Blogs

SITC 2023 Science Coverage (Nov. 3)

By SITC Communications posted 11-03-2023 00:00

  

Neutrophils are important for the destruction of heterogenous tumors in mouse models of melanoma

878. T cell immunotherapies recruit and activate neutrophils to eliminate tumor antigen escape variants

Daniel Hirschhorn (Weill Cornell Medical and Graduate Schools, New York, NY, USA) presented a study investigating the mechanisms of immune-mediated destruction of heterogeneous tumors that contain antigen-loss variants. Selective pressures of the immune system create heterogeneity and antigen-loss variants within tumors, and this heterogeneity often leads to resistance to T cell-targeted immunotherapies.  Mouse models of Trp1+ B16 melanoma were treated with a combination treatment of cyclophosphamide (CTX), anti-OX40 antibody, and Trp1-specific CD4+ T cells. Combination treatment promoted killing of Trp1+ tumors as well as bystander killing of tumors that did not express Trp1 or MHC. Depletion of Trp1-T cells early (Day 3) during therapy abrogated the anti-tumor response against heterogenous tumors, but anti-tumor activity was not affected if depletion of Trp1-T cells occurred later (13 days), suggesting the endogenous cell population was participating in the killing of cancer cells. Flow cytometry confirmed that neutrophils are recruited to the tumors after administration of combination therapy, and neutrophil extracellular traps (NETs) were observed in tumors of mice treated with combination therapy. Depletion of neutrophils with anti-LysG antibodies after combination therapy led to higher rates of tumor escape and decreased survival rates, compared to mice in which neutrophils were not depleted. This pattern was also observed among mice treated with immune checkpoint blockade and with anti-OX40 antibodies, indicating that neutrophils are required for elimination of heterogeneous tumors. Tumor-killing neutrophils exhibited unique transcriptional profiles, with upregulated expression of genes involved in neutrophil activation and migration such as CD40 and CD177. To determine whether reactive oxygen species were necessary for neutrophil-mediated killing of tumor cells, iNOS knockout mice were generated, and combination therapy was not sufficient to kill heterogeneous tumors in iNOS knockout mice. Neutrophils from wild type mice, but not iNOS knockout mice, destroyed melanoma cells in vitro, indicating iNOS was partially necessary for neutrophil anti-tumor activity. To apply these findings to clinically relevant scenarios, samples from patients who had been treated with anti-OX40 for head and neck cancers in the neoadjuvant setting exhibited high levels of NETs after treatment, and similar results were observed in patients who had been treated with immune checkpoint blockade. These data led to the development of a model of immunotherapy promoting heterogenous tumors: immune checkpoint blockade or OX-40 antibodies promote infiltration of T cells into tumors and T cell-mediated destruction of cancer cells. Chemokines and cytokines secreted by the activated T cells recruit and activate neutrophils to kill escape variants in an iNOS-dependent manner. These findings underscore the potential of combination therapies to engage multiple types of immune cells, such as neutrophils, in overcoming antigen loss variants and destroying heterogenous tumors.

Anti-tumor activity of CAR macrophages and transforming endogenous myeloid cells in vivo with CAR mRNA

1514. In vivo CAR-M: redirecting endogenous myeloid cells with mRNA for cancer immunotherapy 

Bindu Varghese (Carisma Therapeutics, Natick, MA, USA) presented two studies of chimeric antigen receptor (CAR) macrophages and monocytes (CAR-M) to overcome barriers to immunotherapy in solid tumors. Genetically redirected CAR macrophages can destroy tumor cells, secrete cytokines to activate dendritic cells and T cells, and present neoantigens to T cells, controlling tumor growth and improving long-term survival. HER2-targeting CAR macrophages control tumor growth in mouse models of HER2+ tumors, improve survival rates, and increase infiltration of T cells into the tumor, indicating the remodeling of the tumor microenvironment. Mice previously treated with HER2-targeting CAR macrophages (CT-0508) rechallenged with HER2- tumors, rejected tumor growth, while treatment-naïve mice exhibited tumor growth, suggesting CAR macrophages can protect against tumor recurrence, even in the absence of the original tumor antigen. Phase 1 clinical trials of CT-0508 as monotherapy and in combination with pembrolizumab are in progress.

To generate functional CAR-M in vivo, a novel strategy was used to deliver CAR messenger RNA (mRNA) encapsulated in lipid nanoparticles (LNPs) to generate CAR-M. Macrophages and monocytes engineered with CAR-encoding mRNA/LNP in vitro exhibit high CAR expression and antigen specific killing of tumor cells in three-dimensional tumor spheroids. CAR-Ms engineered with mRNA/LNP produce inflammatory cytokines and chemokines, and this activity is specific to CAR expression. CAR-macrophages engineered with mRNA/LNP have increased M1 and decreased M2 markers upon antigen stimulation. Administration of CAR mRNA/LNP to mice in vivo led to transfection of macrophages, monocytes, and dendritic cells, and transfection of human pBMCs in vitro primarily transformed myeloid cells. When CAR mRNA/LNP was administered to a HER2+ lung cancer model in mice with human immune system, robust tumor shrinkage was specifically observed in mice receiving HER2 CARs, and administration of CAR mRNA/LNP also suppressed metastatic pancreatic tumor growth of the highly invasive HER2 PANC1 cell line. CAR mRNA/LNP also cleared liver metastases and reduced lung metastases. Animals treated with CAR mRNA/LNP repeatedly was well tolerated. CAR-M are highly functional and show promising anti-tumor activity in mouse models of multiple cancers. CAR mRNA/LNP represents a novel off-the-shelf platform to develop anti-tumor myeloid cells in vivo that has potential to target numerous cancer antigens and indications and increase access to targeted cellular therapies against cancer.

Manufacturing CAR cells in the presence of inosine increases anti-tumor activity

275. Inosine endows CAR T cells with features of increased stemness and anti-tumor potency

Dorota Klysz (Stanford University School of Medicine, Palo Alto, CA, USA) reported a study of adenosine, inosine, and their role in CAR T cell exhaustion. Previous studies of the HA (GD2) CAR T model, in which GD2-targeting CAR clusters on the surface of human T cells in the absence of antigen, mimics chronic antigen exposure have indicated tonic signaling results in CAR T cell exhaustion. CD39 and CD73, two ectoenzymes involved in adenosine production, are expressed by exhausted CAR T cells. Adenosine is known to be a major mediator of T cell exhaustion and immune suppression. In order to determine whether CD73+ CD39+ CD8+ T cells mediate immune suppression via adenosine, transmembrane-bound adenosine deaminase (ADA-TM), which metabolizes adenosine to inosine, was overexpressed. HA (GD2) CAR T cells overexpressing ADA-TM exhibited a higher frequency of stem- and central-like memory as well as significant transcriptional changes. Because overexpression of ADA-TM also induces increased levels of inosine in addition to decreased levels of adenosine, HA (GD2) CAR T cells were manufactured in the presence of high inosine concentrations (HA-INO). The presence of inosine induced transcriptional changes in CAR T cells, including downregulation of genes in glycolytic pathways and upregulation of genes in glutaminolysis. HA-INO cells exhibited decreased glycolytic activity, increased mitochondrial activity, and HA-INO cells were more resistant to oxidative stress, compared to cells manufactured in high glucose concentrations. CAR T cells cultured in the presence of inosine curbed tumor growth and promoted survival in mouse models of cancer in vivo, and inosine induced high expression of CD62L, indicating stemness, as well as increased secretion of interferon gamma2, indicating anti-tumor activity. The conclusions of this study, that inosine induces metabolic changes in T cells that increased stemness and promote anti-tumor activity, indicate that manufacturing CAR T cells in the presence of inosine has potential to improve CAR T cell quality, reduce CAR T cell exhaustion, and improve patient outcomes. Work has begun to apply these observations to CAR T cell development and to investigate the feasibility of including inosine in a clinical, large-scale manufacturing process.

Modifying the microbiome with fecal filtrate to improve response to immunotherapy

1328. The use of fecal filtrate transplant to enhance response to immune checkpoint blockade

Golnaz Morad (The University of Texas MD Anderson Cancer Center, Houston, TX, USA) presented a study of microbial modification with fecal filtrate and its effects on response to immune checkpoint blockade (ICB). While multiple studies have shown that patients with ICB-refractory cancer benefit from fecal microbiota transplantation (FMT), FMT is associated with limitations, including reliance on donors, rigorous screening, and the risk of transferring unknown pathogens. In an attempt to identify strategies with lower risk and better cost efficiency, sterile filtrates were prepared from stool samples from melanoma patients who were complete responders (CR) and those who did not benefit from ICB (non-responder, NR). Germ free mice received NR or CR FMT or NR or CR filtrate, were injected with BP melanoma tumor cells and then treated with anti-PD-L1. Mice that received FMT or filtrate from CR had an improved response to anti-PD-L1 treatment, suggesting microbial derivatives may be sufficient for inducing a response to ICB. While treatment with NR FMT or NR filtrate was associated with poor response, mice that received filtrate from NR responded better than NR FMT, suggesting bacteria in FMT may abrogate response to ICB. Treatment with CR filtrate was associated with significantly higher levels of B-derived cells and macrophages/monocytes in the colon and increased NK cells in the tumor, compared to CR FMT, suggesting CR filtrate can alter the tumor microenvironment. CR filtrate was associated with enrichment in immunoglobulins and proteins from 4 major bacterial species (mostly Bacteroides) compared to NR filtrate, suggesting the CR microbiome induces a B cell response.  While these results are currently being validated in studies of larger NR and CR cohorts, the data suggest that acellular components of FMT may confer improved response to ICB, and synthetic bacteria and/or their associated byproducts may someday be used to optimize the microbiota of patients with cancer, improving their response to immunotherapy.

Safety and efficacy of a novel CTLA-4 antibody for metastatic non-small cell lung cancer

599. Single-agent safety and activities of target-preserving anti-CTLA-4 antibody gotistobart (ONC-392/BNT316) in PD-(L)1 resistant metastatic NSCLC and population PK analysis in patients with solid tumors

Kai He (James Cancer Center, The Ohio State University, Columbus, OH, USA) presented results from one arm of a phase I-II of gotistobart (ONC-392), the first-in human study of a next-generation target-preserving anti-CTLA-4 monoclonal antibody as monotherapy for patients with non-small cell lung cancer (NSCLC) that progressed on PD-(L)-1 therapy. Gotistobart is a pH-sensitive antibody that dissociates from CTLA-4 in the acidic environment of the endosome, allowing recycling of the antibody and CTLA-4, thus CTLA-4 is preserved and re-expressed on the cell surface, preventing autoimmunity and toxicities. Preclinical studies in humanized mice indicate gotistobart is more effective and less toxic than ipilimumab. Gotistobart exhibited pharmacokinetics similar to other monoclonal antibodies used as cancer immunotherapy. 34% of patients who received at least one dose of gotistobart experienced an immune-related adverse event (irAE) of grade 3 or 4. 30% of patients who received a dose of 10 mg/kg x 2, followed by 6 mg/kg every three weeks experienced an irAE of grade 3 or 4, which is lower than reported rates for other CTLA-4 inhibitors.  Among 27 evaluable patients, eight patients responded to therapy (overall response rate 29.6%), with one complete response and seven partial responses.  The median duration of response was greater than 6 months, and the disease control rate was approximately 70%. Gotistobart shows promising anti-tumor activity among patients with NSCLC that is resistant to PD-(L)-1 inhibitors and a manageable safety profile. A larger, phase 3 study comparing gotistobart to docetaxel in patients with metastatic NSCLC whose disease progressed on anti-PD-(L)1 therapy is in progress.

Inhibition of mitochondrial citrate export is associated with lipid accumulation and T cell exhaustion

944. Obese, not starving: mitochondrial citrate export drives lipid accumulation and dysfunction in exhausted T cells

Kellie Spahr (University of Pittsburgh, Pittsburgh, PA, USA) reported a study of the relationship between lipid accumulation and loss of function in exhausted T cells. Prior studies have indicated that exhausted T cells have low mitochondrial mass, caused by fewer and smaller mitochondria, suggesting that T cell exhaustion can be driven by mitochondrial stress, caused by the immunosuppressive tumor microenvironment. CD8+ T cells also accumulate large lipid stores as they progress towards exhaustion. Citrate was investigated as a factor in lipid accumulation, and it was hypothesized that as cells lose mitochondrial function, citrate that would go through the TCA cycle in the mitochondria may instead be exported to cytosol to fuel lipid synthesis. An in vitro model of T cell exhaustion was used in this study, where chronic TCR stimulation and oxygen levels were manipulated to drive T cells into exhaustion. 1, 2, 3-benzenetricarboxalate (BTA) was used to inhibit the activity of the SLC25A1 citrate transporter in mice, T cells were isolated and chronically stimulated in hypoxic conditions to simulate T cell exhaustion. Cells treated with BTA expressed significantly lower levels of PD-1 and TIM-3, compared to cells with functional citrate transport, suggesting blocking mitochondrial citrate export reduced T cell exhaustion. Tumor-specific T cells in which Slc25a1, the gene encoding the mitochondrial citrate transporter, was deleted (knockout) were co-transferred with control T cells in tumor-bearing hosts. After 11 days, tumor-infiltrating T cells were removed, and knockout T cells had accumulated fewer lipids than the control T cells, suggesting deletion of the mitochondrial citrate transporter reduced lipid accumulation in CD8+ T cells. Knockout cells also expressed inhibitory receptors PD-1 and TIM-3 at lower levels. Expression of CD39, another marker of exhaustion was also reduced in the knockout cells that co-expressed PD-1 and Tim-3. Knockout T cells also exhibited increased cytokine production in tumor draining lymph nodes compared to control T cells, but no differences were detected in tumor-infiltrating T cells. Knockout T cells expressed increased levels of Slamf6, suggesting deletion of the mitochondrial citrate transport slows but does not completely stop differentiation towards an exhausted state. Results from this study suggest a correlation between citrate export, lipid accumulation, and T cell exhaustion. Follow-up studies will more precisely define the effects of deletion of mitochondrial citrate export on carbon utilization and determine whether the stored lipids could fuel the function of exhausted T cells.

Combining Zika virus oncolytic therapy with immune checkpoint blockade generates durable responses against glioblastoma in mice

463. Targeting a novel myeloid checkpoint Siglec-15 in GBM generates an extremely durable response with Zika virus oncolytic therapy

Ashwani Kesarwani (Washington University, Saint Louis, MO, USA) reported results from preclinical studies of combining Zika virus (ZKV) with immune checkpoint blockade for mouse models of glioblastoma (GBM). Previous studies have shown ZIKV preferentially infects and kills glioma stem cells relative to differentiated tumor progeny or normal brain cells. Intratumoral injection of ZKV in some GBM mouse models improves survival to 40% compared to control (0%), and this efficacy of ZKV is dependent on CD8+ T cells. ZKV also increases the number of myeloid cells in the brain tumor. In this study, ZKV was combined with therapies targeting Siglec-15 to improve efficacy. Siglec-15, a novel immune checkpoint that suppresses T cell activation, is expressed on myeloid cells and human GBM cells. ZIKV combined with anti-Siglec-15 antibodies or with Siglec-15 knockouts did not improve survival in the immune checkpoint blockade-resistant SB25 mouse model of GBM but including anti-PD1 with ZKV in a Siglec-15 knockout mouse improved survival to approximately 60%. Using an anti-Siglec-15 antibody instead of a knockout improved survival to 70%. One week post-treatment, mice that had received combination treatment exhibited significantly lower levels of myeloid cells in the tumor microenvironment. ZIKV also significantly activated CD8 T cell cytotoxicity, and this cytotoxicity was enhanced by addition of anti-PD-1 and anti-Siglec-15. Mice that had been cured of GBM with combination therapy were rechallenged with SB28 on the other side of brain. All mice that had been previously cured with combination therapy survived rechallenge, while all treatment-naïve mice died within 40 days. Cured mice exhibited significantly higher levels of CD8+ T cells and CD8+ memory T cells, suggesting combination therapy may provide long-term protection against GBM.

Permalink