Nicotinic acid: the broad-spectrum lipid drug. A 50th anniversary review

Authors


Lars A Carlson, Konung Gustaf V:s forskningsinstitut, Karolinska universitetssjukhuset Solna, SE 171 76 Stockholm, Sweden.
(fax: 46 8 253976; e-mail: lars.a.carlson@bredband.net).

Abstract.

Nicotinic acid has, like the Roman God Janus, two faces. One is the vitamin. The other is the broad-spectrum lipid drug. The Canadian pathologist Rudolf Altschul discovered 50 years ago that nicotinic acid in gram doses lowered plasma levels of cholesterol. From the point of view of treatment of the dyslipidaemias that are risk factors for clinical atherosclerosis nicotinic acid is a miracle drug. It lowers the levels of all atherogenic lipoproteins – VLDL and LDL with subclasses as well as Lp(a) – and in addition it raises more than any other drug the levels of the protective HDL lipoproteins. Trials have shown that treatment with nicotinic acid reduces progression of atherosclerosis, and clinical events and mortality from coronary heart disease. The new combination treatment with statin-lowering LDL and nicotinic acid-raising HDL is reviewed. A basic effect of nicotinic acid is the inhibition of fat-mobilizing lipolysis in adipose tissue leading to a lowering of plasma free fatty acids, which has many metabolic implications which are reviewed. The very recent discovery of a nicotinic acid receptor and the finding that the drug stimulates the expression of the ABCA 1 membrane cholesterol transporter have paved the way for exciting and promising new 50 years in the history of nicotinic acid.

The double-faced nicotinic acid

Like the Roman God Janus nicotinic acid has two faces. One is the vitamin potent in milligram doses, the other is the broad-spectrum lipid drug potent in gram doses. The vitamin effects of nicotinic acid were demonstrated in the early parts of the 20th century. In the middle of the 20th century the Canadian pathologist Rudolf Altschul discovered that nicotinic acid lowered plasma cholesterol of rabbits [1]. Altschul extended these early observations with detailed experimental studies showing that nicotinic acid not only lowered serum cholesterol but also inhibited lipid deposits and atheromatosis in the cholesterol-fed rabbit [2]. In the landmark study of 1955 Altschul et al. [3] reported that nicotinic acid in gram doses lowered plasma cholesterol in normal as well as hypercholesterolaemic subjects. Of considerable interest is that nicotinamide did not affect the plasma lipid levels. This is a remarkable observation as both nicotinic acid and nicotinamide, chemically quite alike, are nutritionally equivalent and known as vitamin B3. Both the acid and the amide are precursors to the coenzyme nicotinamide adenine dinucleotide which is a major electron acceptor in the oxidation of fuel metabolites. The unexpected difference between nicotinic acid and nicotinamide may be due to the fact that whilst nicotinic acid is a powerful inhibitor of fat-mobilizing lipolysis in adipose tissue, this property is not shared by nicotinamide [4]. The inhibition of lipolysis in adipose tissue resulting in a decrease in plasma free fatty acids (FFA) has been suggested to be a basic mechanism for the lipid effects of nicotinic acid [5]. Recently a receptor for nicotinic acid has been discovered [6–8], which will be discussed below. Of particular interest is that in line with the different metabolic properties of nicotinic acid and nicotinamide, the former, but not the latter, is a ligand for the nicotinic acid receptor.

Nicotinic acid is a potent lipid-modifying drug and has been named ‘the broad-spectrum lipid drug’ [9]. However, its usefulness has been limited by side effects, particularly the flush. Nevertheless, there is a renewed interest for its clinical use, partly based on results from clinical trials demonstrating a significant reduction in atherosclerotic cardiovascular complications and total mortality, and partly on the availability of a new formulation of nicotinic acid (NiaspanTM, KOS pharmaceuticals, Miami, FL, USA) with less side effects. Against this background it was felt to be timely for a review on nicotinic acid after 50 years in use as a lipid drug focusing on its clinical and metabolic effects.

Early studies in hypercholesterolaemia

At the time when Altschul had made his discovery it was beginning to be recognized that hypercholesterolaemia carried with it a high risk for cardiovascular disease (CVD), in particular coronary heart disease (CHD). Altschul and colleagues in USA were now beginning to consider to lower plasma cholesterol in order to reduce the risk for CHD. Nicotinic acid was then the only effective drug available for cholesterol lowering.

Nicotinic acid – the primeur of the lipid drug market

Already 1 year after the epoch-making publication of Altschul and coworkers, a study from the Mayo clinic by Parsons et al. [10] reported on the effect of 3 g of nicotinic acid daily for 12 weeks. In seven patients with familial hypercholesterolaemia, plasma cholesterol on average decreased by 16% from 9.2 to 7.7 mmol L−1. In this early study it was also shown by lipoprotein electrophoresis that the cholesterol reduction obtained with nicotinic acid was due to a decrease in β-lipoproteins, i.e. LDL, and that the ratio of β- to α1-cholesterol (LDL to HDL) decreased from 9.0 to 5.6. In 1959, Parsons and Flinn reported on the effect of nicotinic acid in 44 patients, the majority having a plasma cholesterol above 7 mmol L−1, treated with 3–6 g day−1 of plain [immediate-release (IR)] or modified nicotinic acid for 56 weeks. On average, the cholesterol reduction was around 20% and the ratio of β- to α1-lipoprotein cholesterol decreased from 5.0 to 2.8 [11]. Nicotinic acid increased α1-lipoprotein (HDL) cholesterol by 44%, from 1.1 to 1.5 mmol L−1. In this trial it was once again shown that nicotinamide in high doses did not lower plasma cholesterol. Importantly, Parsons and Flinn also reported that subcutaneous cholesterol deposits (xanthoma tuberosum) were reduced during treatment with nicotinic acid. This is the first report of a lipid-modifying drug that reduces xanthomata, an effect that might reflect removal of cholesterol from tissues by means of HDL-mediated reverse cholesterol transport (RCT).

Nicotinic acid: an inhibitor of FFA mobilization

After Altschul's discovery of the cholesterol-lowering properties of nicotinic acid the basic mechanism for this effect was searched for. Decreased hepatic cholesterol synthesis, increased cholesterol oxidation, a relation to vasodilation and a vitamin effect were amongst candidate mechanisms studied. However, none of the possibilities studied could explain the cholesterol-lowering effect.

In the late 1950s, the plasma FFA, mobilized from adipose tissue, were discovered simultaneously at the Rockefeller Institute in New York [12] and at NIH, Bethesda [13]. It was soon thereafter demonstrated by Laurell in Lund that injected labelled FFA was rapidly recycled as esterified fatty acids in plasma of rats [14]. Furthermore, it was shown in man that a single dose of intravenously injected 14C-labelled FFA was rapidly incorporated into liver triglycerides with a maximum activity after about 15 min [15]. This was followed by increasing activity in the plasma triglycerides with maximum specific activity after 60–90 min [15]. A detailed study of the conversion of plasma FFA into plasma triglycerides was then published by Havel [16]. It was thus established that FFA are immediate precursors of hepatic and subsequently plasma triglycerides transported in VLDL, the predecessor of the cholesterol-rich LDL. We therefore tested whether nicotinic acid might lower plasma FFA, thereby reducing the fatty acid supply for hepatic synthesis of VLDL, as an explanation of its lipid effects.

Nicotinic acid lowers plasma FFA by inhibiting mobilization of FFA from adipose tissue

We found that nicotinic acid, 200 mg per os, within minutes lowered the arterial plasma concentration of FFA in fasting human subjects (Fig. 1) [5]. The lowering was followed by a rebound within 1 h. In the same study it was shown that pretreatment with nicotinic acid almost completely inhibited the marked rise in FFA caused by infusion of noradrenaline (Fig. 2) without affecting the cardiovascular response. In further studies it was shown that the length of depression of FFA levels by nicotinic acid is dose dependent, the depression by 1 g of plain nicotinic acid per os having a duration of 3 h [17]. This has been one of the reasons for administering IR nicotinic acid four times daily [9] in order to have a sustained lowering of FFA during daytime. As there is a tachyphylaxis for the vasodilation (flush) caused by nicotinic acid already after 1 or 2 weeks regular treatment we also tested whether the inhibitory effect on FFA mobilization persisted after long-term treatment with nicotinic acid. The inhibition of the noradrenaline-induced increase in plasma FFA by nicotinic acid remained unchanged after 6 months of regular treatment in man [18].

Figure 1.

Effect of nicotinic acid on plasma free fatty acids (FFA). Concentration of plasma FFA in blood in response to oral administration of 200 mg of nicotinic acid given at 40 min followed by 100 mg at 105 and 165 min (arrows). Values from three healthy, fasting male subjects. Reproduced from Carlson and Orö [5] with permission from Blackwell Publishing Ltd.

Figure 2.

Effects of nicotinic acid on noradrenaline-stimulated rise in plasma free fatty acids (FFA) and blood pressure. Noradrenaline infused intravenously (black bars) to anaesthetized, fasting dogs. Left panel: two control experiments. Right panel: three experiments with pretreatment with nicotinic acid, given in 10 doses of 100 mg kg−1 at the arrows before the second infusion. Reproduced from Carlson and Orö [5] with permission from Blackwell Publishing Ltd.

Nicotinic acid acts in adipose tissue

It was suggested that the lowering of plasma FFA and the inhibition of catecholamine-induced FFA increase in plasma FFA were due to the effects on mobilization of FFA from adipose tissue [5]. In vitro studies with rat epididymal fat pads showed that nicotinic acid prevented the release of FFA from adipose tissue by inhibiting lipolysis, measured as glycerol release [4]. Figure 3 shows the increase in release of FFA and glycerol from adipose tissue caused by noradrenaline and that nicotinic acid at concentrations of 10−5 and 10−4 mol L−1 inhibits this increase.

Figure 3.

Effect of nicotinic acid on noradrenaline-stimulated lipolysis in adipose tissue. Release of glycerol (left bar) and free fatty acids (FFA) (right bar) from adipose tissue incubated in vitro. Each bar represents the mean ± SEM of five incubations. Reproduced from Carlson [4] with permission from Blackwell Publishing Ltd.

The hypothesis that the rapid reduction in the plasma concentration of FFA induced by nicotinic acid is due to the inhibition of mobilization of FFA from adipose tissue requires that nicotinic acid is immediately taken up by this tissue. In fact it was shown that nicotinic acid promptly bound to adipose tissue. After intravenous administration of 3H-labelled nicotinic acid to mice, autoradiography showed that this drug had a unique tissue distribution [19]. Figures 4 and 5 show the distribution of radioactivity 5 min after intravenous injection of 131I-labelled albumin and 3H-labelled nicotinic acid respectively. The macromolecule is mainly present in blood as expected whilst nicotinic acid had almost completely left the bloodstream. The highest concentration of labelled nicotinic acid was present in adipose tissue in various locations (subcutaneous, omental, pericardial and perirenal) and in kidneys. These findings support the hypothesis that the effect of nicotinic acid on FFA is due to a direct action of the drug in adipose tissue. The onset of the effect on plasma FFA is rapid as is the uptake of nicotinic acid in adipose tissue.

Figure 4.

Whole-body (mice) autoradiogram. Distribution of radioactivity 5 min after intravenous injection of 131I-labelled albumin. Reproduced from Carlson and Hanngren [19] with permission from Elsevier.

Figure 5.

Whole-body (mice) autoradiogram. Distribution of radioactivity 5 min after intravenous injection of 3H-labelled nicotinic acid. Reproduced from Carlson and Hanngren [19] with permission from Elsevier.

The nicotinic acid receptor

The rapid uptake of nicotinic acid in adipose tissue and its preferential distribution and accumulation in this tissue, described 40 years ago [19], has recently been explained by the demonstration of a specific, high-affinity receptor for nicotinic acid which is highly expressed in adipose tissue [6–8]. The role of the nicotinic acid receptor for the lipid-metabolic effects of nicotinic acid was recently discussed by Karpe and Frayn [20]. Of considerable interest is that nicotinamide, which does not share the lipid-metabolic effects of nicotinic acid, is not bound by this receptor. The endogenous ligand for the receptor is, however, not known.

Immediate metabolic effects of nicotinic acid related to FFA

Plasma FFA are continuously mobilized from adipose tissue with a daily turnover of about 100 g. They have a central role in the energy metabolism of the body, supplying fatty acids for immediate oxidation. They affect other lipid transport systems such as the lipoproteins as precursors to VLDL triglycerides. Furthermore, they exert, besides the lipid-metabolic effects, a profound influence on glucose and insulin homeostasis. The rapid inhibition of FFA mobilization by nicotinic acid can therefore be expected to immediately affect various metabolic processes.

Immediate effects on whole-body and whole-heart metabolism.  Free fatty acids transport fatty acids for immediate oxidative purposes corresponding to at least 1000 calories per day. We therefore wondered what is the effect of inhibition of FFA turnover on whole body and whole heart metabolism?

Whole body

Healthy, fasting young adult men were given nicotinic acid intravenously [21]. Plasma FFA fell from 700 to 240 μmol L−1, the oxygen consumption remained constant before and after the administration of nicotinic acid, the rectal temperature remained unchanged and the RQ rose from 0.75 to 0.82. Heart rate and blood pressure did not change. The conclusion was that whole body metabolism remained unchanged after suppression of FFA mobilization but the oxidative metabolism was shifted from fat to carbohydrate.

Whole heart

Nicotinic acid given to fasting, healthy male volunteers, lowered FFA from about 800 to 200 μmol L−1, changed the myocardial metabolism from predominantly using FFA as a substrate to increased utilization of carbohydrates, as determined from the arterial minus coronary sinus difference in concentration of FFA and of glucose, without changes in the extraction of oxygen [22, 23]. The OER% [oxygen extraction ratio, i.e. the fraction (%) of myocardial oxygen uptake that would be utilized if the extracted substrate was fully oxidized] fell for FFA from about 50% to 10% whilst the OER for glucose increased from 25 % to 50%. The RQ increased from 0.75 to 0.90. In summary, inhibition of FFA mobilization by nicotinic acid did not change whole-heart metabolism but switched its oxidative metabolism from lipids to carbohydrates.

Immediate effects on tissue lipids in the normal state.  After a single injection of nicotinic acid to fasted rats the plasma FFA levels were reduced up to 4 h and the concentration of plasma cholesterol and triglycerides lowered up to 6 h [24]. The triglyceride content of liver, heart muscle and red fibre skeletal muscle was decreased without any change in the cholesterol content of the tissues [24]. The reduction in triglyceride pools in various tissues probably reflects the consumption of these pools for oxidative purposes in a situation characterized by significantly diminished availability of FFA for oxidation.

Immediate effects in emotional stress.  Nicotinic acid inhibits stress-induced lipid changes. Healthy human subjects were exposed to a 2-h standardized emotional stress causing, in comparison with a control group, increases in FFA, plasma triglycerides, blood pressure and urinary excretion of catecholamines. One group exposed to the stress was pretreated with nicotinic acid which inhibited the rise of both FFA and triglycerides but had no effect on the rise in blood pressure or on the increased urinary excretion of catechols [25].

Immediate effects in the diabetic state.

Plasma FFA levels are elevated in diabetes [26, 27] due to increased mobilization from adipose tissue [28]. High levels of plasma FFA may contribute to several of the metabolic abnormalities in diabetes as reviewed in a Minkowski lecture [29]. Important examples are hyperglycaemia and insulin resistance due to a combination of the so-called Randle effect [30] and increased gluconeogenesis and glycogenolysis [31, 32], hypertriglyceridaemia, fatty liver and increased ketone body formation due to a direct precursor relationship. Acute metabolic effects of nicotinic acid in diabetic patients are a slight decrease in blood glucose and an immediate reduction in hepatic ketone body production by more than 90% [33]. In rats made insulin deficient by treatment with insulin antibodies, nicotinic acid produces an immediate fall in plasma FFA, a small decrease in blood glucose and a pronounced fall in the increased levels of β-hydroxybutyric acid as well as a lowering of plasma and liver triglycerides [34].

Nicotinic acid and plasma lipids

From cholesterol to different types of hyperlipidaemias

At the time when the effects of nicotinic acid were explored clinically in the 1960s the field of clinical lipidology was undergoing a dramatic expansion from dealing only with total blood cholesterol to comprise the six types of hyperlipidaemia of the Fredrickson/WHO classification system [35, 36] presented in Table 1. Although this classification is more than 30 years old, it is still in use as a convenient way for describing hyperlipidaemic phenotypes.

Table 1.  The Fredrickson/WHO classification of hyperlipoproteinaemias [36]
TypeLipid increasedLipoprotein increased
  1. LDL, low-density lipoprotein; VLDL, very-low-density lipoprotein; IDL, intermediate-density lipoprotein. aMassive elevation.

ITriglyceridesaChylomicrons
II ACholesterolLDL
II BCholesterol and triglyceridesLDL and VLDL
IIICholesterola and triglyceridesaBeta-VLDL (IDL and chylomicron remnants)
IVTriglyceridesVLDL
VTriglyceridesa and cholesterolChylomicrons and VLDL

Nicotinic acid and plasma lipids in different types of hyperlipidaemia

We summarized our first experience with nicotinic acid in the various Fredrickson/WHO types of hyperlipidaemia in 1973 and showed that nicotinic acid not only lowered plasma cholesterol but also plasma triglycerides and was effective in the hyperlipidaemias of types II–V (Table 2) [37]. With the standard dose given females responded with a more marked reduction in both cholesterol and triglyceride levels than males, which, in part, may be due to their 20% lower body weight (Table 2).

Table 2.  Illustration of the effect of nicotinic acid on plasma lipids in hyperlipidaemia types II–V [37]
Type of HLPaII AII BIIIIVV
LipidCholTGCholTGCholTGCholTGCholTG
  1. Effect of 1 g of plain nicotinic acid three times daily for 1 month on fasting plasma concentrations of cholesterol and triglycerides in the different types of hyperlipoproteinaemia. Average values for the entire cohort: Mean body weight before and after 1 month was 78.1 and 77.7 kg for men (n = 130) and 63.4 and 63.4 kg for women (n = 58) respectively. Neither average values for transaminases nor for fasting blood glucose were changed significantly by the treatment. Uric acid (μmol L−1) rose from 276 to 359, and from 241 to 329 for men and women, respectively, with both increases being highly significant. aCut-off points set in the 1960s: cholesterol, 7.8 mmol L−1; triglycerides, 2.0 mmol L−1. bMean values (mmol L−1).

Men
 Pretreatmentb121.79.02.9169.37.76.92081
 Decrease (%)8292140506216557394
Women
 Pretreatmentb121.79.62.48.25.4
 Decrease (%)253625422659

This early study demonstrated that nicotinic acid not only lowered cholesterol but also triglycerides, which percentage-wise were more lowered than cholesterol. In particular, the effectiveness in the rare types III and V, with their massive elevations of lipids, is noteworthy.

Nicotinic acid and plasma lipoproteins

New methods for plasma lipids and lipoproteins

From the beginning of the 20th century until the 1950s cholesterol was the only plasma lipid that was determined and used in clinical practice. In the 1950s methods for triglyceride estimation were introduced in the clinic. Methods for analysis of plasma lipoproteins were generated [38]. An important breakthrough for lipoprotein analysis in clinical practice was the development of simple precipitation techniques for the determination of the levels of LDL and HDL cholesterol.

Dyslipidaemia as a risk factor for CVD

Dyslipidaemia is present when the concentration of one or more of the plasma lipoproteins is abnormal. Of note is that the dyslipidaemias encompass both hyperlipidaemia and hypolipidaemia.

LDL cholesterol was soon recognized as an important risk factor for CHD [for a review, see 39]. In the 1960s and 1970s the role of other components of dyslipidaemia, including raised levels of triglycerides [40, 41] and low levels of HDL cholesterol [42], were shown by observational studies to be risk factors for CHD. Prospective studies from Framingham later on confirmed that both high triglycerides [43] and low HDL cholesterol [44] are risk factors for CHD.

Effects of nicotinic acid on plasma lipoproteins

VLDL and LDL.  When the evaluations of the effects of nicotinic acid were extended from plasma lipids to lipoprotein analyses it was soon demonstrated that, as expected, cholesterol lowering was, to a great extent, due to a lowering of LDL cholesterol, as demonstrated by Parsons already in the 1950s, and that the lowering of triglycerides was almost entirely caused by lowering of VLDL. In type IV hyperlipidaemia and particularly in types III and V the reduction in the grossly elevated VLDL (and chylomicrons) certainly contributes to the reduction in total plasma cholesterol.

Small, dense LDL.  The LDL fraction is heterogenous and comprises lipoprotein particles of different sizes. Small, dense LDL are considered to be the most atherogenic LDL particles [45], carrying a high risk for clinical atherosclerosis [46].

Immediate-release as well as prolonged-release (PR) nicotinic acid not only lower the total LDL cholesterol but also reduce the amount of small, dense LDL particles [47].

HDL.  In the 1950s Parsons and Flinn [11] had already shown that treatment with nicotinic acid not only lowered total and LDL cholesterol but also increased the concentration of cholesterol in the α1-lipoprotein fraction, i.e. HDL cholesterol. This finding did not attract much attention as the protective properties of HDL for CHD had not yet been recognized, and all attention in the lipid field was focused on cholesterol-lowering. In a study on the effect of nicotinic acid treatment of hyperlipidaemic patients, HDL cholesterol rose by 50% but, as shown in Fig. 6, the subfraction HDL2, the large HDL particles, increased almost by 100% [48]. It is now generally accepted that nicotinic acid is the most powerful drug for raising the concentration of HDL, in particular, the subspecies HDL2.

Figure 6.

Effect of nicotinic acid on plasma lipids, lipoprotein cholesterols (mmol L−1) and Lp(a) (mg protein L−1). Mean ± SEM from 31 dyslipidaemic subjects treated with 4 g of nicotinic acid daily for 6 weeks. Figures indicate mean pretreatment levels. C, cholesterol; TG, triglycerides. Reproduced from Carlson et al. [48] with permission from Blackwell Publishing Ltd.

Lp(a).  Lipoprotein (a) [Lp(a)] is an independent risk factor for all major forms of clinical atherosclerosis [49]. Despite this fact, relatively little attention is given to a patient's level of Lp(a). This is, in part, due to the fact the usual lipid-lowering components such as diet, fibrates or statins do not affect elevated plasma concentrations of Lp(a) [50]. However, nicotinic acid has a pronounced lowering effect on elevated levels of Lp(a). We treated hyperlipidaemic patients with 4 g of plain nicotinic acid for 5–7 weeks [48]. The striking effect on the plasma concentration of Lp(a) is shown in Fig. 6. The level was, on average, decreased by nearly 40%.

Nicotinic acid and special dyslipidaemias

Dyslipidaemia of diabetes.  Diabetes is associated with a cardiovascular mortality which is about two to four times that of nondiabetic subjects. Many risk factors contribute to clinical atherosclerosis in diabetic patients. However, dyslipidaemia and hypertension play a particularly important role. Diabetic dyslipidaemia is characterized by elevated plasma levels of triglycerides and low levels of HDL cholesterol but without major changes in LDL cholesterol [51].

The lipid-modifying properties of nicotinic acid makes this drug tailored for treatment of diabetic dyslipidaemia. However, there has been some concern about the use of nicotinic acid in diabetes because of reports of worsening of the diabetic condition during treatment with nicotinic acid. These reports have, however, been based mainly on uncontrolled case reports with small numbers of outpatients. Controlled, randomized studies have not substantiated this concern about nicotinic acid in diabetes. The Arterial Disease Multiple Intervention Trial (ADMIT) is a large placebo-controlled randomized study of long-term treatment with nicotinic acid, 3 g day−1 for 1 year [52]. Of the 468 patients randomized, 125 had diabetes. Average blood glucose of both the diabetic and nondiabetic patients had increased marginally in response to nicotinic acid whilst HbA1c was unaffected. Plasma triglyceride levels decreased about 25% in diabetic and nondiabetic subjects. HDL cholesterol rose by 29% by nicotinic acid treatment in both patient groups. The authors concluded that despite current recommendations against the use of nicotinic acid in diabetes, lipid-modifying doses of nicotinic acid can be used safely in patients with stable, controlled diabetes. In the HDL Atherosclerosis Treatment Study (HATS), a trial treating patients with low HDL cholesterol with the combination of nicotinic acid 2–4 g daily with a statin for 4 years (see below), a subsample had diabetes (n = 25) [53]. The glycaemic control slightly deteriorated for the diabetic patients during the first 8 months of treatment, but thereafter blood glucose returned to pretreatment levels and remained stable for the remainder of the study. Once again it was concluded that treatment with nicotinic acid, in this study in combination with a statin, is effective, safe and well tolerated in patients with diabetes.

Dyslipidaemia of the metabolic syndrome.  The metabolic syndrome is defined as a clustering of risk factors for CVD almost always occurring in overweight/obese individuals. Patients with the syndrome have a high risk for development of events of CVD. The characteristic dyslipidaemia of the syndrome is elevation of triglyceride levels and reduction in HDL cholesterol. As in the case of the diabetic dyslipidaemia, nicotinic acid is tailored for treatment of the dyslipidaemia of the metabolic syndrome. The metabolic syndrome is associated with a high risk for CHD; the global prevalence is alarmingly high (20% to 40%) and the incidence is increasing with age and time.

Hyperlipidaemia type V and pancreatitis

The pronounced forms of type V hyperlipidaemia, triglyceride levels above 20 mmol L−1, are sometimes associated with frequent recurrent attacks of pancreatitis. The treatment of choice for the dyslipidaemia and the recurrent attacks is nicotinic acid. In my experience doses of 10–20 g daily of plain nicotinic acid are required. With this treatment the triglyceride levels drop to below 5 mmol L−1 and the attacks disappear completely.

The unique effects of nicotinic acid on plasma lipoproteins

The nicotinic acid lipid story started with Altschul's discovery 50 years ago that nicotinic acid in gram doses reduced the cholesterol content in blood. Today's knowledge on the multifaceted effects of nicotinic acid on plasma lipoproteins has advanced enormously since then and is summarized in Table 3. Nicotinic acid has the unique property of modifying all lipoprotein levels in a way beneficial from the point of view of risk for CVD. Nicotinic acid not only lowers the levels of all atherogenic lipoproteins – LDL and VLDL with subfractions as well as Lp(a) – but also raises the concentration of the protective HDL, particularly the subfraction HDL2, and uniquely reduces elevated levels of chylomicrons. The lipid-modifying effects of nicotinic acid are indeed tailored for the dyslipidaemias of diabetes and the metabolic syndrome as well as for supplementing the LDL-lowering effect of statins with raising of the levels of protective HDL.

Table 3.  Summary of the effects of nicotinic acid on plasma lipoprotein classes [9] Thumbnail image of

Nicotinic acid in combination treatments

The ultimate goal for treatment of dyslipidaemia is to achieve optimal levels for the individual patient's lipoproteins. This is today for the greater part focused on obtaining low levels of LDL cholesterol, which is the only lipoprotein for which numerical goals have been set in the guidelines. The risk associated with other lipoprotein abnormalities such as low HDL cholesterol, hypertriglyceridaemia and elevated Lp(a) have not led to formulation of treatment goals, but only to setting of arbitrary ‘risk limits’ for the two former lipoproteins. The increasing awareness of the risk for CVD associated with low HDL cholesterol and for the protective role of this lipoprotein for manifestations of clinical atherosclerosis [54] has focused interest on the use of nicotinic acid as an HDL-raising drug in combination with statins and other drugs primarily lowering LDL cholesterol [55].

The well-documented LDL cholesterol-lowering effects of statins and their beneficial effects on prevention of clinical atherosclerosis have made them drugs of first choice in the treatment of high LDL cholesterol. However, in the landmark studies with LDL lowering by statins in patients with high risk for CHD, such as 4S (Scandinavian Simvastatin Survival Study) [56] and HPS (Heart Protection Study) [57], the reduction in CHD events and mortality in high-risk subjects was never better than 20–40%. Evidently, there is a need for improved treatment of the remaining 60–80% of the high risk population affected by CHD events, despite treatment with LDL-lowering statins. An option for the improvement of the lipid-modifying treatment with statins is to increase the levels of HDL cholesterol and decrease those of triglycerides as the statins only have moderate effects in this regard. For this purpose, nicotinic acid, with its striking HDL-raising and triglyceride-lowering properties, is an ideal drug to combine with statins. Indeed, a dual component, single tablet containing a statin and an extended (PR) formulation of nicotinic acid (Advicor, KOS pharmaceuticals) is registered for lipid modification in the USA.

Table 4 summarizes the lipid effects of nicotinic acid in combination with resins (bile acid sequestrants) or statins in six studies. Treatment with the combination of nicotinic acid with high doses of resin resulted in an increase in HDL cholesterol of around 40% and a concomitant lowering of LDL cholesterol by 30–40% and triglycerides by around 20%. The more patient-friendly combination of nicotinic acid with statins in moderate dose increased HDL cholesterol by nearly 30% and lowered LDL cholesterol and triglycerides by 25–40% and 20–40% respectively. The combination of nicotinic acid with statins was well tolerated in all studies.

Table 4.  Illustration of the effect of combined treatment with nicotinic acid and a resin or a statin
ReferenceNicotinic acid (dose day−1)C/S (dose day−1)TG (%)LDL (%)HDL (%)
  1. C, colestipol; S, statin; IR, immediate-release; SR, slow-release. Percentage change of plasma triglycerides (TG), LDL and HDL cholesterol in dyslipidaemic patients. Values indicate percentage change from pretreatment level. aDose titration to goal for HDL and LDL. Modified from [41] and extended.

Blankenhorn et al. [59]IR 3–12 gC 30 g−16−39+35
Brown et al. [60]IR 0.25–4 gC 15–30 g−29−32+43
Jacobson et al. [96]IR <3.0 gS (fluvastatin) 20 mg−30−40+28
O'Keefe [97]IR 3.0 gS (pravastatin) 40 mg−42−25+29
Gardner [98]IR 1.5 gS (lovastatin) 20 mg−19−30+27
Brown et al. [53]SR 2 gS (simvastatin)−36−42+26
IR 3–4 ga10−a   

Nicotinic acid and lipoprotein metabolism

The lowering of VLDL and LDL by nicotinic acid may be explained by the inhibition of mobilization of FFA from adipose tissue (see Fig. 7). The triglyceride-rich VLDL particle is synthesized in the liver with one molecule of apolipoprotein B (apo B) on the surface and triglycerides – the main constituent – and cholesterol esters in the core. FFA is an important precursor of triglycerides [14–16]. The ripe particle is secreted into plasma where it undergoes delipidation by the action of capillary-bound lipoprotein lipase. By this mechanism, VLDL successively loses triglyceride fatty acids from the core, which are taken up by peripheral cells and the liver. The particle shrinks. Eventually the VLDL particle, which retains it apo B molecule on the surface and its cholesterol esters in the core, turns into an LDL particle equipped with one molecule of apo B. Through the action of the LDL receptor the LDL particle may then by taken up by either peripheral tissues, including arteries, as a cholesterol donator, or by hepatocytes. LDL may also be metabolized to small, dense or modified (oxidized, glycosylated) LDL particles (Fig. 7). If the hepatic supply of FFA is reduced by the action of nicotinic acid the triglyceride synthesis in the liver is decreased, the VLDL triglyceride precursor pool diminished and the result will be a reduction in the secretion of VLDL leading to diminished plasma levels of VLDL and subsequently of LDL.

Figure 7.

Illustration of the role of free fatty acids (FFA) as a precursor to the VLDL–LDL transport chain. Plasma FFA are important precursors of liver triglycerides which are the precursor pool for VLDL triglycerides that constitute the major part of the core of VLDL. Nicotinic acid reduces the flow of FFA into the liver by the inhibition of FFA mobilization from adipose tissue and hence the formation first of the liver triglyceride pool which leads to a reduced synthesis of VLDL due to lack of core material. Reproduced from Carlson and Riccardi [95] with permission from Science Press.

Whilst the lowering of plasma triglycerides (VLDL) and cholesterol (LDL) can be explained by the inhibition of mobilization from adipose tissue of FFA, other mechanisms have been looked for to clarify the mechanisms behind the HDL-raising effect of nicotinic acid.

Nicotinic acid and atherosclerosis

The obvious need for evaluation of the clinical effects of treatment with nicotinic acid – as monotherapy or combined therapy – has been hampered by its side effects, particularly the flush, which has caused doctors to avoid prescribing this drug. Nevertheless, few long-term trials have been carried out for studying the effect of nicotinic acid on either atherosclerosis and/or clinical atherosclerosis, i.e. manifestations such as myocardial infarction, peripheral arterial atherosclerosis and stroke.

Pioneer study in peripheral arterial disease

The earliest study on the effect of nicotinic acid on atherosclerosis in man was carried out in the 1960s by Öst and Sténson [58] in patients with peripheral arterial disease. Patients with intermittent claudication were treated, on average, for 16 months with 3 g of plain nicotinic acid day−1. Plasma cholesterol dropped from a mean value of 9.4 to 7.1 mmol L−1. Repeated femoral arteriographies performed in 29 legs showed regression of the wall changes in the femoral artery in four legs. The authors stressed that they do not have ‘an acceptable control series’ but refer to a large series of clinically similar cases with femoral atherosclerosis given standard treatment, which they had studied with the same angiographic techniques without seeing any regression at all of wall changes with time.

These results showing regression of atherosclerotic lesions by treatment with nicotinic acid are historical because in the 1960s atherosclerotic lesions were considered to be irreversible. This conception was not changed until evidence was obtained from animal experiments in the 1970s showing that atherosclerotic lesions indeed could regress. Later on randomized, controlled studies in man in the 1980s provided proof of reversal of atherosclerosis by means of angiography.

Nicotinic acid and coronary atherosclerosis

The Cholesterol-lowering Atherosclerosis Study (CLAS) enrolled men with coronary artery bypass grafts in the late 1980s [59]. The patients were randomized to treatment with nicotinic acid (3–12 g day−1) combined with the bile acid sequestrant colestipol, 30 g day−1, or to diet/placebo. Quantitative coronary angiography after 2 years of treatment showed regression of lesions in 16% of patients on lipid treatment, but only in 2% of those on diet/placebo. LDL cholesterol and triglycerides had decreased by 39% and 16%, respectively, whilst HDL cholesterol rose to 35%. At a 4-year follow-up, new, progressing atherosclerotic lesions occurred in 14% and 45% in the nicotinic acid/colestipol group compared with 40% and 75% in the diet/placebo group. Furthermore, whilst 38% in the control group had lesion progression in the bypass graft, progression was only demonstrated in 16% of patients in the actively treated group.

The Familial Atherosclerosis Treatment Study (FATS) randomized men with significant coronary atherosclerosis and high levels of apo B, i.e. increased levels of LDL and/or VLDL, to either conventional treatment (diet plus placebo) or to two groups both receiving colestipol 30 g day−1 and in addition, either nicotinic acid up to 4 g daily or lovastatin 40 mg day−1 [60]. The duration of the trial was 2inline image years. In the nicotinic acid/colestipol group LDL cholesterol and triglycerides were reduced by 32% and 29%, respectively, whilst HDL cholesterol increased by 43%. Corresponding changes for the statin/colestipol group were 46%, 9% and 15%. The rates of regression of coronary lesions in the niacin/colestipol, the statin/colestipol and the diet/placebo groups were, respectively, 39%, 32% and 11%. The corresponding figures for lesion progression were 25%, 31% and 46%.

Nicotinic acid in combination withstatins

The ARBITER 2 study recruited patients on stable treatment with statins, average LDL cholesterol 2.3 mmol L−1, mean duration of statin treatment 5 years [61]. They all had CHD and low HDL cholesterol. They were randomized to the addition of placebo or PR nicotinic acid (Niaspan, see below) 1 g day−1 in combination with statin treatment for 1 year. Nicotinic acid increased HDL cholesterol by 21%. Carotid intima-media thickness progressed in the placebo/statin group but remained unchanged in the nicotinic acid/statin group, i.e. a beneficial effect of Niaspan.

Nicotinic acid and CHD

There are two early long-term studies from the ‘prestatin era’ on the effect of nicotinic acid on CHD events. In the Coronary Drug Project (CDP) men having survived a myocardial infarction were randomized to treatment with 3 g of nicotinic acid or placebo daily for 6.5 years. Major CHD events, nonfatal myocardial infarction and cerebrovascular episodes were reduced, but there was no effect on mortality [62]. However, in the 15 year post-trial follow-up of CDP, total mortality had been reduced by 11% for those given nicotinic acid [63]. In the Stockholm Ischaemic Heart Disease Secondary Prevention Study all survivors of a myocardial infarction at a major Stockholm hospital were randomly entered into one control group and one group given nicotinic acid 3 g and clofibrate 2 g daily for 5 years. Total and CHD mortality were reduced by 26% and 36%, respectively, by the lipid-modifying treatment compared with the control group [64].

The HDL Atherosclerosis Treatment Study is a recent secondary prevention study with nicotinic acid and a statin, in patients with low HDL cholesterol and normal LDL cholesterol [53]. The patients were randomized to a double-blind controlled trial for 4 years with placebo or nicotinic acid + statin (N + S). Nicotinic acid was given as a slow-release preparation escalated to 2 g day−1. Patients who had not reached the goal of an increase in HDL cholesterol by 0.25 mmol L−1 were switched to plain nicotinic acid 3–4 g daily. Simvastatin was started with 10 mg daily and increased if, necessary, to reach a goal of 2.3 mmol L−1 for LDL cholesterol. LDL cholesterol and triglycerides decreased by 42% and 36%, respectively, whilst HDL cholesterol increased by 26%. The primary composite cardiovascular end-point occurred in 24% of placebo patients and in 3% of N + S patients, a reduction by 90% in CVD events. The average coronary stenosis grade increased by 3.9% in the placebo group but decreased by 0.4% in the N + S group. The beneficial effect obtained by nicotinic acid and simvastatin is better than that obtained in previous trials with monotherapy with statins. The authors conclude that if the observed reduction in major coronary events (90%) with the combination of nicotinic acid and a statin is confirmed, this treatment regime would represent a substantial advance over current practice.

In the FATS trial, clinical cardiovascular events occurred in 10 of 52 patients in the control group but only in 5 of 94 patients subject to the intensive combined lipid-modifying therapy, the relative risk being 0.27 (95% CI 0.10–0.77).

A 10 year follow-up of the CLAS trial showed that CHD events occurred in eight patients in the nicotinic acid/colestipol group and in 21 placebo-treated patients.

There are two recent posthoc analyses on the clinical effects of treatment with nicotinic acid in patients with the metabolic syndrome, as defined by NCEP ATP III criteria [39], both demonstrating reduction in CHD events. Results from the HATS trial showed beneficial effects treating the subgroup with metabolic syndrome with nicotinic acid combined with a statin [65]. Progression of coronary atherosclerosis was reduced by 90% and coronary events by 40% in the group with metabolic syndrome treated with nicotinic acid plus simvastatin compared to that with the group given placebo. In the analysis of the metabolic syndrome patients in CDP, another posthoc evaluation, the rate of recurrent myocardial infarction (in trial, 6 years) for those with the metabolic syndrome treated with nicotinic acid and placebo were 8.8% and 25.6% respectively (relative hazard 0.31) [66]. The corresponding figures for total mortality (9 years follow-up post-trial) were 53% and 64% respectively (relative hazard 0.73). The authors concluded that nicotinic acid reduced nonfatal myocardial infarction and mortality in patients both with and without the metabolic syndrome.

In summary, the long-term clinical studies with nicotinic acid, especially in combination with other lipid-modifying drugs, particularly statins, have demonstrated significant clinical benefits with reduction in CHD events and mortality. In particular, the combination of nicotinic acid and statins with the marked rise in HDL cholesterol, have presented results better than those obtained in the landmark studies with only statins.

Prolonged-release nicotinic acid

Nicotinic acid as a lipid-modifying drug was, from the beginning, used as plain nicotinic acid, i.e. in crystalline, IR form. Side effects, particularly the immediate flushing, has limited the use of IR nicotinic acid. For this reason sustained-release (SR) formulations of nicotinic acid were developed in the hope of diminishing the flush. Unfortunately, however, the SR preparations were associated with hepatotoxicity and have been abandoned in clinical use. A PR preparation of nicotinic acid, in USA called extended-release (ER), with absorption rates between IR and SR preparations was then developed [67]. It has been marketed in the USA as NiaspanTM and has just been registered in Europe. The typical nicotinic acid side effects are not completely absent with the PR preparation. The flush, however, occurs much less frequently and intensely with PR nicotinic acid than with IR nicotinic acid. The PR nicotinic acid is prescribed in doses of around 1–2 g once daily at bedtime to minimize the flushing effect during daytime.

Monotherapy with PR nicotinic acid

Nondiabetic patients.  An early comparison of the lipid effects of PR nicotinic acid 1.5 g at bedtime with the same dose of IR nicotinic acid in over 200 hypercholesterolaemic (mean cholesterol 7.1 mmol L−1) subjects is summarized in Fig. 8 [68]. The lipid effects of the two preparations are similar except for the effect on Lp(a) which appeared to be better with PR nicotinic acid. Both drugs gave rise to small increases in plasma glucose.

Figure 8.

Comparison of the effect of prolonged-release (PR) and immediate-release (IR) nicotinic acid on plasma lipids and lipoproteins. Mean values of percentage change from pretreatment values of 223 dyslipidaemic patients randomized into three 8-week treatment groups: placebo, PR and IR nicotinic acid respectively. C, cholesterol; TG, triglycerides.

Diabetic patients.  In a study comprising patients with type 2 diabetes (n = 148) treated with PR niacin 1 or 1.5 g once at bedtime, Grundy et al. [69] found that diabetic dyslipidaemia considerably improved with an HDL cholesterol increase of 19% and 24%, respectively, and a reduction in triglycerides by 13% and 28%. Fasting blood glucose did not change after treatment for 4 months. HbA1c did not change in the 1 g group but increased marginally in the 1.5 g group from 7.1% to 7.5%. The authors concluded that PR nicotinic acid is an option for the treatment of dyslipidaemia in type 2 diabetes.

Combined therapy with PR nicotinic acid

A summary of the lipid effects of treatment with the combination of PR nicotinic acid, in doses of 1–2 g once daily, with statins is shown in Table 5. It is evident that the combination of these two drugs has an excellent effect on atherogenic lipid profiles with a reduction in both LDL cholesterol and triglycerides between 30% and 45% and an increase in HDL cholesterol by 25–40% with 2 g of PR nicotinic acid. The increase in HDL cholesterol is clearly dose dependent and rises from treatment with 1–2 g. The beneficial effect of the addition of PR nicotinic acid to statins on progression of atherosclerosis in the ARBITER 2 study [61] is described above.

Table 5.  Illustration of the effect of combined treatment with PR nicotinic acid and a statin
ReferencePR nicotinic acid (g day−1)Statin (mg day−1)TG (%)LDL C (%)HDL C (%)
  1. PR, prolonged-release. aFluvastatin 20, lovastatin 20, pravastatin 10 and simvastatin 10.

Guyton and Capuzzi [99]2Severala 10–20−35−36+26
Kashyap et al. [100]2Lovastatin 40−42−45+41
Bays et al. [101]1Lovastatin−29−39+17
240−49−42+32
Capuzzi et al. [102]1Rosuvastatin 40−39−42+17
210−34−36+24
Hunninghake et al. [103]1Lovastatin 20−26−28+21
240−43−42+30

Side effects

The two regularly occurring side effects with nicotinic acid are flush and the increase in uric acid in blood, the former occurring rapidly with the first dose of nicotinic acid, the latter appearing after treatment for some time. Whilst there is tachyphylaxis for the flush, the increase in uric acid remains during treatment. Of note is that nicotinamide, once again in contrast to nicotinic acid, neither gives rise to a flush reaction nor affects blood lipids.

The flush

The intensive skin flush induced by IR nicotinic acid is the side effect which has limited its clinical use as a lipid drug. As little as 50 mg of nicotinic acid orally within minutes gives rise to the characteristic flush. The unfortunate patient who starts treatment with 1 g of nicotinic acid as the initial dose will experience a very pronounced cutaneous vasodilation often over the whole body. The flush always starts in the face, it is deep red, accompanied by an intense feeling of warmth and itching. It may spread to the arms and chest and occasionally down to the legs and feet. The duration is about half an hour. It is very seldom accompanied by a significant fall in blood pressure. Simultaneously with the cutaneous vasodilation there is an increase in forearm blood flow [70], which can be used as an objective measure of the vasodilatory effect of nicotinic acid (Fig. 9).

Figure 9.

Forearm blood flow. Response to the oral administration at 0 min of placebo, 1 g of nicotinic acid with or without pretreatment with 100 mg of indomethacin, given 1 h before nicotinic acid. Mean value ± SEM for seven healthy, fasting males. Reproduced from Kaijser et al. [70] with permission from Taylor & Francis.

Fortunately there is tachyphylaxis for this side effect occurring within a week after regular dosing. In general we have prescribed plain nicotinic acid starting with 250 mg after lunch on day 1, escalating the dose up to 1 g four times daily after 4 days [9] at which time the flush disappears in most patients. However, a patient on treatment with nicotinic acid who stops taking the drug for a couple of days will experience anew the flush on starting taking nicotinic acid again.

As the skin flush (vasodilatation) is similar to that seen after infusion of prostaglandins [71], the occurrence of flush and the increase in forearm blood flow after the administration of nicotinic acid was studied with or without the prostaglandin synthesis (cyclooxygenase) inhibitor indomethacin [70]. After the administration of 1 g of nicotinic acid by mouth to human subjects the flush was paralleled by a fourfold increase in forearm blood flow. Pretreatment with indomethacin markedly reduced the flush as well as the increase in forearm blood flow (Fig. 9), which suggests that the flush is mediated by the release of prostaglandins. Further support for the role of prostaglandins in the vasodilatory effects of nicotinic acid comes from the observation that women taking nicotinic acid had a twofold increase in the urinary excretion of prostaglandin PGF1α [72]. Ingestion of 500 mg of nicotinic acid increased the plasma levels of the PGD2 metabolite 9α, 11β-PGF2 about 500-fold [73]. In the same study, nicotinamide, which does not cause flush, did not have this effect on the PGD2 metabolite. These studies suggest that PGD2 may be the mediator of nicotinic acid-induced flush as in the case of severe flush episodes of systemic mastocytosis.

Flush and other side effects with PR nicotinic acid combined with statins

A study of the safety and compliance of PR nicotinic acid and lovastatin given at bedtime in a tablet containing 500 and 20 mg, respectively, of the two lipid drugs was recently reported from the USA [74]. Slightly more than 4000 dyslipidaemic patients were given one tablet for 4 weeks, thereafter two tablets daily for 8 weeks. Compliance was 77%, flush was reported by 18% and 6% stopped treatment because of the flush. No case of drug-induced myopathy was observed.

Uric acid.  An increase in plasma uric acid levels occurs regularly during treatment with nicotinic acid. The increase is usually moderate in subjects with normal pretreatment levels. However, the increase may be of clinical significance, particularly in patients with joint problems and/or high pretreatment levels of urate. Therefore, urate levels should be checked regularly during treatment with nicotinic acid. Any rise of clinical significance responds well to treatment with allupurinol in doses of 100–300 mg daily.

Gastrointestinal side effects.  Gastritis-like symptoms may occur in 10–20% of patients of sufficient degree to cause discontinuation of treatment with IR nicotinic acid. Gastric secretion was studied in 10 patients with basal secretion of acid. In nine of these, the secretion of acid increased between 10% and 200% in response to 500 mg of nicotinic acid orally [75]. Also, in Pavlov pouch dogs, intravenous infusion of nicotinic acid caused a significant increase in gastric acid output [75]. Further studies are needed to elucidate the role of the stimulation of secretion of acid for gastrointestinal side effects. Gastric/duodenal ulcers have, however, not been reported as a complication to treatment with nicotinic acid.

Rare and minor side effects.  Dry skin is now and then reported during treatment with nicotinic acid. In a few cases acanthosis nigrans may develop and in a few cases retinal oedema may occur, which disappears immediately when treatment is stopped.

The pros for raising HDL for prevention of CHD

The two unique and pronounced lipid-modifying effects of nicotinic acid, in addition to the lowering of cholesterol (LDL) and triglycerides (VLDL), are the raising of HDL cholesterol and the lowering of Lp(a). Both effects are of clinical significance as they lead to a diminished risk for atherosclerotic CVDs.

Turnover studies have indicated that the levels of HDL in blood, to a large extent, are regulated by the fractional catabolic rate (FCR) of apolipoprotein A-I (apo A-I), the major protein component of HDL [76]. Nicotinic acid has been shown to reduce FCR of apo A-I by decreasing hepatic removal of apo A-I without affecting the uptake of cholesterol esters from the HDL particles into the liver [77]. This mechanism may be one of the explanations for how nicotinic acid raises plasma levels of HDL. An additional mechanism that may raise the levels of HDL cholesterol is the stimulation of the expression of the membrane cholesterol transporter ABCA 1 protein induced by nicotinic acid.

The protective role of HDL in the development of clinical atherosclerosis comes from several pieces of evidence.

  • • Increasing HDL cholesterol levels are strongly associated with decreasing incidence of CHD.
  • • Low HDL cholesterol is common in CHD patients.
  • • HDL promotes RCT by transporting cholesterol from tissues to the liver.
  • • Trials increasing HDL cholesterol levels have decreased incidence of CHD.
  • • Infusion of ‘synthetic HDL’ stimulates RCT, increases cholesterol elimination from the body and reduces coronary artery atheroma volume.

CHD risk and HDL

The classic paper of ‘The Miller brothers’ paved the way for the concept that low levels of HDL cholesterol ‘may accelerate the development of atherosclerosis, and hence IHD, by impairing the clearance of cholesterol from the arterial wall’ [42]. Many observational studies have shown that low HDL cholesterol levels are present in high frequency in survivors of myocardial infarction. Two important prospective cardiovascular studies, The Framingham study in USA [44] and the PROCAM study in Europe [78], clearly showed that low HDL is a risk factor for the occurrence of CHD. Data from prospective epidemiological studies in USA have indicated that an increase in the concentration of HDL cholesterol by 0.025 mmol L−1 (1 mg dL−1) is associated with a 2–3% decrease in CHD risk [79].

Protective effects of HDL, particularly in RCT

HDL has a number of effects contributing to its property as a protective factor for clinical atherosclerosis. The most important of these is the role of HDL in RCT. RCT starts with the cholesterol esters in the foam cells in the arteries (the hallmark of atherosclerosis) and ends up with the faecal steroids (elimination of cholesterol from the body). In this process the ABCA1 cassette binder transports intracellular cholesterol to the cell membranes where it is taken up by lipid-poor HDL particles which then undergo a number of complex processes in blood, involving circulating enzymes and hepatic lipoprotein receptors, whereby they mature into HDL2-particles which deliver the cholesterol to the liver via the SR-B1 receptor.

In addition to its major role in RCT, HDL stimulates other processes that may contribute to its protective actions such as protection against inflammation, protection against oxidation (especially of LDL), protection of endothelial function, protection of NO-production and protection against infection [55].

Trials in which HDL has been raised

There is no trial with lipid-modifying drugs aimed at the prevention of CHD in which there has been a selective increase in HDL without other lipid changes. However, all trials with nicotinic acid described above speak in favour of beneficial effects by raising HDL cholesterol with nicotinic acid. The European Consensus Panel on HDL cholesterol refers particularly to the Veterans Affair HDL Intervention Trial [80, 81], as evidence for the protective effects of HDL [55]. Furthermore, in the recent ARBITER 2 study, PR nicotinic acid in combination with statins increased HDL cholesterol by 21% and slowed progression of atherosclerosis whilst LDL cholesterol was unchanged and triglycerides reduced only marginally [61].

Raising HDL by infusion of ‘synthetic HDL’

There is no lipid-modifying drug that raises HDL cholesterol without a concomitant lowering of LDL cholesterol and triglycerides. It is therefore difficult to obtain specific proof for the protective role of HDL from clinical studies using lipid-modifying drugs. In order to prove that increased levels of HDL lead to a stimulation of RCT with an increased elimination of cholesterol from the body, we infused ‘synthetic HDL’. A single dose of the cholesterol-free ‘synthetic HDL’ complex when infused into subjects with low HDL cholesterol increased the plasma levels of HDL cholesterol by 30%, suggesting a mobilization of tissue cholesterol into the circulation [82]. ‘Synthetic HDL’ was also infused into patients with familial hypercholesterolaemia. Within hours the plasma levels of HDL cholesterol increased, indicating mobilization of cholesterol from tissues into blood [83]. Furthermore, during several days after the infusion the faecal excretion of sterols rose by more than 30% in response to this infusion of ‘synthetic HDL’ particles. In another study, patients with acute coronary syndromes received five weekly infusions of the mutant apo A-I Milano as a liposome–phospholipid complex [84]. The effect on atheroma volume was evaluated by intracoronary ultrasound. The atheroma volume decreased by 1.06% in the group receiving the apo A-I Milano–phospholipid complex (P < 0.02), whilst it increased by 0.14% in the placebo group. This study thus showed a significant regression of atheroma volume as a result of the infusion of ‘synthetic HDL’, which presumably had increased plasma levels of HDL cholesterol.

These two studies, in combination with a number of in vitro experiments unequivocally showing a removal of cholesterol from cholesterol-loaded tissues by HDL or apo A-I liposome preparations [85], clearly suggest that raising HDL can be of benefit in the treatment and prevention of atherosclerosis by stimulating RCT and diminishing atheroma volume.

Beyond plasma lipid modification

Nicotinic acid and thrombogenic factors

In the late 1950s it was shown that there is an increase in plasma fibrinolytic activity after intravenous injection of nicotinic acid with maximum activity occurring in 10–15 min [86]. Nicotinamide, once again in contrast to nicotinic acid, had no such effect. Nicotinic acid added in vitro to blood plasma had no effect. The rapid effect in vivo indicates that nicotinic acid influences the fibrinolytic system directly or possibly by means of its prompt effect on plasma FFA. The activation of plasminogen to plasmin, the enzyme that ‘lyses’ fibrin, is the central step in fibrinolysis. The activation occurs by the tissue plasminogen activator (tPA). tPA is inactivated by plasminogen activator inhibitor 1 (PAI-1), which is synthesized by the liver, adipose tissue and endothelial cells. Nicotinic acid decreases PAI-1 expansion from Hep G2 cells as a result of decreased PAI-1 synthesis [87]. PAI-1 is a risk factor for CHD and is closely associated with hypertriglyceridaemia [88–90]. We therefore studied the effect of treatment with nicotinic acid on PAI-1 plasma levels and other haemostatic factors in men with hypertriglyceridaemia [91]. The main results (Table 6) show the usual unique effect of nicotinic acid on plasma lipids with an increase in HDL cholesterol and a decrease in Lp(a). PAI-1 and tPA were reduced by 31% and 15%, respectively, by nicotinic acid. An additional new finding was that nicotinic acid lowers the plasma levels of fibrinogen by 15%. The reduction in fibrinogen by nicotinic acid has been confirmed both with IR nicotinic acid [92] and PR nicotinic acid [93]. In conclusion, nicotinic acid has favourable effects on two thrombogenic risk factors – PAI-1 and fibrinogen.

Table 6.  Effect of treatment with 4 g daily of IR nicotinic acid for 6 weeks on lipoprotein levels and haemostatic function values [91]
 VLDL TG (mmol L−1)LDL C (mmol L−1)HDL C (mmol L−1)Lp(a) (mg L−1)PAI-1 (U ml−1)tPA (U ml−1)α2-AP (U ml−1)Fibrinogen (g L−1)
  1. IR, immediate-release; PAI-1, plasminogen activator inhibitor-1; tPA, tissue plasminogen activator. α2-AP, α2-antiplasmin.

Pretreatment concentration4.14.80.8836934151.13.6
Treatment change (%)−49−10+27−30−31−15−12−15

Nicotinic acid and cholesterol removal from tissues

The enhancement of RCT by nicotinic acid through the raising of HDL cholesterol was discussed above. However, recent in vitro studies have suggested that nicotinic acid has additional effects that may stimulate RCT. In human macrophages nicotinic acid increases the expression of the ABCA1 protein, which transports intracellular cholesterol to the outer leaflet of the plasma membrane where cholesterol can be rapidly taken up by lipid-poor apo A-I HDL particles circulating in the bloodstream [94]. This is the first phase of RCT which causes tissue removal of cholesterol for subsequent elimination from the body via the bile as faecal steroids. This newly discovered effect of nicotinic acid, independent of the lipid effects, may, in concert with the HDL-raising effect of nicotinic acid, lead to a stimulation of RCT and thereby increase the use of nicotinic acid, alone or in combination with statins, in the prevention and treatment of clinical atherosclerosis.

Fifty years with nicotinic acid: the broad-spectrum lipid drug

A unique combination of intense research interest for the role of cholesterol in atherosclerosis and a good portion of serendipity made the Canadian pathologist Altschul discover the cholesterol-lowering effect of nicotinic acid 50 years ago. Within a couple of years Parsons at the Mayo clinic was a forerunner in describing the complex actions of nicotinic acid on blood lipids, showing that the drug, on the one hand, lowered LDL cholesterol, but, on the other, raised HDL cholesterol.

The broad spectrum of effects on plasma lipoproteins has been reviewed. Nicotinic acid, in a dose-dependent way, lowers triglyceride-rich VLDL and cholesterol-rich LDL and their subclasses, including IDL and small, dense LDL, as well as Lp(a). Furthermore, nicotinic acid is the most potent HDL-raising pharmaceutical. The mechanisms for these lipid/lipoprotein effects are discussed. The combination of nicotinic acid with other lipid-modifying drugs is reviewed, from the early combination with cholestyramine and fibrates to the up-to-date use in combination with statins. The combination of nicotinic acid and statins is well tolerated, is very effective as an aggressive treatment for dyslipidaemia, resulting in pronounced lowering of LDL cholesterol and triglycerides and a marked elevation of HDL cholesterol. As reviewed, plain as well as PR nicotinic acid are well tolerated in type 2 diabetes.

Nicotinic acid rapidly lowers plasma levels of FFA due to the inhibition of their mobilization from adipose tissue. The role of reduced FFA levels as the basic mechanism for the lowering of VLDL and subsequently LDL is discussed. Other metabolic results of the inhibition of FFA mobilization, such as the immediate effects on glucose homeostasis, ketone body production and tissue lipid concentrations are reviewed. These may be relevant in conditions of increased, excessive mobilization of FFA such as diabetes, obesity and stress.

The recent finding of a high-affinity, G-protein-coupled nicotinic acid receptor present particularly in adipose tissue is briefly discussed. It is noteworthy that nicotinamide, which neither has the lipid effects nor the flush effect, does not bind to this receptor. The discovery of the receptor may be the end of the beginning of our understanding of the lipid actions of nicotinic acid.

A summary of long-term trials with nicotinic acid, alone or in combination with other lipid-modifying drugs, is given. In essence, they have shown beneficial clinical effects with reduction in coronary atherosclerosis and prevention of CHD events, CHD mortality and strokes. Of particular importance is that the addition of PR nicotinic acid in combination with ongoing treatment with statins retards the otherwise progressive atherosclerosis of the carotid arteries.

The relationship of the flush induced by nicotinic acid with production and release of prostaglandins is analysed.

Conflict of interest statement

I have no conflict of interest to declare except for having worked with nicotinic acid for more than 40 years.

Ancillary