Advertisement
Canadian Journal of Cardiology

Simplified Canadian Definition for Familial Hypercholesterolemia

Open AccessPublished:May 18, 2018DOI:https://doi.org/10.1016/j.cjca.2018.05.015

      Abstract

      Familial hypercholesterolemia (FH) is an autosomal codominant lipoprotein disorder characterized by elevated low-density lipoprotein cholesterol (LDL-C) and high risk of premature atherosclerotic cardiovascular disease. Definitions for FH rely on complex algorithms that are on the basis of levels of total or LDL-C, clinical features, family history, and DNA analysis that are often difficult to obtain. We propose a novel simplified definition for FH. Definite FH includes: (1) elevated LDL-C (≥ 8.50 mmol/L); or (2) LDL-C ≥ 5.0 mmol/L (for age 40 years or older; ≥ 4.0 mmol/L if age younger than 18 years; and ≥ 4.5 mmol/L if age is between 18 and 39 years) when associated with at least 1 of: (1) tendon xanthomas; or (2) causal DNA mutation in the LDLR, APOB, or PCSK9 genes in the proband or first-degree relative. Probable FH is defined as subjects with an elevated LDL-C (≥ 5.0 mmol/L) and the presence of premature atherosclerotic cardiovascular disease in the patient or a first-degree relative or an elevated LDL-C in a first-degree relative. LDL-C cut points were determined from a large database comprising > 3.3 million subjects. To compare the proposed definition with currently used algorithms (ie, the Simon Broome Register and Dutch Lipid Clinic Network), we performed concordance analyses in 5987 individuals from Canada. The new FH definition showed very good agreement compared with the Simon Broome Register and Dutch Lipid Clinic Network criteria (κ = 0.969 and 0.966, respectively). In conclusion, the proposed FH definition has diagnostic performance comparable to existing criteria, but adapted to the Canadian population, and will facilitate the diagnosis of FH patients.

      Résumé

      L’hypercholestérolémie familiale (HF) est une maladie autosomique codominante caractérisée par un taux élevé de cholestérol à lipoprotéines de faible densité (cholestérol LDL) et un risque élevé de maladie cardiovasculaire athéroscléreuse prématurée. Les définitions de l’HF reposent sur des algorithmes complexes basés sur les concentrations de cholestérol total ou de cholestérol LDL, les caractéristiques cliniques, les antécédents familiaux et les analyses de l’ADN souvent difficiles à obtenir. Nous proposons une nouvelle définition simplifiée de l’HF. Pour un diagnostic définitif d'HF, il faut : 1) un taux élevé de cholestérol LDL (≥ 8,50 mmol/l) ou 2) un taux de cholestérol LDL ≥ 5,0 mmol/l (pour les 40 ans et plus); ≥ 4,0 mmol/l (pour les moins de 18 ans); ≥ 4,5 mmol/l (de 18 à 39 ans) lorsque associé à au moins l’une des caractéristiques suivantes : 1) un xanthome tendineux; ou 2) une mutation causale de l’ADN observée dans les gènes LDLR, APOB ou PCSK9 chez le propositus ou les parents de premier degré. L’HF probable concerne les sujets qui ont un taux élevé de cholestérol LDL (≥ 5,0 mmol/l) et qui montrent la présence d’une maladie cardiovasculaire athéroscléreuse prématurée chez le patient ou les parents de premier degré, ou un taux élevé de cholestérol LDL chez les parents de premier degré. Les seuils de cholestérol LDL ont été déterminés à partir d’une importante banque de données qui regroupait > 3,3 millions de sujets. En vue de comparer la définition proposée aux algorithmes actuellement utilisés (c.-à-d. le registre de Simon Broome et le Dutch Lipid Clinic Network), nous avons réalisé les analyses de concordance chez 5987 individus du Canada. La nouvelle définition de l’HF concorde très bien avec les critères du registre Simon Broome et du Dutch Lipid Clinic Network (κ = 0,969 et 0,966, respectivement). En conclusion, la définition proposée de l’HF possède une performance diagnostique comparable aux critères existants, tout en s’adaptant à la population canadienne, et facilitera le diagnostic des patients atteints d’HF.
      Familial hypercholesterolemia (FH) is an autosomal dominant genetic lipoprotein disorder; the more common heterozygous form is characterized by a low-density lipoprotein cholesterol (LDL-C) > 95th percentile for age and sex within a family. Affected individuals might show clinical manifestations (premature corneal arcus, xanthomas, xanthelasmas), although these are seen less frequently in modern practice. FH is underdiagnosed and undertreated, in part because existing diagnostic criteria are complex and not widely used outside of specialty clinics. The most commonly used diagnostic algorithms for FH are the Simon Broome Register (SBR) and the Dutch Lipid Clinic Network (DLCN) criteria, which incorporate LDL-C, clinical signs, and family history of premature atherosclerotic cardiovascular disease (ASCVD) and an elevated LDL-C > 95th percentile in a first-degree relative to generate a score that leads to classification of either “definite,” “probable,” or “possible” FH (Supplemental Tables S1 and S2). Detection of a pathogenic DNA mutation in an FH-related gene in a proband leads to a diagnosis of “definite FH.” There are important limitations to the currently used algorithms: the clinical manifestations are infrequent; the baseline LDL-C (untreated) level is often unavailable because of the use of lipid-lowering therapies; and, family history is sometimes unavailable or unreliable. DNA testing is not readily available and not always concordant with the FH phenotype. Despite the complexities, diagnosis is important because untreated FH leads to premature ASCVD, whereas early identification and treatment can normalize risk.
      • Nordestgaard B.G.
      • Chapman M.J.
      • Humphries S.E.
      • et al.
      Familial hypercholesterolaemia is underdiagnosed and undertreated in the general population: guidance for clinicians to prevent coronary heart disease: consensus statement of the European Atherosclerosis Society.
      Heterozygous FH has a prevalence of approximately 1:250
      • Akioyamen L.E.
      • Genest J.
      • Shan S.D.
      • et al.
      Estimating the prevalence of heterozygous familial hypercholesterolaemia: a systematic review and meta-analysis.
      and might be higher in populations with founder effects, as in the province of Québec. The homozygous form is rare and constitutes an orphan disease. Age of onset of ASCVD can vary considerably in FH subjects and in addition to sex, depends on the severity of the mutation and other risk factors. The increase in ASCVD risk remains across a broad range of elevated LDL-C levels and is at least sixfold higher even in the absence of documented FH-causing mutations. Currently used criteria are difficult to use. We therefore propose to redefine FH on the basis of simplified criteria as a genetic condition characterized by marked elevations in LDL-C and risk of early onset ASCVD (Supplemental Table S3). We provide Canada-specific LDL-C cut points and a validated calculation for an imputed LDL-C, on the basis of the type and intensity of lipid-lowering therapy.
      • Ruel I.
      • Aljenedil S.
      • Sadri I.
      • et al.
      Imputation of baseline LDL cholesterol concentration in patients with familial hypercholesterolemia on statins or ezetimibe.

      Methods

      Results

      Screening criteria for FH

      The 95th percentile cut points for LDL-C are shown in Supplemental Figure S2; frequency distribution according to age and sex is shown in Supplemental Table S4. Overall, the 95th percentile for the population was 5.0 mmol/L in men and in women. The 95th percentile value for LDL-C in men younger than 18, 18-39, and older than 40 years were 3.67, 4.79, and 5.08 mmol/L, respectively. In women, these were 3.70, 4.27, and 5.18 mmol/L, respectively. We therefore selected the LDL-C cut points of ≥ 4.0 mmol/L for men and women younger than 18 years, ≥ 4.5 mmol/L for ages 18-39 years, and ≥ 5.0 mmol/L for subjects 40 years of age and older (Supplemental Table S3). These LDL-C levels constitute an obligatory major criterion for the diagnosis of FH and should be confirmed on repeat testing.
      Along with the DLCN criteria, an LDL-C ≥ 8.5 mmol/L has > 99% specificity for a diagnosis of FH in genetically confirmed patients.

      FH criteria: major

      Xanthomas, corneal arcus and xanthelasmas

      The prevalence of cutaneous manifestations of FH has decreased markedly in the statin era. In 268 new FH patients diagnosed according to the DLCN or SBR criteria examined in the Québec City Lipid clinic, CHU de Québec-Université Laval, and the Chicoutimi Hospital Lipid Clinic after 2012, only 20% had tendon xanthomas and none had premature corneal arcus or xanthelasmas (Supplemental Fig. S1). However, tendon xanthomas, which are highly specific of FH in subjects with genetic high LDL-C, are included in the DLCN and SBR criteria as a major clinical diagnostic criterion. Corneal arcus after age 45 and xanthelasma are not specific for FH and were not considered in the proposed definition of FH.

      DNA mutation

      The presence of a known pathogenic mutation in the LDLR, APOB, or PCSK9 genes is a major criterion for FH. The availability of next-generation sequencing now allows the rapid and unbiased molecular diagnosis of FH using exome sequencing of the LDLR, APOB, or PCSK9 and capture large insertion/deletion copy number variants in the LDLR gene. The FH diagnostic algorithm is shown in Figure 1. We do not recommend nor mandate DNA analysis systematically for all patients.
      Figure thumbnail gr1
      Figure 1Canadian definition for the clinical diagnosis of familial hypercholesterolemia (FH). ASCVD, atherosclerotic cardiovascular disease; LDL-C, low-density lipoprotein cholesterol. * Secondary causes of high LDL-C should be ruled out (severe or untreated hypothyroidism, nephrotic syndrome, hepatic disease [biliary cirrhosis], medication, especially antiretroviral agents); LDL-C ≥ 4.0 mmol/L for age younger than 18 years; and LDL-C ≥ 4.5 mmol/L for age 18 years to younger than 40 years. ** Causal DNA mutation refers to the presence of a known FH-causing variant in the LDLR, APOB, or PCSK9 gene on the basis of the presence of the variant in ClinVar, The Human Gene Mutation Database (HGMD), or Western Database of Lipid Variants (WDLV) databases, in the proband or a first-degree relative. FH diagnosis in a patient with a DNA mutation but normal LDL-C levels is unclear. Yearly follow-up of the proband is suggested and cascade screening of family members should be initiated. Note: In any case, cascade screening should be implemented; treatment decision should be at the discretion of the treating physician.

      FH criteria: minor

      There are 2 minor criteria: (1) a family history of elevated LDL-C > 95th percentile, according to the LDL-C criteria outlined in the Screening Criteria for FH section in a first-degree relative, according to age; and (2) a history of ASCVD in the proband or in a first-degree relative younger than 55 years for men or younger than 65 years for women. A diagnosis of “definite FH” is on the basis of the LDL-C criterion and 1 major criterion. “Probable FH” is on the basis of the LDL-C criterion and 1 minor criterion. “Severe hypercholesterolemia” refers to the LDL-C criterion (> 95th percentile), but without major or minor criteria for FH.

      Sensitivity/specificity analyses

      Agreement analyses were carried out using data from 2 large clinical databases in Canada and Australia, comparing the performance of the Canadian definition with that of SBR and the DLCN. Table 1 shows the sensitivity and specificity values for each set of data, the positive and negative predictive values, as well as the Cohen κ coefficient. Using the SBR criteria for comparison, the Canadian definition achieved 99.7% sensitivity and 98.9% specificity in the largest data set from Chicoutimi, Quebec, composed of 5987 subjects. Compared with the DLCN definition, the Canadian definition achieved 100% sensitivity and 98.8% specificity. The new Canadian definition of FH showed excellent agreement with the SBR and DLCN criteria, with κ coefficients of 0.969 and 0.966, respectively (P < 0.0001). Similar results were obtained in the Australian population, with the Canadian definition of FH showing excellent agreement with the SBR criteria (κ = 0.966) and the DLCN criteria (κ = 0.834; P < 0.0001 for both).
      Table 1Agreement between proposed Canadian definition of FH and Simon Broome Register and DLCN criteria
      Canadian definition vs Simon Broome RegisterCanadian definition vs DLCN
      Canadian database (n = 5987)Australian database (n = 947)Canadian database (n = 5987)Australian database (n = 947)
      Sensitivity (95% CI), %99.7 (99.2-99.9)99.3 (97.6-99.9)100 (99.6-100)80.8 (76.5-84.6)
      Specificity (95% CI), %98.9 (98.6-99.2)98.2 (96.8-99.0)98.8 (98.4-99.1)100 (99.4-100)
      Positive predictive value (95% CI), %95.3 (93.8-96.4)96.1 (93.3-98.0)94.5 (93-95.8)100 (98.8-100)
      Negative predictive value (95% CI), %99.9 (99.8-100)99.7 (98.9-100)100 (99.9-100)88.6 (85.9-91)
      Cohen κ coefficient0.9690.9660.9660.834
      P< 0.0001< 0.0001< 0.0001< 0.0001
      Sensitivity, specificity, and positive and negative predictive values as well as the Cohen κ coefficients were obtained from the comparison of the Canadian FH definition against the Simon Broome Register and DLCN criteria.
      CI, confidence interval; DLCN, Dutch Lipid Network Criteria; FH, familial hypercholesterolemia.

      Discussion

      This new definition of FH showed excellent agreement with the most widely used FH criteria, the SBR and DLCN criteria, and is well adapted to the Canadian population. The risk of developing ASCVD in mutation carriers with high LDL-C has been shown to be markedly elevated; identification and early treatment of subjects with FH has been shown to normalize life expectancy. Compared with normolipidemic individuals, ASCVD risk is increased 6-fold when LDL-C is > 5 mmol/L vs noncarriers having LDL-C levels < 3.4 mmol/L and up to 22-fold when a pathogenic DNA FH-causing mutation is present.
      • Khera A.V.
      • Won H.H.
      • Peloso G.M.
      • et al.
      Diagnostic yield and clinical utility of sequencing familial hypercholesterolemia genes in patients with severe hypercholesterolemia.
      This is likely related to higher cumulative lifetime vascular exposure to atherogenic low-density lipoprotein particles. In this article, we propose a novel definition of FH and online or downloadable applications that should facilitate diagnosis (www.circl.ubc.ca).
      We acknowledge limitations to this scheme but this simplified definition will provide physicians and health care professionals a reliable way to diagnose FH and to initiate treatment and cascade screening in affected patients so that appropriate treatment is initiated early and might prevent cardiovascular events and deaths. There is no “gold standard” for a definition of FH and therefore, comparison with existing diagnostic criteria are necessarily limited. We recognize that our LDL-C cut points are arbitrary and that the imputed LDL-C represents the average response to lipid-lowering agents. However, the new LDL-C cut points will minimize the underdiagnosis of FH in young adults as is the case in other criteria such as the SBR criteria.
      Some subjects with a causal mutation in the LDLR, APOB, or PCSK9 genes might have an LDL-C < 95th percentile. Nevertheless, a subject with a causal mutation in the LDLR, APOB, or PCSK9 genes remains at elevated ASCVD risk and preventive therapies must be considered. DNA testing for FH is not widely available in Canada, might not detect all types of variants, and is costly. Although a DNA diagnosis is not mandated for a diagnosis of FH, it should be considered in “probable FH” or “severe hypercholesterolemia” cases, when this might influence therapeutic decisions, especially in younger subjects. Approximately 20% of FH patients have a polygenic form of the disease. These patients would not meet the DNA criterion, but might meet the LDL-C and ASCVD criteria, and still require aggressive treatment including a possible need for PCSK9 inhibitors.
      Treatment decisions should be at the discretion of the physician and the patient and should follow the 2014 Canadian Cardiovascular Society position statement on FH,
      • Primary P.
      • Genest J.
      • Hegele R.A.
      • et al.
      Canadian Cardiovascular Society position statement on familial hypercholesterolemia.
      and the 2016 Canadian Cardiovascular Society guidelines for the management of dyslipidemia (www.onlinecjc.ca/article/S0828-282X(16)30732-2/pdf). The proposed definition for FH will also be particularly useful as a guide to select patients suitable for genetic testing, which is becoming more widely available. Because of the worldwide prevalence of FH, this new definition might be useful in countries other than Canada. The absence of positive genetic testing does not imply lack of risk in patients with LDL-C > 95th percentile, and these individuals still require active treatment to reduce their risk. The opportunity for clinicians to initiate cascade screening from an index patient is a very cost-effective method to identify new patients and initiate treatment and might prove more effective than broad cholesterol screening in childhood.

      Funding Sources

      This study was funded through FH Canada (www.FHCanada.net) with unrestricted grant support from Sanofi, Amgen, Pfizer, Aegerion, and Valeant. Parts of this study were also supported by the Institute for Clinical Evaluative Sciences and an operating grant from the Institute of Circulatory and Respiratory Health-Canadian Institutes of Health Research Chronic Diseases Team (grant number TCA 118349). The Institute for Clinical Evaluative Sciences is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care.

      Disclosures

      Briefly, Z.A., J.B., J.C.G., and P.R. have collaborated with Amgen and Sanofi; D.B. with Amgen; L.R.B. and G.A.F. with Sanofi, Amgen, Akcea, and The Medicines Company; P.C. with Merck, Pfizer, Atrium Innovations, and Kaneka; D.G. with Aegerion, Amgen, Akcea/Ionis, AstraZeneca, Chiesi, DalCor Pharma, Esperion, GlaxoSmithKline, Gemphire, Pfizer, Regeneron, Sanofi, and UniQure; M.G. with Valeant, Sanofi, Amgen, and The Medicines Company; R.A.H. with Aegerion, Akcea/Ionis, Amgen, Boston Heart Diagnostics, Gemphire, Pfizer, Regeneron, Sanofi, and Valeant; G.B.J.M. with Sanofi, Amgen, Novartis, Janssen, Novonordisk Boehringer-Ingelheim, Merck, AstraZeneca, and Bayer; B.W.M. with Janssen, Mezzion, and Kowa; G.F.W. with Sanofi, Regeneron, Gemphire, Amgen, and Kowa; and J.G. with Sanofi, Amgen, Pfizer, Aegerion, Valeant, Novartis, Merck, and Eli Lilly. Full disclosures are listed in the Disclosures section of the Supplementary Material.

      Supplementary Material

      References

        • Nordestgaard B.G.
        • Chapman M.J.
        • Humphries S.E.
        • et al.
        Familial hypercholesterolaemia is underdiagnosed and undertreated in the general population: guidance for clinicians to prevent coronary heart disease: consensus statement of the European Atherosclerosis Society.
        Eur Heart J. 2013; 34: 3478-3490a
        • Akioyamen L.E.
        • Genest J.
        • Shan S.D.
        • et al.
        Estimating the prevalence of heterozygous familial hypercholesterolaemia: a systematic review and meta-analysis.
        BMJ Open. 2017; 7e016461
        • Ruel I.
        • Aljenedil S.
        • Sadri I.
        • et al.
        Imputation of baseline LDL cholesterol concentration in patients with familial hypercholesterolemia on statins or ezetimibe.
        Clin Chem. 2018; 64: 355-362
        • Khera A.V.
        • Won H.H.
        • Peloso G.M.
        • et al.
        Diagnostic yield and clinical utility of sequencing familial hypercholesterolemia genes in patients with severe hypercholesterolemia.
        J Am Coll Cardiol. 2016; 67: 2578-2589
        • Primary P.
        • Genest J.
        • Hegele R.A.
        • et al.
        Canadian Cardiovascular Society position statement on familial hypercholesterolemia.
        Can J Cardiol. 2014; 30: 1471-1481