Safety and clinical activity of invikafusp alfa, a selective T cell activator, against antigen-rich solid tumors
1470. START001: A Phase 1/2 study of Invikafusp alfa (STAR0602), a first-in-class TCR β chain-targeted bispecific antibody, as monotherapy in patients with antigen-rich solid tumors resistant to anti-PD(L)1
James Gulley (National Cancer Institute, Bethesda, MD, USA) presented results from the STARt-001 trial, phase 1/2 study investigating invikafusp alfa, a dual T cell agonist for the treatment of anti-PD(L)1-resistant, antigen-rich solid tumors. Invikafusp alfa is a selective T cell activator consisting of a selective Vβ T cell receptor (TCR) agonist linked to a wild-type IL2 moiety. Vβ6/10 TCR-expressing T cells are the most-represented T cell subset among tumor infiltrating lymphocytes (TILs), thus invikafusp alfa may represent a novel approach to treating tumors that do not respond to immune checkpoint blockade. Previous studies indicate that invikafusp alfa reprograms the TIL compartment with more CD8 memory T cells than anti PD-1 immune checkpoint inhibitors, and in mouse models of colorectal cancer, reinvigorating Vβ TILs promotes potent and durable anti-tumor activity. Phase 1 of the study followed a standard 3+3 dose escalation design. Patients with unresectable, locally advanced, or metastatic solid tumors that were antigen-rich (i.e. high tumor mutation burden (TMB-H), microsatellite instability high/mismatch repair deficient (MSI-H/dMMR) or virally associated) participated in the study. 35 patients, representing 16 different types of antigen-rich tumors, received invikafusp alfa intravenously every two weeks. 54.3% of patients had received four or more prior lines of therapy, and 82.9% of patients had received prior anti-PD-(L)1 therapy. Expansion of peripheral Vβ6/10 TCR-expressing T cells occurred in a dose-dependent manner, with optimal expansion corresponding with the recommended phase 2 dose (RP2D) of invikafusp alfa, 0.08 mg/kg. Expansion of Vβ6/10 TCR-expressing T cells, both CD8+ and CD4+ T cells, lasted through day 14 of the study and occurred across all 16 tumor types. Most adverse events were Grade 1 or 2 and were consistent with the mechanism of action of invikafusp alfa. No Grade 4/5 treatment-related adverse events were observed. The disease control rate (DCR) for the entire patient population was 50% and 32% among patients who had received prior anti-PD-(L)1 treatment. Tumor shrinkage was observed in 7 of 14 patients who received the optimal biologic dose range (0.08-0.12 mg/kg), and 2 patients, both with TMB-H, microsatellite stable colorectal cancer, experienced confirmed partial responses (PRs). Among patients with TMB-H tumors, the DCR was 59%, and tumor shrinkage was observed in 4 different tumor types. Invikafusp alfa represents a new class of T cell agonists that promotes anti-cancer immunity through the selective activation of Vβ6/10 TCR-expressing T cells, and data from the dose escalation phase of the study suggest it is well-tolerated and exhibits anti-tumor activity in a heavily pre-treated patient population. A phase 2a expansion study of invikafusp alfa monotherapy for TMB-H solid tumors, MSI-H solid tumors, and TMB-H and/or MSI-H/dMMR colorectal cancer has opened and is enrolling patients.
Using artificial intelligence to assess patient samples and identify predictive biomarkers in patients receiving immunotherapy for rare tumors
1207. Artificial intelligence-powered assessment of tumor microenvironment in pre-treatment and on-treatment biopsies informs treatment outcomes to pembrolizumab in patients with rare tumors
Mohamed H. Derbala (The University of Texas MD Anderson Cancer Center, Houston, TX, USA) reported results from a study of the predictive role of the tumor microenvironment (TME) and tumor contents (TC) in rare tumors. 259 baseline and 248 on-treatment biopsies from 84 patients with rare tumors treated with pembrolizumab were evaluated. Tumor infiltrating lymphocyte density (iTIL) and tumor content (TC; the ratio of cancer area to the total of cancer area, stroma, and background tissue) were quantified from H&E slides with an artificial intelligence (AI) powered whole-slide image analyzer. Tumor types were subgrouped as having high iTIL or low iTIL at baseline. In the high baseline iTIL subgroup, higher baseline iTIL was associated with favorable progression free survival (PFS) compared to low baseline iTIL (p=0.04). Among the 83 patients with matched baseline and on-treatment biopsies, an increase in iTIL of one or more and a decrease in TC by 0.3 or more were associated with longer PFS and better overall survival (OS). Patients who exhibited both an increase in iTIL and decrease in tumor content exhibited significantly longer PFS (p=0.0031) and better OS (p=0.009). Results from this study underscore the potential benefits of using AI and other biomarkers to predict patient response to immunotherapy, ultimately informing the development of personalized treatments and future immunotherapy trials
Characterization of the LIRG1-VISTA axis, an emerging immune checkpoint target
855. LRIG1 engages ligand VISTA to impede T cell activation and enable tumor immune evasion
Hieu Ta (Cleveland Clinic Foundation, Cleveland, OH, USA) presented a study identifying LRIG1, a regulator of T cell function, as a novel binding partner of V-domain immunoglobulin suppressor of T cell activation (VISTA), an emerging immune checkpoint target, and the effects of this interaction on anti-cancer immunity. VISTA is expressed on immune cells, including T cells and dendritic cells, as well as a variety of human tumors, where it is often highly upregulated. Prior studies have indicated that VISTA expression on tumor cells impairs anti-tumor immunity by interacting with an unknown receptor to impair a T cell-mediated anti-tumor response. A selective proteomic proximity labeling assay using tyramide identified the transmembrane tumor suppressor LRIG1 as binding partner. Interaction of LRIG1 with VISTA occurred in vitro and in vivo, VISTA-LRIG1 interaction could occur in cis and in trans, and LRIG1 was found to be expressed exclusively on activated T cells. LRIG1 knockout T cells exhibited enhanced T cell signaling, proliferation, and survival. Mice with T cell-specific LRIG1 deletion exhibited superior anti-tumor response compared to mice with T cells expressing full length LRIG1, and tumor control was associated with a decrease in quiescent cytotoxic T lymphocytes (CTLs) along with an increase in progenitor and memory-like CTLs. LRIG1 expression in CD8+ tumor-infiltrating lymphocytes was significantly associated with resistance to immune checkpoint blockade. Results from this study suggest that interaction of LRIG1 and VISTA on activated T cells results in decreased cytokine production and decreased T cell survival, ultimately promoting tumor growth. Results from this study provide new insights to the mechanisms underlying VISTA-mediated impairment of T cells and support the development of new cancer immunotherapies targeting the VISTA-LRIG1 axis.
A novel mouse model to identify drivers of immune-related adverse events
1175. Deciphering immunotherapy's challenges: A novel mouse model for immune related adverse events
Dipyaman Patra (University of Pittsburgh, Pittsburgh, PA, USA) presented the development and characterization of a preclinical mouse model of immune related adverse events (irAEs), termed the Lifestyle Induced Susceptibility to irAEs (LISA). This mouse model meets an unmet need in understanding the mechanisms and management of irAEs. Currently, irAEs are often managed with steroids and other immunosuppressants that can affect the efficacy of immunotherapy, It was hypothesized that common mouse lab strains do not develop irAEs because they are exposed to high fiber diet and a controlled microbiome. LISA mice were fed a Western Diet consisting of higher fats and simple carbohydrates, potentially changing metabolites and the microbiome leading to the development of irAEs. Control mice (fed a standard diet) and LISA mice were subjected to three weeks of anti-PD1 therapy. Unlike the control mice, LISA mice exhibited systemic onset of irAEs including inflammation within the skin, colon, and liver. Single cell RNA sequencing indicated an upregulation of inflammatory CD4+ regulatory T cells (Tregs) in affected tissues. These inflammatory Tregs expressed inflammatory markers including interferon (IFN) gamma and became unstable, losing their immunosuppressive capacity. during irAEs. Both the Western diet and the LISA mice microbiome contributed to Treg dysfunction. Female LISA mice and older LISA mice with higher inflammatory markers were more prone to irAEs. Western diet and immune checkpoint blockade were found to drive gut dysbiosis In LISA mice, indicated by an increase in Bacteroides species. Microbiome differences between male and female LISA mice were also found to play a role in susceptibility to irAEs. Lifestyle intervention, the switching of LISA mice from a Western diet to a regular diet at the beginning of anti-PD-1 treatment brought about normal levels of CD4 T regs and IFN gamma production. In addition to meeting an unmet need in the field of immuno-oncology research, development and characterization of a new mouse model of irAEs has identified new drivers of irAE prevalence and severity and potential approaches like lifestyle intervention, to prevent and manage irAEs.
Elucidating the mechanisms underlying exercise-enhanced anti-cancer immunity
1253. Exercise-induced microbiota vitamin B9 metabolite enhances CD8 T cell antitumor immunity and promotes immunotherapy efficacy
Catherine Phelps (University of Pittsburgh School of Medicine, Pittsburgh, PA, USA) presented a study to identify specific mechanisms underlying exercise-enhanced anti-tumor immunity. Previous studies have established that a sedentary lifestyle increases the risk of cancer and that exercise is associated with enhanced anti-tumor immunity and improved response to immune checkpoint inhibitors for some cancers like melanoma. Mouse models of immune checkpoint inhibitor (ICI)-resistant BRAF V600E melanoma tumors ran 1 hour each day for 5 days per week on a mouse treadmill. Exercised mice exhibited enhanced CD8+ T cell-mediated tumor control compared to sedentary mice. Exercised germ-free mice and exercised mice on antibiotics did not exhibit enhanced tumor control, indicating exercise-induced anti-cancer immunity in this mouse model was dependent on an intact germ-free microbiota. Fecal microbiota transplant (FMT) from exercised mice to sedentary mice significantly improved anti-tumor immunity in the sedentary mice, indicating that the exercised microbiota is sufficient for tumor suppression. Metabolomics on feces of mice indicated that exercise changed the metabolic output of microbiota by increasing bacterial folate (vitamin B9) metabolism, and formate, a short-chain fatty acid and known intermediate of folate metabolism, was enriched in the cecum and serum of exercised mice. Oral administration of formate to the mouse melanoma model was sufficient to promote CD8+ T cell effector function, restrain tumor growth, and enhance immunotherapy efficacy. Nuclear factor erythroid 2-related factor 2 (Nrf2) signaling was found to be required for exercise-mediated anti-tumor immunity. Mice colonized with E. coli that could not produce formate exhibited decreased-levels of anti-tumor immunity and FMTs from formate-high human donors was superior in suppressing tumor growth in the mouse model, compared to FMTs from format-low donors. These results support the development of a model in which exercise changes the microbiome, leading to increased formate synthesis, which promotes CD8+ T cell-mediated anti-melanoma immunity in an Nrf2-dependent manner. More studies are needed to develop potential therapies that enhance this pathway and improve the efficacy of immune checkpoint inhibitors against melanoma.
Tumor draining lymph nodes as a source of lymphocytes for adoptive cellular therapies for solid tumors
1099. Harnessing lymph nodes for polyclonal multipotent T cell therapy in solid tumors
Tatiana Delgado Cruz (Weill Cornell Medical College, New York, NY, USA) presented a study investigating tumor draining lymph nodes (TLDNs) as potential reservoirs of T cells for autologous adoptive cellular therapies like tumor infiltrating lymphocyte (TIL) therapy and chimeric antigen receptor (CAR) T cell therapy. While these cellular therapies have been successful in treating specific cancers, they present unique challenges. For instance, T cells for TIL therapy, derived from tumors, are often exhausted and differentiated. T cells for CAR T therapy are isolated from peripheral blood and are usually naïve and less differentiated. Previous studies have indicated that memory progenitor CD39- CD69- T cells with a stem -like phenotype are associated with TIL efficacy. Because draining lymph nodes are a reservoir of antigen-specific stem cell-like T cells, it was hypothesized that TDLNs may be reliable reservoirs of T cells for adoptive cellular therapies. Lymphocytes were harvested from TDLNs, peripheral blood, and tumors of 3 patients with non-small cell lung cancer (NSCLC). T cell receptor and single cell sequencing indicated that 34% of tumor-specific expanded T cells from the tumor were also present in TDLNs. T cell clones from TLDNs exhibited a less exhausted and more activated phenotype compared to paired T cell clones from the tumor. T cells from TLDNs also exhibited progenitor-like transcriptional profiles. To determine whether the less exhausted T cells from TLDNs could serve as a source for cellular engineering with CARs, TDLN-derived T cells were compared with T cells from peripheral blood. T cells derived from TLDNs could be maintained ex vivo without IL-2. Certain T cell memory cells, including CD8+ PD-1+ CXCR5+ stem-like T cells were found in TDLNs but not peripheral blood. T cells from TDLNs and peripheral blood were transduced with an ICAM-1 targeting CAR. TDLN-derived T cells demonstrated transduction efficiency and expansion levels comparable to T cells derived from peripheral blood. TDLN-derived CAR T cells expressed higher levels of transcription factors associated with persistence and cytotoxicity, compared to peripheral blood-derived CAR T cells, which expressed genes associated with lower responsiveness and apoptosis. Both types of CAR T cells exhibited similar tumor-killing capacities and similar anti-tumor activity in vivo against mouse models of NSCLC. These findings indicate that TDLNs, which can be easily harvested from many tumor types, represent a diverse source of polyclonal T cells for adoptive cellular therapies, and may have the potential to enhance clinical outcomes of TIL therapies and CAR T therapies for solid tumors.
Patient-specific and tumor-specific factors affecting the growth and reactivity of tumor infiltrating lymphocytes
447. Clinical factors associated with growth and reactivity of tumor infiltrating lymphocytes for adoptive cell transfer in patients with metastatic epithelial cancers
Alexandra Gustafson (National Cancer Institute, Bethesda, MD, USA) presented a retrospective review analyzing factors related to the growth and reactivity of adoptive cellular therapy with tumor infiltrating lymphocytes (TIL) in patients with epithelial cancers. TIL growth was measured as the number of days for half of tumor fragments to reach four or more confluent wells in a 24-well plate. TIL reactivity was measured as neoantigen-induced release of interferon gamma in cocultures and 4-1BB upregulation on flow cytometry. The study analyzed all metastatic tumor resections performed for the purpose of TIL analysis that took place at the National Cancer Institute from August 2014 to September 2023. Patient characteristics, including radiation history and exposure to immune checkpoint inhibitors and/or chemotherapy as well as tumor characteristics were collected. 47% of samples were from colorectal cancer, 22% from breast cancer, and 11% from pancreatic cancer. Based on samples from 291 patients, prior therapies did not affect TIL growth. Viable TILs were more likely to be obtained from the lungs and less likely from the liver. Tumor size and number of tumors did not affect TIL growth, and the median TIL growth time was 21 days. Prior exposure to immune checkpoint inhibitor therapy significantly reduced the likelihood of obtaining reactive TILs (p=0.03). Correlations of TIL reactivity with the site of tumor harvest were similar to those observed for TIL growth: tumors from the lung were more likely to yield reactive TILs (p=0.01), and tumors from the liver were less likely to yield reactive TILs (p=0.04). These results indicate that while tumors of the lung are more likely to generate reactive TILs, TILs can be harvested from all sites investigated in this study. These results also suggest that heavily pretreated patients are good candidates for autologous adoptive cellular therapy, tumors from patients with prior exposure to immune checkpoint blockade may not be immunogenic. Additional analyses are ongoing, including a study correlating the site of TIL harvest with patient response.