Chronic hepatitis B is the leading cause of liver cancer worldwide, but most people with it never develop the disease. Two independent loci, found in two separate studies, each separate who does from who doesn't — one confirmed six times over in more than 3,700 people, the other backed by lab experiments pointing to a tumour-suppressor gene.
Solid lines are connections this site curates. Dashed lines mean the two ends share a research paper — worth knowing, and not a claim that one explains the other.
Hepatocellular carcinoma is the main form of primary liver cancer, and chronic infection with hepatitis B virus (HBV) is its leading cause worldwide. But most people living with chronic HBV — the great majority — never develop liver cancer. That gap is what makes a narrower genetic question worth asking: not what causes liver cancer in general, but among people already chronically infected, what separates those who go on to develop it from those who do not.
The study behind this page asked exactly that question in 355 chronic HBV carriers with hepatocellular carcinoma against 360 chronic HBV carriers without it, all of Chinese ancestry. It found one clear answer: rs17401966, an intronic variant in KIF1B on chromosome 1p36.22.
What sets this finding apart is the scale of its confirmation. The association was tested and confirmed across five further, independent samples — 1,962 additional cases, 1,430 controls, and 159 family trios. Across all six studies combined, the joint odds ratio was 0.61, meaning the variant is protective — carriers have lower odds of developing hepatocellular carcinoma than non-carriers, among people already chronically infected with HBV. That degree of independent replication, six studies rather than one, is unusually strong for a cancer-susceptibility locus, and it is what makes a single position worth its own page.
The region tagged by this variant also sits near two other plausible genes, UBE4B and PGD, and the study's authors note that pathways involving any of the three — KIF1B, UBE4B or PGD — could plausibly be involved in how this cancer develops. KIF1B itself is a motor protein gene with roles that have separately been linked to nerve cell survival and to other cancer types, which makes it a biologically plausible candidate, though this study's own evidence is about the DNA association, not about which of the three genes actually does the work.
A separate genome-wide association study, run in a two-stage design across 2,413 hepatocellular carcinoma cases and 2,794 HBV-positive controls from a high-incidence region of southern China, found a second answer at a different part of the genome: rs55718051, at the 4p14 locus near FAM114A1 (odds ratio 0.73, 95% CI 0.67–0.80).
This study went further than a DNA association. Laboratory experiments showed the variant sits in a region that regulates how much FAM114A1 protein a cell makes, likely through an interaction with another protein, FOXA1. When the researchers experimentally reduced FAM114A1 in liver cancer cells, the cells took on more cancer-like behaviour — evidence that FAM114A1 normally acts as a brake on tumour growth, and that losing it is what raises risk, rather than the variant adding something harmful. That functional follow-up is a different, more mechanistic kind of evidence than a DNA association alone provides.
This locus and the KIF1B locus above come from two separate genome-wide association studies, run in different cohorts and tested independently rather than jointly. Both point to genes plausibly involved in restraining tumour growth, which is a pattern worth noting — not, on its own, evidence that the two loci interact.
2026-06-25 · Multi-ancestry genome-wide association meta-analysis of hepatocellular carcinoma identifies 15 risk loci including MAP3K9, DHRS1, MTTP, and 8q24.21. HGG Advances. 2026. PMID:42357869
This site's own hepatocellular carcinoma (HCC) coverage is specific to HBV-related cases (one variant, UBE4B). This much larger study looked at all-cause HCC -- from any underlying liver disease, not only hepatitis B -- across 15 cohorts totaling 17,697 cases and 2,715,683 controls, finding 15 genome-wide significant germline loci including GCKR, MTTP, ADH5, MAP3K9, GABPB2 and 8q24.21 (nearest gene MYC), plus 2 more from transcriptome- and regulome-wide analysis. A key mechanistic finding: MAP3K9, TERT and GABPB2 variants act independently of cirrhosis, and MEN1, MYC and TERT are all part of the same beta-catenin pathway transactivation complex -- meaning inherited HCC risk substantially overlaps with the biological pathway cirrhosis itself drives, not a separate mechanism layered on top. Fine-mapping identified 11 likely causal coding variants, including a specific missense change in GCKR (p.Leu446Pro). Six loci, including variants in the HLA region, showed significantly different effects between East Asian and European ancestries -- a genuine ancestry-specific finding rather than a single universal risk profile. The study also found that population-level HCC genetic architecture overlaps with both fatty liver disease and viral liver disease, and that genetic predisposition to lower BMI is itself associated with higher HCC risk. This site's UBE4B variant on the HBV-HCC page is not among this study's 15 loci; the two findings describe different (though overlapping) patient populations, so this entry documents the broader all-cause study at the condition level rather than conflating it with the HBV-specific variant already on the page.
2026-01-01 · Integrating plasma proteomes with genome-wide association data for causal protein identification in hepatocellular carcinoma: A bidirectional Mendelian randomization study. Medicine. 2026. DOI:10.1097/md.0000000000047315
This study used two-sample Mendelian randomization to test whether specific blood proteins have a causal relationship with hepatocellular carcinoma (HCC) risk, combining published plasma-proteome GWAS (35,559 healthy people) with an HCC GWAS. It identified 17 proteins with a causal association at nominal significance (P<0.05): 16 positively associated with HCC risk (TMCC3, METTL1, SNRPF, KRT19, MED4, RFNG, IL26, NRXN1, MSH2, CLCA2, AKT2, CRYZL1, RDH16, CSF3, CPA4, COPS7B) and one, EPHA2, inversely associated. Pathway analysis implicated JAK-STAT, PI3K-AKT and chemokine signaling. Two important caveats the authors themselves note: the underlying HCC GWAS had only 168 cases, a small number for this kind of analysis, and the P<0.05 threshold used is nominal significance, not the much stricter genome-wide significance bar (P<5x10^-8) this site otherwise requires for a standalone variant page. The authors explicitly call for experimental validation before treating these as established causal factors. This site's HBV-related hepatocellular carcinoma page carries 1 variant; none of these 17 proteins are currently represented, and this entry is recorded with the sample-size and significance-threshold caveats intact rather than as confirmed findings.
This finding applies specifically to people already chronically infected with hepatitis B — it is a question about who, within that group, is more or less likely to develop liver cancer. It says nothing about liver cancer risk in someone without chronic HBV, and it is not a substitute for hepatitis B vaccination, which prevents the infection that makes this whole question relevant in the first place.
What actually manages this risk. Regular monitoring — liver ultrasound and blood tests, typically every six months — is the standard of care for people with chronic hepatitis B, specifically because it can catch hepatocellular carcinoma early, when treatment is most effective. That monitoring schedule is decided by a doctor based on established risk factors (viral load, liver damage, family history, age), not by this or any other single genotype.
This site's chronic hepatitis B page covers the genetics of who becomes chronically infected and how the infection itself progresses — a different, earlier question from the one this page addresses.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about HBV-Related Hepatocellular Carcinoma comes down to these specific, well-studied positions — not a diagnosis.
FAM114A1 · rs55718051
See detailed info →Databases, guidelines and references
This variant is protective on average, not a screening test, and regular liver monitoring — ultrasound and blood tests roughly every six months — is what actually catches hepatocellular carcinoma early in people with chronic HBV. That schedule is set by a doctor based on established risk factors, not on this genotype.
No. This finding is specifically about who, among people already chronically infected with HBV, is more or less likely to develop liver cancer. It says nothing about liver cancer risk in someone without chronic hepatitis B infection.
Most genome-wide findings are reported once and replicated, if at all, in one further sample. This one was tested across five independent samples beyond the original discovery, totalling well over 3,000 additional people, and held up across all of them — a much stronger form of evidence than a single study provides.
No, and it cannot. This finding is only relevant to people who are already chronically infected with hepatitis B. Vaccination prevents that infection from happening in the first place, which is a separate and earlier line of defence.
No — they are two separate genome-wide association studies, run in different southern Chinese cohorts, that each independently identified a different locus associated with HBV-related liver cancer risk. Both are listed here because they concern the same disease, not because the studies found them together.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.