Screening strategies

Professional paediatric organisations recommend selective instead of universal screening strategies. The strategies proposed by the American Academy of Pediatrics (AAP) and the American Heart Association (AHA)4 5 7 are summarised in Table 1.

These strategies are quite similar in that screening is recommended when a family history of dyslipidaemia or premature CVD (≤55 years of age for men and ≤65 years of age for women) is present. However, several studies showed that self-reported parental history is not particularly useful when identifying children and adolescents with raised low-density lipoprotein cholesterol (LDL-C) levels (sensitivity 25–93%, specificity 21–80%).17 Therefore, applying this criterion is not much more effective than general population screening. Even for children with FH, who would be expected to have a positive family history of CVD, it was shown that using family history to determine who to screen for cholesterol measurement would have failed to detect a substantial number of children with FH.18 Furthermore, this approach is time-consuming, is not well defined, and is not standardised.

The use of body mass index (BMI) alone or combined with family history as a criterion, instead of family history alone, increases the sensitivity of this screening method, although it decreases specificity.19 However, it is unlikely that using BMI would increase the sensitivity of screening children with FH, the group most likely to benefit from screening.

Screening in Europe

Several different screening strategies are currently used across Europe. In this section, we will describe the active screening strategies in five countries, summarised in table 1.

In the Netherlands, there is an ongoing cascade screening programme to identify patients with FH, in which all first degree family members of an index patient are identified and offered a DNA test. Cascade screening then is offered to more distant relatives using the inheritance pattern across the pedigree. To date, >20 000 individuals with FH carrying more than 550 different mutations have been identified, which is estimated to be 50% of the total number of FH patients in the Netherlands. A literature review reported that screening of the relatives of index cases with FH is the most cost-effective screening strategy.20

In Norway, as in the Netherlands, all genetic FH tests are performed by a single centre which has been working closely with the Lipid Clinic in Oslo. The total number of FH patients has been estimated at 16 000 and approximately 5000 individuals with >140 different mutations have been identified so far. A cascade screening program has been established,21 but as the national health system in Norway has not provided funding for other areas of the country, laboratory investigations will remain dependent on the Lipid Clinic and the Medical Genetic Laboratory of the University of Oslo.

The Slovenian Pediatric Society decided to initiate general cholesterol screening in preschool children at the age of 5 in 1995. Since then, more than 15 000 children have undergone screening annually. A blood sample is taken at a routine examination and total cholesterol (TC) values are measured. If TC values are elevated, a fasting lipid profile is performed. According to national guidelines, the child is referred to University Children's Hospital in Ljubljana if TC levels are >6 mmol/l or >5 mmol/l and there is a positive family history of premature CVD. The data of these children are used for research and from 2011 onwards selective genetic screening for FH will also be carried out.

In Italy selective screening is delegated to general practitioners or paediatricians who follow Italian guidelines.22 Children considered at risk and suspected of having an inherited lipoprotein disorder (showing a positive family history of premature CVD or exhibiting dyslipidaemia with TC >6.2 mmol/l, HDL-C <0.9 mmol/l or triglyceride (TG) >3.4 mmol/l) are invited to attend lipid units where they undergo biochemical and genetic analysis. However, only a limited number of children complete this program, accounting for less than 10% of the total number of children with possible FH. To bridge the gap between the number of children detected and the number theoretically affected, an extensive program of recruitment, diagnosis and registration is currently in progress. This program is aimed at promoting awareness of the benefits of screening, extending this practice throughout Italy, defining the genetic mutations involved in FH (LDLR, ApoB, PCSK9) and providing an Italian register of genetically detectable lipoprotein disorders.

In the UK, the National Institute of Healthcare and Clinical Excellence (NICE) has recommended a family cascade testing approach for FH, starting with adults who have a clear (preferably genetic) diagnosis. The NICE is regarded as the main organisation that promotes evidence based clinical practice in the UK. Testing of children in families known to have a clear diagnosis of FH is recommended from the age of 10 years. Recently, this approach has been fully implemented (2010) in Wales, which is one of the devolved countries within the UK.23

Wald and colleagues recently proposed the alternative strategy of screening children in the UK when they visit their general practice for routine vaccination at about 15 months of age. In children aged 1–9 years, the use of TC levels as a cut off yields a false positive rate of 0.1% (1.53 multiples of the mean) and would identify 88% of cases with FH. Measurement of cholesterol levels in the parents of an affected child would determine which parent was affected. Applying the rule that the parent with the highest cholesterol levels has FH gives a detection rate of above 96%.24

Screening tools

Once children are selected for screening, lipid levels are measured. A raised LDL-C level is the trigger for initiating treatment and determining therapy goals in children5 7 and is generally calculated according to the Friedewald formula. However, because this is dependent on TG levels, a fasting sample is required. The advantage of a TC measurement is that it is not necessary to obtain a fasting sample. In the Bogalusa Heart Study, increased TC levels detected elevated LDL-C levels with 44–50% sensitivity and 90% specificity.25

It was shown that the apoB/apoA-1 ratio measured in adolescents (12–18 years) was superior to the LDL/HDL ratio for predicting increased IMT in adulthood.26 A high apoB/apoA-1 ratio indicates the ratio between pro- and anti-atherogenic lipoprotein particles. Therefore, measuring these apolipoproteins might have additional value in risk assessment and screening in children.

An important issue is the accuracy of paediatric lipid levels to identify children at increased risk. Approximately 40–55% of children with elevated lipid levels will continue to have elevated lipid levels in adulthood.27 On the other hand, subjects with multifactorial/polygenic causes of hypercholesterolaemia often exhibit normal lipid levels during childhood. Therefore, those with a strong family history of premature CVD in particular should be retested after puberty. At least two measurements are necessary to ensure that the mean measurement reflects the true value to within 10%. Children with very high LDL-C levels (>4.1 mmol/l (>160 mg/dl)) are likely to have substantially lower levels (mean decrease 0.5–0.75 mmol/l) at re-examination, which is attributable to regression to the mean.28 Furthermore, as childhood is a time of rapid growth and development, especially during puberty, and cholesterol is involved in the structure of cellular membranes and hormones, lipids vary with age among children.29 This must also be taken into account when interpreting cholesterol levels.

The criteria for a clinical diagnosis of FH are LDL-C levels above the 95th percentile for age and gender, a family history positive for CVD, and physical symptoms of cholesterol deposits in the skin, eyes or tendons known as xanthelasmas, arcus corneae and tendon xanthomas, respectively, which are typical for FH although rare in children.30 Once there is a clinical diagnosis, a definite diagnosis may be made using DNA based mutation screening methods. Once the causative mutation in the index patient has been identified, reasonably cheap molecular testing in relatives is possible. Molecular testing is simple in countries with only a few different mutations causing FH. On the other hand, in the majority of countries with full-blown genetic heterogeneity, DNA sequencing could fail to identify a mutation, with detection rates ranging from only 20% to 90%.31,–,35 However, as mentioned previously, extreme hypercholesterolaemia in children is less likely than in adults to be polygenic, which allows a higher mutation detection rate. In fact, in a recently published study it was found that in 95% of children with an FH phenotype, a functional gene mutation could be found in the LDL receptor or ApoB genes.36

Identification of the FH-causing mutation in a family permits a definite diagnosis, and the diagnostic problem caused by the overlap in cholesterol levels between the general population and the FH patient is thereby eliminated. A single test, performed once only, will ascertain FH status, and early diagnosis in these children, at least, becomes a reality.

Neonatal screening has been examined in some studies, but none followed abnormal results with mutation analysis, which makes it difficult to determine the value of such screening.27 The usefulness of IMT measurement as a screening tool in children must be further explored.