Dose-Response Characteristics of Cholesterol-Lowering Drug Therapies: Implications for Treatment

Methods

To evaluate the effectiveness and toxicity of lipid-lowering agents at various doses, a MEDLINE search was done for the period of January 1975 through November 1995. Articles were evaluated if they combined subject headings of hyperlipidemia or hypercholesterolemia and one of the following cholesterol-lowering agents: statins (lovastatin, pravastatin, simvastatin, and fluvastatin), bile acid sequestrants (including cholestyramine and colestipol), or niacin. Abstracts were reviewed and received further consideration if a clinical trial that used a cholesterol-lowering drug as monotherapy at two or more doses for hypercholesterolemia was described (Table 1, Table 2, and Table 3). Further inclusion criteria were 1) the clinical trial included at least 10 adult patients per treatment group, 2) the treatment period was at least 1 month in duration, 3) the drug dose did not depend on whether target lipid levels were achieved, and 4) the study design was adequately described. Patients were generally excluded from these trials if their triglyceride levels were greater than 2.8 to 4.5 mmol/L (250 to 400 mg/dL) at baseline and their mean triglyceride levels at study entry were less than 2.5 mmol/L (220 mg/dL).

To evaluate the effectiveness of drug combinations for the reduction of LDL cholesterol levels, we did a MEDLINE search for clinical trials that included at least two of the following classes of lipid-lowering drugs: statins, sequestrants, and niacin. In addition to the criteria stated above, responses in LDL cholesterol levels had to be reported when the two drugs were used separately and in combination (Table 4). Studies reporting the effect of low-dose combination therapy on LDL cholesterol levels were considered separately (Table 5).

We evaluated controlled clinical trials that used combination therapy to lower cholesterol levels to determine whether the reduction is consistent with an additive and independent effect for each drug [45]. To estimate the additive effects of two agents (drug A and drug B), assuming that both act independently, we used the following two-step formula:

1. (initial LDL cholesterol level) x (1-x) = (LDL cholesterol level with monotherapy), where x is the percentage of reduction in the LDL cholesterol level after use of drug A, divided by 100.

2. (LDL cholesterol level with monotherapy) x (1-y) = (LDL cholesterol level with combination therapy), where y is the percentage of reduction in the LDL cholesterol level after use of drug B, divided by 100.

For example, if drug A reduced the LDL cholesterol level by 25%, drug B reduced the LDL cholesterol level by 30%, and the LDL cholesterol level at baseline was 5 mmol/L (193 mg/dL), then the percentage of reduction of these two agents together can be calculated as follows:

1. (5.0) x (1 −0.25) = 3.75

2. (3.75) x (1 −0.30) = 2.63

The expected reduction from these two drugs combined would be 47.4%, a reduction of 2.37 mmol/L from a baseline value of 5.0 mmol/L.

Role of Dietary Therapy

The importance of dietary therapy in the management of hypercholesterolemia has recently been reviewed [46]. Modest reductions of 3% to 10% in the LDL cholesterol level are frequently achieved by persons in the United States who adopt the step I cholesterol-lowering diet devised by NCEP [46-49]. Somewhat greater reductions may be achieved by persons who can adhere to the step II diet, which allows less dietary intake of saturated fat and cholesterol [50]. The ability of an overweight patient to lose weight may be an important predictor of LDL cholesterol response among patients with hypercholesterolemia who follow the step I diet [47, 48]. Because reductions of more than 10% to 15% are uncommon with dietary therapy, patients with moderate to severe hypercholesterolemia are unlikely to achieve the target level with dietary therapy alone. For example, among patients with moderate hypercholesterolemia who received an aggressive dietary intervention program, only 7% achieved the target level and thus avoided therapy with cholesterol-lowering drugs [51]. However, the individual patient's response to diet varies and is not easily predicted; therefore, dietary intervention should be attempted for most patients before drug therapy is considered [8]. Although implementation of the NCEP step I diet will obviate the need for drug therapy in few patients, close adherence may allow the reduction of drug doses or may avoid the need to use several cholesterol-lowering agents together to achieve the target LDL cholesterol level.

Dose-Response Characteristics of Cholesterol-Lowering Drugs

Statins

Fluvastatin, lovastatin, pravastatin, and simvastatin are the most effective agents for lowering levels of LDL cholesterol. These drugs are well tolerated; fluvastatin, pravastatin, and simvastatin can be taken once daily at bedtime, and lovastatin can be taken twice daily with meals, particularly when the dose exceeds 20 mg. The initial recommended doses are 10 mg for simvastatin and 20 mg for the other statins. Hepatotoxicity and myopathy, the major adverse effects associated with these agents, are unusual and only rarely require cessation of therapy. Rash, heartburn, and sleep disturbance have occasionally been noted but are uncommon. Statins have much smaller effects on triglyceride and high-density lipoprotein (HDL) cholesterol levels. In patients who do not have hypertriglyceridemia, statins reduce triglyceride levels by approximately 10% to 25%. Statin therapy increases HDL cholesterol levels by 5% to 10%, but responses vary.

Studies that evaluate dose-response relations for the statins are reviewed in Table 1 and show that increasing the dose of each statin to more than 20 mg produces only small incremental reductions in the LDL cholesterol level. The expanded clinical evaluation of lovastatin (EXCEL) trial [12], which is the largest dose-response study of a statin, randomly assigned 8245 patients to receive 20 mg, 40 mg, or 80 mg of lovastatin for 48 weeks. The 20-mg dose achieved 60% of the maximum LDL cholesterol-lowering effect that was seen with the 80-mg dose. Tuomilehto and colleagues [24] found little increase in efficacy when they increased the dose of simvastatin from 10 mg (which reduced LDL cholesterol levels by 30%) to 40 mg (which reduced LDL cholesterol levels by 40%). Another trial [15] found that pravastatin reduced LDL cholesterol levels by 19% at 10 mg, 25% at 20 mg, and 27% at 40 mg. Similar findings have been reported for fluvastatin, although the maximum reduction achieved at the highest recommended dose (80 mg) is less than that seen with other statins. These studies show that the effectiveness of the statins does not increase proportionally with dose and that lower doses capture most of the expected efficacy. Approximately two thirds of the expected maximum response can generally be expected with only one quarter of the highest dose.

Significant toxicity with the statins is unusual and (although the toxicity of lovastatin was shown to be dose-related in the EXCEL trial) is not clearly related to dose. In EXCEL [12], 4% of patients receiving 80 mg of lovastatin but only 1% of patients receiving 20 mg of lovastatin developed abnormalities of liver function (defined as aminotransferase levels > three times the upper limit of normal) [12]. Six percent of patients receiving 80 mg of lovastatin and 2% of patients receiving 20 mg of lovastatin developed symptoms of severe myalgia. High doses of other statins may be associated with even less toxicity. Studies comparing responses to various doses of simvastatin, pravastatin, and fluvastatin do not report a higher incidence of adverse effects at higher than at lower doses, but the power to detect these differences is smaller than in the EXCEL trial [16-1820-24, 38]. In two large, controlled clinical trials with more than 5 years of follow-up [3, 6], the rates of adverse events and drug discontinuation for simvastatin at doses of 20 mg or greater and pravastatin at doses of 40 mg were equal to those in groups receiving placebo [3, 6]. These data indicate that statins are well tolerated and rarely cause significant hepatic or muscle toxicity, even at higher doses.

In summary, statins have a nonlinear dose-response relation, and most of their effectiveness is preserved at lower doses. Statins at higher doses may further reduce LDL cholesterol levels by 5% to 10% and are well tolerated; they have an adverse event rate similar to that of placebo. To contain cost, statins may be started at low doses and increased gradually until an effective response is achieved. However, if cost is not a concern (for example, if the cost of each tablet is the same for any dose), starting therapy at a higher dose to achieve maximum effectiveness can also be considered, particularly in the setting of moderate to severe hypercholesterolemia.

Bile Acid Sequestrants

Levels of LDL cholesterol are effectively lowered by bile acid sequestrants. These agents bind cholesterol and bile acids but are not absorbed during transit through the gastrointestinal tract. The recommended dose is 2 to 6 scoops daily (cholestyramine resin, 1 scoop = 4 g; colestipol hydrochloride, 1 scoop = 5 g). Although no systemic toxicity is associated with the use of these agents, sequestrants commonly induce unpleasant gastrointestinal side effects, including constipation, dyspepsia, flatulence, and eructations, all of which limit patient acceptance. In addition, patients frequently find that available preparations of sequestrants are unpalatable. For these reasons, long-term compliance with sequestrants is poor, and only about 50% of patients continue therapy for longer than 1 year [52]. Sequestrants can also cause hypertriglyceridemia, particularly in patients who have an elevated blood triglyceride level before therapy.

Fewer dose-response trials have been done with sequestrants than with statins. The trials that have been done [27-32] show a similar nonlinear dose response: Increasing the dose by 2 to 3 scoops daily does not substantially decrease LDL cholesterol levels further. In a randomized, double-blind trial, Superko and colleagues [27] found that 1 scoop of colestipol hydrochloride reduced LDL cholesterol levels by 16%, 2 scoops reduced them by 23%, and 3 scoops reduced them by 27%. More than 50% of the maximum effectiveness of the drug was achieved with less than one third of the highest dose. In studies using higher doses of sequestrants [29, 31], 3 scoops daily provided three quarters of the lowering effect that was achieved by 6 scoops, suggesting little benefit from increasing the dose beyond 3 scoops daily.

Bile acid sequestrants prescribed at higher doses are associated with substantial adverse gastrointestinal side effects. In the large Lipid Research Clinics Primary Prevention Trial [5], more than 1900 patients were randomly assigned to take 3 packets (1 packet is equal to 1 scoop) of cholestyramine resin twice daily during a 7-year period; by the end of the first year of the study, 39% of patients reported constipation, 32% had abdominal gas, and 27% reported heartburn. All of these side effects occurred substantially more often in the treatment group than in the placebo group. Only half of the enrolled patients reported taking 5 or more packets daily at the conclusion of the study. Trials that have studied various doses of bile acid sequestrants have been too small to accurately categorize toxicity by dose. Available clinical data documenting poor long-term tolerance to sequestrants does not show whether lower doses are associated with improved long-term adherence. However, clinical experience suggests that lower doses are better tolerated and may improve patient compliance.

In conclusion, sequestrants have a nonlinear dose-response relation and retain most of their effectiveness at lower doses. As few as 2 to 3 scoops daily reduce LDL cholesterol levels by 15% to 25%. Higher doses of sequestrants (4 to 6 scoops daily) have poor long-term compliance rates because of their sandy, gritty taste and gastrointestinal side effects. Therefore, if lower doses of sequestrants are prescribed, their efficacy will generally be retained, the rate of drug discontinuation may be reduced, and cost will be limited.

Niacin

In addition to lowering LDL cholesterol levels, niacin (nicotinic acid) lowers triglyceride levels and increases HDL cholesterol levels. Unfortunately, niacin is commonly associated with side effects (particularly symptoms of vasodilation, such as flushing, pruritus, and rash) that limit patient acceptance. These disturbing cutaneous side effects often abate if niacin therapy is continued [8]. Less commonly, niacin causes myalgias, dyspepsia, heartburn, hyperglycemia, gouty arthritis, acanthosis nigricans, and elevations in hepatic aminotransferase levels. Even more rarely, niacin may induce visual changes associated with cystoid maculopathy [53]. Largely because of these adverse effects, discontinuation rates for niacin therapy after 1 year have been reported to approach 50% in the managed care setting [52] and were high even in a lipid disorder specialty clinic [54].

Partially effective strategies have been developed that minimize cutaneous vasodilatory effects. These strategies include gradually increasing the dose from as little as 250 mg daily to as much as 2 to 3 g daily over 6 to 8 weeks while the dose is titrated against the occurrence of flushing. Vasodilatory effects can also be minimized by using aspirin, taking niacin with meals, and avoiding simultaneous consumption of hot beverages [8]. However, even with these strategies, discontinuation rates remain high, especially with higher doses [55, 56]. Sustained-release niacin preparations, designed for slow release into the gastrointestinal-lumen, produce less flushing but otherwise have a toxicity profile similar to that of immediate-release niacin [33, 34, 54]. At equal doses, sustained-release niacin is probably more potent and more toxic than immediate-release niacin [33, 57]. In particular, sustained-release niacin may be more likely than immediate-release niacin to cause severe hepatotoxicity, especially at doses greater than 2 g [58]. For this reason, recent expert opinion [8] has suggested that sustained-release niacin should have a limited role in the management of hypercholesterolemia and that the dose of this agent should be limited to 2 g.

Because niacin is associated with significant toxicity that limits compliance, it is important to understand whether effectiveness is compromised with lower doses. Although the results of many dose-response studies are available for the statins, few controlled clinical trials of dose-response for niacin have been published. However, these studies [33-35] suggest that the response to niacin is linear and directly proportional to the dose used; that is, doubling the dose approximately doubles the reduction obtained, within the dose range studied (Table 3). McKenney and colleagues [33] sequentially increased the dose of niacin from 0.5 g to 3 g in a randomized, controlled clinical trial comparing immediate-release with sustained-release niacin. They reported that 1 g of immediate-release niacin reduced LDL cholesterol levels by 6%, 2 g reduced levels by 16%, and 3 g reduced levels by 22%. The ability to lower LDL cholesterol levels was similar for the two types of niacin, although sustained-release niacin was more potent and lowered LDL cholesterol levels by 12% at 1 g, 33% at 2 g, and 50% at 3 g. Using a different preparation of sustained-release niacin, Keenan and colleagues [35] reported an 11% reduction in LDL cholesterol levels with 1 g daily and a 26% reduction with 2 g daily. Knopp and colleagues [34] randomly assigned patients to receive immediate-release or sustained-release niacin and found that immediate-release niacin reduced LDL cholesterol levels by 14% at 1.5 g and 27% at 3 g. In that study, evaluation of the response to sustained-release niacin was limited by poor compliance that resulted from use of the higher dose.

Niacin is one of few drugs that can substantially increase HDL cholesterol levels. Levels of HDL cholesterol lower than 0.91 mmol/L (35 mg/dL) are considered a major risk factor for coronary heart disease [8, 59], and current recommendations support the use of drugs that improve LDL and HDL cholesterol levels in patients with hypercholesterolemia who have low HDL cholesterol levels [8]. Several studies have evaluated the effect of niacin on the HDL cholesterol level in patients with hypercholesterolemia. McKenney and colleagues [33] sequentially increased the dose of immediate-release niacin every 6 weeks and found that the HDL cholesterol level increased by 25% among patients treated with 1 g daily, 31% among patients treated with 2 g daily, and 35% among patients treated with 3 g daily, suggesting that a nonlinear relation exists and that most of the increase occurs at the lower doses [33]. Knopp and colleagues [34] reported that HDL cholesterol levels increased 14% during a 1-month lead-in phase with immediate-release niacin at 1.5 g; increasing the niacin dose to 3.0 g for an additional 5 months elevated HDL cholesterol levels by 24% above the baseline value. This sequential-dose study did not address the possibility that 1 month may not be enough time to achieve maximum increases at a given dose. In controlled clinical trials that evaluated dose-response relations for sustained-release niacin, elevations in HDL cholesterol levels were approximately 10% to 15% and were similar at doses of 1.5 and 3.0 g [33-35]. In the two randomized studies that have compared immediate-release with sustained-release niacin [33, 34], the immediate-release formulation more effectively increased HDL cholesterol levels. The effect of 1 g of sustained-release niacin has varied in other clinical trials: This preparation has increased HDL cholesterol levels by less than 5% [41, 60], by 5% to 15% [61, 62], and by more than 15% [63-66]. This variability may depend in part on the preparation of sustained-release niacin used. In contrast, studies of immediate-release niacin at similar doses more consistently report increases greater than 15% [40, 55, 56, 67-69]. Although they are not conclusive, these findings suggest that immediate-release niacin may increase HDL cholesterol levels in doses as low as 1.5 g and that additional dose increases may produce only small incremental improvements. Further, immediate-release niacin may achieve greater and more consistent increases than are produced by most sustained-release preparations.

Relatively little information exists about the relation between dose-related and adverse effects for niacin. At initial doses as low as 500 mg daily, approximately half of 23 patients receiving niacin had transient symptoms of vasodilation, which often resolved despite continued escalation of the niacin dose. Liver function test results and glucose levels increased proportionately with dose (particularly for the sustained-release niacin preparation); 39% of patients were unable to tolerate an increase to 3 g of immediate-release niacin, and 79% were unable to tolerate an increase to 3 g of sustained-release niacin [33]. In another study [68], 35% of patients were unable to tolerate titration of niacin to a total dose of 4.5 g daily because of adverse side effects. In clinical practice, adherence rates for longer periods may be even worse [52]. On the basis of these data and our own observations [55], it appears that higher doses of niacin are associated with greater risk for toxicity and higher rates of discontinuation.

In summary, the dose-response curve of both immediate-release and sustained-release niacin for LDL cholesterol levels appears to be relatively linear, and major reductions (that is, reductions of 15% in LDL cholesterol levels) are more commonly achieved with daily doses of 1.5 g or greater. On the other hand, most of the elevation in HDL cholesterol levels occurs with 1.5 g of niacin. Therefore, relatively low doses may still favorably affect the ratio of LDL to HDL cholesterol, which may be a better predictor of coronary heart disease than are LDL cholesterol levels alone [70]. Doubling the dose of niacin from 1.5 g to 3.0 g would probably reduce LDL cholesterol levels but not necessarily increase HDL cholesterol levels. Because adherence rates potentially improve with a lower dose of niacin, an initial dose of 1.0 g to 1.5 g of immediate-release niacin daily (increased gradually from 100 mg to 300 mg daily for several weeks) has been recommended [8]. This lower dose of niacin dose may suffice, particularly in persons with mildly elevated LDL cholesterol levels and low HDL cholesterol levels. If this initial dose is well tolerated and further reductions in LDL cholesterol levels are desired, additional dose increases can be considered. The patient should be instructed to return to a lower dose if serious adverse effects occur.

Additive Effect of Lipid-Lowering Drugs

When an initial dose of a single agent reduces LDL cholesterol levels but not to the extent desired, the clinician must decide whether to increase the dose or to add a second agent. The relative benefit of combination drug therapy depends on whether the response to combination drug therapy is less than additive, additive, or synergistic [greater than additive]. If the effect is less than additive, increasing the dose of an agent may be more appropriate than adding a second agent. On the other hand, if the effect of several agents on LDL cholesterol levels is synergistic, then using several agents together may be more effective than increasing the dose of a single drug. Because of the nonlinear relation between response of the LDL cholesterol levels and dose of statins and sequestrants, an additive effect at a lower dose predicts that combination therapy would be more effective than a higher dose of a single agent. We reviewed studies that evaluated the effectiveness of combination therapy compared with that of monotherapy to determine whether 1) the effect of combined drug therapy on the response of LDL cholesterol levels is additive and 2) low-dose combination drug therapy reduces LDL cholesterol levels more effectively than does high-dose monotherapy.

Table 4 shows the results of six studies [36-41] that evaluate the lowering effect of two agents individually and in combination. Two studies [36, 38] had multiple dosing groups, allowing several evaluations of predicted compared with observed responses. Three studies evaluated combination therapy with statins and niacin and found that the predicted reduction in LDL cholesterol levels was almost identical to the additive effect that would be expected from the use of either agent alone. Three studies evaluated statins and sequestrant combinations and found the LDL cholesterol reductions to be additive. Predicted and observed responses of LDL cholesterol levels for combination therapy were similar (mean ± SE, −40.0%± 2.0% compared with −39.7%± 2.6%; difference, 0.3% ± 1.1%; P > 0.2 by paired t-test) and highly correlated (r = 0.92, P < 0.001) (Figure 1).

Because the dose-response curve for statins and sequestrants used as monotherapy is not linear, adding low doses of these agents together might be as effective as much larger doses of either medication used alone. This hypothesis has been tested in three studies [42-44] (Table 5). Schrott and colleagues [44] randomly assigned patients to receive either 40 mg of lovastatin as monotherapy, 20 mg of lovastatin in combination with 5 g of colestipol, or 20 mg of lovastatin in combination with 10 g of colestipol. Similar reductions in LDL cholesterol levels (38% below baseline values) were seen in both the group receiving 40 mg of lovastatin and the group receiving 20 mg of lovastatin and 5 g of colestipol; 20 mg of lovastatin and 10 g of colestipol produced the greatest reductions in LDL cholesterol levels (48% below baseline values). These low doses of sequestrants were particularly well tolerated; adherence rates were greater than 90%. Denke and colleagues [42] found that 8 g of cholestyramine plus 5 mg of lovastatin reduced LDL cholesterol levels as much as did monotherapy with 20 mg of lovastatin. Hagen and colleagues [43] similarly reported that 20 mg of fluvastatin plus 1 scoop of cholestyramine lowered LDL cholesterol levels as much as did monotherapy with 40 mg of fluvastatin. These studies suggest that in addition to minimizing toxicity, combinations of low-dose sequestrants and statins produce reductions in LDL cholesterol levels greater than or equal to higher doses of either drug used alone.

Treatment Strategies

Overview

In this section, we outline a therapeutic approach for the management of elevated LDL cholesterol levels in patients who do not have hypertriglyceridemia (those with a blood triglyceride level <2.82 mmol/L [250 mg/dL]) and in whom dietary intervention has not produced the desired therapeutic result. This approach is derived from observations of the relation between drug dose, effectiveness, and toxicity, as previously discussed. It takes advantage of the dose-response relations of drugs that lower LDL cholesterol levels, the additive effects of combination therapy, and the benefits of lower niacin and sequestrant doses to minimize toxicity. Because adherence to therapy may improve when drug regimens are kept relatively simple [71, 72], we emphasize increasing the dose of existing agents rather than introducing additional lipid-lowering drugs unless increasing the dose is likely to result in considerable toxicity, cost, or both. These suggestions provide a general therapeutic framework; specific patient characteristics (such as the presence of diabetes or renal insufficiency) may necessitate tailoring therapy to the needs of the individual patient.

General recommendations about lipid-lowering therapy should be considered when drug therapy is initiated [8]. First, appropriate dietary therapy should be maintained when lipid-lowering drug therapy is started. Second, appropriate laboratory testing should be done to exclude secondary causes of hypercholesterolemia (including diabetes, hypothyroidism, the nephrotic syndrome, and obstructive liver disease). Third, niacin should be used cautiously (if at all) in patients with diabetes, and sequestrants should not be used in patients with hypertriglyceridemia (that is, a blood triglyceride level >2.82 mmol/L [250 mg/dL]). Fourth, a full lipid profile, including measurements of HDL cholesterol, triglyceride, and LDL cholesterol levels should be obtained approximately 6 weeks after beginning or changing drug therapy; laboratory tests should also be done to monitor for drug toxicity. If the subsequent lipid profile is not consistent with the expected response to the medication, then the lipid profile should be repeated and the two measurements averaged. Fifth, if a medication does not improve the lipid profile (<15% decrease in LDL cholesterol levels or <15% increase in HDL cholesterol levels), a second agent should be substituted rather than added.

Selection of the Initial Agent

Moderate to Severe Hypercholesterolemia

If LDL cholesterol levels are greater than 20% to 25% above target levels (as they are with moderate to severe hypercholesterolemia), statins at moderate doses (such as 20 mg daily) are suggested as initial agents because they reliably produce large reductions, are well tolerated, and approach the cost-effectiveness of niacin when used to treat moderate to severe hypercholesterolemia [11].

Mild Hypercholesterolemia

For elevations in LDL cholesterol levels that are 20% above target levels or less, several reasonable alternatives to statin therapy are available. More aggressive nonpharmacologic therapy can be considered, including further efforts to encourage adherence to diet, weight reduction, and initiation of an exercise program. Pharmacologic therapy with either sequestrants or niacin can also be considered in this setting; monotherapy with either agent may be sufficiently effective to achieve target levels.

Combination Drug Therapy

In patients with moderate to severe hypercholesterolemia, initial management with a statin may fail to achieve the target level of LDL cholesterol. In this case, the clinician has to choose whether to increase the statin dose or add a second agent. Because higher statin doses have been associated with low toxicity, this decision could reasonably be influenced by the cost of each option. If the cost to the patient increases proportionately with the dose of statin, then adding low-dose niacin or sequestrants may be more cost-effective than increasing the dose of statin. On the other hand, if the cost of statin does not increase appreciably as the dose is increased, then increasing the statin dose may be the preferred option. (At the same time, however, it is important to recall that the added cost of increased doses must be borne by someone; in this case, by payors other than the patient.) The clinician should be aware that most of the anticipated response may have already been achieved at the lower dose level. If the target levels are still not attained after the highest doses of statin have been prescribed (40 mg with simvastatin or pravastatin; 80 mg with fluvastatin or lovastatin), then the addition of niacin or sequestrants can be considered.

Patients with a Low Level of High-Density Lipoprotein Cholesterol

The HDL cholesterol level should be taken into account during the decision to add either niacin or sequestrants as a second agent in patients with moderate to severe hypercholesterolemia [8]. The addition of immediate-release niacin in low doses (1.0 g to 1.5 g daily) is particularly appropriate if HDL cholesterol levels are less than 0.91 mmol/L (35 mg/dL) because this dose is likely to be well tolerated and to effectively increase HDL cholesterol levels. If the LDL cholesterol levels remain above target levels despite the addition of niacin, then increasing the statin to the highest dose (if this has not already been done) or adding low-dose sequestrants (1 to 3 scoops taken daily with breakfast, dinner, or both) can be considered. Because further increasing the dose of niacin may be associated with a high frequency of adverse effects that limit long-term adherence, this approach may be the last option for further reducing LDL cholesterol levels.

Patients with a Normal or High Level of High-Density Lipoprotein Cholesterol

If the HDL cholesterol level is greater than 0.91 mmol/L (35 mg/dL), low-dose sequestrants are preferable to low-dose niacin as a second agent. Compared with low-dose niacin, lower doses of sequestrants probably produce greater reductions in LDL cholesterol levels and may be better tolerated. For reductions beyond those achieved with moderate-dose statins and low-dose sequestrants, the statin dose can be increased to maximum doses (if this has not already been done). This provides effective lowering of LDL cholesterol levels but allows treatment with only two lipid-lowering drugs. However, if target levels are still not attained, the addition of niacin as a third agent can be considered, particularly if the patient is willing to attempt the higher doses of niacin that are necessary to achieve optimal reductions in the LDL cholesterol level (1.5 g to 4.0 g of immediate-release niacin or 1.0 g to 2.0 g of sustained-release niacin). To improve patient compliance and tolerance, the clinician should titrate the dose of niacin carefully and monitor side effects and lipid response.

Conclusions

Statins, bile acid sequestrants, or niacin, alone or in combination, are likely to be required in addition to dietary therapy in many patients with elevated LDL cholesterol levels and in some with reduced HDL cholesterol levels to lessen risk for cardiovascular disease. On the basis of dose-response characteristics, we propose a strategic framework to guide the use of these drugs alone or in combination. This strategy emphasizes the benefits of using lower doses of drugs in combination to enhance efficacy, improve long-term adherence, and reduce the overall cost of therapy. Judicious use of cholesterol-lowering medications alone and in combination should facilitate the efforts of clinicians to optimally reduce LDL cholesterol levels and to minimize cardiovascular morbidity and mortality rates.

Dr. Hiatt: Milwaukee Veterans Administration Medical Center, 5000 National Avenue, Milwaukee, WI 53295.