Abstract Hepatocellular carcinoma (HCC) is the most common type of liver cancer. Although increasing immune checkpoint blockade (ICB) inhibitors have been applied in HCC clinical trials, the immunosuppressive tumor microenvironment (TME) restricts therapeutic responses to a small subset of patients. The triggering receptor expressed on myeloid cells-2 (TREM2) plays a critical role in counteracting inflammation and maintaining metabolic fitness in myeloid cells. Recently, we performed single-cell RNA sequencing (scRNA-seq) on tumor biopsies from a phase II clinical trial of pembrolizumab on advanced hepatitis B virus (HBV)-related HCC patients (NCT03419481) and identified a subset of tumor-associated macrophages (TAMs) over-expressing TREM2, which were enriched in non-responders following therapy. Consistently, our syngeneic ICB-sensitive and resistant HCC mouse models verified that Trem2+ TAMs were adaptively increased after anti-PD-1 treatment in resistant tumors. We observed that TREM2+ TAMs enriched in lipid-laden TME and transferred fatty acids to tumor cells. Trem2 deficiency in myeloid cells reduced the lipid level of tumor cells and re-sensitized them to anti-PD-1 therapy. Citation Format: Zhixian Liang, Xiaohang LONG, ZHEWEN XIONG, PATRICK WONG, Siyuan HUANG, Siyun Chen, Yiling Zhang, Lingyun ZHANG, Chunning Leung, Saiming Ngai, Stephen Lam Chan, Alfred S. L. Cheng. Macrophages promote immunotherapy resistance of hepatocellular carcinoma through increasing tumor cell lipid availability [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3429.

The management of hepatocellular carcinoma (HCC) has undergone radical change over the past decade. Immunotherapies now dominate the treatment of advanced-stage disease and are increasingly being evaluated in perioperative and intermediate-stage settings. However, in some instances, positive phase III trials have not translated into adoption by guidelines or regulatory agencies, highlighting the need to harmonize and update the current standards for trial design and end points. In response to these challenges, four scientific societies - the European Association for the Study of the Liver (EASL), the American Association for the Study of Liver Diseases (AASLD), the International Liver Cancer Association (ILCA) and the American Society of Clinical Oncology (ASCO) - appointed representatives to develop a consensus recommendations document addressing current unmet needs and future challenges in HCC trial design and end points. Through a modified Delphi process, 102 consensus statements were developed across several distinct domains: surveillance; early-stage, intermediate-stage and advanced-stage disease; transplant-related contexts; regulatory considerations; and emerging topics. The document rigorously defines target populations, stratification factors, control arms and benchmarks for expected clinical benefit. Collectively, this consensus is intended to provide a dynamic, evidence-based, multisociety roadmap for optimizing trial design, accelerating therapeutic development and improving clinically meaningful outcomes in patients with HCC.

540 Background: START-FIT (sequential transarterial chemoembolization, stereotactic body radiotherapy, plus anti-PD-L1 and anti-CTLA-4 immunotherapy) shows promise in locally advanced HCC. We evaluated its activity using anti-PD-L1 and anti-CTLA-4. Methods: Patients aged ≥18 with unresectable HCC (≥5cm, ≤3 nodules, Child-Pugh A-B7) were enrolled. Exclusions included extrahepatic metastasis and main portal vein (VP4) or IVC (VV3) invasion. Patients underwent TACE, 5-fraction SBRT (27.5-40 Gy), then Tremelimumab (300 mg) and Durvalumab (1500 mg) starting 7 days post-SBRT, every 4 weeks. Primary endpoint: ORR by mRECIST 1.1; secondary endpoints: PFS, OS, LC, TRAEs. Results: From 2020 to 2024, 33 patients (44 tumors) were enrolled; median age 67 (50–82), 91% male. Median tumor size was 11.1 cm (5.6–24.7), with 73% (24/33) showing macrovascular invasion (hepatic vein n=14, portal vein n=5, both n=5). Median follow-up was 21 months (5.6–53). The ORR was 72.7% (24/33; 95% CI 57.5–87.9%), with 42.4% (14/33) achieving CR and 30.3% (10/33) PR. Stable disease was seen in 9.1% (3/33), PD in 12.1% (4/33). Two patients were not evaluable. Among CR patients, 42.4% (14/33) were under surveillance; 21.2% (7/33) converted to curative treatments (resection n=3, ablation n=4). The 18-month LC was 96.6% (95% CI 90.1–100%), PFS 61.8% (44.1–79.4%), OS 90% (79.4–100%). Grade ≥3 TRAEs occurred in 36.3% (12/33), mainly transient AST/ALT elevations (n=5). Four patients (12.1%) had grade ≥3 immune-related adverse events. Conclusions: START-FIT using STRIDE is safe and effective in locally advanced unresectable HCC resulted in 42% CR rate with an additional of 21% patients converted to curative surgery. Clinical trial information: NCT 04988945 . Patient and tumor characteristics. N=33 Age, median (range), years 67 (50-82) SexMaleFemale 30 (90.9%)3 (9.1%) ECOG performance status01 29 (87.9%)4 (12.1%) Aetiology of liver cirrhosisHepatitis BMultiple etiologiesCryptogenic 26 (78.8%)3 (9.1%)4 (12.1%) Child-Pugh scoreA5A6 28 (84.8%)5 (15.2%) Albumin Bilirubin Score (ALBI)Grade 1Grade 2 19 (57.6%)14 (42.4%) BCLC stageA-BC without extra-hepatic spread 9 (27.3%)24 (72.7%) Reasons of Unresectable Inadequate liver remnant volume, poor ICG, or unable to achieve R0 resection, or bothBCLC stage B beyond up-to-sevenBCLC stage C without extra-hepatic spread 8 (24.2%)1 (3.1%)24 (72.7%) Tumor vascular invasionNoYesBranched portal vein invasionHepatic vein invasionBoth portal vein and hepatic vein invasion 9 (27.3%)24 (72.7%)5 (15.2%)14 (42.3%)5 (15.2%) Number of lesion(s)12-3 22 (66.7%)11 (33.3%) Size of largest lesion, median (range), cm 5.6 (11.1-21.8) Sum of largest diameters of lesions, median (range), cm 11.1 (5.6-24.7) Baseline AFP, nmol/L≤ 400 ng/ml> 400 ng/ml 23 (69.7%)10 (30.3%)

Figure S1 shows time course of HBV DNAa and ALT levels, duration of treatment, and initiation of new antiviral therapy in participants with clinically resolved HBV at baseline who experienced HBV reactivation during the study

477 Background: Tumor recurrence is frequent after surgical resection and local ablation in participants (pts) with hepatocellular carcinoma (HCC). There remains an unmet need for standard-of-care adjuvant therapy to prevent disease recurrence and improve overall survival. The phase 3, randomized, double-blind KEYNOTE-937 (NCT03867084) study evaluated the efficacy and safety of pembrolizumab vs placebo as adjuvant therapy in pts with HCC after surgical resection or local ablation with curative intent. We present results of the third interim (IA3) analysis of KN937. Methods: Eligible pts were aged ≥ 18 years with confirmed HCC, complete response after surgical resection or local ablation, ECOG PS 0 - 1, and Child-Pugh liver class A. Pts with past or ongoing hepatitis C or controlled hepatitis B virus infection could enroll after meeting prespecified criteria. All pts were randomized 1:1 to pembrolizumab 200 mg or placebo IV Q3W until disease recurrence, unacceptable toxicity, intercurrent illness, withdrawal, or up to 17 cycles of pembrolizumab or placebo. Randomization was stratified by geographic region, prior local therapy (resection vs ablation), recurrence risk, and alpha-fetoprotein level at diagnosis. The primary endpoints were recurrence-free survival (RFS) by imaging (BICR) or pathology, and overall survival (OS). Safety was a secondary endpoint. Key exploratory endpoints included distant metastases-free survival (DMFS) and time to recurrence. The data cut-off date was Mar 20, 2025. Results: A total of 959 pts were randomized (476 pembrolizumab; 483 placebo). At IA3, median follow-up was 50.7 months (range, 31.6 – 69.1). Median RFS was not statistically different with adjuvant pembrolizumab vs placebo (46.7 mo vs 45.5 mo; HR 1.06; 95% CI, 0.88-1.26; P=0.719 [P-value boundary for significance of 0.0187]), with 48 mo RFS rate of 50% vs 50%. As the RFS hypothesis was not met, OS was not tested per multiplicity and the study will not proceed to final analysis. Median OS was not reached (NR) in either arm (HR 1.08; 95% CI, 0.81-1.43; P=0.704 [nominal]), with 48 mo OS rate of 79% vs 81%. Median DMFS was NR (95% CI, 58.7 to NR) with pembrolizumab vs NR (95% CI, 59.0 to NR) with placebo (HR 0.98; 95% CI, 0.77-1.24), with 48 mo DMFS rate of 71% vs 70%. Median time to any recurrence was 52.5 mo vs 50.1 mo. Grade ≥3 adverse events (AEs) occurred in 32% vs 22% pts, respectively, with grade ≥3 drug-related AEs occurring in 14% vs 5% of pts. There were no drug-related deaths. Conclusions: Adjuvant pembrolizumab versus placebo did not demonstrate improvement in RFS in pts with HCC after surgical resection or local ablation. Clinical trial information: NCT03867084 .

541 Background: In HIMALAYA (NCT03298451), STRIDE (Single T Regular Interval D) improved overall survival vs sorafenib at the 3-yr and 5-yr data cutoffs (DCOs), with manageable safety. We assess baseline characteristics and safety outcomes in STRIDE-treated HIMALAYA participants (pts) who were T-rechallenged. Methods: Pts were randomized to STRIDE (T 300 mg once + D 1500 mg every 4 weeks [Q4W]), D monotherapy (1500 mg Q4W), or sorafenib (400 mg BID). Pts who, by investigator assessment, were benefitting from STRIDE but had evidence of progressive disease (of target or non-target lesions), could receive one additional dose of T 300 mg after ≥6 cycles of D. Efficacy was not assessed due to implicit survivorship bias in the cohort. In this post-hoc exploratory analysis, baseline characteristics at initial T dosing, treatment (tx) exposure, and safety are assessed in STRIDE-treated pts who were T-rechallenged. Outcomes are assessed at 3 yrs (primary DCO: Aug 27, 2021) and 5 yrs (DCO: Mar 1, 2024) of follow-up. Results: At the 3-yr DCO, 30/393 (8%) pts were T-rechallenged; at the 5-yr DCO, 34/393 (9%) pts were T-rechallenged (64 yrs median age, 85% male, 27% Asian; similar to overall study population). At the 3-yr DCO, the median (range) tx duration of D was 16.5 (5.5–42.6) months (mo) for T-rechallenged pts vs 5.5 (0.4–42.6) mo for all STRIDE pts. At the 5-yr DCO, the median (range) tx duration of D was 16.6 (5.5–72.0) mo for T-rechallenged pts vs 5.5 (0.3–72.0) mo for all STRIDE pts; the median (range) time from tx start to T rechallenge was 12.8 (4.9–53.4) mo; 44% of pts were rechallenged ≤12 mo of starting tx. Overall, T-rechallenged pts had similar safety outcomes to the overall STRIDE population. At the 3-yr DCO, adverse events (AEs) of any causality occurred in 97% of T-rechallenged pts vs 97% of all STRIDE pts (Table). No AEs with outcome of death occurred in T-rechallenged pts. Immune-mediated AEs (imAEs) occurred in 33% of T-rechallenged pts; 10% had imAEs requiring high-dose steroids. Among T-rechallenged pts, 17% had a Grade 3 or 4 treatment-related AE (TRAE) at any time (10% before and 7% after rechallenge). Serious TRAEs occurred in 6% of T-rechallenged pts vs 18% of all STRIDE pts, with no new serious TRAEs between the 3-yr and 5-yr DCOs. Conclusions: Tx was well-tolerated by T-rechallenged pts under the protocol conditions. Despite a limited cohort size, the baseline characteristics and safety outcomes were in line with the overall STRIDE population. Clinical trial information: NCT03298451 . n (%) T-rechallenged pts(n=30*) All STRIDE pts(n=388*) Any AE 29 (96.7) 378 (97.4) Grade 3/4 AE 12 (40.0) 196 (50.5) AE with outcome of death 0 30 (7.7) imAE 10 (33.3) 139 (35.8) Any TRAE 22 (73.3) 294 (75.8) TRAE pre-rechallenge 21 (70.0) - TRAE post-rechallenge 9 (30.0) - Serious TRAE by 3-yr DCO 2 (6.7) 68 (17.5) Serious TRAE by 5-yr DCO 2 (5.9) † 68 (17.5) *In the safety analysis set (all pts who received ≥1 dose of study tx). † n=34 at 5-yr DCO.

HBV-related outcomes in evaluable participants with HBV infectiona.

Cell-free DNA (cfDNA) end motifs serve as fragmentomics biomarkers for cancer. Prior studies primarily focused on 5 ' ends, whereas 3 ' ends were overlooked due to artifactual modification in existing sequencing protocols. We utilized single-stranded library preparation ("2-end sequencing") to assess the native 5 ' and 3 ' end motifs (EM5 and EM3, respectively). Additionally, we demonstrated diagnostic power from the nucleotide motifs located immediately upstream and downstream of 5 ' and 3 ' ends, named pre-end motifs (PREMs) and post-end motifs (POEMs). These fragmentomics markers collectively achieved an area under the curve (AUC) of 0.95 for hepatocellular carcinoma (HCC) detection. Fragmentomics-based methylation analysis of 3 ' ends (3 ' FRAGMA) improved detection of HCC (AUC: 0.97). We further developed "4-end sequencing" to interrogate both ends of both strands of a double-stranded cfDNA molecule, enhancing fragmentomics-based cancer detection. Holistic end profiling adds to the armamentarium of liquid biopsy and sheds light on the biology of cfDNA fragmentation.

Summary of adverse events by HBV infection status (APaT population)

Baseline demographics and disease characteristics by HBV infection status (ITT population).

Background & Aims:Intrahepatic cholangiocarcinoma (ICC) with multiple liver tumours is staged as T2 disease in the 8th edition of the AJCC staging system. However, tumour multiplicity is frequently associated with poor prognosis in ICC. We aimed to elucidate the prognosis of ICC with multiple tumours and the impact of the number of liver tumours on survival. Methods:Patients who were diagnosed with ICC with solitary liver tumours or multiple liver tumours between May 2008 and December 2019 at three tertiary referral centres in Hong Kong, Seoul and Shanghai were enrolled. Kaplan-Meier analyses and multivariable Cox proportional hazards models were used to analyse the association between tumour multiplicity and overall survival (OS). Results:A total of 849 patients were included in the analysis. Among the 563 patients without distant metastasis, 434 (77.1%) and 129 (22.9%) had solitary and multiple liver tumours at baseline, respectively. The median follow-up period was 91.9 months (IQR 69.6-131.5). Median OS for patients with a solitary liver tumour, multiple liver tumours, and distant metastasis was 39.5 months (95% CI 34.4-49.2), 14.8 months (11.4-20.3), and 7.1 months (95% CI 6.1-8.3), respectively (log-rank p <0.001). Among N0M0 patients with multiple liver tumours, those with 2-3 tumours and those with ≥4 tumours had a median OS of 36.9 months (16.0-NA) and 15.5 months (11.0-23.7), respectively. In multivariable analysis, N0M0 patients with ≥4 liver tumours had similar survival as those with N1 disease (adjusted hazard ratio 1.01, 95% CI 0.58-1.75). Conclusions:Tumour multiplicity in ICC is associated with inferior OS compared with solitary liver tumours, particularly in patients with ≥4 tumours. Future ICC staging systems should consider the prognostic significance of tumour number. Impact and implications:Our research provides critical insights into the prognosis of intrahepatic cholangiocarcinoma (ICC) with multiple liver tumours, highlighting the significant impact of tumour multiplicity on survival outcomes. This work addresses the gap in the current staging system, which does not adequately reflect the poor prognosis associated with multiple tumours, especially with the presence of ≥4 tumours. Incorporating these insights into clinical practice could aid in treatment decision-making, risk stratification, and potentially improve patient care and outcomes in ICC management.

Tumor cells metabolically adapt to the nutrient-deprived tumor microenvironment (TME). However, the metabolic plasticity underlying immune-checkpoint blockade (ICB) adaptation remains unclear. Here, we report that tumor cells exploit macrophage efferocytosis to metabolically counteract immune-checkpoint targeting. Serial tumor biopsies from patients with ICB-resistant hepatocellular carcinoma (HCC) demonstrate heightened tumor cell fatty acid uptake (FAU) with concomitant up-regulation of TREM2+ lipid-associated macrophages (LAMs) in lipid-laden TME. Myeloid-specific Trem2 deficiency and anti-TREM2 antibody abolish fatty acid-dependent energy production in ICB-resistant tumor cells, resensitizing them to ICB via epigenetic TME remodeling. Mechanistically, TREM2+ LAMs recycle fatty acids to tumor cells via efferocytosis-derived extracellular vesicles, thereby promoting H3K36 acetylation-associated activation of MYC and TGF-β signaling. Single-cell spatial analysis supports TREM2+ LAM efferocytosis in the epigenetic immune evasion of patients with ICB-resistant HCC. As high TREM2+ LAMs correlate with FAU and ICB non-responsiveness in multiple human cancers, our study identifies a common metabolic vulnerability for combinatorial immune-checkpoint targeting.

553 Background: Worsening liver function is a major challenge in the management of uHCC. The optimal subsequent treatment strategy for patients with CP-B liver function at progression on first-line Atezo-Bev remains unclear. Methods: This multinational, multicenter retrospective study included pts with uHCC and CP-B liver function at progression on first-line Atezo-Bev from five institutions in Korea, Hong Kong, Taiwan, and Singapore between July 2016 and March 2025. For efficacy analyses of subsequent therapy, only pts with CP scores of 7 or 8 were included. Post-progression survival (PPS) was defined from the time of progression on Atezo-Bev to the death of any cause. Results: A total of 146 pts (median 64 year-old; 89% male, 11%) were included. At progression, CP scores were 7 (51%), 8 (34%), and 9 (15%). Of these, 84 pts (57.5%) received subsequent therapy; sorafenib in 42 (50%) and lenvatinib in 39 (46%), while 62 (43%) received best supportive care (BSC) only. Median PPS was 5.4, 2.3, and 2.5 mo in pts with CP score 7, 8, or 9, respectively (p<0.001). In pts with CP scores 7–8, those who received subsequent therapy had significantly longer PPS compared those with BSC (7.3 vs. 2.0 mo; HR 0.37, p<0.001). Among pts receiving second-line therapy, lenvatinib was significantly associated with longer PFS compared with sorafenib (3.5 vs 1.7 mo, p=0.013), although OS did not differ (6.0 vs 4.4 mo, p=0.56). In pts with CP-B and albumin-bilirubin (ALBI) grade 2, lenvatinib led to significantly longer PFS compared with sorafenib (4.2 vs. 1.4 mo, respectively; p=0.01). No significant difference was observed among pts with CP-B and ALBI grade 3 (1.2 vs 1.9, p=0.87). Conclusions: Among pts with CP-B liver function at progression on first-line Atezo-Bev, PPS differed according to the degree of liver dysfunction. Subsequent systemic therapy, particularly lenvatinib, might be associated with improved outcomes in this population.

Surgery after response to first-line chemoimmunotherapy may offer a potential curative option for patients with locally advanced biliary tract cancer (BTC) initially considered unresectable. However, robust real-world evidence in this setting remains limited. We retrospectively analyzed patients with locally advanced BTC treated with first-line cisplatin, gemcitabine, and durvalumab (CGD) across 55 centers in 12 countries. Endpoints included overall survival (OS), progression-free survival (PFS), disease-free survival (DFS) among resected patients, objective response rate (ORR), and the prognostic impact of surgery. A multivariable logistic regression model was developed to predict surgical conversion using baseline clinical and laboratory variables. Among 1358 screened patients, 219 had locally advanced disease and were included in the analysis. Median follow-up was 14.5 months. ORR was 34.6%, with median PFS and OS of 10.0 and 20.7 months, respectively. Twenty-four patients (10.9%) underwent surgical resection after systemic therapy. Median OS was 22.5 months in resected patients versus 19.9 months in those who did not undergo surgery. Median DFS after resection was 15.8 months. Pathological lymph node involvement was independently associated with shorter DFS. Larger baseline tumor size correlated with higher odds of radiological response but not with OS. An exploratory elastic-net model retained four baseline variables and showed moderate discriminative ability for surgical conversion, with an apparent AUC of 0.72 and an optimism-corrected AUC of 0.65. First-line CGD enabled secondary resection in approximately 11% of patients with locally advanced BTC. A simple baseline model may support patient selection for conversion surgery, although prospective validation is needed.

Systemic therapy for advanced HCC has evolved from a linear, stepwise algorithm into a multidimensional and truly polyhedral process shaped by treatment response, liver function, and access to therapy. This review integrates evidence from phase II and III trials and real-world studies, including more than 4000 patients across different global regions, to outline how sequencing beyond first-line immunotherapy is redefining clinical practice. After immune-based combinations, tyrosine kinase inhibitors remain active and feasible options, achieving a median overall survival of ~10-11 months, while preservation of liver function consistently determines treatment eligibility and postprogression survival. Incorporation of the Barcelona Clinic Liver Cancer concept "upon response" refines classification for patients achieving major responses or conversion and aligns staging with contemporary treatment dynamics. The CUSE framework-encompassing complexity, uncertainty, subjectivity, and emotion-reframes sequencing as an adaptive, patient-centered, and multifaceted process that integrates toxicity, feasibility, and patient values alongside efficacy. By structuring diverse clinical inputs into a unified interpretive model, CUSE supports multidisciplinary reasoning and enhances conceptual coherence in treatment sequencing. Conversion therapy and "drug-off" strategies following complete or major responses represent emerging frontiers, emphasizing that sequencing in HCC now extends beyond second-line therapy. Additionally, preserving hepatic function and applying polyhedral, data-informed decision frameworks are key to sustaining eligibility and optimizing outcomes across the continuum of care.

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer, accounting for up to 80% of all cases. Most patients present at an advanced stage and are unsuitable for curative treatments. In the past 5 years, immunotherapy combination has superseded tyrosine kinase inhibitors (TKI) as the standard first-line therapy for advanced HCC. These therapies include the combination of anti-PD-(L)1 with an anti-CTLA-4 or partner with anti-VEGF; these agents offer unprecedented high rates of response and survival. Unfortunately, only about 20%-30% of patients respond to first-line immunotherapy combination, and about 50% of them would develop disease progression at 6 months. In patients with progression on immunotherapy, recent prospective studies support the efficacy and safety of multiple TKIs. To further improve the efficacy of systemic therapies, novel therapeutic strategies are actively being investigated, such as with the addition of a third immune checkpoint inhibitor. Furthermore, there is increasing interest to incorporate locoregional therapies in patients with advanced disease. Three Phase 3 randomized studies (EMERALD-1, LEAP-012, and TALENTACE) have recently demonstrated survival benefits with the combination of trans-arterial chemoembolization (TACE) and immunotherapy, over TACE alone. The higher response rates brought by combining locoregional therapies and systemic therapies have enabled the possibility of downstaging and conversion. In addition, cellular therapy has shown promise in early phase studies, demonstrating potential to expand the use of immunotherapy in HCC beyond immune checkpoint inhibitors. In this review, we provide an overview of the current treatment landscape and emerging therapeutic strategies for advanced HCC.

Abstract Although outcomes for patients with advanced hepatocellular carcinoma (HCC) have improved with the advent of immune-checkpoint inhibitor (ICI)-based therapies, only approximately 30% of patients achieve an objective response. With the rapid development of single-cell technologies, increasing attention has been paid to immunosuppressive cell types in the tumor microenvironment (TME) from a transcriptomic perspective; however, the contribution of genetic variation to immunotherapy resistance remains largely unexplored. Here, we employed single-cell multiomics of 14 treatment-naïve and ICI-resistant HCC patients to investigate cell type-specific genetic regulation underlying chromatin accessibility and gene expression within the HCC TME. High-quality single-cell ATAC-seq data were generated for 155,700 cells and annotated into 11 major cell types, including B cells, CAFs, CD4+ T cells, CD8+ T cells, dendritic cells, endothelial cells, macrophages, monocytes, NK cells, Tregs, and tumor cells. Leveraging genotype information inferred from single-cell ATAC-seq, RASQUAL systematically mapped 23,329 chromatin accessibility-associated quantitative trait loci (caQTLs) across all cell types. We found a wide range of cell number-normalized caQTLs in different cell types (0.01 to 0.22). Notably, tumor cells exhibited the highest frequency of caQTLs, suggesting greater genetic susceptibility within their regulatory landscape. Colocalization with liver cancer-related GWAS signals, combined with peak-to-gene association analyses, identified enhancer-like regulatory elements that may modulate genes involved in tumor progression and immunotherapy resistance. In summary, our study delineates the cell type-specific genetic control and regulatory architecture of HCC TME and provides mechanistic insights into the molecular basis of immunotherapy resistance. Acknowledgement: This study is supported by Li Ka Shing Foundation and CUHK Strategic Seed Funding for Collaborative Research Scheme. Citation Format: Siyuan Huang, Xiaohang Long, Stephen Lam Chan, Alfred Sze-Lok Cheng, . Single-cell multiomics analysis reveals cell type-specific genetic regulatory programs underlying immunotherapy resistance in hepatocellular carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1503.