DNA sequencing in familial hypercholesterolaemia: the next generation

This editorial refers to ‘Next-generation sequencing to confirm clinical familial hypercholesterolemia’, by L.F. Reeskamp et al., pp. 875–883. Familial hypercholesterolaemia (FH) is a common disorder with serious public health consequences, characterized by elevated plasma low-density lipoprotein (LDL) cholesterol.1 Heterozygous FH affects 1 in 250–300 adults in most populations,1,2 and usually results from an inherited pathogenic rare variant—or mutation—in the LDLR, APOB, or PCSK9 genes.1 Most FH patients are undiagnosed.1 This care gap is a lost opportunity for prevention, since FH patients bear a high lifetime risk of atherosclerotic cardiovascular disease (ASCVD). For instance, the risk of premature ASCVD for someone with LDL cholesterol >5 mmol/L and a pathogenic FH mutation is increased by 22-fold compared with a mutation-negative person with LDL cholesterol <3.4 mmol/L.3 Similar increments in risk have been reported among other mutation-positive patients with elevated LDL cholesterol.4 Early statin treatment can reduce lifetime ASCVD risk to near normal levels.1 Furthermore, positive diagnosis of FH should trigger family cascade screening, since half of first-degree relatives will also be affected.1 But what is the best way to diagnose FH patients? Diagnostic aids for FH, e.g. the Dutch Lipid Clinic Network (DLCN) algorithm, rely primarily on physical findings, personal and family history, together with LDL cholesterol level bundled into a score for ‘possible’, ‘probable’, or ‘definite’ FH status.5 There is also the option of entering genetic test results, but until recently DNA testing for FH was restricted to academic research laboratories. Furthermore, traditional Sanger sequencing of DNA was costly and laborious, and would miss certain types of pathogenic mutations.6 However, breakthroughs in next-generation DNA sequencing (NGS) technology now permit rapid, high-throughput, and cost-effective genetic analysis.6 Today, patients with suspected FH can have testing by a targeted NGS panel to scan key regions of not just the three major FH-causing genes, but also ‘minor’ FH genes, such as LDLRAP1, ABCG5, ABCG8, APOE, and LIPA, which occasionally cause an FH-like phenotype.7 In theory, NGS technology should improve the diagnostic yield and improve patient care. The lipidology community in the Netherlands has historically punched above its weight in the FH field. The Dutch registry system is unrivalled and serves as a role model for care of FH patients and families. DNA analysis has been part of the FH diagnostic process for >20 years in Holland. Initially, this was accomplished with Sanger sequencing, but in the past 5 years, this was superseded by targeted NGS. These remarkable resources enable testing of practical hypotheses of DNA analysis in FH patients. For instance, in this issue of European Journal of Preventive Cardiology, Reeskamp et al.8 compared the yield of mutation-positive Dutch patients using a targeted NGS panel to the mutation detection in the past using Sanger sequencing. The results were intriguing. Between 1999 and 2015, ∼30% of patients referred for genetic testing for FH had a pathogenic DNA variant found by Sanger sequencing. Since May 2016, NGS was used to study 1528 patients with at least possible FH based primarily on LDL cholesterol >5 mmol/L. The authors found that 227 patients—or 14.9%—were FH mutation-positive. Patients had many mutation types primarily in the LDLR (∼80% of all mutations), APOB, and PCSK9 genes. There were also clinical differences according to the mutation type, e.g. heterozygotes for large copy number variants involving the LDLR gene had higher LDL cholesterol levels than those with other mutations. Notably, 4.8% had a heterozygous variant in one of the five minor genes. Finally, in patients <44 years old, detection of mutation-positive cases decreased from 40.4% in 1999 to 19.5% in 2018. This informative study confirms that a priori clinical suspicion of FH increases the positive yield from NGS-based testing.4 For instance, among unselected ASCVD patients with LDL cholesterol >4.9 mmol/L from the general population, only ∼2% were FH mutation-positive.3 Reeskamp et al.8 documented a 15% yield of mutation-positive cases, once the diagnosis of FH was considered, which reflects enrichment compared with the general population. Furthermore, subgroups with probable or definite FH—driven largely by higher LDL cholesterol levels—were ∼50% mutation-positive, emphasizing the importance of pre-test likelihood in making this diagnosis. We have reported similar yields of mutation-positive cases in our Canadian FH cohort,8 and in addition observed >90% mutation-positivity among patients with LDL cholesterol level >8 mmol/L. But why has the yield of mutation-positive cases decreased by half in the Netherlands since 1999? This is unlikely related to technical issues, since if anything NGS is more accurate than Sanger sequencing. The authors suggest that the referring clinicians currently adhere less strictly than in the past to criteria when sending samples for confirmatory NGS analysis. Also, it is possible that the pool of contemporary Dutch hypercholesterolaemia patients has been depleted of monogenic FH cases, who were systematically referred more frequently during the Sanger sequencing era. What are the implications for the clinician who suspects their patient might have FH? We believe that if timely, cost-effective, reimbursed, accredited NGS-based DNA testing is available, it is worth discussing with the patient.9 Critically, given ongoing unresolved technical issues regarding mutation pathogenicity and test interpretation, as well as potential ethical implications for family members and insurability, an experienced genetic counsellor should ideally be available to discuss the appropriateness of NGS-based DNA testing for FH.10 Once a pathogenic variant is found, the potential benefits include: (i) enabling a definitive FH diagnosis; (ii) increased motivation for family cascade screening; (iii) eased access in some jurisdictions to newer medications such as PCSK9 inhibitors; and (iv) evidence for improved adherence to medical advice.1 However, NGS-based genotyping is not essential to guide appropriate intervention advice, since this can also be provided based on clinical and biochemical grounds alone.11 For instance, while odds for ASCVD events compared with normolipidaemic subjects are 22-fold increased in mutation-positive patients with LDL cholesterol >5 mmol/L, the odds are still six-fold increased in mutation-negative patients with LDL cholesterol >5 mmol/L.3 This risk is sufficiently high that the benefits of preventive treatment should be discussed. Furthermore, ∼25% of first-degree relatives of such patients have hypercholesterolaemia,11 which would support biochemical cascade screening. Many mutation-negative individuals with LDL cholesterol >5 mmol/L have polygenic susceptibility7; however, scientific and practical limitations preclude recommending routine testing of polygenic hypercholesterolaemia at present.10 Finally, we anticipate that machine-learning approaches will enhance the clinical utility of the DLCN algorithm and others.12 The opinions expressed in this article are not necessarily those of the Editors of the European Journal of Preventive Cardiology or of the European Society of Cardiology. J.L. is supported by the Canadian Institutes of Health Research (Doctoral Research Award) and the Schulich School of Medicine and Dentistry (Cobban Student Award in Heart and Stroke Research). R.A.H. is supported by the Jacob J. Wolfe Distinguished Medical Research Chair, the Edith Schulich Vinet Research Chair in Human Genetics, and the Martha G. Blackburn Chair in Cardiovascular Research. R.A.H. has also received operating grants from the Canadian Institutes of Health Research (Foundation award), the Heart and Stroke Foundation of Ontario (G-18-0022147). Conflict of interest: R.A.H. reports consulting fees from Acasti, Aegerion, Akcea/Ionis, Amgen, HLS Therapeutics Novartis, Regeneron, and Sanofi. J.L. has no disclosures.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.