Background: HCC is increasing rapidly in incidence worldwide driven by a rise in chronic liver disease including non-alcoholic steato-hepatitis (NASH). Most pts are not suitable for curative or loco-regional treatments and may be candidates for systemic therapies. Immune checkpoint inhibitors combined with VEGF inhibition is a standard of care in HCC. However, a meta-analysis of 3 phase III randomised trials of PD-1 or PDL-1 inhibitors (n > 1,600 pts) with HCC suggests that pts with NASH-related HCC treated with PD-1/PDL-1 inhibitors had reduced overall survival compared with other aetiologies. Neutrophils expressing the chemokine receptor CXCR2, crucial to neutrophil recruitment in acute-injury, are highly represented in NASH-HCC. In NASH-HCC murine models, lacking response to immune-checkpoint inhibitors, AZD5069 (CXCR2 inhibitor) in combination with anti-PDL-1 suppressed tumor burden and extended survival, accompanied by an increase in tumor-associated neutrophils which switched from a pro-tumor to anti-tumor progenitor-like neutrophil phenotype. We propose that inhibition of CXCR2 may potentiate the efficacy of anti-PDL-1 inhibition in pts with HCC. Methods: In this multi-centre (n = 10) study, pts with biopsy-confirmed HCC, PS ECOG < 1, Child-Pugh A, < 1 prior systemic therapies, receive 1 of escalating doses of AZD5069 (bid, po daily) with Durvalumab (1.5 gm iv on day 1) in 28-day cycles for up to 2 years to determine the recommended phase II dose using a Keyboard design, followed by an additional cohort of pts to determine the anti-tumor efficacy of this combination using a Simon’s two-stage design (min 18, max 35 pts; target objective response rate > 30%; unacceptable response rate < 10%). Dose limiting toxicities (DLTs) are assessed during cycle 1. Disease assessments are performed 8-weekly (12-weekly after 1 year). The 1st dose cohort has been completed with no DLTs. The 2^nd dose cohort opened to recruitment in September 2022. Exploratory studies (blood; pre- & on-treatment tumor and non-malignant liver biopsies) include biomarkers of CXCR2 inhibition (blood); proof-of-mechanism (tumor: expression of CXCR2, PD-L1, PD-1, CD8, CD4, CD66b, CD69); proof-of-mechanism (blood: ctDNA); drug-induced changes of mRNA expression, CXCR2 ligands & signalling pathway genes, T-cell and myeloid cell pathways, neutrophil-associated genes; predictive biomarkers (blood and tumour) include biomarkers of the CXCR2/PD-L1 immune axis; aberrant CXCR2 signalling pathways; proliferation biomarkers and CD10 (neutrophils), CD68 (macrophages), CD103 (T-regs); tumour mutational status. This study is funded by a grant from Cancer Research UK (A29287) and is co-sponsored by University of Glasgow and NHS Greater Glasgow & Clyde. Study sites are supported by the Experimental Cancer Medicine Centre Network. AZD5069 and Durvalumab are provided by Astra Zeneca. Clinical trial information: 2020-003346-36.