 |
|
|
| |
Metabolic Complications of HIV and its Therapies: What Did We Learn in Barcelona?
Written by David Alain Wohl, MD, University of North Carolina AIDS Research and Treatment Center
|
| |
| |
During the six days during which the 14th World AIDS Conference was held an estimated 50,000 people died of AIDS. They died as we in Barcelona listened to government leaders call for 'scaling-up' efforts to combat the spread of HIV, although reports of such efforts from their own countries were scarce. They died while results from studies of a new antiretroviral class made headlines, although these agents seem almost as out of reach to those in the developed world as basic antiretrovirals are to those living in developing nations. They died while we were lectured on why another conference would pass without a vaccine to prevent or treat HIV infection. And, they died as we learned how the pandemic has surged across Eastern Europe, Asia and the Caribbean in the two years since the last conference was held in Durban. Against this backdrop of suffering and disappointment it was easy to ask if body shape changes, elevated lipids and glucose intolerance really matter.
They do. Antiretroviral therapies are beginning to be prescribed in developing nations and their use in these areas will almost certainly grow exponentially, albeit belatedly. Complications associated with these drugs are sure to follow. For those living in the parts of the world thirsting for HIV treatment these complications may seem trivial compared with devastating opportunistic infections (of what concern is hypertriglyceridemia when young people are dying from cryptococcal meningitis?), yet, we the privileged have learned that these complications can impact the acceptance and adherence of these medications. For us complications are a frustrating reminder of the fragility of our newfound success in controlling HIV. Viral loads drop but lipids rise and cheeks sink. People start needing insulin. Others combat lipid levels that break indoor records. Whether in Malawi or Miami body shape and other metabolic problems present dangerous obstacles to delivery of HIV care and need to be reckoned with. If we understand metabolic problems associated with HIV now we will have the best hope of avoiding them wherever these drugs are used.
At the conference there was some movement forward in the complex arena of metabolic complications of HIV and its treatment. As people with HIV live longer on therapy we are starting to get a sense of the longer-term problems they are confronting. In addition, treatment trials that incorporate rigorous metabolic evaluations are beginning to report results, identifying drugs that do and do not tend to cause metabolic disturbances. Below are some of the more noteworthy presentations from the oral and poster sessions:
The specter of increased cardiovascular disease (CVD) with antiretroviral exposure is a concern that transcends gross domestic product. At this conference we heard some of the best data to date suggesting an increased risk of CVD among persons treated with HIV drugs. In a large prospective study that it seems incredibly no one knew about until it was presented, 1551 HIV+ treatment naive patients in Italy were randomized to couple two nucleosides with either a protease inhibitor (PI) or a non-nucleoside reverse transcriptase inhibitor (NNRTI) (G. Barbarini, et al; abstract WeOrB1307). Subjects were followed for development of angina and myocardial infarction (heart attack) over approximately 3 years. Not unexpectedly, more subjects had elevations of triglyceride and total cholesterol in the PI arm and there were increased rates of "lipodystrophy" in this group compared to those assigned to an NNRTI. More surprising, however, was the differential rate of CVD between the arms. There were 35 cases of CVD, including 12 heart attacks, in the PI arm versus 3 CVD diagnoses in the NNRTI arm, including one myocardial infarction. The rate of CVD among the PI randomized arm was much higher than that of the general population and was associated with development of lipodystrophy, male gender, increased lipids, high fibrinogen levels and smoking. (Note that in this Italian cohort, almost 85% of the subjects were smokers. In fact, smoking was probably universal, as patients were not classified as smokers unless they smoked more than 15 cigarettes per day!) The high rate of smoking among the cohort when added to other CVD risk factors may well have accelerated the development of heart disease in this study. When pressed for details after the presentation, the author of the paper ominously stated that as many as 15% of the CVD cases were women, that all cases were less than 48 years old and that the diagnosis of CVD was made with the development of symptoms of angina and/or scheduled EKGs - therefore, increased scrutiny and bias for CVD in the PI arm seems unlikely.
The relationship between PI use and myocardial infarction was also described in data from the Centers for Disease Control and Prevention (CDC) supported HIV Outpatient Study (HOPS) (Holmberg S, et al.; abstract TuPeB4494). This prospective dynamic cohort of HIV-infected patients seen at 9 clinics in 7 geographically diverse U.S. cities provided data on 5672 ambulatory HIV+ patients seen in follow-up between 1993 and 2001. Clinic and hospital records were examined for diagnoses of heart attack, angina or stroke. During the study period, as PI use increased so too did the rates of myocardial infarction. Of the 3247 patients receiving PIs, 19 developed a heart attack compared to two of the 2423 patients not treated with PIs. In a multivariate analysis that adjusted for traditional CVD risk factors such as hypertension, smoking, diabetes, age, gender and dyslipidemia, the association between PI use and myocardial infarction remained (hazard ratio=7.43 [95% CI 1.0,53.6], p=0.047). Although heart attacks in this cohort were relatively uncommon, the results seem to validate the concerns of many clinicians and patients that PI exposure is a risk factor for CVD.
These data stand in contrast to findings from a Veteran's Administration (VA) Cohort presented at the 9th Conference on Retroviruses and Opportunistic Infections (CROI) earlier this year that found no association between increased PI use and CVD (See Dr. Judy Aberg's report from CROI on the NATAP website). The studies were different in their methodologies and populations and this may explain their discordant findings. The VA study looked for temporal associations between PI uptake and rates of CVD related admissions or death whereas the HOP study compared the CVD rates of patients on PI and non PI containing regimens. Additionally, there were more CVD events in the VA cohort (~2%) compared to in the HOPS group (~0.4%). Could the chronically higher level of CVD within the VA have obscured any contribution made by PIs? A contribution of these agents to heart disease seems likely and is probably mediated through their effects on lipids and glucose intolerance. Minimization of correctable CVD risk factors (smoking cessation, lipid and glucose lowering interventions and exercise) may ameliorate much of the effect of PIs.
Antiretroviral clinical trials also yielded interesting metabolic data. The 48-week results of the Gilead 903 were presented during a late breaker session. In this study 600 treatment naive subjects from North America, Western Europe, Brazil and the Caribbean were randomized to either tenofovir+3TC+efavirenz or d4T+3TC+efavirenz. Most notable in this trial was the extremely high viral suppression rate in each arm (82% in the tenofovir arm, 81% in the d4T arm). Likewise, the overall rate of adverse events was not different between study regimens. What distinguished the study arms however, was the increase in triglycerides and cholesterol during the study period. The mean change from baseline in triglycerides was 74 mg/dL in the d4T assigned subjects compared to no mean change in those randomized to tenofovir (p<0.001). Total cholesterol increased in both arms but more so and more quickly in the d4T group (53 mg/dl versus 25 mg/dL, p<0.001). These results add further evidence of a role of d4T in HIV-associated dyslipidemia - such as was suggested by the results of the study of d4T versus AZT when combined with nelfinavir presented by Kumar P, et al. at the 9th CROI, (abstract #33).
The adverse effect of even 'baby dose' ritonavir on lipids was reported in a retrospective review conducted by Grace McComsey of Case Western Reserve University (abstract ThPeB7319). Lipids of patients prescribed either ritonavir at doses of 100-133 mg twice daily in combination with other PIs were compared to those receiving nelfinavir. In total 69 patients received ritonavir and 74 nelfinavir. Baseline characteristics, including lipid profiles were not substantially disparate between the arms. Importantly, about a third of subjects had triglyceride levels >200 mg/dL and a quarter had cholesterol levels >200 mg/dL at baseline. At 3 and 6 months following the initiation of therapy, triglyceride levels had increased from baseline by 100 mg/dL at 3 months and 73 mg/dL at 6 months in the ritonavir boosted arm compared to slight decreases in triglycerides in the nelfinavir arm at these time points. Total cholesterol increased mildly in both the ritonavir boosted and nelfinavir arms to about the same extent (~15-20%). Bottom line: low dose ritonavir in combination with other PIs does contribute to hypertriglyceridemia compared to nelfinavir.
Lastly, atazanavir. This PI was serendipitously discovered to be lipid friendly in early clinical trials and save for a nagging tendency to cause asymptomatic jaundice is touted as a kinder gentler once a day PI. Whether switching from another PI to atazanavir would lead to improvements in lipid levels was studied in a multinational rollover study sponsored by the manufacturer of the drug and presented as a late breaker poster by Rob Murphy of Northwestern University in Chicago. In this study subjects previously enrolled in another trial and assigned atazanavir continued on this agent at either 400 mg daily or 600 mg daily depending on their original randomization while those assigned nelfinavir were switched to atazanavir 400 mg daily. All subjects also received d4T and 3TC. A total of 346 subjects were studied of whom 63 switched from nelfinavir to atazanavir. At baseline, the mean total cholesterol level of the subjects on nelfinavir was, not surprisingly higher at 213 mg/dL, while the atazanavir treated subjects had levels of 185 mg/dL. LDL cholesterol was also higher at baseline in the nelfinavir arm (138 mg/dL versus 112 mg/dl in the combined atazanavir arms). Fasting triglyceride levels were 156 mg/dL in the nelfinavir arm, 151 mg/dL in the atazanavir 400 mg arm and 135 mg/dL in the atazanavir 600 mg arm. After 12 weeks following the switch from nelfinavir to atazanavir, total cholesterol levels declined to 175 mg/dL (-16%), LDL levels dropped to 104 mg/dL (-21%), triglycerides fell to 108 mg/dL (-28%) and HDL levels increased by 5%. There were no clinically significant changes in lipid profiles of those continuing on atazanavir. Jaundice developed in 5% of the subjects switching from nelfinavir and 10% had grade 3 or higher hyperbilirubinemia. These are promising results. For patients with a history of CVD or at high risk for CVD and in need of a PI, this drug provides great promise. How the bilirubin/jaundice problem will play in the real world will be seen.
What does all this mean? Accumulated evidence indicates that lipid abnormalities resulting from HIV therapies in combination with other traditional risk factors place HIV-infected patients at some increased risk of cardiovascular disease. Protease inhibitors seem to not hold a monopoly on dyslipidemia and data presented indicate nucleosides, especially d4T, can play a role. In recognition of the magnitude of metabolic problems and its impact on long term patient health and treatment tolerance and adherence, much work is being done to understand its causes and develop prevention and treatment strategies. Hopefully, as these drugs finally become available worldwide, research focusing on the metabolic complications of HIV and its treatment will be conducted in resource challenged areas so that many of the problems we have encountered can be avoided.
|
|
| |
|
|
|
|
|
