The classical nosological entity called familial hypercholesterolemia had been part of the medical literature for a long time before its mechanism was fully unraveled by the seminal work of the 1985 Nobel laureates, Joseph L Goldstein and Michael S Brown. For a monogenic disorder, it is relatively common, its frequency varying from 1 in 500 in most parts of the world to as much as 1 in 80 in regions where a founder effect exists due to factors such as endogamy, consanguinity, geographic isolation and limited genetic admixture. Such regions include, among others, Lebanon, Finland, South Africa (Afrikaners), Canada (French-Canadians) (1.1) and Israel (FH-Sephardic and FH-Lithuania). Its importance stems also from the devastating severe and premature cardiovascular consequences for the affected members (50%) of a family, stressing the need for early diagnosis. As there is treatment that will prevent or markedly delay the complications of this condition, early intervention may result in a normal life. In other words, it is a treatable killer and worldwide efforts have therefore been made to identify individuals and families at risk. MEDPED, ‘make an early diagnosis, prevent an early death’ is one such initiative (www.cholesterol.med.utah.edu/medped/). In certain communities, it may be a major health problem, disabling or killing many young adults.

The etiology of FH is well established. The metabolic defect (1.2), which leads to a large increase in plasma levels of LDL-cholesterol (LDL-C) (two to three times that of normal subjects), is attributed to mutations of the LDL receptor gene that result in a decreased number or total absence of LDL receptors or, alternatively, in expression of dysfunctional receptors. This ubiquitous receptor allows uptake and degradation in the liver of LDL, the major lipoprotein carrier of circulating cholesterol (mostly cholesteryl esters), allowing excretion of the latter into the bile (as free cholesterol). It is essential for bringing cholesterol, a major constituent of membranes, to cells. The defect causes a considerable delay of LDL clearance from plasma. Normal subjects catabolize about 45% of their LDL pool per day, whereas the fractional catabolic rate is 25-30% for heterozygotes and 15% for homozygotes. There is also a delayed clearance of intermediate-density lipoproteins (IDL), which represent ‘remnant’ lipoproteins and an increased conversion of IDL into LDL. Recent studies by Tremblay and co-workers have shown that there is also a 50% increased production rate of verylowdensity lipoprotein (VLDL) apolipoprotein B (apoB) – 100 in heterozygotes and 109% in homozygotes. This finding sheds light on an old debate. The LDL particles are large and buoyant in this condition because there is excess cholesterol associated with apoB, the main apolipoprotein associated with these cholesterol-rich particles. There are over 800 mutations of the LDL receptor gene (LDLR), which is located on the short arm of chromosome 19 (19p13.1-13.3) (1.2). They have been indexed continually in the UK since October 1996 on a dedicated website (www.ucl.ac.uk/fh/) and in France since April 1998 (www.umd.necker.fr/). A classification of the various defects of the LDLR gene has also been established (Hobbs et al. 1990). The abnormality may involve synthesis in the endoplasmic reticulum, transport of newly synthesized receptors to the Golgi complex, transport to the cell surface, clustering in the surface-coated pits or binding affinity for LDL, depending on the portion of the receptor gene that is affected (1.1). For practical purposes, it is useful to know if the receptor activity is impaired (receptor defective) or not functional at all (receptor negative). Severity and resistance to treatment are greater in the latter. This is especially important for homozygotes and can be determined by identifying the mutation or testing LDL receptor activity in cultured skin fibroblasts or blood mononuclear cells.

The diagnosis of heterozygous FH (Table 1.1) is based essentially on:

Detection of affected individuals is more difficult in childhood. The family history, combined with a blood sample for LDL-C is most useful in children. In women, manifestations are delayed by about 10-15 years compared with men (1.9). Evolution of tendon xanthomas may be assessed and followed using standardized X-ray techniques (1.10), ultrasonography or magnetic resonance imaging. The xanthoma size correlates with the duration and severity of the disease (1.11). Achilles tendon xanthomas are prone to sporadic inflammation, causing painful acute tendinitis (1.12).

Because LDL-C level is the best biochemical marker of FH, the most threatening and most directly linked to the causal defect, it remains the centre of attention. FH was classified in the Fredrickson era among subjects presenting a ‘type IIa’ lipoprotein phenotype (isolated hypercholesterolemia), and ‘type IIa’ became wrongly synonymous with FH. Indeed, the hypercholesterolemia may occasionally be associated with hypertriglyceridemia (becoming Fredrickson’s ‘type IIb’). This associated hypertriglyceridemia may be due to a second gene defect, another medical condition or environmental factors, or may be part of the defect in a particular family (triglyceride-rich remnant lipoproteins are cleared to some extent by the LDL receptor). The diagnosis may be established in most cases without resorting to determination of the genetic defect(s) using molecular biology techniques. This task, when required, is easier in communities where a founder effect exists, because only a few mutations may explain a majority of cases (1.1). When a doubt exists, some specialized lipid clinics may be able to help with identification of the gene defect.

The differential diagnosis must include other causes of xanthoma tendinosum since they constitute quite a reliable diagnostic criterion; these include lesions that can be mistaken for xanthoma tendinosum (Table 1.2). Few hereditary dyslipidemias apart from heterozygous FH and autosomal dominant hypercholesterolemia (FH3) have such high levels of cholesterol except the other monogenic forms discussed below and familial dysbetalipoproteinemia (type III), but in this condition, LDL-C measured directly is not elevated. Severe familial combined hyperlipidemia with isolated hypercholesterolemia (lipoprotein phenotype IIa) may rarely have LDL-C levels similar to those found in the lower distribution of FH, but tendon xanthomas have not been reported. Atherosclerotic vascular disease in the family tends to occur later in life than in FH and the variation in lipoprotein phenotype in first-degree relatives is typical. Among the secondary forms of hypercholesterolemia the most likely to be confused with FH are primary biliary cirrhosis, nephrotic syndrome and hypothyroidism, all potentially associated with very high levels of LDL-C.

Treatment almost always necessitates the addition of a statin (an HMG-CoA reductase inhibitor) to a cholesterollowering diet. Often, large doses of statin are needed and combination therapy with a resin (a bile acid absorption inhibitor) or with ezetimibe (a cholesterol absorption inhibitor) may be necessary. Various forms of LDL apheresis have been used in some very severe heterozygous FH refractory to drug therapy.

The homozygous (inheritance of two identical defective genes) or double heterozygous (two different gene defects at the same locus) forms are extremely rare (approximately 1 in 300 000 to 1 in 1 000 000). The clinical picture is so dramatic that the diagnosis is rarely missed (1.13,1.14,1.15,1.16,1.17). Xanthomas are diverse (tuberous, planar, tendinous, xanthelasma), extensive, ubiquitous (friction sites, elbows, knees, popliteal space, palms, plantar aponeurosis, gluteal crease) and may be present at birth. LDL-C levels may be four to six times the upper limit of normal and the type of mutation may also influence the plasma levels (1.18). Coronary death can occur as early as two years of age and the affected patients, whether male of female, rarely live beyond the third decade. One typical complication, in addition to myocardial ischemia and infarction, is aortic stenosis (1.19), which is sometimes seen in heterozygotes with severe disease. Differential diagnosis must include ARH and a condition reported in the past as pseudo-homozygous hypercholesterolemia (1.20). However, some of these cases may have had an ARH defect, especially when they responded well to dietary or statin treatment. Statins have a modest effect that is enhanced by combination with ezetimibe and LDL apheresis. Probucol, a major antioxidant, now withdrawn from the market but still available in Japan, has been reported to reduce the size of tendon xanthomas in homozygous FH. ‘Last resort’ treatments have been used, including porto-caval shunts, gene therapy and liver transplantation, with all but the latter having little success.

It is worth remembering that in FH, and especially in the homozygous form, the first concern of the doctor should be the accelerated atherosclerosis that accompanies this condition. The autopsy specimen presented in 1.21 showing severe aorto-femoral atheroma and aneurysmal weakening of the wall reminds us of the consequences of failing to intervene early and aggressively in these cases. A sense of urgency should always be uppermost in the mind of the doctor.

FDB was identified and its etiology determined in 1985-1986 by Grundy and colleagues at the University of Texas in Dallas and Mahley and co-workers at the Gladstone Research Foundation Laboratories in San Francisco. It is an autosomal dominant monogenic disorder due to point mutations in the APOB gene (1.22). This very large gene (43 kb, 29 exons) was mapped to the distal short arm of chromosome 2 (2p23-p24) in 1985-1986 by investigators from several laboratories (including Knotts, Chan, Law, Deeb and their co-workers). These mutations impair the affinity of apoB, the ligand, to its receptor, the LDL receptor (1.23), hence the synonym of ‘familial ligand-defective apoB-100’ (FLDB). Five mutations in exon 26 of APOB may cause this condition, Arg₃₅₀₀→Gln (the first common mutation identified), Arg₃₅₀₀→Trp, Arg₃₅₃₁→Cys, and Arg₃₄₈₀→Trp (Sweden) and Thr₃₄₉₂→Ile (Poland). A sixth recently reported mutation His₃₅₄₃→Tyr is four times more frequent than the R3500Q variant (for amino acid nomenclature see www.chem.qmul. ac.uk/iupac/AminoAcid/AA1n2.html#AA1). It appears to be associated with a variable but moderate degree of LDL-C elevation and a reduced apoB-100 fractional catabolic rate. An Asn₅₃₁₆→Lys mutation of APOB has little impact on the lipoprotein profile but changes apoB conformation. FDB is inherited as an autosomal dominant trait with incomplete penetrance or variable phenotypic expression.

The prevalence of FDB varies widely from country to country. In Caucasians from the USA and Europe, it averages 1 in 500 to 1 in 700. It is high in Switzerland (1 in 209 to 1 in 230) and Poland (1 in 250) and rare in Mediterranean countries. It has not been found at all in the Turkish or Finnish populations, or among hypercholesterolemic Japanese or Israelis. From prevalence and haplotype studies, Miserez and Muller at the Basel University Clinics speculated that the common mutation originated from Celtic ancestors in a region between Lake Geneva, the Jura mountains and the Rhine (in the northwestern part of Switzerland the prevalence of this mutation is 1 in 114), perhaps as early as the Mesolithic period (6000-10 000 years ago) (1.24). This hypothesis is consistent with a previous study from Myant and colleagues who used a combined molecular and population genetic approach to estimate the age of the mutation to be 6000-7000 years.

The impaired ligand-receptor interaction (20-30% of normal binding to fibroblasts) results in delayed clearance of defective LDL particles with a residence time of LDL–apoB 3.6 times longer than that of normolipidemic controls (8.2 vs. 2.3 days). This is associated with decreased production of LDL and enhanced removal of the apoE-containing VLDL, as demonstrated by Schaefer and co-workers in 1997 in a subject homozygous for the common mutation R3500Q (1.25). The residence time of VLDL–apoB is also increased, but that of VLDL-apoE is reduced since apoE becomes the favoured ligand to clear particles via the LDL receptor and LDL receptor-related protein (LRP). In addition, LDL isolated from these subjects has increased susceptibility to oxidation. The molecular mechanism whereby the apoB-100 mutations cause the phenotype was unravelled by Borén and co-workers in 2001. Arginine at residue 3500 is essential for normal receptor binding. The carboxyl terminus of apoB-100 is necessary for mutations affecting this arginine at residue 3500 to disrupt LDL receptor binding. Borén and colleagues drafted a model illustrating that Arg3500 interacts with Trp4369 and facilitates the conformation of apoB-100 required for normal receptor binding of LDL; the carboxyl terminal of apoB-100 interacts with the backbone of apoB-100, which in turn wraps around the LDL particle (1.26).