Face to Face with Lipoatrophy An Interview with David Nolan
By Nelson Vergel
HIV-related lipoatrophy (fat loss under the skin) is a clinical problem that affects many people living with HIV. Lipoatrophy can cause substantial loss of buttock tissue, veiny legs and arms, and facial wasting. Lipoatrophy can happen alone or in combination with lipohypertrophy (fat accumulation) in the visceral (organ) and dorsocervical (back of the neck) area. These body changes, with or without blood level alterations of cholesterol, triglycerides, lactic acid, glucose, and insulin, is called HIV-related lipodystrophy syndrome.
Many of these body changes occur normally with aging, but being HIV positive seems to accelerate their development. Many observational cohorts and studies to determine the causes and potential treatments of lipodystrophy have been performed since 1997, the first year when we realized that living longer with HIV may be accompanied with side effects like body changes. Even though there is not yet a consensus-based case definition for lipodystrophy, there are many things that we have learned since then. One is the potential impact of nucleoside thymidine analogs like d4T (Zerit, stavudine) and AZT (zidovudine) on subcutaneous (under the skin) fat wasting. Studies in which patients were switched from these drugs to abacavir (Ziagen) or tenofovir (Viread) have produced encouraging results in reversing lipoatrophy, even if at a slow rate. Other studies looking at using insulin sensitizers like rosiglitazone (Avandia) to reverse lipoatrophy have produced conflicting results. It seems that the use of Avandia in combination with Zerit or AZT does not allow for regrowth of subcutaneous fat tissue. Fortunately, good news came to patients with lipoatrophy in the U.S. when the FDA approved Sculptra (polylactic acid) injections in August 2004 for the treatment of HIV related facial lipoatrophy. Although this option does not treat the root cause of the problem, it can give patients the hope of restoring a healthy appearance once again.
David Nolan is a clinician and researcher based at Royal Perth Hospital in Western Australia, where he works at the Centre for Clinical Immunology and Biomedical Statistics. He has a broad range of interests relating to HIV and its treatment, with a particular emphasis on lipoatrophy where he has explored associations between NRTI therapy, mitochondrial toxicity and fat tissue pathology. He has published more than 30 research papers and reviews, covering the broad topic of lipodystrophy as well as related topics including mitochondrial toxicity syndromes, metabolic complications of HIV protease inhibitor therapy, hyperlactatemia syndromes, genetic susceptibility to abacavir and nevirapine hypersensitivity reactions, and the effects of HIV and treatment on bone density.
Nelson Vergel: Can HIV itself cause lipoatrophy or mitochondrial dysfunction in treatment-naive patients?
David Nolan: There is no evidence that HIV infection itself can cause lipoatrophy. It has been known for some time that severe immune deficiency and AIDS-defining illnesses can be associated with "AIDS wasting," but this is quite different in terms of the effects on body composition. Lipoatrophy specifically affects the fat tissue just under the skin surface (subcutaneous fat), leading to fat loss that is particularly noticeable over the legs, buttocks and face. It is also notable that the lean body mass (i.e., muscle tissue predominantly) is unaffected in the case of lipoatrophy, and indeed muscle mass often improves in the presence of lipoatrophy.
On the other hand, AIDS wasting associated with untreated HIV predominantly affects lean body mass while having less of an effect on fat tissue. The only exception here is that women with advanced HIV disease tend to lose fat as well as muscle, although again this has different characteristics. Loss of fat in AIDS wasting in women tends to be generalized (i.e., the same all over the body) as with any weight loss, while part of the reason that lipoatrophy is so noticeable in both men and women is the way in which fat loss is so "unevenly spread" over the body. For women in particular, the preferential loss of fat over the legs and buttocks associated with lipoatrophy is quite unusual.
Your lipoatrophy data on AZT shows only about half the adipocyte (fat cell) depletion that is asssociated with d4T. Do you think the depletion of fat cells and mitochondrial DNA/function warrants removing AZT from the treatment guidelines of industrialized countries for first-line therapy?
This is a very interesting question and one where there is no definitive answer yet. The adipocyte depletion findings tell some of the story, although the clinical data are probably more important in showing how significant the effect of AZT is in terms of lipoatrophy risk. Overall, the severity of fat loss — and the risk of clinical lipoatrophy — is approximately halved with zidovudine treatment compared to stavudine. This means that in "ball park" terms, the risk of developing noticeable fat loss is 15 percent to 20 percent after 3 to 5 years of zidovudine treatment.
My own view (and again there is no consensus opinion available as yet) is that the impact of zidovudine treatment is not sufficient to minimize its use for first-line therapy in industrialized countries. There are a number of reasons:
The fat loss associated with zidovudine treatment tends to be fairly slowly progressive, typically beginning after the first year of treatment with the greatest risk of fat loss between 12-36 months. In this instance, if those taking AZT (and those clinicians treating them) are aware of the risk of lipoatrophy and actively monitor for early signs (i.e., fat loss that particularly affects the legs, buttocks and/or face, and which is often not associated with any loss of weight), then treatment can be altered if and when this complication develops. We know from the results of "NRTI switching" studies involving the use of abacavir or tenofovir that fat loss is at least halted, and often improves, when the NRTI drugs are changed.
There are host and disease factors that also contribute to the risk of lipoatrophy when AZT or d4T are used (although, as stated above, these risk factors only operate when these drugs are used). Therefore, the risk of lipoatrophy with AZT treatment is likely to be substantially reduced in those who are younger than 35 years of age, who are of non-white racial origin, and who start treatment when the CD4+ T cell count is greater than 100–200.
Finally, I think that part of the reason that AZT has remained in continuous clinical use since 1987 is that it has a favorable resistance profile in terms of its interaction with other NRTI drugs. The available data for newer "backbone" NRTI drugs such as tenofovir and abacavir is certainly very promising, but there are still some questions regarding both their long-term effectiveness and (to a lesser extent) the potential for tenofovir to cause renal and/or bone toxicity.
For those who want to start or remain on a non-protease inhibitor combo, AZT seems to have a protective role in the selection for K65R and L74V/I in nucleoside combos. How can patients or doctors balance this fact with the potential lipoatrophy effects of AZT?
Again, a very interesting question. I think it is fair to say that any triple-NRTI drug regimen needs to include AZT to have a good chance of success, given the poor outcomes that have been associated with NRTI combinations such as tenofovir/abacavir/lamivudine or tenofovir/didanosine/lamivudine (very poor efficacy), or stavudine/didanosine/abacavir (poor efficacy and high toxicity) [reviewed in 1]. In these regimens, it appears that these drugs all tend to favor the emergence of a similar pattern of resistance mutations, so that there is very little barrier to the emergence of K65R or L74V/I. AZT appears to counteract this effect, as it "pulls in another direction," actually becoming more potent when these resistance mutations start to emerge. This means that the virus has more difficulty becoming broadly resistant to both AZT and the other NRTI's (most commonly lamivudine and abacavir) within the NRTI combination.
As stated above, I think the risk of lipoatrophy associated with AZT can be managed in a rational way that minimizes the chances of developing problematic fat loss, and in this particular case (when triple NRTI therapy is desired) there really is no effective alternative drug at present.
Do you think the move in international and U.S. guidelines for delayed treatment at lower CD4 counts will increase the incidence of lipodystrophy in the HIV population?
I think there is strong evidence that delaying treatment until CD4+ T cell counts fall below 200 increases the risk of a number of drug-related toxicities, including lipoatrophy (if using d4T or AZT) and neuropathy (which can be associated with HIV itself as well as the use of d4T or ddI) [reviewed in 2]. There is also a concern that low HDL-cholesterol levels associated with more advanced HIV disease may predispose patients to metabolic complications and the potential for a greater risk of cardiovascular disease . I would argue that treatment should be initiated at higher CD4+ counts (i.e., 200–350) and should be focused on:
Assessing each individual case on its merits. This means taking into account the patient's characteristics and finding the best "match" in terms of HIV therapy. These considerations may include the toxicity profiles of individual HIV drugs, as well as the tolerability of the HAART regimen. The main goal is to provide the best possible chance of achieving 100 percent adherence to therapy.
And, assessing and monitoring those side-effects that are specifically associated with the HIV drugs being used, so that these complications can be picked up early and managed appropriately.
The TARHEEL study results seem to point to the intriguing fact that even though mitochondrial DNA (mtDNA) levels rebounded after d4T therapy was discontinued, mitochondrial function did not recover. Do you think that there may be some permanent mitochondrial damage even after patients switch from d4T or AZT to non-thymidines?
I think the mitochondrial toxicity at a cellular level is reversible, as it is likely that the toxicity of these drugs is mediated specifically by their effects on mitochondrial DNA depletion. The more worrying problem is that severe lipoatrophy represents a profound loss of adipocytes (fat storing cells) within the fat tissue through cell death (apoptosis). The TARHEEL study showed very nicely that you can "turn off" this process of cell death by switching from d4T to a non-thymidine NRTI (i.e., far fewer cells are killed), but you are then left with the problem of having to replace the cells that have been lost. While the non-thymidine NRTI's are obviously not toxic to fat tissue, they do not actively encourage new cells to grow, and treatments that may be anticipated to help this process of regeneration (such as rosiglitazone) have not performed well to date. This problem is further complicated by the fact that lipoatrophic fat tissue contains a large number of macrophages, inflammatory cells that are probably there to "mop up" the adipocytes and their stored fat after cell death, and these may also inhibit the growth and development of new fat cells.
What is your opinion about the use of micronutrients (carnitine, coenzyme Q-10, thiamine, riboflavin) or the use of uridine to reverse or prevent loss of mitochondrial DNA or function in the presence of thymidine analogs?
With regard to the micronutrients you mention, there is no evidence that they improve any clinical outcomes if used as a preventive treatment. They have been used in the management of lactic acidosis (usually in an intensive care setting) although, because these events are so rare, the benefits are not really known. Uridine treatment (in the form of a sugar cane extract called "Mitocnol" or "Nucleomaxx") has shown a lot of promise as a preventive treatment that may limit the toxicity associated with either stavudine or zidovudine without compromising the effectiveness of these drugs against HIV, but these results have been obtained primarily from the laboratory. Clinical experience with this extract is minimal at present and the cost of treatment is about $100 per month. I'm sure more information will be available in the next year or so, which will be watched with interest.
We have seen several studies that show improvement in limb fat after switching patients from AZT or d4T to abacavir or tenofovir. Unfortunately, no one has really been able to quantify facial fat in those studies. Do you think that patients with moderate to severe facial wasting may see improvements after this switch? Some of your data show loss of fat cells due to macrophage activation. Do you think people may not have enough fat cells for facial lipoatrophy reversal?
This is an important area, but one where it has been difficult to obtain good data. Facial lipoatrophy is obviously one of the most stigmatizing aspects of lipodystrophy, and this is an area where people really want to see improvement following NRTI switching. We do see improved facial appearance associated with NRTI switching, although in general the more severe the initial lipoatrophy (prior to switching) the more limited the improvement. This is a frustrating aspect of the syndrome, as it is a case of "first affected, last to improve." This means that less affected areas (such as the arms and trunk) tend to improve before more badly affected areas such as the face and legs. Also, because the process of "regrowing" populations of fat cells seems to be slow, improvements happen over years rather than months.
What is your personal opinion about the fact that AZT and d4T are still the most widely used drugs for the treatment of HIV in the developing world?
My own opinion is that we need to put this issue on the agenda and keep it there — particularly in the case of stavudine, where there is a substantial concern about toxicity issues. There is a rationale that stavudine may be better tolerated in non-white populations — and certainly this is an anecdotal opinion that is expressed by clinicians who look after mainly African-American patients. However, the potential toxicity of these drugs needs to be assessed carefully and quickly in the developing world, before the development of severe and widespread complications. We have already seen the burden of disease associated with severe lipoatrophy in our own communities — this history should not be reproduced in developing countries.
Besides lipoatrophy, what other long-term health implications does having decreased mitochondrial DNA or impaired function mean to someone with HIV?
In truth, probably very few implications. The effects of NRTI drugs on mitochondrial function appear to be very tissue-specific, so in the case of lipoatrophy it is very likely that the "damage" is limited specifically to fat tissue. For example, it is notable that stavudine treatment — which has been consistently associated with severe mitochondrial DNA depletion in adipocytes — has no significant effects on mitochondrial DNA in blood cells. Also, the mitochondrial toxicity associated with NRTIs appears to be readily reversible, in the sense that mitochondrial DNA depletion goes away quickly after ceasing/switching NRTI therapy.
Will we ever have data to show whether or not ddI is implicated in lipoatrophy?
Some historical data exists suggesting that ddI treatment is not associated with lipoatrophy. For example, an early study of dual NRTI therapy by Thierry Saint-Marc (published in 1999) included a number of patients receiving didanosine in both stavudine (d4T/ddI = 13/27, 48%) and zidovudine treatment groups (AZT/ddI = 13/16, 81%) that were well-matched for NRTI therapy duration. In this study, use of stavudine remained the most significant risk factor for lipoatrophy (relative risk 1.95 compared with zidovudine), while no significant didanosine effect could be demonstrated . More recently, results from the FTC-301A study also suggest that the once-daily NRTI drug emtricitabine (FTC) compares favorably with stavudine (each combined with ddI-EC and efavirenz) in terms of lipoatrophy risk over 72 weeks (n=571), while maintaining equivalent efficacy and improved overall tolerability. Average loss of fat was noted only in the stavudine group, despite the fact that ddI was used in both study arms . More long-term data are awaited in this study.
In your experience, does tenofovir have the same or different benefits when it comes to lipoatrophy reversal/prevention as abacavir? How about when it comes to lipids?
We have observed the same "non-toxic" effects on adipocyte's mitochondrial DNA levels for these drugs, and the study data certainly suggest that tenofovir and abacavir are not associated with risk of lipoatrophy (both with an incidence of <3% over 3 years). In this respect, neither drug will have a particular advantage. With regard to lipids, the "lipid lowering" effect of NRTI switching appears to basically represent the effect of removing stavudine from the HAART regimen (in NRTI switching strategies) or of removing PI drugs such as indinavir, rather than any special attribute of tenofovir or abacavir. There was some attempt to differentiate the effects of abacavir and tenofovir on lipid/lipoatrophy outcomes in presentations at this year's 12th Retrovirus Conference, but I think these will come to nothing once other confounding factors are taken into account. From the "lipodystrophy" point of view, including both metabolic and lipoatrophy outcomes, these are both good drugs.
What makes subcutaneous fat so different from visceral or dorsocervical fat? Why do nucleoside analogs not affect visceral fat? Why does insulin resistance seem to expand visceral and dorsocervical fat but not subcutaneous fat?
Probably the best way to conceive of the difference between subcutaneous fat and these other fat depots is that subcutaneous fat is the ideal storage site for dietary fat, while visceral fat tends to function as an "overflow" system if the fat-storing capacity of the subcutaneous fat tissue is exceeded.
This means that subcutaneous fat responds very effectively to the insulin stimulus that is associated with eating a meal. What happens in this transition from a "fasting" to a "fed" state, therefore, is that:
fat tissue stops breaking down its stores of triglyceride, (where it has been used as an energy source during a period of fasting);
fat tissue starts to take up dietary fatty acids very efficiently (so that it stores energy in the form of triglyceride for later use while the dietary sugars and protein provide an immediate source of metabolic fuel). Overall, about 50 percent of a dietary intake of fat gets stored in this manner.
The whole idea of this process is that fat stores are created while there is food around, so that there is an energy source available for a period of fasting (e.g., overnight, for those who don't have late-night snacks!). As you can imagine, this is a system that has evolved over thousands of years — when fast food hadn't been invented and the next meal wasn't always guaranteed!
In this context, visceral fat is different in that it doesn't respond to insulin so strongly — which means that it doesn't generally compete with subcutaneous fat tissue for the storage of dietary fat. It is also much more "labile," in that fatty acids are also released back into the circulation much more readily from visceral fat than from subcutaneous fat. Its purpose is probably to create a short-term store of fat that can be used again quickly, without running the risk of letting excess fatty acids build up in tissues such as muscle and liver where they can be quite damaging.
So, from a metabolic point of view, there are a couple of important points to make here. One is that you can create excess visceral fat by eating more fat than your subcutaneous stores can cope with. The second is that "insulin resistance" — which means that tissues (including subcutaneous fat but also liver and muscle) don't react to the presence of insulin appropriately — reduces the ability of subcutaneous fat to store fat in the most efficient way. This leads to the accumulation of fatty acids in all the wrong places (muscle, liver) and also creates a reservoir of fatty acids (from visceral fat) that is released at all the wrong times (e.g., even when you've just eaten and there is already plenty of metabolic fuel available).
How does this relate to lipodystrophy?
Visceral fat accumulation (around the organs) is one part of the "Metabolic Syndrome" that also includes insulin resistance and dyslipidemia, and these three elements cluster together quite strongly.
Excessive dietary intake of saturated fats and sugars contributes to the development of visceral obesity by exceeding the capacity of the subcutaneous fat to store dietary fat in an appropriate way. Lack of exercise also leads to fat not being "burned" as a source of energy. This means that you don't have to have HIV infection and/or PI therapy to get these problems — and indeed about 30 percent of U.S. adults are affected by Metabolic Syndrome.
Once visceral obesity is established it makes it harder to "recover" from insulin resistance, as there is always a source of fat that must also be "burned" along with dietary fat before the system can go back to efficient functioning.
One of the least understood areas is whether having lipoatrophy also contributes to the risk of insulin resistance and visceral obesity. This would make some sense, as having less and/or poorly functioning subcutaneous fat would be likely to make it easier to "overload" the capacity of this organ to store fat. Some recent data from the Netherlands [van Wijk JP, et al J Clin Endocrinol Metab. 2005 Mar 22; Epub ahead of print] support this possibility, indicating that more severe lipoatrophy increases the risk of insulin resistance. I think this is an example of how we need to think carefully about how to look after the health of those who are affected by severe lipoatrophy into the future.
Getting back to the original question, it is not really known why stavudine and zidovudine don't affect visceral fat in the way that they cause lipoatrophy. One explanation may be that visceral fat is intrinsically more resistant to mitochondrial toxicity, because (1) it doesn't rely so much on energy-requiring processes such as triglyceride synthesis, and (2) because it expresses "anti-apoptotic" proteins (one is called cIAP) that make visceral fat cells less susceptible to mitochondrial toxicity and subsequent cell death. At this stage, no one has collected fat samples from this fat depot in HIV-infected patients.
Besides serving as caloric storage and protection against cold weather, what other function does subcutaneous fat have?
This is currently a booming area in medical research, as it becomes increasingly recognized that subcutaneous fat actually functions as an active metabolic organ rather than as an inert storage site. There are many examples of how this plays out, but the case of adiponectin might be a good starting point. Adiponectin is basically a hormone that is released only from subcutaneous fat, which profoundly influences the way that fatty acid metabolism is regulated by the body as a whole. When subcutaneous fat tissue is healthy (and not affected by insulin resistance) it releases increased amounts of adiponectin into the system, which acts as a signal for the body to efficiently "burn" the fat that is present in muscle and liver. This means that adiponectin acts to protect against insulin resistance. However, when subcutaneous fat is not functioning properly, adiponectin levels go down and this then becomes part of the problem of insulin resistance — fat is allowed to build up in the wrong places and is not efficiently burned.
There is a lot more to this story, but to summarize, it is certainly true to say that subcutaneous fat is an integral player in metabolism generally — and is likely to be just as important to "metabolic health" as muscle or liver.
To contact Nelson Vergel: www.facialwasting.org.
1. Arribas JR. The rise and fall of triple nucleoside reverse transcriptase inhibitor (NRTI) regimens. J Antimicrob Chemother. 2004;54:587–92.
2. Keswani SC, Pardo CA, Cherry CL, Hoke A, McArthur JC. HIV-associated sensory neuropathies. AIDS. 2002;16:2105–17.
3. El-Sadr WM, Mullin CM, Carr A, Gibert C, Rappoport C, Visnegarwala F, Grunfeld C, Raghavan SS. Effects of HIV disease on lipid, glucose and insulin levels: results from a large antiretroviral-naive cohort. HIV Med. 2005;6:114–21.
4. Saint-Marc T, Partisani M, Poizot-Martin I, Bruno F, Rouviere O, Lang JM, Gastaut JA, Touraine JL. A syndrome of peripheral fat wasting (lipodystrophy) in patients receiving long-term nucleoside analogue therapy. AIDS. 1999;13:1659–67.
5. Saag MS, Cahn P, Raffi F, Wolff M, Pearce D, Molina JM, Powderly W, Shaw AL, Mondou E, Hinkle J, Borroto-Esoda K, Quinn JB, Barry DW, Rousseau F; FTC-301A Study Team. Efficacy and safety of emtricitabine vs stavudine in combination therapy in antiretroviral-naive patients: a randomized trial. JAMA. 2004;292:180–9.