Niacin in HIV-Infected Individuals with Hyperlipidemia Receiving Potent Antiretroviral Therapy
Clinical Infectious Diseases
August 1 2004;39:419-425
Marisa T. Gerber,1 Kristin E. Mondy,1 Kevin E. Yarasheski,2 Henning Drechsler,1 Sherry Claxton,1 John Stoneman,1 Debra DeMarco,1 William G. Powderly,1 and Pablo Tebas1
Divisions of 1Infectious Diseases and 2Endocrinology, Department of Internal Medicine, Washington University School of Medicine, Saint Louis, Missouri
"...This small, open-label pilot study evaluated whether ER-niacin could be an effective option for treatment of antiretroviral therapy--associated dyslipidemia...
"...Our findings indicate that ER-niacin may be a useful option for lipid-lowering therapy for HIV-infected people with normal glucose tolerance. ER-niacin therapy significantly reduced serum levels of triglycerides, non-HDL cholesterol, and total cholesterol... niacin therapy induced a form of insulin resistance in the basal state... these effects did not lead to frank diabetes during the 14-week course of therapy... our study found that ER-niacin did not cause cutaneous flushing to a limiting degree in our study participants when given with 325 mg of aspirin. Flushing did not necessitate discontinuation of ER-niacin therapy for any of the patients. None of the subjects in this small pilot study developed significant elevations in transaminase levels or myopathy... Our findings indicate that ER-niacin may be a useful option for lipid-lowering therapy for HIV-infected people with normal glucose tolerance..."
Background. Extended release (ER)--niacin therapy, which has been associated with reduced glucose tolerance in human immunodeficiency virus (HIV)--seronegative individuals, has not been evaluated in the HIV-infected population.
Methods. This open, prospective trial evaluated the safety and efficacy of ER-niacin therapy for antiretroviral therapy--associated dyslipidemia. Fourteen individuals received ER-niacin at maximum doses of 2000 mg per day for 14 weeks.
Results. Significant reductions in serum levels of triglycerides (P = .02), total cholesterol (P = .005), and non-HDL cholesterol (P = .04) were seen after ER-niacin therapy. Seven of 11 subjects were glucose intolerant after ER-niacin therapy; for 3 of these subjects, this was a new finding.
At baseline, the subjects had considerable hypertriglyceridemia (median triglyceride level, 489 mg/dL; range, 341--715 mg/dL). All 14 of the subjects had fasting triglyceride levels of >200 mg/dL. The median total cholesterol level was 245 mg/dL (range, 165--537 mg/dL) LDL-c levels were only mildly elevated (median, 121 mg/dL). Only 4 of 14 subjects had baseline LDL-c levels of >160 mg/dL. Non-HDL cholesterol levels were >190 mg/dL in 10 (71%) of 14 patients (median, 209 mg/dL).
After 14 weeks of niacin therapy, median triglyceride levels had decreased by 34% (P < .02), and median total cholesterol levels had decreased by 14% (P < .05). No significant changes were observed in LDL-c and HDL-c values. However, the median non-HDL cholesterol level decreased by 19% (from a median of 200 mg/dL to a median of 177 mg/dL; P < .04). After the end of therapy, only 5 (42%) of 12 subjects had levels of non-HDL cholesterol of >190 mg/dL. In subjects who were receiving lipid-lowering agents prior to enrollment on this trial, ER-niacin therapy for 14 weeks lowered triglyceride and non-HDL cholesterol levels to values similar to those achieved with the lipid-lowering agents used previously.
b-Cell sensitivity to basal glucose levels increased significantly without concomitant increase in overall glucose disposition indices. The values for the homeostasis model of insulin resistance index increased significantly (P = .005).
Glucose tolerance and insulin sensitivity were evaluated before and after ER-niacin therapy in 11 of the 14 subjects. In addition to the 2 subjects who discontinued ER-niacin therapy prematurely, 1 subject did not undergo an OGTT after completing 14 weeks of therapy. At baseline, 6 (43%) of 14 individuals were glucose intolerant (i.e., they had a 2-h OGTT value of >140 mg/dL). After 14 weeks of ER-niacin therapy, 7 (64%) of 11 subjects were glucose intolerant. For 3 subjects, this was a new finding. None of the subjects developed overt diabetes.
The calculated overall insulin sensitivity (i.e., the minimal model index of the data from the entire 5-h OGTT) was not significantly reduced after 14 weeks of ER-niacin therapy. However b-cell sensitivity to the basal glucose level and basal insulin secretion rate were increased significantly. As a result, the disposition index was unchanged (table 4). On the basis of the homeostasis model of insulin resistance (HOMA-IR) values, we found that fasting insulin sensitivity was significantly reduced after ER-niacin treatment.
The subjects receiving protease inhibitor--based antiretroviral regimens tended to have higher HOMA-IR values at baseline, compared with subjects not receiving protease inhibitor--based regimens. This trend persisted after 14 weeks of ER-niacin therapy: HOMA-IR values worsened to a similar degree in both groups.
Conclusion. ER-niacin's role in the treatment of antiretroviral therapy-associated dyslipidemia requires further evaluation, but the results of this pilot study indicate that it is safe and tolerated and provides a valuable treatment option.
The introduction of HAART has resulted in significant reductions in the morbidity and mortality associated with HIV infection. Unfortunately, the use of HAART has also been associated with significant adverse effects, including dyslipidemia and insulin resistance, which are known risk factors for the development of atherosclerotic disease in the general population. Although the significance of these risk factors remains undefined in the HIV-infected population, there is preliminary evidence of an increased risk of myocardial infarction associated with antiretroviral therapy in HIV-infected individuals. Although dyslipidemia has been described in association with both antiretroviral therapy and HIV disease itself, it is more common and more severe when it occurs in association with HAART. Current options for treatment of antiretroviral therapy--associated dyslipidemia are limited by drug-drug interactions and low efficacy.
The main advantage of niacin over other agents that decrease lipid levels (hereafter, "lipid-lowering agents") is its beneficial effect on levels of all lipoprotein fractions, including high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), triglycerides, and lipoprotein-a. However, the use of traditional, immediate-release niacin has been limited by adverse effects, including cutaneous flushing and potential insulin resistance. A new once-daily formulation of extended-release (ER) niacin (Niaspan; Kos Pharmaceuticals) has been shown to have a favorable hepatic safety profile and a reduced incidence of cutaneous flushing in the HIV-negative population. Although niacin is not a new medication, ER-niacin may be a useful new option for antiretroviral therapy--associated dyslipidemia because of its ability to both increase HDL-c levels and lower triglyceride levels.
We conducted a prospective pilot study of ER-niacin treatment in HIV-infected individuals with significant dyslipidemia. Our aim was to evaluate its therapeutic potential, tolerability, and adverse-effect profile, focusing on insulin resistance, because several studies in the HIV-seronegative population have reported reduced glucose tolerance in association with niacin therapy.
Current treatment options for dyslipidemia in the HIV-infected population are not ideal. There are significant drug interactions between antiretroviral medications and lipid-lowering agents, such as statins. ER-niacin therapy is an appealing alternative to evaluate in this population because of its beneficial effect on levels of all lipoprotein fractions. Because of the potential for glucose intolerance and decreased insulin sensitivity, as described in several studies in the HIV-negative population, the Adult AIDS Clinical Trials Group did not recommend the use of niacin as a first-line agent for HIV infected patients concurrently treated with protease inhibitors.
To evaluate this concern regarding insulin resistance, we used an OGTT test with frequent glucose, insulin, and C-peptide measurements, as described elsewhere. We observed an increase in the basal insulin secretion rate caused by enhancement of the sensitivity of b-cells to stable basal glucose levels, which was not accompanied by significantly enhanced glucose disposition. These findings imply that niacin therapy induced a form of insulin resistance in the basal state. This was consistent with the observed significant increase in fasting HOMA-IR values that was due primarily to an increase in fasting insulin concentration. Approximately one-third of our patients were taking protease inhibitors. Their values for insulin sensitivity parameters tended to be lower at baseline but did not appear to worsen disproportionately more than did insulin sensitivity values in the patients taking nonnucleoside reverse-transcriptase inhibitor--based therapy. Although our findings are consistent with previous findings regarding niacin-induced insulin resistance in HIV-negative subjects, these effects did not lead to frank diabetes during the 14-week course of therapy.
Another limitation to the widespread use of traditional, immediate-release niacin has been intolerance of cutaneous flushing. In agreement with the findings of studies of HIV-negative populations, our study found that ER-niacin did not cause cutaneous flushing to a limiting degree in our study participants when given with 325 mg of aspirin. Flushing did not necessitate discontinuation of ER-niacin therapy for any of the patients. None of the subjects in this small pilot study developed significant elevations in transaminase levels or myopathy.
Our findings indicate that ER-niacin may be a useful option for lipid-lowering therapy for HIV-infected people with normal glucose tolerance. ER-niacin therapy significantly reduced serum levels of triglycerides, non-HDL cholesterol, and total cholesterol. In contrast to prior studies in HIV-seronegative individuals, our study did not find a significant change in LDL-c and HDL-c levels after niacin therapy. However, values for both parameters were only mildly abnormal at baseline, and the maximum dose of ER-niacin (2000 mg per day) was only taken for 4 weeks. Continued increases in the HDL-c level were observed after 16 weeks of full-dose niacin therapy (2000 mg/dL) in HIV-seronegative subjects.
Voluntary changes in dietary intake may account for some of the favorable lipid profile changes. However, the lipid profile changes that were observed during ER-niacin therapy (i.e., study weeks 4--18) were greater than the changes that might be expected for dietary intervention alone (decreases of 5--9 mg/dL for total cholesterol, and increases of 1--4 mg/dL for triglycerides). Furthermore, total cholesterol, triglyceride, and non-HDL cholesterol values increased significantly after ER-niacin therapy was discontinued (i.e., during study weeks 18--22), despite ongoing dietary counselling. Therefore, we feel confident that the lipid profile changes observed during study weeks 4--18 predominantly reflect the actions of niacin. Although little toxicity from ER-niacin therapy was observed, the number of subjects studied was small, and the duration of the study was short. The fact that the study population was exclusively male may limit the generalizability of the study results, given that differences between the sexes with respect to HIV-related dyslipidemia are currently not well defined.
This small, open-label pilot study evaluated whether ER-niacin could be an effective option for treatment of antiretroviral therapy--associated dyslipidemia. Currently available options for treatment are limited by drug-drug interactions and low efficacy. A trial comparing fenofibrate and pravastatin for the treatment of HAART-related elevations in LDL-c and triglyceride levels did not reduce lipid levels to NCEP-recommended levels with single-agent therapy in most subjects. Although statins and fibrates have been used to treat antiretroviral therapy--associated hyperlipidemia, monotherapy with these commonly used agents may not be effective. Furthermore, the dual use of fibrates and statins is limited by concerns regarding skeletal-muscle toxicity in a population already at elevated risk for myopathy.
The purpose of this pilot study was to examine the tolerability and efficacy of ER-niacin therapy for people living with HIV-related dyslipidemia. The ER formulation was well tolerated in our subjects, and it should be evaluated in larger, placebo-controlled trials that can define its role as a therapeutic agent for antiretroviral therapy--associated dyslipidemia. The use of ER-niacin in patients with glucose intolerance or diabetes may be limited, but our findings indicate that ER-niacin therapy may be a useful option for lipid-lowering therapy for the HIV-infected population, as dual therapy or as monotherapy. Due to its lack of interaction with hepatic metabolism, ER-niacin might be a promising agent for combination lipid-lowering therapy. The findings of our study justify a larger, more extensive study of the use of this agent in the management of HIV-related dyslipidemia.
The study population consisted of HIV-infected individuals recruited from the AIDS Clinical Trials Unit and the Infectious Diseases Outpatient Clinic at Washington University School of Medicine (St. Louis, MO) from July 2001 through May 2002. Fourteen HIV-infected individuals with dyslipidemia (defined as an 8-h fasting triglyceride level of >200 mg/dL and/or a directly measured LDL level of >130 mg/dL) were consecutively enrolled. The Human Studies Committee of the Washington University Medical Center approved the study. All patients gave written informed consent.
The inclusion criteria were as follows: age >18 years, HIV seropositivity, receipt of HAART for at least 6 months with a stable regimen for at least 4 weeks prior to study entry, and willingness to adhere to a National Cholesterol Education Program (NCEP) Step I diet. Patients were excluded if they had a history of toxicity to niacin; diagnosis of diabetes mellitus, gout, or peptic ulcer disease; or were currently receiving treatment for active opportunistic infections or malignancy. We also excluded subjects with a history of cirrhosis, substance abuse, or pregnancy or a with present unwillingness to practice birth control.
The subjects were followed up for a total of 22 weeks: a 4-week lead-in period, a 14-week study period, and a 4-week "washout" period. At enrollment (i.e., 4 weeks before baseline lipid and cholesterol analysis), all eligible patients received counseling regarding a NCEP Step I diet. Reinforcement of dietary advice was given by a registered dietician or a study nurse at all study visits. Treatment with all lipid-lowering agents was discontinued for at least 4 weeks prior to baseline lipid and cholesterol analysis.
Niacin therapy was initiated at a dosage of 500 mg per day at week 4. The dose was increased to 1000 mg per day after 2 weeks of therapy and subsequently titrated upwards by 500 mg every 4 weeks, to a maximum of 2000 mg per day, in an attempt to achieve triglyceride values of <200 mg/dL or NCEP goals for LDL-c. Participants were advised to take 325 mg aspirin 30 min prior to the ingestion of each niacin dose, as needed, for prophylaxis against cutaneous flushing. After 14 weeks of therapy, ER-niacin therapy was discontinued. The subjects were then observed for another 4 weeks while they were not receiving ER-niacin therapy. A fasting lipid profile, a complete metabolic panel, and the uric acid level, CD4 cell count, and HIV load were determined at study weeks 0, 4, 10, 14, 18, and 22. Cholesterol levels, including directly measured LDL-c levels and triglyceride levels, were measured using commercially available immunoassays (Roche Diagnostics).
A 5-h, 11-time-point oral glucose tolerance test (OGTT) was performed at baseline and after 14 weeks of ER-niacin therapy (i.e., at study weeks 4 and 18). After a 75-g glucose challenge (Dexicola 75; Fisher HealthCare), values for insulin sensitivity and secretion parameters were derived from glucose, insulin, and C-peptide levels measured during the 5-h OGTT. The OGTT adequately assesses b-cell function and insulin sensitivity, and a clinical classification for glucose tolerance is derived from the OGTT results using American Diabetic Association criteria. Insulin secretion indices were calculated as described elsewhere using the minimal model of C-peptide secretion and kinetics. A global index of insulin secretion was calculated, which is based on the b-cell sensitivity to glucose. It represents the average increase in the insulin secretion value to greater than the basal value, expressed according to the average glucose-stimulus level; its components are the ratio of the basal insulin secretion level to the basal glucose level (i.e. b-cell sensitivity) and to the stored insulin secretion level measured during the OGTT. The minimal model of glucose disposal was applied to glucose and insulin concentrations measured during the OGTT, and an estimate of insulin sensitivity (SI) was derived for each subject. To normalize insulin secretion to the magnitude of insulin sensitivity, the disposition index was calculated as SI x F, where F is b-cell sensitivity to glucose. This parameter is similar to the disposition index derived from the results of a frequently sampled intravenous glucose tolerance test . Insulin concentrations were determined using a double-antibody radioimmunoassay and C-peptide and glucagon concentrations were determined by radioimmunoassay (Diabetes Research and Training Center Radioimmunoassay Core Laboratory, Washington University School of Medicine, St. Louis, MO). During the OGTT, plasma glucose concentrations were measured using a YSI 2300 Stat Glucose Analyzer (YSI Life Sciences).
Baseline characteristics of patients--
A total of 14 patients were enrolled. All were men, and all but one was white. The median CD4 count was 424 cells/mm3; 11 of 14 subjects had HIV RNA loads of <400 copies/mL. Eight were receiving nonnucleoside reverse-transcriptase inhibitor--based antiretroviral regimens and 6 were taking protease inhibitor--based regimens. According to American Diabetic Association criteria, 50% of patients were glucose intolerant. Four (29%) of the patients were taking statin therapy at study enrollment; 2 of these were taking combination therapy with fibrates. The majority of patients were current or former smokers. One-half of the subjects had either a family or personal history of cardiovascular disease.
Tolerability and adverse effects--
Twelve of 14 (86%) patients completed the study. Two patients discontinued ER-niacin therapy prematurely: 1 patient when he was hospitalized for exacerbation of preexisting heart failure, and 1 patient after hospitalization for pneumonia. The ER-niacin dosage was titrated to 2000 mg per day for all 12 subjects who completed the trial. Although 6 (43%) of 14 of the subjects reported sporadic niacin-related adverse effects (i.e., cutaneous flushing, itching, or headaches), they described an improvement between weeks 10 and 14 of niacin therapy. None of the 12 patients requested discontinuation of therapy as a result of these adverse effects during the study. Most of the adverse effects were well controlled with aspirin (325 mg once daily). No pattern could be discerned with regard to predisposition to the adverse effects in this small study. No symptoms of myopathy were reported. There was no increase in mean transaminase levels during the course of the study, and there were no significant transaminase elevations of more than twice the upper limit of normal in any study participant. There were also no significant changes in uric acid level, CD4 cell count, or viral load. None of the subjects changed antiretroviral regimens during the course of the study.