Overestimation of Pathogenic Variants in Idiopathic Hypogonadotropic Hypogonadism and Kallmann Syndrome

When re-scored by ACMG standards, 26% of reported "causative" IHH/Kallmann variants were benign — and ClinVar confirmed only 18% as pathogenic

Journal: Endocrine Practice | Published: 2026-01-30 | Type: Review | PMID: 41621652 Authors: Grater L, Hawkins Z, Ben-Mahmoud A, Kim H-G, Layman LC (Medical College of Georgia / Rutgers / Qatar Biomedical Research Institute) Funding/COI: Not listed; authors declare no conflicts of interest

Summary

Idiopathic hypogonadotropic hypogonadism (IHH) and Kallmann syndrome are rare disorders in which GnRH signaling fails, causing absent or delayed puberty and infertility. The genetics literature has accumulated variants across 50+ genes labeled as causative — but this review finds the label is frequently wrong. When 933 variants from 273 publications were re-classified using standardized ACMG criteria, a quarter came back benign or likely benign, and another quarter were merely variants of uncertain significance.

Claims

• 273 publications yielded 933 variants reported as causative across 27 OMIM-established IHH/Kallmann genes
• VarSome reclassified 444/933 (47.6%) as pathogenic/likely pathogenic (P/LP), 249 (26.7%) as variants of uncertain significance (VUS), and 240 (25.7%) as benign/likely benign (B/LB)
• ClinVar was harsher: only 171/933 (18.3%) P/LP; 492/933 (52.7%) had no ClinVar entry at all
• 84.3% of VarSome-classified VUS were missense variants — the category most prone to misclassification
• 37.2% of VarSome P/LP calls were also missense, raising questions about how many of those will be downgraded as evidence accumulates
• Pre-2015 (pre-ACMG guideline) publications account for a disproportionate share of the overclaimed variants

Study Quality

This is a systematic literature review, not primary genetic data — it re-analyzed existing variant calls against current ACMG/AMP criteria using two independent databases (VarSome and ClinVar). The methodological core is sound: they searched OMIM-confirmed causative genes only, screened for variants explicitly labeled causative by the original authors, and applied the 2015 ACMG classification framework consistently. The pre/post-2015 stratification is a useful design choice that lets the authors separate legacy noise from more recent overclaiming.

The significant caveat is that VarSome and ClinVar disagreed substantially (47.6% vs. 18.3% P/LP rate), which the authors acknowledge but don't fully resolve. This discordance is itself informative — it shows the evidence base is genuinely contested — but it means readers shouldn't treat either number as definitive.

Red Flags

• No functional validation data: the review relies entirely on computational reclassification, not bench experiments confirming or refuting variant pathogenicity
• ClinVar's 52.7% "no entry" rate means the most troubling variants — those never scrutinized by the community — are invisible to this analysis
• Funding source not reported; for a paper challenging a large body of published genetics work, transparency matters
• The review covers only 27 OMIM-established genes, leaving out the wider set of 50+ genes in circulation — the problem may be even larger than reported
• Missense variant overcalling is a known systemic issue in genetics; this paper documents it but doesn't propose a fix beyond "use ACMG guidelines," which apparently wasn't preventing the problem

Strengths

• Large variant corpus: 933 variants across 273 publications is a substantive dataset for a rare disease
• Dual-database approach (VarSome + ClinVar) exposes the real-world inconsistency in how the field classifies these variants
• Pre/post-guideline stratification directly tests whether the field is improving — and gives a baseline for future audits
• Restricting scope to OMIM-confirmed genes makes the pathogenicity bar explicit and defensible
• No COI, which matters when the finding is that a lot of published genetics work is wrong

Verdict

This paper does something the genetics field needs more of: it takes a stack of published "causative" variants and asks whether they actually hold up. The answer — roughly one in four don't, and another one in four are uncertain — is a direct challenge to how IHH and Kallmann syndrome are being genetically characterized in the clinic. The ClinVar/VarSome discordance is uncomfortable but honest; the authors don't paper over it. The limitation is that computational reclassification is not the same as functional evidence, so this is a flag, not a verdict. Still, for a rare condition where a genetic diagnosis can drive major clinical decisions, the takeaway is clear: the literature has overclaimed pathogenicity, and clinicians relying on pre-2015 papers should treat those variant calls with skepticism.