Highlights
• In a French case series of 896 patients with uveal melanoma, 23 carried a germline pathogenic or likely pathogenic MBD4 variant.
• This corresponded to an estimated relative risk of 31.44 (95% CI, 18.18-53.00; P < .001) compared with the non-Finnish European reference population in gnomAD.
• The findings strengthen the case for routine or broad germline MBD4 testing in patients with uveal melanoma, not only for risk stratification but also for genetic counseling and possible therapeutic implications.
• The study materially extends prior evidence and suggests that MBD4 is a clinically relevant hereditary susceptibility gene in uveal melanoma, at least in the French population.
Background and Clinical Context
Uveal melanoma is the most common primary intraocular malignancy in adults, but it remains a rare cancer overall. Its rarity has historically limited the power of germline studies, and most patients have been considered to have sporadic disease. Even so, the recognition of inherited susceptibility syndromes has grown steadily, especially in patients with early onset disease, multiple primary tumors, or a family history of melanoma or other cancers.
MBD4 has recently emerged as one of the most important hereditary candidates in this space. The gene encodes methyl-CpG binding domain protein 4, a DNA glycosylase involved in base-excision repair. In biologic terms, MBD4 helps correct deamination-induced mismatches at methylated CpG sites. Loss of function can therefore increase mutational burden and alter tumor biology, providing a plausible mechanistic bridge between inherited variants and oncogenesis.
Prior reports suggested that monoallelic germline pathogenic or likely pathogenic variants in MBD4 confer an increased risk of uveal melanoma, with an earlier estimate of approximately 9.15-fold increased risk. The present study by Le Ven and colleagues was designed to refine that estimate by expanding the number of tested patients in a real-world institutional series.
Study Design
Design and setting
This was a case series conducted at Curie Institute. Germline targeted sequencing of MBD4 was offered to every new patient with uveal melanoma from February 2021 to September 2025. The analysis included 896 individuals in total, including 319 patients who had been evaluated previously.
Population and comparator
The case population comprised patients with uveal melanoma seen at a French tertiary cancer center. For comparison, the investigators used non-Finnish European participants from the Genome Aggregation Database (gnomAD) as the reference population. This is a practical approach when estimating enrichment of rare germline variants in a disease cohort, although it does introduce some caveats regarding ancestry matching, ascertainment, and phenotypic characterization of controls.
Exposure and endpoint
The exposure of interest was the presence of a monoallelic germline pathogenic or likely pathogenic MBD4 variant. The main outcome was prevalence of these variants in the uveal melanoma cohort and the relative risk of developing uveal melanoma compared with the reference population.
Key Results
Among 896 patients with uveal melanoma, 23 were found to carry a germline pathogenic or likely pathogenic MBD4 variant. This prevalence is notable for a rare ocular cancer and substantially higher than would be expected in the general population.
Using the non-Finnish European gnomAD population as a comparator, the estimated relative risk of developing uveal melanoma among carriers was 31.44, with a 95% confidence interval of 18.18 to 53.00. The association was highly statistically significant on a 2-sided Fisher exact test (P < .001).
Two aspects of this result deserve emphasis. First, the point estimate is materially higher than the previously reported 9.15-fold increase. Second, even the lower bound of the confidence interval remains large, supporting a robust association. Although risk estimates derived from case-enrichment studies should not be interpreted identically to those from prospective population cohorts, the signal here is strong enough to be clinically meaningful.
From a prevalence standpoint, 23 of 896 corresponds to roughly 2.6% of the tested cohort. In practical terms, that means about 1 in 39 unselected patients in this institutional series carried a clinically actionable germline MBD4 finding. For a rare inherited predisposition in a rare cancer, this is a substantial yield.
Why These Findings Matter Clinically
Implications for germline testing
The study supports a broad rather than highly restrictive testing strategy for MBD4 in patients with uveal melanoma. If variant prevalence had been exceedingly low, testing might reasonably be confined to patients with suggestive personal or family histories. Here, however, the detection rate appears high enough to justify routine consideration of germline evaluation, particularly in centers managing ocular oncology patients.
This matters because identifying a hereditary predisposition affects more than etiologic labeling. It can shape post-diagnosis counseling, cascade testing for relatives, and surveillance considerations for associated malignancies depending on the evolving phenotype linked to MBD4 pathogenic variants. It may also help explain unusual tumor biology and influence eligibility for research protocols.
Potential therapeutic relevance
The authors specifically note therapeutic implications. While the abstract does not detail treatment outcomes or response biomarkers, this point is biologically credible. DNA repair defects can alter mutational processes, and MBD4-deficient tumors have been associated in prior literature with a hypermutator phenotype in some contexts. In a disease such as uveal melanoma, where systemic treatment options remain limited and outcomes in metastatic disease are often poor, any inherited marker with possible therapeutic relevance deserves attention.
At present, clinicians should be careful not to overstate immediate treatment consequences from germline MBD4 status alone. Still, the finding strengthens the rationale for integrating germline testing into multidisciplinary care and for studying whether MBD4-associated tumors show distinct sensitivity to immune or DNA damage-related therapeutic approaches.
Implications for family counseling
A positive MBD4 result may have consequences for relatives who could themselves be carriers. Genetic counseling should therefore be embedded in the testing pathway. Counseling should address variant interpretation, uncertainty about penetrance outside uveal melanoma, psychosocial implications, and the current limits of evidence regarding optimal surveillance for unaffected carriers.
Biological Plausibility
MBD4 is involved in the repair of G:T and G:U mismatches arising from spontaneous deamination of methylated cytosine. When this repair pathway is impaired, mutational accumulation can increase in a characteristic context. This makes MBD4 a plausible cancer predisposition gene on mechanistic grounds, not merely a statistical association.
For uveal melanoma specifically, the biologic story is compelling because this tumor type has traditionally been defined by a relatively narrow set of recurrent somatic alterations and by strong prognostic genomic correlates. Adding a germline DNA repair component helps refine the framework from one centered mostly on tumor prognosis toward one that also includes inherited susceptibility.
Strengths of the Study
The main strength is size. For a rare tumor, a cohort of 896 patients is substantial and provides a stronger basis for prevalence estimation than prior smaller studies. The systematic offering of germline MBD4 testing to every new patient during the study period also reduces one common bias in hereditary cancer research: selective testing only in patients with suggestive histories.
Another strength is the clear clinical question. The investigators did not merely ask whether MBD4 can be found in isolated cases; they tested whether its prevalence is high enough to matter in everyday practice. The answer appears to be yes.
Limitations and Interpretation Cautions
Single-center design
The study was conducted at Curie Institute, a high-level referral center. Referral patterns may enrich for unusual, complex, or younger patients. That said, the strategy of offering testing to every new patient is reassuring and likely mitigates some selection bias.
Use of gnomAD as a reference population
gnomAD is a powerful resource for rare variant frequency estimation, but it is not a disease-free screening cohort. Participants come from multiple studies, and subtle ascertainment differences remain possible. Relative risk estimates based on case series versus population databases should therefore be interpreted as strong evidence of enrichment rather than exact prospective penetrance measures.
Population specificity
The authors conclude that MBD4 is an important predisposing gene in the French population. Whether the same prevalence and effect size apply in other ancestries or healthcare systems remains uncertain. The reference population was restricted to non-Finnish Europeans, which improves internal consistency but limits broad extrapolation.
Incomplete phenotype information in the abstract
The abstract does not provide age at diagnosis, family history patterns, co-occurring cancers, or tumor genomic features among MBD4 carriers. Those details would be clinically valuable because they could help determine whether some patients derive especially high diagnostic yield from testing or whether the phenotype is sufficiently broad to support truly universal screening.
Practical Takeaways for Clinicians
For ophthalmologists, ocular oncologists, medical oncologists, and genetic counselors, this study supports incorporating MBD4 into the standard hereditary workup for uveal melanoma. If a center does not yet have a dedicated pathway, a reasonable next step is to include MBD4 in germline panels offered to patients with newly diagnosed uveal melanoma, ideally with pretest or posttest counseling.
For pathology and molecular teams, the study argues for alignment between somatic and germline testing programs. When tumor features suggest DNA repair defects or unusual mutational signatures, there may be added value in reviewing whether germline MBD4 testing has been performed.
For health systems and policy planners, the data raise a practical question: should broad MBD4 screening be considered standard of care in uveal melanoma? The answer will depend on local resources, reimbursement structures, and the strength of downstream management recommendations, but the diagnostic yield reported here is difficult to ignore.
Research Priorities
Several questions remain open. Prospective multi-institutional and multiethnic studies are needed to refine penetrance estimates and to determine how risk varies by variant class and ancestry. It will also be important to define the broader cancer spectrum associated with monoallelic MBD4 pathogenic variants and to establish evidence-based surveillance recommendations for carriers.
On the therapeutic side, translational studies should assess whether MBD4-associated uveal melanomas have distinct mutational signatures, immune profiles, or drug vulnerabilities. If so, MBD4 testing could evolve from a counseling tool into a marker with direct treatment relevance.
Conclusion
This study provides strong confirmatory evidence that germline pathogenic or likely pathogenic MBD4 variants are enriched in patients with uveal melanoma and confer a markedly increased relative risk compared with the general population reference. In this French series, approximately 2.6% of patients were carriers, and the estimated relative risk reached 31.44. For a rare ocular cancer, that is a clinically significant signal.
The implications are immediate: MBD4 should now be regarded as an important hereditary predisposition gene in uveal melanoma, with consequences for patient testing pathways, family counseling, and potentially future therapeutic stratification. The next phase of work should focus on external validation across populations, penetrance estimation, and translation of this genetic insight into measurable patient benefit.
Funding and ClinicalTrials.gov
The abstract provided does not report funding details. No ClinicalTrials.gov registration number is reported in the abstract, which is consistent with an observational case series rather than an interventional trial.
References
1. Le Ven A, Villy MC, Le Mentec M, Houy A, Dubois d’Enghien C, Matet A, Malaise D, Bubien V, Lortholary A, Ait Omar A, Stoppa-Lyonnet D, Cassoux N, Rodrigues M, Golmard L, Stern MH, Colas C. Prevalence of the Predisposing Gene MBD4 for Uveal Melanoma. JAMA Ophthalmology. Published online April 30, 2026. PMID: 42060275.
2. Lek M, Karczewski KJ, Minikel EV, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285-291.
