Phase II study of nivolumab (nivo) with relatlimab (rela) in patients (pts) with first-line advanced melanoma: Early on-treatment major pathologic response on biopsy

9514

Background: A phase II study of nivo and rela was designed to evaluate the antitumor activity and mechanism of this combination and components for first-line treatment of pts with advanced melanoma. Pts received lead-in treatment with 1 cycle of nivo (480mg IV q4wk), rela (160mg IV q4wk), or nivo-rela followed by combination therapy. We assessed the effect of each lead-in treatment on immune-related pathological response (irPR) at 4-wk biopsy to develop early biomarkers of antitumor response. Methods: Core biopsy of an index lesion was performed at baseline and after 4 wk on-treatment. Immune characteristics of pathological response were assessed on H&E sections, including presence of tumor-infiltrating lymphocytes (TIL), neovascularization, proliferative fibrosis, plasma cells, and lymphoid aggregates. irPR score was calculated as described by Stein JE et al Ann Oncol 2019, from 0 (no irPR features) to 3 (major pathologic response on biopsy [MPRbx], ≤10% residual viable tumor). We assessed the association between irPR and radiological response (RECIST v1.1) at 4-wk evaluations. Results: The current cohort includes 22 pts, median age = 67, male = 13. Pts were randomized to nivo = 7, rela = 7, and nivo-rela = 8 lead-in groups. Two pts had no irPR evaluation due to early progression and unscorable tumor. Among 20 evaluable pts, proliferative fibrosis, neovascularization, plasma cells, brisk TIL, and lymphoid aggregates were identified in 50%, 35%, 26.3%, 25%, and 5% of cases, respectively. Lead-in nivo (n = 2/6), rela (n = 0/6), and nivo-rela (n = 3/8) resulted in irPR = 3 in 25% of pts. Radiological response was identified as partial response (PR) = 1/22 (4.5%), stable disease (SD) = 12/22 (54.5%), and progressive disease (PD) = 9/22 (41%). Among pts with PD, 44% received rela-, 33% nivo-, and 22% nivo-rela- lead-in. Pts with irPR score = 3 had radiological PR = 1, SD = 3, and PD = 1 at 4wks. No association was found between MPRbx and radiological response at 4 wks. Conclusions: Four-wk MPRbx may serve as an early biomarker of treatment response in advanced melanoma. Lead-in treatment resulted in MPRbx of 25% and was greatest with nivo-rela lead-in. Correlations between 4 wk MPRbx and later radiological responses, survival and other endpoints will be made at completion of trial accrual. Clinical trial information: NCT03743766. [Table: see text]

653 The immune landscape of primary and recurrent adolescent and young adult bone sarcomas

Background

Pediatric patients with metastatic or recurrent bone sarcomas have poor survival, necessitating new therapies. Studies of immunobiology in pediatric bone sarcomas have focused on analysis of samples from surgical local control, when patients are receiving immunosuppressive chemotherapy, with little data available from relapse. Here, we sought to leverage transcriptomic and imaging approaches in tandem to characterize the immune landscape of primary and recurrent Ewing sarcoma and osteosarcoma and to identify new therapeutic avenues for these patient populations.

Methods

Single-cell RNAseq (scRNAseq; 10X Genomics) was performed on sorted CD45+ cells from paired peripheral mononuclear cells (PBMC) and tumor infiltrating leukocytes (TIL) from freshly resected bone tumor specimens in the setting of pre-treatment or >6 months post-chemotherapy in the setting of disease relapse. Multiplexed immunofluorescence (mIF) analysis for CD45, DAPI, CD4, CD8, CD68, CD20, and FOXP3 was also performed on FFPE tissue from the same tumors. CIBERSORTx was used in conjunction with TARGET-OS bulk RNAseq from osteosarcoma tumors to infer cell type frequencies. Expression of ligands and receptors in primary versus relapsed disease was assessed from scRNAseq using CellTalker.

Results

We analyzed a total of 29,993 cells from 20 donors (4 paired PBMC/TIL from osteosarcoma, 4 paired samples from Ewing sarcoma, 4 healthy donor PBMC) by scRNAseq. A total of 5 TIL samples were from primary disease sites and 3 were from metastatic sites. We identified major immune populations by canonical expression profiles and used transcriptional profiles from TIL to derive a signature matrix for CIBERSORTx. In 88 osteosarcoma samples from TARGET-OS, we found that higher frequencies of CD14+CD16+ macrophages were associated with better survival (HR:0.28, p = 0.01). scRNAseq from our cohort revealed expression of CXCL12, CCL7, and CCL3L1 by CD14+CD16+ macrophages, suggesting this macrophage population may drive tumor immune infiltration. In both Ewing sarcoma and osteosarcoma, mIF revealed greater numbers of tumor-infiltrating immune cells in the setting of relapse versus primary tumors. scRNAseq analysis revealed higher levels of interferon-gamma expressing CD8+ T cells and CD4+ regulatory T cells in relapsed versus primary disease, suggesting that recurrent tumors may be more immunogenic.

Conclusions

Although pediatric bone sarcomas are typically considered “immunologically cold”, our transcriptomic and imaging approaches revealed a role for a myeloid cell subset in overall survival and increased immune infiltration and T cell activation in recurrent disease. These data suggest specific immunotherapeutic avenues should be tailored to both primary and recurrent disease to improve outcomes in pediatric bone sarcoma.

Ethics Approval

Human specimens were collected with written informed consent under the IRB approved STUDY19030108 and the IRB approved Musculoskeletal Oncology Biobank and Tumor Registry.

839 Microbiota-specific T follicular helper cells drive tertiary lymphoid structure formation and anti-tumor immunity in colorectal cancer

Background

Colorectal cancer (CRC) is one of the most common and deadly cancers in the US, and the survival rate for advanced cases is poor. While immunotherapy has revolutionized cancer treatment, CRC remains largely unresponsive, with only ~6% of patients responding to anti-PD1. Specific microbiome signatures are associated with anti-PD1 response in melanoma patients; however, the underlying mechanism remains unclear. While the microbiome in cancer patients has been extensively studied, the endogenous immune response to these microbes and the subsequent effects on cancer immunity remain unstudied. Most microbes reside within the gut, and bacteria that adhere to the intestinal epithelium can stimulate bacteria-specific immune responses. Therefore, we hypothesized that the microbiome, especially adherent, immunogenic bacteria, may support anti-tumor immunity through activation of local microbiota-specific T cells.

Methods

Using a carcinogen-induced mouse model of CRC, we sought to determine the impact of microbiome modulation on the anti-tumor immune response. We colonized tumor-bearing mice with Helicobacter hepaticus (Hhep) and assessed tumor burden, survival, and immune infiltration. Lymphocytes were isolated from the tumor and surrounding tissue when tumors were terminal (12 weeks). We utilized TCR transgenic mice and MHC class II tetramers to track the spatial and transcriptional Hhep-specific T cell response through 5’ single cell RNAseq, flow cytometry, and spectral immunofluorescence.

Results

Hhep colonization in tumor-bearing mice led to decreased tumor burden and significantly improved survival. Interestingly, colonization induced activation of Hhep-specific T follicular helper cells (TFHs) that supported formation of mature peri- or intra-tumoral tertiary lymphoid structures (TLS). The presence of TLS led to increased infiltration of cytotoxic lymphocytes (T and NK cells) within the tumor core. Surprisingly, the anti-tumor response was dependent on CD4+ T and B cells but not CD8+ T cells. Using TFH KO mice, we found that Hhep-specific CD4+ T cells were both necessary and sufficient to drive TLS maturation and anti-tumor immunity.

Conclusions

Here, we demonstrate that addition of a single bacterial species after tumor formation leads to a reduction in CRC tumor burden and increased survival through TLS maturation. This microbiome-dependent remodeling of the tumor microenvironment is driven by Hhep-specific TFH cells that are both necessary and sufficient for tumor control, demonstrating for the first time that microbiota-specific T cells contribute to anti-tumor immunity. Overall, these findings suggest that microbiome modulation and the subsequent microbiota-specific CD4+ T cell response may represent a new variety of immunotherapies for cancers that remain resistant to checkpoint blockade.

Intractable Coronavirus Disease 2019 (COVID-19) and Prolonged Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Replication in a Chimeric Antigen Receptor-Modified T-Cell Therapy Recipient: A Case Study

A chimeric antigen receptor-modified T-cell therapy recipient developed severe coronavirus disease 2019, intractable RNAemia, and viral replication lasting >2 months. Premortem endotracheal aspirate contained >2 × 1010 severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA copies/mL and infectious virus. Deep sequencing revealed multiple sequence variants consistent with intrahost virus evolution. SARS-CoV-2 humoral and cell-mediated immunity were minimal. Prolonged transmission from immunosuppressed patients is possible.

Abstract PR03: The tumor microenvironment (TME) in head and neck squamous cell carcinoma (HNSCC): Investigating new aspects of known cell types using single-cell RNA sequencing

While immunotherapy has been a significant improvement to traditional head and neck squamous cell carcinoma (HNSCC) treatment, the response rate is only 15-20%. The resistance to immunotherapy is mediated by intrinsic and extrinsic factors produced in the tumor microenvironment (TME). Single-cell RNA sequencing can be employed to study heterogeneous populations (e.g., stroma, cancer, and immune cells) within the TME. We hypothesize that this transcriptomic mapping of single cells will improve our understanding of the complex interactions between the different cellular components of the TME. For this study, matched peripheral blood leukocytes (PBL) and tumors from treatment-naive patients were obtained from the operating room and were processed immediately. Following dissociation of tumors into single-cell suspensions, samples were sorted into CD45+ (tumor-infiltrating leukocytes: TIL) and CD45- (tumor and associated stromal) cells. 10x Genomics 3’ single-cell libraries were generated and sequenced on a NextSeq500 (Illumina). Cells from all samples were aggregated and normalized using the CellRanger pipeline. Downstream bioinformatic analyses were performed using the Scanpy package. 103,006 single cells that passed filtering criteria with a median of 1,105 genes per cell were analyzed. From these populations we identified 30 different clusters, of which 22 clusters were formed by PBL and TIL while CD45- nonimmune cells formed 8 clusters. By extracting and subclustering each cell type individually, subpopulations were identified and characterized. In the immune cell subsets, therapeutically important cell activation substates (e.g., activated regulatory T cells and exhausted CD8 T cells) were observed. Gene set enrichment analysis identified novel pathways between activated and exhausted lymphocytes. Among epithelial cells, transcriptomic profiles of HPV+ and HPV- specimens clustered separately and displayed a surprising degree of heterogeneity that was also seen using pseudotime analyses. The impact of patient-specific differences in epithelial cell profiles on the quality of immune infiltrate was evaluated by modeling interactions between epithelial and immune cell populations. Differences were observed in lymphocyte phenotypes based on HPV status. In conclusion, the TME of patients with HNSCC consists of heterogenous cell populations of immune, stroma, and cancer cells. These can be characterized transcriptomically using single-cell RNA sequencing, which allows to identify diverse cell states. In particular, we uncovered differences in tumor cell populations between HPV+ and HPV- patients, while the nontumor stromal compartment was similar.

Citation Format: Cornelius Kürten, Aditi Kulkarni, Xueer Chen, Lazar Vujanovic, Anthony Cillo, Xinghua Lu, Robert L. Ferris. The tumor microenvironment (TME) in head and neck squamous cell carcinoma (HNSCC): Investigating new aspects of known cell types using single-cell RNA sequencing [abstract]. In: Proceedings of the AACR-AHNS Head and Neck Cancer Conference: Optimizing Survival and Quality of Life through Basic, Clinical, and Translational Research; 2019 Apr 29-30; Austin, TX. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(12_Suppl_2):Abstract nr PR03.

IL-6, IL-8 drive LAG3/PD1 immune suppression on effector and naïve, peripheral blood CD8+ T cells in cancer patients

Cancer patients off therapy and with normal white blood cell counts are often at greater risk for infections, immune dysregulation, progressive disease or reactivation of viral infections. However, the exact mechanism of this systemic immunosuppression in cancer patients is not fully understood. We hypothesize that cancer patients may have systemic immune suppression via cytokine-driven IR expression in all CD8+ T cells subsets, including naïve cells.

PBL were obtained from healthy donors and treatment-naïve NSCLC, HNSCC, and melanoma patients. IR (i.e. LAG3, PD1, CTLA4, etc) expression was assessed on CD8+ T cells, CD4+ T cells, and regulatory T cells. Plasma cytokine concentrations were compared by Luminex. Autologous micro-stimulation assays were performed on peripheral CD8+ or CD4+ T cells with antigen presenting cells plus or minus IR blockade.

CD8+ T cells, including naive CD8+ T cells, from cancer patient PBL contain elevated total LAG3 expression which correlated with worse state, worse survival, and elevated expression of other IRs. Further, CD8+ T cells from these patients had decreased proliferation, which was rescued with the addition of anti-LAG3 or anti-PD1. Plasma from these patients had significantly elevated levels of cytokines that can signal via STAT3 (i.e. IL-6, IL-8), which were independently found to increase total IR expression in healthy donor, naïve CD8+ T cells.

Patients with cancer have systemic elevations IL-6 and IL-8 leading to increased IR expression in naïve, peripheral CD8+ T cells making them poised for exhaustion even before TCR binding. These findings suggest that cytokine combined with IR blockade may play a role in reversing PD1 blockade resistance.

Tumor infiltrating B cells co-localize with CD4 T effector cells within organized tertiary lymphoid structures to present antigen and educate the anti-tumor immune response in human primary tumors

Only 20% of cancer patients produce a durable response to current immunotherapies. Thus, a need exists to develop additional therapeutic strategies to treat these patients, which includes evaluation of other tumor infiltrating immune cells that could further augment the T cell response. Tumor infiltrating B cells (TIL-B) represent a possible target for immunotherapy due to their crucial role in the immune response. In fact, current data demonstrate that detection of TIL-Bs within tertiary lymphoid structures (TLS) correlate with increased survival in patients with solid tumors and enhanced response to anti-PD1 immunotherapy. We hypothesize that TIL-Bs help generate potent, long-term immune responses against cancer by presenting tumor antigens to CD4 TILs within TLS.

We observed increased numbers of activated TIL-Bs in the primary tumors of head and neck squamous cell carcinoma and non-small cell lung cancer patients. We further assessed the transcriptional signature and location of TIL-Bs within the TME and demonstrated that they are adjacent to Th1 and Tfh CD4+ tumor infiltrating lymphocytes (TILs) within organized, germinal center containing TLS. Further, we generated an antigen presentation assay in vitro, and we observed increased CD4+ TIL responses when TIL-Bs presented autologous tumor antigens. These data suggest that TIL-Bs influence the function of CD4+ TILs in patient tumors within TLS. Ultimately, results from this study will increase effective targeting of TIL-Bs within patient primary tumors.