3.1 Apparent dominant mode of inheritance in a family with familial combined hyperlipidemia (FCH). This unique pedigree originates from a hyperlipidemic (HLP, marked by a dot on the diagram) father with the type IV phenotype and a hyperlipidemic mother with the type IIa phenotype (low-density lipoprotein cholesterol [LDL-C] in green >90th percentile). Both have elevated LDL–apoB (value of LDL-B in red) and both are presumed to originate from a familial combined hyperlipidemia family. At 63 years of age both have manifestations of atherosclerosis: myocardial infarction (MI), coronary artery disease (CAD) or peripheral vascular disease (PVD). The type IV is labelled B (IV B) because of the hyperapobetalipoproteinemia.

This putative FCH × FCH mating is further borne out as all descendants of this couple have hyper-apoB defined by levels of total apoB ≥120 mg/dl (LDL-B + VLDL-B) or LDL–apoB ≥100 mg/dl. Multiple phenotypes are present in first-degree relatives. Five subjects have hypertriglyceridemia, defined by a fasting plasma value ≥150 mg/dl (1.7 mmol/l) (figures in blue). Three normolipidemic descendants have only hyperapobetalipoproteinemia (NB). This is compatible with a dominant mode of inheritance (as opposed to a co-dominant mode) because none of the eight offspring has a more severe lipoprotein phenotype than the others although the likelihood of including a homozygote is high (¼). The vertical transmission is obvious. Furthermore, the offspring range in age from 29 to 38 years of age, and there is a trend for the triglyceride levels to be higher in hypertriglyceridemics with ageing compatible with the reported delayed expression of this trait. The apoE phenotype is provided. It has been shown that the E4/3 phenotype (father and the four sons) may be associated with higher LDL-C levels than the E3/3 phenotype.

3.2 Hyper-apoB discriminates for coronary artery disease beyond low-density lipoprotein-cholesterol (LDL-C). This study was carried out in consecutive patients undergoing diagnostic cardiac catheterization for coronary artery disease (CAD). Thirty-one were ‘free of CAD’ (<25% stenosis) (blue cohort marked Normal) and 59 had CAD (≥50% stenosis; blinded evaluation, orange group). The mean ± SD for LDL-C was 2.9 ± 0.8 mmol/l (112 ± 30 mg/dl) and 3.46 ± 0.7 mmol/l (134 ± 27 mg/dl), respectively. A third group of 40 CAD subjects with ‘type II hyperlipoproteinemia’ representing essentially patients with familial hypercholesterolemia (FH) and an LDL-C ≥4.91 mmol/l (190 mg/dl) was used for comparison; the mean LDL-C in this group was 6.65 ±1.4 mmol/l (250 ± 55mg/dl, red). Although there is an overlap (grey zone) between normal and CAD subjects, discriminant analysis showed that LDL–apoB (LDL-B) most clearly separated the two groups compared with cholesterol, LDL-C or triglycerides.

Although assessment of disease severity has the limitations of angiographic evaluation, this work supports the importance of plasma LDL–apoB to separate patients at risk for CAD in the presence of ‘normal’ LDL-C. Among the hyper-apoB patients most would fit the definition of familial combined hyperlipidemia (FCH). LDL-B levels in these patients reached values observed in familial hypercholesterolemia (FH) with much higher LDL-C concentrations. The upper limit of normal for LDL-C when this work was published was 4.91mmol/l (190 mg/dl). Afterwards it was brought down to 4.18mmol/l (160 mg/dl), and later to 3.36mmol/l (130mg/dl). Even with these new cut-offs, part of the CAD sample remains in the upper left quadrant of hyperapoB without hypercholesterolemia. Setting the limit of LDL-C in secondary prevention of CAD to 2.6 mmol/l (100 mg/dl) may include most of these hyper-apoB subjects and also most of the normal subjects. The apoB level is particularly important to know when identifying subjects at risk in an FCH family when triglycerides and total or LDL-C are normal (NB phenotype). The mean levels of LDL-B are given for the three different groups of patients illustrated on the left side of the graph. This figure was constructed from figures in the article by Sniderman AD et al. (1980). Proc Natl Acad Sci USA, 77: 604, linking the data point at the extremes of the distribution for each sample.

3.3 Metabolic defect in familial combined hyperlipidemia (FCH). In FCH there is a delayed clearance of postprandial chylomicron remnants (dotted red line on the left), with increased circulating levels of free fatty acids (FFA) and apoB-48. There is also reduced triglyceride synthesis from FFA in adipose tissue (see 3.4) resulting in an enhanced flux of FFA to the liver for triglyceride synthesis. The major metabolic defect in FCH is an overproduction of very-low-density lipoprotein (VLDL)-apoB by the liver leading to increased plasma VLDL and apoB concentrations. The VLDL particle size and lipid content are smaller than in familial endogenous hypertriglyceridemia (FEHTG). The increase in VLDL may account for the hypercholesterolemia as well as for the hypertriglyceridemia (IIb) but the final lipoprotein phenotype may depend on variations in the fate and or composition of the VLDL particles.

Effective lipolysis coupled with reduced clearance of LDL may result in isolated hypercholesterolemia (IIa). In contrast, reduced lipoprotein lipase (LPL) activity (LPL ± ↓, dotted red line on the top right) — which occurs in as many as 50% of cases of FCH — with effective LDL clearance may favour the hypertriglyceridemic phenotype (IV), sometimes with traces of chylomicrons (type V). Phenotype variation over time may result from the transient effect of modulating environmental or genetic factors. The LDL particles formed tend to be small, dense and numerous. This is determined in part by the concentration and size of triglyceride-rich lipoproteins as there is an inverse relationship between triglycerides and LDL particle size (r = —0.71, P < 0.001, Vakkilainen J, Jauhiainen M, Ylitalo K, Nuotio IO, Viikari JSA, Ehnholm C, Taskinen MR [2002]. LDL particle size in familial combined hyperlipidemia: effects of serum lipids, lipoprotein-modifying enzymes, and lipid transfer proteins. J Lipid Res, 43: 598-603. Small dense LDL (sdLDL) particles are made with the concerted action of cholesteryl ester transfer protein (CETP) and hepatic lipase (HL). Although an impaired LDL removal may be a secondary component in FCH pathophysiology (^* with dotted red line), one family has been reported in which the FCH clinical and biochemical phenotype was essentially associated with a defect at this step in the absence of VLDL–apoB overproduction. For more details regarding the metabolic scheme and abbreviations, see 1.2 and 2.12.

3.4 Putative involvement of adipose tissue in familial combined hyperlipidemia (FCH). In healthy subjects, there is a fine balance between free fatty acid (FFA) uptake and release by adipose tissue which reflects lipid storage and lipolysis, respectively, two physiologically important components of energy balance. In FCH, impairment of both lipid storage and lipolysis have been postulated. (1) After a fat meal in FCH, there is no major change in lipoprotein lipase (LPL) activity but increased serum free fatty acid (FFA) levels and flux to the liver with a transient increase in ketone bodies (0—8 hours). In the post-absorptive period (8—24 hours), FFA and ketone bodies rapidly decrease in subjects with FCH compared with normal subjects. This was ascribed by Meijssen et al. (2000), from Erasmus University, to inadequate incorporation of FFA into triglycerides in adipocytes coupled with a diminished release of FFA by hormone-sensitive lipase (HSP). This scenario is supported by other lines of evidence. (2) Acylation-stimulating protein (ASP), a complement C3a derivative, C3a-des-Arg, made in adipose tissue (with the concerted action of adipsin [factor D], complement factors C3 and B), promotes triglyceride (TG) synthesis by stimulating acyl-CoA:diacylglycerol acyl-transferase (DGAT), inhibits hormonesensitive lipase (HSL) and favours glucose uptake in adipocytes. ASP is a lipid-storage hormone. Some of its autocrine and paracrine effects are mediated by its receptor C5L2 (an orphan G protein-coupled receptor) and enhanced by insulin. ASP activity is impaired in FCH and ASP-mediated triglyceride synthesis is reduced in adipose cells from FCH subjects as shown by Cianflone and co-workers in Montreal. They have also shown that in fibroblasts from FCH patients (patients with ‘hyperapobetalipoproteinemia’) the specific binding of ASP to a single class of surface receptors was reduced in proportion to reduction in triglyceride synthesis (apart from evidence of reduced receptor number, it is not established whether there is impaired receptor-ligand interaction or ASP resistance in FCH). This results in a diversion of FFA from lipid storage to liver uptake for triglyceride and very-low-density lipoproteins (VLDL) synthesis. The major source of C3, the precursor of ASP, is the liver. An increase in circulating C3 has been reported in FCH. (3) There is also a defect in the ability of insulin to suppress free fatty acid release from adipose tissue in FCH. (4) This adds to the problem of impaired activity of HSL leading to impaired lipolysis as reported by Reynisdottir and co-workers, from Huddinge Sweden, in 1995 in work carried out with subcutaneous adipose tissue from FCH patients, (although, in contrast, there has been no demonstration of reduced HSL gene expression nor linkage with the HSL gene in affected Finnish subjects). An inhibition of ASP receptor by an expanding triglyceride pool was postulated. (5) Adiponectin (AQ) is another hormone formed in adipose tissue with protective effects against atherosclerotic cardiovascular disease. van der Vleuten and colleagues in Nijmegen, the Netherlands, found that adiponectin is decreased in FCH independent of waist circumference and insulin resistance. It is an independent predictor of the atherogenic lipoprotein profile of FCH, including high triglyceride levels, low HDL-C levels and amount of circulating small dense LDL. Several properties of adiponectin may account for this. Adiponectin enhances fatty acid oxidation (FA ox) in the circulation and in skeletal muscle, through activation of AMP-activated kinase (AMPK). There is also a strong relationship between adiponectin levels and LPL activity. This is consistent with the fact that plasma apoB concentrations are inversely related to adiponectin levels and that adiponectin correlates positively with apoB catabolism (but not with apoB production) (3.5). Furthermore, adiponectin increases glucose (Glu) uptake by skeletal muscle and enhances insulin sensitivity. Therefore a reduction in adiponectin would enhance insulin resistance, which may occur in FCH. (6) Finally, FFA may also promote insulin resistance by increasing gluconeogenesis in the liver. Increased glucose production stimulates the pancreas to secrete more insulin, which may lead to increased fasting insulin and insulin resistance coupled with defective insulin action (see point 3). This can lead to reduced FFA trapping and further increase in circulating FFA thereby creating a vicious cycle.

3.5 Relationship between adiponectin and very-low-density lipoprotein (VLDL)-apoB levels or catabolism. Ng and co-workers (2005) investigated the relationship of plasma adiponectin concentrations with VLDL-apolipoprotein B (apoB)-100 kinetics in men using stable isotopes. They showed that in multivariate analysis, plasma adiponectin concentration was the most significant predictor of plasma VLDL–apoB concentration (P = 0.001) and, with total or subcutaneous adipose tissue mass, was an independent predictor of VLDL–apoB catabolism (P < 0.001). The associations using a stepwise regression model including HOMA (homeostasis model assessment) score, age and total adipose tissue mass as co-variates are depicted in the left panel for apoB levels and the right panel for apoB catabolism (n = 41). There was no association of adiponectin with apoB production. In contrast with adiponectin, insulin resistance, measured with the HOMA, was a significant predictor of apoB production (P < 0.05). Interestingly, leptin, resistin, interleukin-6 and tumour necrosis factor a were not associated with any of the measured variables. Reproduced with permission from Ng TWK et al. (2005). Adipocytokines and VLDL metabolism: independent regulatory effects of adiponectin, insulin resistance, and fat compartments on VLDL-apolipoprotein B-100 kinetics? Diabetes, 54: 795-802.

3.6 Discrete clinical signs in familial combined hyperlipidemia (FCH): xanthelasma and arcus corneae. Clinical signs when present tend to be discrete in FCH. Besides manifestations of atherosclerosis that might emerge eventually, such as arterial bruits, angina pectoris, transient ischemic attacks or other forms of expression of tissue ischemia, xanthelasma of the eyelids or arcus corneae may be the only clinical clue to the existence of FCH. This figure (the arrow is pointing to the arcus) is of a 46-year-old patient with hypercholesterolemia (phenotype IIa).

3.7 Discrete manifestations of familial combined hyperlipidemia (FCH): arcus corneae and xanthelasma. Corneal arcus and xanthelasma are not always as obvious as in 3.6. Instead of fully encircling the cornea (see 1.3), FCH is often manifested by a discrete lower (as above) or upper corneal crescent, always leaving a rim of clear cornea between the lipid deposit and the sclera. It may be easily missed if it is not looked for in particular using a light, especially for upper crescents, which are masked readily by the upper eyelid. They tend to occur more frequently in the elderly. In the past a distinction was made between arcus juvenilis and arcus senilis, the latter considered a degenerative ocular manifestation. In black people, arcus corneae are routinely seen in the absence of dyslipidemia. In this figure, a unique small xanthelasma is also present. Half of the subjects with xanthelasma also have dyslipidemia. Presence of either of these is an excellent clue for dyslipidemia.

3.8 Syringoma is often mistaken for xanthelasma associated with dyslipidemia. The top left panel shows typical xanthelasmas of the eyelids and of the loose skin under the lower eyelid in a 38-year-old patient with familial combined hyperlipidemia (FCH; type IIa phenotype). They are typically planar, slightly raised, occurring in patches of various sizes, coalescing with a yellow to yellow-orange discoloration. The other three panels represent syringomas which, in contrast, are puncta of smaller sizes, are not found in patches or coalescing, and are more pinkish than yellow. These are small tumours, usually benign, of the sweat glands. The lower left panel shows typical syringomas in a 51-year-old woman, referred for hypercholesterolemia and xanthelasma of the eyelids. In fact she had histologically confirmed syringomas of the eyelids and not xanthelasma and hyperalphalipoproteinemia. Twenty-two years later her lipoprotein profile had changed very little: total cholesterol 5.98 mmol/l (231 mg/dl), low-density lipoprotein-cholesterol (LDL-C) 3.41 mmol/l (132 mg/dl), triglycerides 0.93 mmol/l (82 mg/dl), and HDL-C 2.15 mmol/l (83 mg/dl). She never developed clinical manifestations of atherosclerosis. The two panels on the right show a 33-year-old women who was referred to the authors’ lipid clinic in 1972 for ‘xanthelasma’ and hyperlipidemia discovered fortuitously during a routine clinical examination. Further evaluation revealed that she had syringoma of the eyelids and not xanthelasma. The modest rise in cholesterol levels to 6.85 mmol/l (265 mg/dl) and triglycerides to 2.74 mmol/l (243 mg/dl) was not due to a familial diathesis but was secondary to oral contraceptives (norethindrone 1 mg/mestranol 50 μg). Lipids normalized on withdrawal and the dyslipidemia recurred on rechallenge. The positive family history of a premature death at the age of 45 of a maternal aunt due to coronary causes, which had suggested a familial disease turned out to be that of an unrelated spouse of a family member. Correct interpretation of the clues suggestive of dyslipidemia is a key element of the diagnostic process.

3.9 Type V secondary to the P207L LPL mutation in a familial combined hyperlipidemia (FCH) family. In this family the proband is a 57-year-old man with FCH. On the paternal side, four of his relatives had had a myocardial infarction and/or a cerebrovascular accident between the ages of 45 and 79. He himself had a cerebral thrombosis following a left carotid endarterectomy at the age of 44 after a few episodes of amaurosis fugax. His 55-year-old spouse was found to have hypertriglyceridemia (type IV). His daughter first consulted at the age of 21 years, five years after her father’s first visit to the authors’ lipid clinic. A diagnosis of familial hyperchylomicronemia had been made at the age of 8 months. She was followed regularly thereafter and was on a very low fat/no alcohol diet. Thus, she was able to avoid having pancreatic complications or developing eruptive xanthomas. Homozygosity for the P207L mutation of the LPL gene was established at the age of 19. At the authors’ lipid clinic her lipoprotein phenotype was found to be type V. Chylomicrons ranged from traces on lipoprotein electrophoresis to a typical creamy layer on top of a diffusely milky plasma (image on the left). Her triglycerides tended to increase over seven years of follow-up and fluctuated widely from 3.83 mmol/l to 21.7 mmol/l. The recent lipoprotein analysis given above shows the phenotypic profile of type V in which the very-low-density lipoproteins (VLDL) fraction predominates (right panel). Her 23-year-old brother is normolipidemic, but his total plasma apoB is 121 mg/dl, which is compatible with the presence of an FCH susceptibility gene at this age (type NB). Red, mixed hypertriglyceridemia; yellow, hypercholesterolemia; blue, hypertriglyceridemia; black, hyperapobetalipoproteinemia.

3.10 Mixed hypertriglyceridemia (type V) in a familial combined hyperlipidemia (FCH) family. There have been several reports of mixed hypertriglyceridemia (lipoprotein phenotype V) occurring in FCH families. This figure concerns one of the early reports (Kobori et al. 1986). The proband (arrow) was a 16-year-old girl with recurrent episodes of abdominal pain. As is often the case, laparotomy was carried out, which revealed that the pain was secondary to pancreatitis. Later on, the plasma was noted to be milky, which revealed the cause and she was referred to a lipid clinic where the family survey was done. This case is interesting because, first, the diagnosis of FCH was supported by raised levels of apoB in all hyperlipidemic cases, including the proband and the other hypertriglyceridemic subjects; there were two different phenotypes in the family (IIa and IV); there were discrete clinical manifestations (19 relatives were examined; there were no tendon xanthomas but corneal arcus was present in the father) and the presence of two normolipidemic subjects with hyper-apoB (NB phenotype). There was no overt coronary artery disease in this family with relatively young members (the oldest in generation II was 58 years and in generation III was 30 years). Second, there was a reduction in post-heparin lipolytic activities in the proband: LPL of 0.8 μM FFA/h/ml (normal: 2.5 ± 0.9) and HL 3.2 μmM FFA/h/ml (normal: 6.9 ± 3.1), which is compatible with the authors’ experience and that of others that mutations of the LPL gene may segregate independently in some FCH families. Secondary causes of dyslipidemia were excluded and there was no apoCII deficiency. The proband had the common E3/3 apoE phenotype. Her lipoprotein profile indicated endogenous hypertriglyceridemia (type IV) when her plasma triglycerides were lowered. There were no cases of dysbetalipoproteinemia (type III) in the family. Total cholesterol (dark orange figures, orange-yellow shading) and triglycerides (blue figures, blue shading) are given in mmol/l and in mg/dl. They are given in mg/dl in the original article. The lipoprotein phenotype is given at the top of the line of each pedigree member; note that ‘NB’ phenotype is the authors’ interpretation, the plasma apoB was 116 mg/dl for the 58-year-old and 90 mg/dl for the 20-year-old relative. The age is given in the lower right corner of the square for males and the circle for females. Redrawn and modified from Kobori S et al. (1986). A kindred of familial combined hyperlipidemia (FCHL) with proband showing type V hyperlipoproteinemia. Jpn J Med, 25: 306-312.