The systemic response to osteoarthritis influences immunotherapy resistance in melanoma

Introduction: Despite unprecedented outcomes associated with immune checkpoint blockade (ICB), most cancer patients either do not respond or relapse after a partial response. The mechanisms responsible for heterogeneous ICB responses are poorly understood. Acute and chronic stressors are known to influence cancer development. The induction of senescent tumor cells and their senescence-associated secretory phenotype (SASP) within pancreatic cancer has been shown to remodel tumor vasculature, increasing ICB responsiveness. Further, clinical studies have reported patients with osteoarthritis (OA), a disease associated with senescent cell (SnC) accumulation, have slower solid tumor progression and a lower risk of cancer-related mortality. OA, along with other forms of arthritis, has also been identified as an immune-related adverse event (irAE) from ICB therapy, and patients that experience irAEs often demonstrate heightened tumor response to ICB therapy. Thus, by understanding the impact of cellular senescence and other systemic responses induced by OA and ICB on tumor growth we seek to elucidate the characteristics of resistance and develop therapies to overcome resistance. Anterior cruciate ligament transection (ACLT), a model of post-traumatic OA, leads to SnC accumulation in the joint and systemic SASP-mediated effects. Here, we investigate the immunological and stromal changes arising with distant ACLT-induced OA and ICB within non-responding murine melanoma models.

Materials and

Methods: ACLT was performed prior to subcutaneous B16F10, B16-OVA, or Yumm1.7 inoculation. Once tumors were palpable, ICB treatment consisted of 4x intraperitoneal-injected 5mg/kg αPD1 monotherapy or αPD1/αCTLA4 combination therapy. Tumor growth was monitored via caliper measurements. Mice were sacrificed once tumors were >1500mm3, as calculated using ellipsoid approximation (½WxWxL). Serum was collected at day 18 post tumor inoculation for proteomic profiling through Seer Bio’s Proteograph platform. B16-OVA tumors were collected at day 18 post inoculation for flow cytometry to quantify the frequency of infiltrating CD4+, CD8+, and tumor-specific CD8+ (OVA-tetramer) cells. Tumors were collected at day 18 post inoculation and at survival criteria for histological assessment. Immunostaining of p16 (SnC), CD31 (vasculature), PDGFRβ (pericytes), ICAM-1 (vascular integrity), CD8, CD4, FoxP3 (regulatory T cells), F480 (macrophages) were performed. QuPath analytical software quantified tumor vessel number and size, and HALO image analysis quantified the frequency and relevant proximity of the aforementioned cell types throughout the tumor. For SnC deletion experiments, transgenic INKATTAC mice were used with 5 consecutive days of high dose (10mg/kg) AP20187 followed by 2x/week low dose (2mg/kg) AP20187 for the entirety of the experiment. Two-way ANOVA with post-hoc Tukey test determined significant changes in tumor growth, flow cytometric analysis, and histologic quantification.

Results, Conclusions, and Discussions: Our data demonstrate that OA-induced by ACLT surgery, either 1-or 7-weeks prior to tumor injection, significantly increases ICB responsiveness in non-responsive murine melanoma models (B16F10 and Yumm1.7). OA ipsi- and contra-lateral to the tumor similarly increase ICB sensitivity, highlighting the systemic response of OA (Fig 1A). We performed histological and immuno-staining within B16F10 tumors to elucidate changes in vascular integrity (CD31, PDGFRβ, ICAM-1) and senescence (p16) due to OA and/or ICB. We found that OA increases the extent of tumor vascularization, vascular integrity, and frequency of p16+ SnCs surrounding vasculature within ICB-treated tumors (Fig 1B, 1C). In addition, we observed an increase in T-cell infiltration with OA+ICB, including a higher frequency of tumor-specific T cells using the B16-OVA model and OVA-tetramer staining (Fig 1D). Furthermore, proteomic profiling revealed elevated serum-levels of pro-inflammatory cytokines and angiogenic factors, such as VEGF, in OA+ICB-treated mice; similar findings of elevated VEGF serum levels were found in human patient ICB responders that experience arthritis-irAE. Our findings suggest that the response to OA induces SnC accumulation within the tumor, which remodels the tumor vasculature to promote immune cell infiltration and confer better ICB responses. We show that SnCs are critical in this system as when p16+ SnCs were deleted in the INKATTAC transgenic mouse model, tumors from OA mice grew significantly faster than without SnC deletion (Fig 1E). In addition, the vascular remodeling within the tumor to promote immune infiltration observed with OA was reversed with SnC deletion (Fig 1F). In summary, this work explores the senescent-vascular-immune axis to reveal factors that influence favorable responses to ICB to help predict non-responders and develop sensitizing therapies. Future work seeks to identify the systemic factors of OA that promote ICB sensitivity and design combination therapies to overcome ICB resistance.

Acknowledgements (Optional):