icon-    folder.gif   Conference Reports for NATAP  
  10th Intl Workshop on Adverse Drug Reactions and Lipodystrophy in HIV,
London, UK Nov 6-8 2008
Back grey_arrow_rt.gif
Report from the 10th Intl Workshop on Adverse Drug Reactions and Lipodystrophy in HIV
  London, UK Nov 6-8 2008
Michael Dube, MD, University of Southern California
Do levels of C-reactive protein (CRP) help predict if HIV patients will have heart attacks, like they do in the general population?
Perhaps the most intriguing presentation was from Dr. Virginia Triant from Harvard, who reported on blood C-reactive protein (CRP) levels and myocardial infarction (heart attack) risk (Abstract O-05). CRP is really in the news again in recent weeks with the results of the JUPITER trial being reported. In JUPITER, patients in the general population who had high CRP levels but relatively normal cholesterol levels had about a 50% reduction in the risk of MI or stroke when they received the cholesterol-reducing statin drug rosuvastatin (NEJM 2008). In the general population, statin class drugs reduce CRP levels in addition to their well-known beneficial effects on cholesterol levels. Generally speaking, levels of CRP in the blood reflect the amount of inflammation going on in the body. CRP itself is involved in some of the atherosclerosis-producing processes in the blood vessel wall. Increased CRP levels have been proven to be associated with increased MI risk in the general population for many years, but the blood test still is not generally used in managing cardiovascular risk or considered to as a routinely modifiable risk factor for MI. What has not been known is if higher CRP levels predict greater MI risk in patients with HIV. Patients with HIV may have increased CRP levels due to HIV infection itself, other infections, and more cigarette smoking than in the general population.
Prior reports from this group showed an approximately 75% increase in the rate of MI among HIV-infected patients as compared to non-HIV patients. In this large database study, increased CRP was associated with significantly increased MI risk in both the HIV-infected and uninfected patients. Among the minority of patients that had CRP levels measured, 59% of HIV-infected subjects had high CRP while 39% of those without HIV were elevated. The investigators were able to control for a variety of conditions that contribute to MI risk, but not all. In adjusted analyses, both HIV infection and a high CRP level were independently associated with about a doubling of the risk of MI. Unfortunately, the nature of the database that they researched retrospectively did not contain reliable smoking data - so they went back and looked at a subset of around 400 subjects' charts and found similar smoking rates among the groups with high CRP and low CRP, suggesting that the differences they observed were not due to smoking alone being more prevalent in the HIV group.
This was a huge database, but only a small minority of subjects actually had CRP drawn. A variety of CRP assays were used with different normal ranges, so they were unable to say precisely what level of CRP indicates a significantly increased risk among the HIV patients. It is also likely that CRP levels were being done for reasons other than evaluating CVD risk, so it raises the question did HIV patients have the test drawn because they were ill from some other reason? Regardless, this study is an important step forward, answering the question: is increased CRP associated with increased MI risk in HIV patients too? YES. However, it is still very much open what level of CRP should we be worried about, and if we find elevated levels in an HIV-infected patient then what if anything we should be doing about it. It may not be appropriate to extend the findings of a study like JUPITER to HIV-infected patients, because statin drugs may not have a similar effect in HIV. Conversely, giving statins to patients with HIV may have an even more profoundly beneficial effect than the drugs had in the general population. Until there are controlled studies large enough to answer these questions in HIV-infected individuals, we are left with only speculation as to what level of CRP justifies intervention and then what sort of intervention should be used when. That said, I believe it is likely that statins will ultimately prove to be effective at reducing CRP levels in HIV.
Are kids on ART at particular risk for CVD complications?
The implications of increased lipids, insulin resistance, and lipodystrophy on children and their long-term CVD risks are unknown. Ross from Rainbow Babies and Children in Cleveland (abstract O-07) reported that carotid IMT, a marker for cardiovascular disease risk in adults, actually decreased in HIV-infected children over a year. However, it is not known what a normal carotid IMT is in children, or what happens to the measure over time as healthy children grow normally. Nonetheless, it was encouraging that these ART-treated kids did not appear to be experiencing serious early signs of subclinical CVD.
Why does abacavir use cause more CVD?
The reasons behind why abacavir use is associated with increased risk of MI remains unclear and was a topic of considerable discussion. Jens Lundrgen from Denmark reviewed the results from the D:A:D study and the recently published results from the SMART study which appear to confirm an approximately 90% increased risk of MI when patients are currently using abacavir in their antiretroviral regimens. Importantly, he emphasized again that the bulk of this increased risk is primarily borne by those individuals who are already at high risk for cardiovascular disease. There is, therefore, general agreement that it is those with multiple pre-existing risk factors such as family history of MI, high cholesterol, cigarette smoking, hypertension, and diabetes for whom we must be most cautious with the use of abacavir.
Podzamczer from Barcelona (abstract O-08) reported results of a trial comparing tenofovir/FTC (Truvada) with abacavir/3TC (Epzicom) with regards to careful lipid testing. Total and LDL cholesterol were modestly higher and triglycerides were slightly higher with abacavir and the composition of LDL particles was somewhat more atherogenic with abacavir. However the magnitude of the differences was small and unlikely to translate into a significant increase in MI risk. Furthermore, HDL cholesterol actually increased significantly only with abacavir, and the ratio of apoliporotein A1/apolipoprotein B were similar in both treatments (one of the most reliable lipid-related measures of CVD risk). This study only suggests minor lipid disturbances from abacavir that do not appear to translate into clear-cut increases in cardiovascular risk. We clearly need to be looking elsewhere for exactly what it is about abacavir that increases MI incidence.
As covered by Georg Behrens of Hannover, Germany in an excellent plenary presentation, the likely culprit is some form of increased tendency of the blood to clot, increased inflammation, or other dysfunction of the lining of the blood vessels (endothelium) related to use of the drug. Notably, earlier this year at the XVII International AIDS Conference, the SMART investigators reported that abacavir use increased CRP and IL-6 levels (interleukin-6, an inflammatory cytokine associated with insulin resistance and vascular dysfunction). Whether these increases are the true cause of the approximately 90% increase in the risk of MI or cardiovascular disease with abacavir use, or whether they merely are associated with some other undefined underlying problem that is leading to the increased risk, remains a matter of speculation.
Is visceral (intra-abdominal) fat different in patients with lipodystrophy?
Giralt from University of Barcelona in Spain presented interesting data regarding the nature of visceral fat (fat inside the abdomen and around the abdominal organs) in 7 patients with HIV who had lipodystrophy (facial lipoatrophy and some increase in visceral fat) and compared their results with visceral fat from 10 non-HIV infected controls (abstract O-01). In addition, comparison was made with subcutaneous fat (SQ fat, i.e. beneath the skin) in both groups. Prior studies have focused on subQ fat, because that is so much more accessible to biopsy as compared to visceral fat. It has been hypothesized in the past that ART makes subQ fat dysfunctional while the more metabolically-active visceral fat remains more functional, predisposing to visceral fat accumulation (at least in certain susceptible individuals) while subQ fat fails to be able to accumulate fat when excess calories are taken in. However, these investigators reported that mitochondrial function and mitochondrial DNA was reduced in both subQ and visceral fat, suggesting no clear differential effect from ART in the 2 fat depots. Markers of inflammation tended to be increased in both fat depots, but interestingly markers of adipogenesis (formation of new fat) appeared to remain adequate only in the visceral fat - suggesting that visceral fat may be able to better retain/accumulate new fat stores once lipodystrophy is established. This might explain why losing visceral fat may become more difficult in those with established lipodystrophy, at least in those individuals who have lost subQ fat. No particular insight into treatment or prevention of lipodystrophy was suggested by this work, however.
Does treatment with IGF-1 improve visceral obesity?
A small pilot trial of trial of IGF-1/IGFBP-3 was reported by Morrie Schambelan from UCSF. IGF-1 (insulin-like growth factor 1) is the hormone that goes up when growth hormone is administered and is thought to be responsible for all the good and few of the bad effects of hGH. IGFBP-3 is the "binding protein" for IGF-1, and so the formulation they used gives both of them together to make the IGF-1 last longer in the bloodstream than giving IGF-1 by itself. Ten stable ART-treated subjects with big bellies got a full course of treatment: 7 got a low dose and when it was recognized that this was having no effect on visceral fat, twice the dose was given to the next 3. Unfortunately, no decrease in visceral fat was seen. However, overall trunk fat decreased, which could be an advantage even if visceral fat did not. Oddly, although muscle mass and overall insulin sensitivity improved, glucose production by the liver (gluconeogenesis) increased, which was not expected. This increased gluconeogenesis is what happens when the liver tissue becomes more insulin resistant. Lipids also did not improve in this study, like they do with other therapies. The treatment appeared to be well tolerated. While these are very preliminary findings, and higher doses may need to be studied to find a beneficial effect, these studied doses of IGF-1/IGFBP-3 do not appear to have the beneficial effects on visceral fat and lipids that giving growth hormone or tesamorelin (which stimulates growth hormone production) have.