Genetics research

How much prion disease is genetic, and what is the inheritance model?

Between 10 and 15 percent of prion disease cases are attributed to variants in PRNP, the single gene encoding prion protein, and familial forms are inherited in an autosomal dominant pattern, with some cases arising from new variants rather than an affected parent (MedlinePlus Genetics; Goldman and Vallabh 2022). In the surveillance cohorts analysed by Minikel and colleagues, 65 percent of cases had PRNP open reading frame sequencing, and 12 percent of all cases, or 18 percent of sequenced cases, carried a rare variant. The variant classes are missense substitutions, truncating variants, and insertions or deletions in the octapeptide repeat region (codons 50 to 90).

Goldman and Vallabh argue in their 2022 review that the historical subtype names (familial CJD, Gerstmann-Straussler-Scheinker syndrome, fatal familial insomnia) map less cleanly onto genotype than once assumed, and that the umbrella term genetic prion disease may serve clinicians better. They also note that healthy at-risk people may be candidates for future genetically targeted trials, which raises the stakes for access to counselling and testing.

What do population cohorts say about penetrance?

The reference analysis is Minikel and colleagues, Science Translational Medicine 2016, which compared 16,025 prion disease cases against 60,706 population control exomes (ExAC) and 531,575 23andMe research participants. Reportedly pathogenic PRNP missense variants were at least 30 times more common in controls than genetic prion disease prevalence would allow. The authors concluded that PRNP missense variants sit on a risk continuum rather than a causal versus benign dichotomy, with lifetime risks from under 0.1 percent to around 100 percent.

The headline groupings:

• P102L, A117V, D178N and E200K: absent from ExAC and on five or fewer alleles in 23andMe, consistent with up to 100 percent penetrance. These four account for more than half of genetic prion disease cases. The paper notes that survival-analysis estimates of E200K penetrance range from about 60 to 90 percent, which falls inside the confidence intervals.
• M232R, V180I and V210I: enriched in cases but too common in controls for a fully penetrant dominant variant. Best estimates of lifetime risk in heterozygotes were about 0.08 percent for M232R, about 1 percent for V180I and about 7.8 percent for V210I, correlating with the proportion of cases with a family history. Each appears mainly in one population (Japanese for M232R and V180I, Italian for V210I), so stratification adds uncertainty, though the authors argue it cannot explain the whole enrichment.
• Many other reported variants had weak prior evidence of pathogenicity, and the paper estimates that about 4 in 1,000 sporadic cases will carry a rare PRNP missense variant by chance that may be reported as causal.
• Heterozygous loss-of-function variants were found in three healthy older people, offered as the first human genetic data on a 50 percent reduction in PRNP dosage, which bears on the safety of PrP-lowering therapies.

The octapeptide repeat region could not be assessed in ExAC because 76 base pair reads could not resolve the 123 base pair repeat. See penetrance for how these numbers are derived and variants for variant-level detail.

Does genetic prion disease transmit at 50 percent?

A 2025 retrospective analysis from Araba University Hospital in Spain (Kortazar-Zubizarreta and others, European Journal of Neurology) examined 24 pedigrees, 65 nuclear families and 151 offspring of known genotype carrying D178N or E200K. Observed transmission was 67.1 percent for D178N (49 of 73; 95 percent CI 56.3 to 77.9) and 70.5 percent for E200K (55 of 78; 95 percent CI 60.4 to 80.6), 68.9 percent combined, against the Mendelian expectation of 50 percent, using generalised estimating equations to handle within-family correlation. The authors call meiotic drive the most plausible explanation and say the findings challenge the 50 percent assumption used in counselling, including embryo selection strategies.

The limits are substantial and stated: retrospective ascertainment through clinical services, about 18 percent inferred rather than tested genotypes, 71 percent power in the D178N paternal subgroup, and a geographically clustered Basque D178N population. The confidence intervals for maternal and paternal transmission overlap widely and the paper does not formally test them against each other. No replication in another population had been published at the date of this page.

What has genome-wide association found beyond PRNP?

Jones and colleagues (Lancet Neurology 2020) ran a two-stage GWAS in people of European ancestry with 5,208 sporadic CJD cases collected from 1990 to 2014. Forty-one SNPs reached genome-wide significance, with independent replication at three loci: PRNP rs1799990 (codon 129; OR 1.23, 95 percent CI 1.17 to 1.30 in the additive model), STX6 rs3747957 (OR 1.16, 1.10 to 1.22) and GAL3ST1 rs2267161 (OR 1.18, 1.12 to 1.25). The authors read the loci as pointing to intracellular trafficking and sphingolipid metabolism, and note that STX6 risk variants are shared with progressive supranuclear palsy.

Hummerich and colleagues (PLoS One 2024) then asked what influences clinical course. With duration data on 3,773 sporadic CJD cases and age at onset on 3,767, 53 SNPs reached significance for clinical duration, all on chromosome 20, with the top signal at codon 129 (rs1799990, p = 3.45 x 10^-36). No SNP reached significance for age at onset; HS6ST3 was significant in a gene-based test. Codon 129 therefore remains the main common modifier of both susceptibility and survival, and has its own page at codon 129.

What does genetics contribute to biomarkers and presymptomatic cohorts?

Presymptomatic carriers are a small but critical group for biomarker validation. Halbgebauer and colleagues (Neurology 2022) measured blood beta-synuclein and neurofilament light chain in 308 participants, including 129 with prion disease and 8 presymptomatic PRNP variant carriers; beta-synuclein separated prion disease from all other groups with an AUC of 0.97 (95 percent CI 0.94 to 0.99) against 0.91 (0.88 to 0.94) for NfL. Eight carriers is too few to say anything about preclinical change, which is the open question that larger carrier cohorts will need to answer.

How is predictive testing organised, and what is being studied around it?

Practice differs by country. The Australian National CJD Registry describes PRNP testing on DNA from peripheral blood or unfixed brain tissue, run by the Neurogenetics Unit at PathWest in Nedlands, entirely voluntary, with consultation with a genetic service recommended and no price or turnaround listed. It also recommends storing DNA from every suspected CJD patient so that relatives can consider testing later.

An Israeli expert panel (Shir and others, European Journal of Neurology 2026) published a four-step framework for asymptomatic relatives: initial counselling, psychological assessment, a blood sampling appointment and an in-person results session, with no less than one month from counselling to sampling, no testing under 18, careful consideration at 18 to 25, and a recommended delay of at least six months for people in acute distress after a relative’s diagnosis or death. The paper reports that asymptomatic tests at Tel Aviv Medical Center rose from 2 to 3 a year in 2014 to 2017 to 29 in 2024, and that no structured programme for ongoing support of carriers yet existed in Israel. E200K is relatively prevalent among people of Libyan and Tunisian Jewish descent, and the authors estimate asymptomatic heterozygotes in Israel in the hundreds to thousands.

In the UK, Brien and colleagues (Neurotherapeutics 2026) tested a single five-hour acceptance and commitment therapy workshop with a booster call in 29 asymptomatic adults at risk of inherited prion disease, recruited through the National Prion Monitoring Cohort and CJD Support Network. Feasibility criteria were met, 21 of 23 completers rated it acceptable, and psychological quality of life improved in the whole sample, but the trial was non-randomised, unregistered, small and almost entirely White British, so it is a feasibility signal only.

Reproductive options are documented back to a 2014 JAMA Neurology report (Uflacker and others) of in vitro fertilisation with preimplantation genetic diagnosis for a 27-year-old asymptomatic woman from a family with GSS due to F198S: six embryos without the variant were identified, two transferred, and twins born slightly premature were developing normally at 27 months. The authors describe it as the first such case for genetic prion disease.

Does a predictive result affect insurance?

In the UK, the Code on Genetic Testing and Insurance, an agreement between the Department of Health and Social Care and the Association of British Insurers, commits signatory insurers not to request or consider predictive genetic test results, with a single exception for Huntington’s disease when life cover exceeds 500,000 pounds. It was first published in October 2018, is reviewed every three years and was last updated in March 2026. Rules elsewhere differ and are outside this hub.

Open questions this hub tracks

• Replication of the above-50-percent transmission ratio outside the Basque and Spanish cohorts.
• Penetrance estimates for octapeptide repeat insertions, which population exome data could not resolve.
• Whether blood or CSF markers change before symptoms in carriers, which requires carrier cohorts far larger than eight.
• Functional work on STX6, GAL3ST1 and the chromosome 20 duration signal.