Immune Gene Variant HLA-C*04 Linked to Poorer Survival After Liver Cancer Surgery in Hawaii

A single immune-system gene variant may shape how long patients live after surgery for hepatocellular carcinoma. A retrospective study of 84 patients treated surgically in Hawaii between 2012 and 2024 found that carrying at least one copy of the human leukocyte antigen allotype HLA-C*04 was independently associated with worse postoperative overall survival, alongside established clinical factors such as larger tumor size, higher MELD score, and a history of other cancers. The finding, published in Frontiers in Genetics, is explicitly labeled exploratory by its authors, but it points toward an immuno-genomic dimension of liver cancer outcomes that has rarely been examined in this ethnically distinctive population.

Hepatocellular carcinoma is a formidable global health problem. It ranks as the sixth most commonly diagnosed cancer and the third most common cause of cancer death worldwide, and it is endemic in Asian and Pacific Islander populations, groups that together make up the majority of the Hawaiian cohort studied. While systemic and locoregional therapies for advanced disease have improved in recent years, cure rates remain low, and only a minority of patients are diagnosed early enough to qualify for surgical treatment, whether orthotopic liver transplantation or partial hepatectomy. Even among those who undergo curative-intent surgery, recurrence is common, and the preoperative factors that determine survival after surgery remain largely unknown.

The researchers, based at the University of Hawai’i Cancer Center and collaborating institutions, drew on a prospectively maintained single-institution database to identify adults with histologically confirmed hepatocellular carcinoma who underwent surgery between 2012 and 2024. The cohort comprised 43 transplant recipients, whose HLA typing was already performed as part of standard pre-transplant workup, and 41 partial hepatectomy patients, whose HLA allotypes were determined from banked peripheral blood mononuclear cells using a commercial next-generation sequencing genotyping protocol. All patients had complete HLA and clinical follow-up data, and the primary endpoint was postoperative overall survival, measured from the date of surgery to death or last contact.

The cohort reflected the clinical profile of an early-stage surgical population. The median age was 62 years, 20 percent of patients were female, and self-identified ethnicity included 50 percent Asian, 10 percent Pacific Islander, and 19 percent Caucasian. Nearly 70 percent of the 81 patients with available staging fell into BCLC stage 0 or A, the earliest categories, and the median largest tumor measured 3 centimeters. The underlying causes of liver disease were diverse: 42 percent of patients had hepatitis C, 38 percent had hepatitis B, 19 percent had NASH or NAFLD, and 37 percent had a history of alcohol use. Cirrhosis was present in 65 percent, the median MELD score at diagnosis was 8, and the median alpha-fetoprotein level was 9.5 ng/mL. Median overall survival reached 1,667 days after surgery, with 18 deaths recorded over the follow-up period and several patients surviving more than a decade.

To analyze survival, the team used Firth’s penalized Cox proportional hazards regression, a method suited to small samples with sparse events, supplemented by Kaplan-Meier curves and log-rank tests. They screened a wide panel of HLA allotypes across the A, B, C, DR, and DQ loci, classifying patients as positive if they carried at least one copy of a given allele. In univariate analysis, a history of non-HCC primary cancer and higher MELD score were significantly associated with reduced survival, while HLA-C*04 approached significance with a hazard ratio of 2.62. Tumor size over 3 centimeters and alcohol use hovered near the threshold as well.

The multivariable model sharpened the picture. After adjusting for sex, treatment type, cirrhosis, alcohol use, and hyperlipidemia, four variables emerged as independently associated with worse postoperative overall survival: a history of other malignancies (hazard ratio 5.72), a largest tumor exceeding 3 centimeters (hazard ratio 5.34), increasing MELD score (hazard ratio 1.24 per point), and the presence of at least one germline copy of HLA-C*04 (hazard ratio 4.08). No other HLA allotype in the model showed a significant association. The authors note that tumor size, secondary cancers, and MELD score have each been linked to poorer outcomes in previous studies of surgically treated hepatocellular carcinoma, lending plausibility to the model, while the HLA-C*04 signal represents a novel observation for this cancer.

HLA molecules are the immune system’s display platforms. Encoded by some of the most genetically diverse genes in the human genome, they present peptide fragments derived from proteins inside cells to T cells, triggering antigen-specific adaptive immune responses. Because different HLA allotypes bind different repertoires of peptides, inherited variation in these genes can influence both susceptibility to disease and response to therapy. HLA allotypes also cluster by geography and ancestry, which means that HLA-outcome correlations observed in one population may not translate to another, a consideration especially relevant in Hawaii, where the population blends Asian, Pacific Islander, Caucasian, and mixed ancestries.

Why might HLA-C*04 matter in liver cancer? The evidence base is thin but intriguing. HLA-C*04:01 is among the most highly expressed HLA alleles in normal breast and lung tissue, and laboratory work has shown that HLA-C*04 can load and present the GPC3 tumor-associated peptide, a well-known target in hepatocellular carcinoma, inducing an antigen-specific T cell reaction. A prior study in head and neck squamous cell carcinoma found that patients carrying a germline copy of HLA-C*04 had reduced survival, though no group had previously connected the allele to hepatocellular carcinoma outcomes. In the Hawaiian cohort, the 20 patients positive for HLA-C*04, ten transplant and ten resection patients, including three homozygotes, showed higher rates of NASH/NAFLD, hypertension, and hyperlipidemia, the last reaching statistical significance, alongside lower rates of hepatitis C and of multiple tumors. Four of the eight Pacific Islanders in the study carried the allele.

The authors are careful to frame the study as hypothesis-generating rather than definitive. The sample size was limited to a single institution’s experience, and the analysis of numerous HLA allotypes was exploratory. Critically, when the researchers applied Benjamini-Hochberg false discovery rate correction for multiple hypothesis testing, no variable, including HLA-C*04, remained statistically associated with survival, underscoring the risk of false discovery in a study of this scale. Other limitations include the possibility that half the cohort, being immunosuppressed after transplantation, experienced differential effects of HLA-mediated immune control of the cancer, although interaction analysis found that treatment type did not modify the HLA-C*04 association. The study also did not measure protein-level HLA expression, which is known to fall as a mechanism of immune escape in hepatocellular carcinoma, and overall survival over long follow-up may not isolate immune-mediated disease control; the authors suggest that future studies consider recurrence-free or HCC-specific survival as endpoints.

Despite these caveats, the work opens a line of inquiry with practical implications. If HLA allotypes do shape postoperative survival in early-stage hepatocellular carcinoma, they could eventually inform more personalized discussions of the risks and benefits of surgery and guide the development of next-generation immunotherapies that exploit HLA-restricted tumor antigen presentation. The authors call for future studies to test whether HLA allotypes correlate with survival beyond early-stage, predominantly Asian and Pacific Islander populations, and to probe the immunologic mechanisms that might connect HLA-C*04 to outcomes. For now, the message is measured: in one of the most ethnically distinctive surgical liver cancer populations in the world, a common immune gene variant has emerged as a candidate prognostic marker that deserves rigorous confirmation.

Subject of Research: Association of the HLA-C*04 immune allotype with post-surgical survival in hepatocellular carcinoma patients in Hawaii

Article Title: HLA-C*04 and hepatocellular carcinoma post-surgical survival in Hawaii

Article References: Abdul-Ghani, S., Brenner, N., Hernandez, B. Y., Wong, L., Ahn, H. J., & Green, B. L. (2026). HLA-C*04 and hepatocellular carcinoma post-surgical survival in Hawaii. Frontiers in Genetics, 17, Article 1874308. https://doi.org/10.3389/fgene.2026.1874308

Image Credits: AI Generated

DOI: 10.3389/fgene.2026.1874308

Keywords: hepatocellular carcinoma, HLA-C*04, liver transplantation, partial hepatectomy, overall survival, immunogenetics, Hawaii, MELD score, Cox proportional hazards, tumor immunology, Pacific Islander health, cancer genetics