When One Gene Means Many Diseases: Multi-Disorder Genes Drive Lab Disagreement on Variant Pathogenicity

When two clinical genetics laboratories look at the same DNA variant, they do not always reach the same conclusion about whether it causes disease. A new national interlaboratory study published in Genome Medicine suggests that a surprisingly specific culprit lies behind much of this persistent disagreement: genes that are associated with more than one disorder. The study, led by Kuo Zhang and colleagues at the National Center for Clinical Laboratories in Beijing together with collaborators at Shandong University and Tongji University, found that laboratories frequently transferred pathogenic assertions from one disease context to another without adequately considering disease mechanism, inheritance pattern, or disorder-specific validity, producing clinically meaningful discordance that could alter patient management.

The research team conducted a national proficiency-testing exercise involving 128 Chinese clinical genetic laboratories. Each laboratory classified a dataset of 13 germline variants distributed across five clinical scenarios, following the standards of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology, commonly abbreviated as ACMG/AMP. Under this framework, variants are assigned to five categories: pathogenic, likely pathogenic, variant of uncertain significance, likely benign, or benign. Accurate classification requires not only correct application of the evidence criteria but also assignment of pathogenicity to a specific disease entity and a defined mode of inheritance, whether autosomal dominant, autosomal recessive, or semi-dominant.

The headline result was sobering. Full concordance across participating laboratories was achieved for only 6 of the 13 variants, while the remaining 7 showed cross-category discordance that the authors describe as clinically meaningful. In other words, laboratories did not merely disagree on nuances of wording; they placed the same variant in different pathogenicity categories, the kind of discrepancy that can change whether a patient is diagnosed, whether family members are offered testing, and whether specific surveillance or treatment is recommended.

Dissecting the sources of residual disagreement, the authors identified several contributing factors. Some laboratories omitted required ACMG/AMP evidence criteria altogether or applied them inappropriately. Others assigned inconsistent strengths to the criteria they did use, a known difficulty since the framework allows evidence to be weighted at different levels such as supporting, moderate, strong, or very strong. But the most prominent and conceptually interesting pattern of discordance involved disease-specific interpretation of variants in genes associated with multiple disorders, a problem the study places at the center of the interlaboratory discordance debate.

The clearest examples came from variants in four genes: APOB, LZTR1, MYH7, and EPCAM. Each of these genes is linked to more than one clinical condition with distinct mechanisms and inheritance patterns. APOB variants, for instance, can cause familial hypobetalipoproteinemia in some contexts and contribute to hypercholesterolaemia in others, and the pathogenicity of a given variant may differ between these entities. MYH7 is classically associated with hypertrophic cardiomyopathy but also with dilated cardiomyopathy, and a loss-of-function variant that is pathogenic for one condition may behave very differently for the other. LZTR1 is associated with schwannomatosis type 2 among other disorders, and EPCAM variants cause constitutional mismatch deficiency through a mechanism involving the gene’s silencing rather than simple protein disruption. The study found that laboratories frequently carried a pathogenic assertion established for one disorder into another disorder context without re-evaluating whether the underlying evidence actually supported pathogenicity for the disease at hand.

This transfer of pathogenicity across disease boundaries is more than a technical quibble. The ACMG/AMP framework is explicitly disease-anchored: a variant is classified as pathogenic or benign for a specific disorder with a specific mode of inheritance. A variant that is convincingly pathogenic for an autosomal recessive condition may be entirely benign with respect to a dominant disorder involving the same gene, and vice versa. When laboratories ignore this distinction, the resulting report can imply management recommendations that are not supported by the evidence, for example triggering cardiac surveillance for a variant relevant only to a recessive metabolic phenotype, or conversely offering false reassurance in a context where the variant is genuinely dangerous.

Instructively, the study also included a positive control of sorts. Two variants in the GJB2 gene, which is associated with a single well-characterized disorder, nonsyndromic hearing loss, showed substantially higher categorical concordance among the participating laboratories. GJB2 has a well-established, essentially one-to-one relationship between genotype and phenotype, and its variant interpretation has been refined over decades of clinical practice. The contrast between the high concordance for GJB2 variants and the discordance for multi-disease gene variants suggests that the problem is not simply laboratory sloppiness or lack of expertise, but a structural gap in how the classification framework and its supporting resources handle genes with pleiotropic effects.

The findings arrive at a moment when the genomics community is investing heavily in harmonization. Variant curation expert panels and the ClinGen initiative have produced gene- and disease-specific specifications for applying ACMG/AMP criteria, and external quality assessment schemes have driven measurable improvements in criterion-level agreement. The authors acknowledge this progress, noting that criterion-level harmonization has improved, yet their data demonstrate that residual discordance persists precisely where harmonization efforts have not yet reached: the disease-specific layer of interpretation. A laboratory may apply every criterion correctly at the level of individual evidence items and still reach the wrong category if it anchors the classification to the wrong disease entity.

The authors argue that the solution requires extending harmonization beyond gene-level pathogenicity assessment toward mechanism-aware, disease-anchored classification frameworks and reporting practices. In practical terms, this means that variant databases, clinical reports, and curation guidelines should explicitly state which disorder a pathogenicity assertion applies to, and that evidence such as loss-of-function status, population frequency thresholds, and segregation data should be evaluated in the context of the specific disease mechanism. It also means that proficiency testing programs should deliberately include variants in multi-disease genes, since these are exactly the cases where laboratories are most likely to diverge and where educational feedback is most valuable.

For patients, the stakes are concrete. A variant of uncertain significance in a multi-disease gene may be reclassified as pathogenic in one laboratory and likely benign in another, with downstream consequences for reproductive planning, cascade testing of relatives, and clinical monitoring. As exome and genome sequencing becomes a routine part of medical care worldwide, the number of variants interpreted in multi-disease genes will only grow. This study provides a clear empirical demonstration that the next frontier of standardization in clinical genetics is not the individual evidence criterion but the disease context in which those criteria are applied, and it offers a roadmap for the expert panels, laboratories, and reporting standards that must now close the gap.

Subject of Research: Interlaboratory discordance in germline variant classification for genes associated with multiple disorders

Article Title: Variant classification challenges in genes associated with multiple disorders contribute to the persistent interlaboratory discordance

Article References: Zhang, K., Li, J., Wang, W., Zhou, Z., Ji, W., Gao, Y., Zhang, J., & Lin, G. (2026). Variant classification challenges in genes associated with multiple disorders contribute to the persistent interlaboratory discordance. Genome Medicine. https://doi.org/10.1186/s13073-026-01787-8

Image Credits: AI Generated

DOI: 10.1186/s13073-026-01787-8

Keywords: genetics, variant classification, ACMG/AMP, clinical genetics, multi-disease genes, interlaboratory concordance, Genome Medicine, precision medicine, Variant, classification, challenges, genes