Targeted inhibition of IRAK-4 with emavusertib on immune remodeling in metastatic brain melanoma.

e21547 Background: The incidence of metastatic melanoma is on the rise, and 40-60% of patients will develop melanoma brain metastases (MBM). Frontline treatment consists of stereotactic radiosurgery given concurrently with immune checkpoint inhibition (ICI) or BRAF/MEK inhibition in BRAF mutated patients. While this therapy is effective at local control, the immune-suppressive nature of the CNS creates a permissive environment that facilitates distal intracranial relapse in over half of patients within 1 year. An improved understanding of the CNS tumor microenvironment (TME) is critical for designing effective treatments that improve MBM response to ICI. MyD88 is part of the inflammasome complex downstream of toll-like receptor or IL-1R receptor activation, and is aberrantly activated in the MBM TME. While critical for mediating innate immune activation, unrelenting stimulation of this pathway can trigger immune suppression through the reflexive recruitment of suppressive myeloid cells. Emavusertib (CA-4948) is a CNS permeable inhibitor of IRAK-4, a rate-limiting factor in the myddosome cascade. Thus, emavusertib may recondition the immunologic landscape of MBM through inhibition of myddosome induced myeloid infiltration, improving tumor response to ICI. Materials and Methods: Patient MBM tissues were examined by IHC to validate myddosome activation in the TME. Immunologic profiling of preclinical MBM tumors and PBMCs was performed via flow cytometry in response to anti-PD-1 ICI alone and in combination with emavusertib. Geospatial profiling of lymphocyte and myeloid cell distribution in single and combination treated MBM tumors was done using 3D imaging of cleared brain tissue. Overall survival was assessed in response to anti-PD-1 ICI treatment in combination with emavusertib. Results: In addition to promoting lymphocyte recruitment to MBM tumors, anti-PD-1 ICI significantly increases tumor infiltration by myeloid cells. Surface expression of PD-L1 was also elevated on myeloid cells in response to anti-PD-1 ICI. Microglia demonstrated no change in overall PD-L1 expression in response to either therapy, suggesting that peripheral myeloid cells are the principle target of therapy. Concomitant treatment with emavusertib alleviated ICI induced peripheral myeloid recruitment and reduced PD-L1 expression in MBM tumors while preserving effector lymphocyte infiltration. Emavusertib treatment also improved overall survival of mice when treated in combination with anti-PD-1 as compared to anti-PD-1 therapy alone. Conclusions: Emavusertib reconditions the MBM TME by mitigating the recruitment of infiltrating suppressive myeloid cells that likely confer resistance to anti-PD-1 ICI. This improves the anti-tumor efficacy of anti-PD-1 ICI, resulting in increased overall survival. Thus, emavusertib may be an effective adjuvant for improving treatment response to ICI in patients with MBM.