Familial Hypercholesterolemia Genomic Screening

Evidence

Familial hypercholesterolemia (FH) can result from pathogenic variants that disrupt low-density-lipoprotein metabolism. Two primary studies used sequencing to identify such variants, but their sampling frames were almost opposites. A July 2026 regional study retrospectively reviewed exome data collected for other clinical purposes, while a 2025 Egyptian study deliberately recruited people who already had premature coronary artery disease. The resulting carrier proportions describe those particular cohorts; they are not interchangeable estimates of population prevalence.

The July study analyzed exomes from 15,841 unrelated people at the Affiliated Women and Children's Hospital of Ningbo University. The datasets came from patients undergoing trio whole-exome sequencing and couples receiving comprehensive carrier screening between January 2022 and June 2025. Researchers examined LDLR, APOB, and PCSK9 as medically actionable secondary findings and reported variants classified as pathogenic or likely pathogenic under current ACMG criteria.

Across those exomes, the study identified 145 carriers, or 0.92%, and 46 distinct variant sites. LDLR variants accounted for 75.17% of carriers and 37 sites; APOB accounted for 22.76% and six sites; PCSK9 accounted for 2.07% and three sites. Three variants were described as previously unreported: one LDLR frameshift and two APOB frameshifts. The two carriers with those APOB variants had low rather than high LDL cholesterol, which the abstract describes as consistent with familial hypobetalipoproteinemia. That counterexample matters because finding a variant in an FH-associated gene is not enough by itself to infer one uniform lipid phenotype.

The Egyptian study began with a much more clinically enriched group. Investigators recruited 96 people with angiographically confirmed premature coronary artery disease from two tertiary hospitals in Cairo and Alexandria; 94 samples passed sequencing quality checks. They examined seven genes: LDLR, APOB, PCSK9, APOE, LDLRAP1, ABCG5, and ABCG8. The wider panel included recessive hypercholesterolemia genes and genetic conditions that can resemble FH, not only the three genes used for secondary findings in the regional exome study.

Fourteen of 94 successfully sequenced participants, or 14.89%, had pathogenic or likely pathogenic variants. Ten more, or 10.64%, had rare variants of uncertain significance, which were not counted as confirmed genetic diagnoses. Among the 14 positive cases, nine involved LDLR, three involved LDLRAP1, one involved PCSK9, and one involved APOB. Six participants had homozygous variants: three in LDLR and three in LDLRAP1. The authors noted that one-third of the recruited cohort reported consanguinity, but the small, selected design cannot establish how much that feature explains the genotype distribution.

Clinical classification and sequencing also did not identify identical groups. Thirty-eight of the 96 recruits met probable or definite Dutch Lipid Clinic Network criteria, yet only 11 of those 38 carried a pathogenic or likely pathogenic variant. With uncertain variants excluded, the full text reports 78.6% sensitivity, 67.14% specificity, 32.3% positive predictive value, and 94% negative predictive value for the probable-or-definite classification in this cohort. These estimates are study-specific because baseline lipid records were incomplete and many participants were already receiving lipid-lowering medication.

Analysis

The cross-study pattern is an ascertainment effect, not evidence that FH biology is fifteen times more common in one country than another. The broad regional review searched three actionable genes as secondary findings among people sequenced for mixed reasons; the Egyptian pilot started with premature coronary disease and tested seven genes, including LDLRAP1 and phenocopy genes. Enrichment at recruitment therefore raises the chance of finding a relevant variant before any population difference is considered. Panel breadth and classification rules further change what can be counted. The studies do converge on LDLR as the largest category, but the smaller Egyptian cohort contains proportionally more recessive and homozygous findings, while the regional cohort resolves a wider set of LDLR and APOB sites. The useful inference is that a genomic-screening yield is a property of the whole workflow—eligibility, phenotype, ancestry, gene list, sequencing coverage, and variant interpretation—not a portable property of the disease. A stronger comparison would apply the same panel and classification process prospectively to representative cohorts in both settings.

Limitations

The July paper was available to this briefing only as a PubMed abstract. Its cohort composition, age distribution, reasons for sequencing, laboratory pipeline, ancestry structure, validation procedures, and clinical follow-up cannot be independently assessed from the ingested evidence. Because the source population combined trio-exome cases with carrier-screening couples, the reported 0.92% should not be treated as a general-population prevalence estimate. The abstract also does not establish the functional effects of the three previously unreported variants beyond their classifications and observed lipid descriptions.

The Egyptian study was a 94-sequence pilot from two tertiary hospitals, used convenience sampling, and selected participants who already had premature coronary disease. The two hospital groups differed in sex, age at disease onset, diabetes, family history, and lipid measurements. Missing pretreatment LDL values required estimation for some clinical scores. The panel did not assess promoter variants, deep intronic variants, large rearrangements, or all possible lipid genes; family segregation was not performed for uncertain variants. Neither cross-sectional design proves that genomic screening itself changes outcomes, and neither supports an individual diagnosis or treatment decision.