocks and subjected to bulk RNA sequencing. Reads were mapped to the human genome (GRCh38.p14) using STAR and to the KSHV genome (NC_009333) using Salmon. Transcript abundance in units of Transcripts Per Million (TPM) and estimated read counts to each transcript was quantified with Salmon. Clinical and demographic data from the time of sample collection were extracted from medical records. Differential gene expression analysis, PCA using DESeq2 and hierarchical clustering were done in R. Results: The 43 samples analyzed were derived from 25 patients, with 23 (92%) being male. Of the 18 patients with known HIV status, 4 (22%) were HIV-positive. None had a history of organ transplantation. Among the 14 patients with available disease status, 5 (38%) had disseminated disease at the time of sample collection. KSHV gene expression was detected in 42 out of 43 samples. The most consistently expressed genes across all samples were ORF75 (42 samples, proximal to the poly-A tail), ORF72/v-cyclin (42 samples), and ORF73/LANA (39 samples). A pronounced pattern of high expression was observed in the latency-associated region of the KSHV genome, which was conserved across all KS tumors. The three clusters of samples generated from hierarchical clustering exhibited different lytic and latent gene profile (lytic, latent, mixed), which are similar to previous studies. The differentially expressed genes between different tumor morphology or HIV status of the host showed a mix profile of lytic and latent genes. Conclusions: This study highlights the latency-associated region of KSHV as the most consistently and highly expressed region, regardless of baseline characteristics, presenting a potential conserved therapeutic target.

Breast cancer progression is influenced by tumor subtype, metabolic environment, and patient factors, including menopausal status and BMI. In this study, we utilize publicly available data to investigate the prognostic relevance of PPARγ gene expression, a key regulator of lipid metabolism, and its implications across different subgroups. We also examine the proliferation of adipose tissue in patients with various tumor subtypes and phenotypic cohorts. We analyzed RNA-seq data from 1094 primary breast cancer patients in the TCGA-BRCA cohort, stratifying patients by PPARγ expression, menopausal status, and tumor receptor subtype (ER+, HER2-, TNBC) using the UCSC Xena Browser Viewer. Kaplan-Meier analysis revealed that high PPARγ expression (≥6.903 FPKM) was significantly associated with both improved overall and disease-specific survival, particularly in premenopausal patients. Complementing this, we analyzed PET-CT scans from 69 breast cancer patients in the ACRIN-6888 clinical trial, focusing on SUV metrics (mean, max, peak) of a cell cycle tracer, 3'-deoxy-3'-[18F]-fluorothymidine (18F-FLT) in visceral and subcutaneous adipose tissue at the L3/L4 level. Postmenopausal patients had lower visceral SUVmean (0.705 vs 0.776), and patients with ER+ tumors or non-TNBC tumors showed significantly lower SUVpeak and SUVmax of adipose tissue compared to their counterparts (p < 0.05), indicating metabolic/proliferative reprogramming of adipose tissue based on tumor type. Our study provides a deeper understanding of PPARγ as a therapeutic target in breast cancer through lipid metabolism pathways. The changes in adipose tissue proliferation across cohorts demonstrate the strong potential for subtype-specific breast cancer treatment.

Although mRNA technologies have reinvigorated cancer vaccine development, the identification of strong antigens with consistent tumor cell expression and generation of durable antigen-specific CD8+ T cell memory remain key challenges. We identified the Merkel cell carcinoma (MCC) large T antigen (LTA) as an optimal vaccine target, essential for tumor cell survival and immunogenic in a cancer with high unmet clinical need. We developed an mRNA vaccine to MCC-LTA in murine studies and patient samples. We showed that antigen loss develops rapidly and causes resistance in mouse models when immunogenic, but non-essential antigens are targeted. To improve T cell response durability, we co-encoded LTA and IL-7, co-localizing proliferative and memory signals spatially and temporally with antigen exposure. IL-7-containing mRNA vaccines improved antigen-specific T cell expansion, memory differentiation, and tumor control. We propose that the principles of antigen essentiality and memory signal co-encoding may be adapted to improve the efficacy of mRNA therapeutics.

Prospective and between trial comparisons indicate that first-line treatment with immune checkpoint inhibitors improves survival outcomes compared to first-line therapy with combined BRAF and MEK inhibitors in metastatic melanoma containing BRAFV600E/K mutations. Long-term outcomes for BRAF/MEK inhibition after progression on immunotherapy have not been reported. Moreover, clinical variables associated with outcome from treatment with combined BRAF/MEK inhibition were previously identified in the first-line setting but have not been investigated when targeted therapies are administered after progression on immune therapy. We performed a retrospective single institution analysis of 40 metastatic melanoma patients receiving combined BRAF/MEK inhibitors after progression on an anti-PD-1 or ipilimumab plus nivolumab to assess response rate by RECIST 1.1, progression-free and overall survival (PFS and OS). Pretreatment clinical variables were analyzed for association with OS. Ipilimumab/nivolumab was the first-line immunotherapy regimen in 39 patients (97.5%), and BRAFV600E/K mutations were present in 33 (83%) and 7 (17%) patients, respectively. The median OS from start of BRAF/MEK inhibitors was 20.3 months (1.73-106.4+, 95% CI of median 13.3-30.7). Clinical characteristics associated with worse survival prior to starting BRAF/MEK inhibitors included age > 60 years (median OS 14 vs. 28 months; HR 2.5; 95% CI 0.91-6.87, P = .023), ECOG-PS > 2 (median OS 7 vs. 33 months; HR 2.89; 95% CI 0.78-10.76, P = .018), and presence of bone metastases (median OS 9 vs. 52 months; HR 3.17; 95% CI 1.33-7.54, P = .002). These associations with shorter survival maintained their significance on multivariate analysis. If confirmed in larger cohorts, the identified prognostic variables can be used for stratification of patients in future randomized trials.

2637 Background: Merkel Cell Carcinoma (MCC) is a neuroendocrine skin cancer associated with integration of a truncated form of Merkel Cell Polyomavirus Large T Antigen (LTA) in most U.S. cases. LTA is a robust antigen and its expression is required for proliferation of MCC, rendering it an ideal target for an mRNA therapeutic vaccine approach. Methods: B16-F10 murine melanoma cells that express the truncated LTA oncoprotein were generated for use as an in vivoMCC model. In vitro transcription was used to create an optimized mRNA coding for the LTA and was encapsulated in lipid nanoparticles for intramuscular injection. LTA expressing tumor cells were injected subcutaneously in C57BL/6 mice and treated with LTA mRNA or placebo +/- anti-PD1 antibodies. Tumor growth and survival were tracked, and flow cytometry of ex-vivo tumor samples was used to characterize immune infiltration. Separate mice were treated with a three-dose series of LTA mRNA or placebo 60 days prior to tumor cell injection, and tumor growth and survival were measured. MCC patient PBMCs were collected and used as a model for human in vitro vaccination. Monocyte derived dendritic cells were transfected with LTA mRNA or placebo and used to stimulate the remainder of the PBMC pool every seven days. LTA specific T cell proportion, activation markers, cytokine release, and MCC specific tumor cell killing were measured. Results: Treatment with LTA mRNA suppressed tumor growth compared to placebo (day 21 mean volume 47.28 mm3 vs. 575.12mm3 P<.001) and prolonged median survival (46 vs. 24 days P<.001). Combination treatment with anti-PD1 resulted in 100% tumor regression compared to 50% with anti-PD1 only (median survival 117 days) and 0% with placebo (median survival 28.5 days). Flow cytometry of ex vivotumors from LTA mRNA treated mice revealed increased immune infiltration (mean CD45+ 11.56% vs. 7.81% P=.037) with an increased proportion of CD3+ cells (mean 69.74% vs. 57.36% P=.0013) and CD8+ cells with increased cytotoxic markers (MFI GZMB 26785 vs. 13607 P=.0026). Prophylactic vaccination with LTA mRNA resulted in 80% tumor rejection vs. 0% with placebo (median survival undefined vs. 31.5 days P<.0001). Furthermore, in the prophylactic vaccination group there were no detectable tumors until day 60 following tumor challenge. In vitro vaccination of patient PBMCs resulted in expansion of LTA tetramer specific CD8+ T cells by day 10 (0.3% vs. 0.077%). Exposure of treated cells to antigen loaded DCs resulted in increased IFNg release by day 10 (21.78ng/mL vs. 5.35ng/mL P<.0001) and increased specific MCC tumor cell killing on day 14 (mean tumor death 9.14% vs. 4.18% P=.0001). Conclusions: Therapeutic vaccination with LTA targeting mRNA suppresses tumor growth and prolongs survival in vivo by increasing infiltration of cytotoxic immune populations. In vitrohuman vaccination increases the proportion LTA specific CD8+ T cells, IFNg release, and specific killing of MCC cells.

Background Although immune checkpoint inhibition has moved into the frontline setting for advanced and metastatic head and neck cancer (HNC), only a minority of patients experience durable treatment responses.1 Therapeutic resistance is associated with elevated hypoxia and STAT3 signaling and infiltration by myeloid derived suppressor cells (MDSC).2 3 In this study, we tested the hypothesis that resistant phenotypes are driven by tumor cell autonomous innate immune signaling. Methods Gene Set Enrichment Analysis was used to assess transcriptional signatures of immunity in the Cancer Cell Line Encyclopedia (CCLE) and The Cancer Genome Atlas (TCGA). Western blotting was used to assess HIF1α, HIF2α and pSTAT3 expression. ELISA was used to test the expression of cytokines including IL-6, IL-8 and VEGF-A. Inhibitors and blocking antibodies were used to disrupt TLR2, TLR4, RAGE, IL-1R1 and HIF1α signaling in human cell lines. Transplantable murine models of immunotherapy-sensitive (MOC1) and -resistant (MOC2) oral squamous cell carcinoma were used to test the immune effects of inhibiting HIF1α and IL-1α signaling. Flow cytometry was used to assess infiltration of immune populations in murine tumors. Results We identified patterns of constitutive STAT3 and hypoxia signaling in HNC cell lines from CCLE and confirmed that these signatures were enriched in human HNC tumors (figure 1A-C). We confirmed enrichment of cell autonomous pSTAT3 activation and HIF1a accumulation in HNC cell lines compared with non-HNC controls (figure 2A,B). Cell autonomous inflammatory signaling was associated with secretion of IL-6 and IL-8 (figure 2C). We dissected the mechanism of spontaneous inflammatory signaling and demonstrated that it depends entirely on IL-1α signaling and partly on HIF1α and TLR signaling, which are associated with increased expression of innate pattern recognition receptors (figure 2D,E). We next confirmed the presence of cell autonomous inflammatory signaling in the murine MOC2 model and its absence in MOC1 (figure 3A,B). Echinomycin, a HIF1α inhibitor, significantly impaired the growth of MOC2 tumors (figure 3C,D). Echinomycin-treated MOC2 tumors demonstrated enhanced immune infiltration, T and NK cell infiltration and decreased MDSC compared with placebo controls (figure 3E-G). Conclusions This study establishes a basis for self-propagating inflammatory signaling in HNC tumors driven by HIF1α, IL-1α and STAT3. Disruption of this loop by targeting the IL-1α or HIF1A signaling pathways represent promising strategies to overcome therapeutic resistance and induce anti-tumor immunity. Acknowledgements The Yale SPORE in Head and Neck Cancer provided funding critical to the generation of this study. References Burtness B, et al. Pembrolizumab alone or with chemotherapy versus cetuximab with chemotherapy for recurrent or metastatic squamous cell carcinoma of the head and neck (KEYNOTE-048): a randomised, open-label, phase 3 study. Lancet 2019;394:1915–1928. Geiger JL, Grandis JR, Bauman JE. The STAT3 pathway as a therapeutic target in head and neck cancer: Barriers and innovations. Oral Oncol. 2016;56:84–92. Swartz JE. et al. Poor prognosis in human papillomavirus-positive oropharyngeal squamous cell carcinomas that overexpress hypoxia inducible factor-1α. Head Neck 2016;38:1338–1346. Ethics Approval This study involved animals (mice) These studies were performed under IACUC approved protocol number 2023–20307.

Background Merkel cell carcinoma (MCC) is a rare, aggressive neuroendocrine cancer with rapid progression and mortality rates of 33–46%.1 The majority of MCC is caused by Merkel Cell Polyomavirus (MCPyV) with the remainder induced by UV-mediated damage.1 2 Regardless of virus positivity or tumor mutational burden (TMB), immune checkpoint inhibitors (ICIs) are first line treatment for MCC1 with half of patients not responding or developing resistance. Few options exist for those refractory to immunotherapy.3 There is a need to identify the MCC-specific factors driving resistance and to identify alternate molecular targets. Methods 205 MCC tumors were analyzed using next-generation sequencing (592, NextSeq; WES, NovaSeq) and WTS (NovaSeq) (Caris Life Sciences, Phoenix, AZ). TMB was measured by totaling somatic mutations per tumor (TMB-H: >10 mutations/MB). MCPy viral (MCPyV) status was determined for 68 WES profiled cases using a cut-off of 1000 reads after concordance testing with IHC. Immune cell infiltrates were estimated by Quantiseq. Significance was determined using Chi-square and Mann-Whitney U tests and adjusted for multiple comparisons (q-value <0.05). Results The majority (89.3%) of MCPyV-negative MCC tumors were TMB-high (>10 mutations/Mb), with 96.4% having mutations in TP53 and 80.8% in RB1. Other gene mutations included NOTCH1 (37%), KMT2C (28.6%), TERT (17.9%), FAT1 (14.3%), and PIK3CA (14.3%). In contrast, MCPyV-positive tumors were frequently TMB-low (100%) and rarely harbored mutations in TP53 and RB1 (10.3% and 2.6%, respectively). Immune checkpoint gene (CD80, CD86, CD274, PD1, PD1L, and CTLA4) expression was similar between MCPyV-positive and -negative tumors. Estimated NK cell infiltration was significantly higher in MCPyV-negative tumors. MCPyV-negative MCC also had significantly higher expression of a MAPK pathway activation signature (MPAS). Conclusions MCPyV-positive and -negative MCC represent two classes of molecularly distinct tumors and can be differentiated based on their TMB and mutational profile. The significantly increased NK cell infiltration seen in MCPyV-negative MCC represents a potential therapeutic pathway with the efficacy of NK cell-stimulating agents currently under investigation in the Quilt-3.063 trial.4 MPAS up-regulation in MCPyV-negative MCC suggests that MAPK inhibitors could be used as an alternative to ICIs, which is supported by preclinical data.3 5 MCPyV-negative and -positive MCC are distinct tumor subtypes whose molecular and immune cell profiles warrant further investigation to optimize use of current ICIs and identify therapeutic targets. References Park SY, Doolittle-Amieva C, Moshiri Y, Akaike T, Parvathaneni U, Bhatia S, Zaba LC, Nghiem P. How we treat Merkel cell carcinoma: within and beyond current guidelines. Future Oncol. 2021 Apr;17(11):1363−1377. Knepper TC, Montesion M, Russell JS, Sokol ES, Frampton GM, Miller VA, Albacker LA, McLeod HL, Eroglu Z, Khushalani NI, Sondak VK, Messina JL, Schell MJ, DeCaprio JA, Tsai KY, Brohl AS. The Genomic Landscape of Merkel Cell Carcinoma and Clinicogenomic Biomarkers of Response to Immune Checkpoint Inhibitor Therapy. Clin Cancer Res. 2019 Oct 1;25(19):5961−5971. Fang B, Kannan A, Zhao S, Nguyen QH, Ejadi S, Yamamoto M, Barreto JC, Zhao H, Gao L. Inhibition of PI3K by copanlisib exerts potent antitumor effects on Merkel cell carcinoma cell lines and mouse xenografts. Sci Rep. 2020 Jun 1;10(1):8867. ImmunityBio,Inc. QUILT-3.063: A Study of N-803, haNK and Avelumab in Patients With Merkel Cell Carcinoma That Has Progressed After Checkpoint Therapy. ClinicalTrials.gov identifier: NCT03853317. Updated June 18, 2023. Accessed June 27, 2023. Temblador A, Topalis D, Andrei G, Snoeck R. Synergistic targeting of the PI3K/mTOR and MAPK/ERK pathways in Merkel cell carcinoma. Tumour Virus Res. 2022 Dec;14:200244. Ethics Approval This study was conducted in accordance with guidelines of the Declaration of Helsinki, Belmont report, and U.S. Common rule utilizing retrospective, deidentified clinical data in keeping with 45 CFR 46.101(b)(4). Therefore, this study is considered Institutional Review Board (IRB) exempt and no consent was necessary from the subjects.

Background Despite the transformational impact of immune checkpoint blockade (ICB) in melanoma, only approximately half of patients derive long-term survival benefit. T cells with antiviral signatures have the capacity to secrete inflammatory cytokines/chemokines and deliver potent cytotoxic signals ideal in tumor immunity.1 A promising therapeutic target is agonism of the double-stranded RNA (dsRNA) antiviral sensor RIG-I. The novel RIG-I agonist Stem Loop RNA (SLR)14 enhanced inflammatory cytokine release and improved the control of murine tumors by TILs and other immune cell types.2 However, dsRNA-targeting therapeutic strategies like RIG-I agonism have not yet translated into clinical advances for patients. Further, the effects of SLR14 in human tumors are unknown. We tested the hypothesis that SLR14 transforms T cells to a cytotoxic antiviral state in immunologically 'cold' human tumor specimens via type-1 interferon. Methods We obtained 9 surgical resections from primary melanoma tumors and lymph node metastases and made single-cell suspension replicates of tumor and infiltrating immune cell co-cultures. We stimulated with IFNβ, SLR14, αPD-1, αCD3/CD28+αPD-1 or αCD3/CD28 for 42–48 hours. We Flourescently Activated Cell Sorted (FACS) live cells and then barcoded for multiplexed single cell sequencing using 10x scRNAseq. To visualize the transcriptional response and assign differentiation trajectories following stimulation, we applied PHATE (potential of heat diffusion for affinity-based transition embedding), which facilitates visualization of state transitions and Slingshot, which defines lineage relationships. Results Following SLR14 stimulation, this approach revealed shifts in tumor co-culture cell-type proportions (figure 1A) and alterations in the transcriptional phenotypes of stem-like progenitor and terminally differentiated T cell populations, which have recently been described in single cell RNAseq studies.3–6 RIG-I agonism induced new CD4+ and CD8+ populations not observed in positive or negative control conditions (figure 1B,C). In tumor cells, NK cells and T cells, SLR14 stimulation induced expression of canonical IFN-stimulated genes (figure 1D, bottom). In CD8+ T cells, however, we observed specific programs of activation and survival including CD69 (p=0.00003), and IL2RG (p=0.000006) (figure 1D, top). SLR14-induced stem-like CD8+ T cells maintained high levels of IL7R, similar to unstimulated progenitors, but upregulated CD74 (p=0.00004) and IL2RG—suggesting the potential for inflammatory memory formation.7–8 Similarly, SLR14-induced antiviral CD4+ T cells without any significant increase in Foxp3+ Tregs. Conclusions RIG-I agonist SLR14 stimulates tumor infiltrating T cells into antiviral states in tumor-immune co-cultures. References Jiang X, Muthusamy V, Fedorova O, Kong Y, Kim DJ, Bosenberg M, et al. Intratumoral delivery of RIG-I agonist SLR14 induces robust antitumor responses. J Exp Med. 2019;216:2854–2868. Philip M, Schietinger A. CD8+ T cell differentiation and dysfunction in cancer. Nat Rev Immunol. 2022;22:209–223. Miller BC, Sen DR, Al Abosy R, Bi K, Virkud YV, LaFleur MW, et al. Subsets of exhausted CD8+ T cells differentially mediate tumor control and respond to checkpoint blockade. Nat Immunol. 2019;20:326–336. Oliveira G, Stromhaug K, Klaeger S, Kula T, Frederick DT, Le PM, et al. Phenotype, specificity and avidity of antitumour CD8+ T cells in melanoma. Nature. 2021:1–7. McLane LM, Abdel-Hakeem MS, Wherry EJ. CD8 T Cell Exhaustion During Chronic Viral Infection and Cancer. Annu Rev Immunol. 2019. doi:10.1146/annurev-immunol-041015–055318 Richer MJ, Pewe LL, Hancox LS, Hartwig SM, Varga SM, Harty JT. Inflammatory IL-15 is required for optimal memory T cell responses. J Clin Invest. 2015;125:3477–3490. Naik S, Larsen SB, Gomez NC, Alaverdyan K, Sendoel A, Yuan S, et al. Inflammatory memory sensitizes skin epithelial stem cells to tissue damage. Nature. 2017;550:475–480. Doherty EH, Piecychna M, Leng L, Bucala R. Adoptive transfer of a novel MIF receptor (CD74+) expressing memory T cell subpopulation is sufficient to transfer inflammatory arthritis. The Journal of Immunology. 2017;198:156.3–156.3. Ethics Approval Tumors were collected with the approval of the Yale University Institutional Review Board, IRB # 0609001869, approval date: 7/19/2022, and expiration Date: 7/18/2023. Participants gave informed consent before taking part.

Background Merkel Cell Carcinoma (MCC) is a rare cutaneous neoplasm of neuroendocrine origin associated with integration of a truncated form of the Merkel Cell Polyomavirus Large T Antigen (LTA) in ~80% of U.S. cases.1 LTA is both antigenic and requisite for continued proliferation of MCC, making it an attractive target for a therapeutic vaccine strategy.2 Methods Expression of truncated LTA was induced in a B16-F10 melanoma cell line to create an in vivo murine MCC model. Modified mRNA coding for LTA was produced using in vitro transcription and was packaged in lipid nanoparticles for delivery by intramuscular injection. LTA expressing tumors were placed subcutaneously in C57BL/6 mice and treated with LTA vaccine or placebo with or without additional anti-PD1 antibody. Tumor growth and survival were measured and characterization of tumor immune infiltration by flow cytometry was performed. PBMCs and matched tumor cells from MCC patients were used to establish a model of in vitro vaccination. Monocyte derived dendritic cells (moDCs) were generated from the PBMC pool and were transfected with LTA mRNA. The PBMC pool was pulsed with antigen loaded or placebo moDCs every 7 days, and immunophenotypes, cytokine release, and specific tumor cell killing were assessed. Results LTA mRNA treatment resulted in tumor growth suppression compared to placebo by day 21 (mean volume 199.8mm3 vs. 830.0mm3 P=.007 figure 1A) and increased median survival (35.5 vs. 25.5 days P=.004 figure 1A). Combination of vaccine with anti-PD1 outperformed monotherapy with growth suppression on day 21 (mean volume 103.2mm3 vs. 398.7mm3 P=.037 figure 1B). Treated murine tumors exhibited increased immune infiltration (mean CD45+ 11.56% vs. 7.81% P=.037 figure 1C). Among infiltrating lymphocytes, there was an increased proportion of CD3+ cells (mean 69.74% vs. 57.36% P=.0013 figure 1D), and CD8+ cells were more cytotoxic (MFI GZMB 26785 vs. 13607 P=.0026 figure 1E). In vitro vaccination of patient PBMCs resulted in expansion of CD8+ cells by day 35 (42.2% vs. 22.2% P=.0013 figure 2B), and an increase in activation and memory markers (PD1+ 13.9% vs. 3.91% P=.01; CD45RO+ 77.5% vs. 39.6% P=.009 figure 2C). Exposure of treated cells to antigen loaded DCs resulted in increased IFNg release (422.9pg/mL vs. 96.16pg/mL P<.0001 figure 2D) and increased specific tumor killing (Mean fluorescent count 12 vs. 9.5 P=.038 figure 2E). Conclusions Treatment with a novel mRNA vaccine targeting LTA increases infiltration of cytotoxic immune populations, resulting in significant growth suppression and increased median survival. In vitro vaccination increases cytotoxic CD8+ populations, IFNg release, and specific killing of patient derived tumor cells. References Paulson KG, Park SY, Vandeven NA, Lachance K, Thomas H, Chapuis AG, Harms KL, Thompson JA, Bhatia S, Stang A, Nghiem P. Merkel cell carcinoma: Current US incidence and projected increases based on changing demographics. J Am Acad Dermatol. 2018 Mar;78(3):457–463. Houben R, Adam C, Baeurle A, Hesbacher S, Grimm J, Angermeyer S, Henzel K, Hauser S, Elling R, Bröcker EB, Gaubatz S, Becker JC, Schrama D. An intact retinoblastoma protein-binding site in Merkel cell polyomavirus large T antigen is required for promoting growth of Merkel cell carcinoma cells. Int J Cancer. 2012 Feb 15;130(4):847–56. Ethics Approval Tumors were collected with the approval of the Yale University Institutional Review Board, IRB # 0609001869, approval date: 7/19/2022, and expiration Date: 7/18/2023. Participants gave informed consent before taking part. Animal studies were conducted with the approval of the Yale University Institutional Animal Care and Use Committee, protocol #2022–20307

Waterfall plots demonstrating the percent change in sum of RECIST v1.1 target lesions compared to baseline. Twenty-five out of 26 patients were included; 1 patient withdrew consent and no follow-up imaging was available for review. The waterfall plots are color-coded by (A) best overall response (PR, SD, PD), (B) cohort (1-6), and (C) tumor type (melanoma, RCC, NSCLC). The majority of patients who experienced disease progression on the study had progression in non-target lesions, as denoted by as asterisk (*), and therefore the percent change in tumor from baseline by RECIST is less than the 20% disease progression threshold for multiple patients who had PD.

Association between pre-treatment CD4+ T-cell density from available baseline tumor tissue (n=14) and the duration of time on treatment in days.

Metabolic homeostasis is one of the most exquisitely tuned systems in mammalian physiology. Metabolic homeostasis requires multiple redundant systems to cooperate to maintain blood glucose concentrations in a narrow range, despite a multitude of physiological and pathophysiological pressures. Cancer is one of the canonical pathophysiological settings in which metabolism plays a key role. In this study, we utilized REnal Gluconeogenesis Analytical Leads (REGAL), a liquid chromatography-mass spectrometry/mass spectrometry-based stable isotope tracer method that we developed to show that in conditions of metabolic stress, the fasting hepatokine fibroblast growth factor-21 (FGF-21) 1, 2 coordinates a liver-brain-kidney axis to promote renal gluconeogenesis. FGF-21 promotes renal gluconeogenesis by enhancing β2 adrenergic receptor (Adrb2)-driven, adipose triglyceride lipase (ATGL)-mediated intrarenal lipolysis. Further, we show that this liver-brain-kidney axis promotes gluconeogenesis in the renal parenchyma in mice and humans with renal cell carcinoma (RCC). This increased gluconeogenesis is, in turn, associated with accelerated RCC progression. We identify Adrb2 blockade as a new class of therapy for RCC in mice, with confirmatory data in human patients. In summary, these data reveal a new metabolic function of FGF-21 in driving renal gluconeogenesis, and demonstrate that inhibition of renal gluconeogenesis by FGF-21 antagonism deserves attention as a new therapeutic approach to RCC.

9555 Background: Patients with advanced BRAFV600-mutated melanoma are typically treated with immunotherapy in the first-line setting, followed by BRAF/MEKi upon disease progression based on an absolute 20% improvement in 2-year overall survival over initial treatment with BRAF/MEKi in the DREAMseq trial. Our goal was to identify clinical predictors of longer survival for patients treated in our institution with this approach. Methods: We identified 40 patients with BRAFV600-mutated metastatic melanoma treated at our institution from 2011 to 2020 with immunotherapy followed by BRAF/MEKi upon progression. Clinical data were collected and analyzed by Cox regression and Kaplan-Meier methods. Favorable outcome was defined as survival > 2 years (y) after starting BRAF/MEKis. Results: Median follow up was 33 months (m, 3 – 172 m). Median age was 54 y (20 - 83). Most patients were female (n = 24, %). Most patients were initially treated with ipilimumab plus nivolumab (n = 34, 85%), with 13 of these patients (38%) tolerating all 4 cycles of initial ipilimumab. Adverse events of any grade were seen in 28 (70%) patients after first-line immunotherapy, with the most common being hepatitis, colitis, hypothyroidism, rash, and fatigue. Median duration of first-line immunotherapy was 3.5 m (.75 - 42.5 m). Most common sites of progression on immunotherapy were lymph nodes (n = 14, 35%), liver (n = 12, 30%), bone (n = 10, 25%) and brain (n = 10, 25%). Prior to BRAF/MEKis, median ECOG-PS was 1 (0-4) and median LDH was 268 mg/dL (151 - 11,300). Most common BRAF/MEKi regimen was dabrafenib plus trametinib (n = 34, 85%). Adverse events of any grade were seen in 30 (75%) patients, with the most common being fever, fatigue, nausea, and vomiting. Best response to BRAF/MEKi was CR (n = 4, 10%), PR (n = 26, 65%), SD (n = 4, 10%) and PD (n = 6, 15%), and at the data cutoff, 35 (87.5%) patients progressed on BRAF/MEKi. Median duration of BRAF/MEK inhibition was 7 m (0.5 – 106 m). Median survival since starting BRAF/MEKi was 19.2 m (1.7 – 106 m). On multivariable analyses assessing predictors of survival, presence of bone metastases after disease progression on first-line immunotherapy was associated with worse 2-year survival after initiation of BRAF/MEKi (OR 2.5, 95% CI, 0.51-5.6, p = 0.0121). Other factors, such as ECOG-PS 0-2, normal LDH prior to BRAF/MEKi, and age at metastatic diagnosis < 60 years were not significantly associated with longer survival after initiation of BRAF/MEKi. Conclusions: We showed that the presence of bone metastases upon progression on first-line immunotherapy was associated with worse 2-y survival on salvage BRAF/MEKi for patients with BRAFV600-mutated metastatic melanoma. Predictive and prognostic biomarkers for long-term response to both immunotherapy and BRAF/MEKi are needed to optimize treatment strategies and patient outcomes.