A Look at Mitochondrial Toxicity

Dave Gilden

September 25, 1999: A group of French doctors report on five cases of a rare syndrome of cerebral, neurologic, and/or retinal damage in 8 of 2,000 babies born to HIV positive mothers. This large mother-child cohort had received AZT (Retrovir) or AZT/3TC (Epivir) to prevent HIV transmission during pregnancy and delivery. The babies also displayed a variety of other abnormalities in muscle, heart, pancreas, and bone marrow activity. Underlying defects in energy production also were apparent. Two of the babies died after about a year. All were HIV negative.

These incidents are but the latest chapters in a long story. Almost since AZT was introduced in 1987, scientists have recognized that several drugs in the NRTI class are not very specific; they block the growth of HIV but they also can be toxic to human cells, sometimes with life-threatening consequences.

NRTIs including ddI, d4T, lodenosine, and adefovir work against HIV by disrupting the function of the virus's reverse transcriptase enzyme, which converts the HIV gene set into a DNA form that inserts itself into human cell genomes. Nucleoside analog drugs are essentially defective versions of natural nucleosides (precursor compounds). Natural nucleosides are first phosphorylated (chemically reconfigured) into nucleotides and then chained together to form DNA; this is done by both reverse transcriptase and the cells' own enzymes during gene replication. The chemical structures of NRTIs lack the necessary molecular hooks to continue this process; they terminate the chain being constructed by reverse transcriptase and block the infection of new cells.

NRTIs generally do not affect construction of new DNA in the nuclei (center) of somatic (body) cells because the enzymes that drive this process in the cells have a "proofreading" mechanism that snips out the nucleoside analogs if they are inserted into a new DNA chain. There is an exception to this rule, though: the cells' weak point lies in the mitochondria, the energy-producing components of the cell.

Eons ago, at the dawn of evolution, mitochondria apparently were independent bacteria-like cells. At some point, they took up abode in larger, more advanced cells with nuclei and entered into a symbiotic relationship. Today mitochondria retain some autonomy within nearly all nonbacterial cells. These organelles, which are small pouches of deeply folded membranes, break down sugar and fat to create energy that the cell can use for its myriad chemical processes.

Hundreds of mitochondria exist in each cell and replicate independently of the cell's own proliferation. The existing mitochondria are doled out to the daughter cells during mitosis (cell division); the mitochondria then replicate in each cell according to the cell's energy needs. Yet despite their integral role within cells, mitochondria retain certain unique genes that govern the production of a very small amount of proteins and enzymes. Two to ten redundant copies of the mitochondrial genome exist within each mitochondrion in the form of circular double-stranded DNA.

Mitochondrial genes are particularly prone to damage. DNA polymerase gamma, the enzyme that directs replication of the mitochondrial genes, is more primitive than the DNA polymerase enzymes found in the cell nucleus. Polymerase gamma has no "proofreading" function, so little repair of errors occurs when stringing nucleotides together. Nucleoside analogs therefore inhibit DNA polymerase gamma in the same way they do reverse transcriptase.

Mitochondria rely on their genetic redundancy to protect against errors in their genes. Defective DNA coexists and replicates alongside correct DNA, which covers for the defect. Similarly, functional mitochondria can compensate for defective mitochondria within the same cell.

The system breaks down only when damaged mitochondrial DNA reaches a threshold proportion somewhere above 70%. Cells then begin to suffer from energy deficiencies and turn increasingly to anaerobic processes (i.e., without using oxygen) outside the mitochondria. Anaerobic respiration is much less efficient than mitochondria's oxidative phosphorylation (use of oxygen to break down sugar and fat); furthermore, it causes the production of lactic acid, a substance that reduces the normal pH of blood.

Mitochondrial Toxicity of NRTIs

Specific disease syndromes are also connected with rare inherited mitochondrial mutations. Mitochondria-related diseases vary in severity from person to person, and symptoms frequently appear only as a person ages. Tissues such as muscles and nerves, which require high levels of energy, are most often involved. Some of the specific conditions related to inadequately low mitochondrial activity are muscle wasting (myopathy); heart failure (cardiomyopathy); peripheral numbness and pain (neuropathy); generalized loss of the kidney's ability to filter the blood (proximal renal tubular dysfunction or Fanconi-like syndrome); low blood cell counts (anemia, leukopenia, thrombocytopenia, or pancytopenia); swelling and fatty degeneration of the liver (hepatomegaly with steatosis); and pancreatic inflammation (pancreatitis). Fatigue, psychological depression, and high lactic acid levels (lactic acidosis) are more generalized signs.

Researchers reported a connection between myopathy and AZT in 1990. A year later an article was published associating AZT with heart muscle damage in particular. Other signs of mitochondrial toxicity were associated with NRTIs in the years that followed. Why different NRTIs have different effects and why different individuals have varying sensitivity to those effects are currently unknown. The variations between individuals may relate to the mitochondrial mutations already present and to their distribution in various organs, which is not necessarily uniform.

The varying sensitivity of different tissues to specific NRTIs presumably arises from variations in the penetration of those drugs in different cells and the energy requirements of those cells. For example, adefovir selectively and severely affects kidney cells because it binds to a kidney cell transporter molecule (hOAT1). This binding causes adefovir to build up inside the cells lining the walls of renal tubules, where unwanted compounds are transferred from the blood to the urine.

D4T and ddC represent two critical cases in which mitochondrial-associated toxicity -- in these cases, neuropathy -- was noted to be so severe that dose levels had to be greatly restricted during research and development. Both compounds originally tested as highly potent antiretroviral agents. In the case of d4T, trials originally included daily doses as high as 900 mg. Trial participants reported immediate improvements in their health, but two-thirds were forced to quickly abandon the drug due to severe neuropathy.

In the end, the committee monitoring the large d4T parallel track program found that even 40 mg twice daily was too much; about a quarter of the persons taking that dose quit the parallel track program because of neuropathy. The development of neuropathy was noted to be mediated by CD4 cell count, i.e., the condition was more prevalent among people with lower CD4 cell counts. Peripheral neuropathy also developed after an average follow-up of 79 weeks in one-half of the participants in a d4T monotherapy trial who were using 40 mg twice daily. Nonetheless, 40 mg twice a day became the standard dose.

FDA-approved package inserts for all NRTIs carry a prominent warning that describes the potential side effect of liver degeneration. Specifically, the inserts warn: "Lactic acidosis and severe hepatomegaly with steatosis, including fatal cases, have been reported with the use of nucleoside analogs alone or in combination … A majority of these cases have been in women. Obesity and prolonged nucleoside exposure may be risk factors." This hepatic degeneration resembles "acute fatty liver" of pregnancy, a maternal syndrome that threatens both mother and fetus and results from an inherited fetal inability to oxidize fatty acids in the mitochondria.

These warnings about liver degeneration are based on isolated incidents that were reported to occur after drug approval. It is difficult to say how frequent they are now. A 1989-1994 review of persons seen at Johns Hopkins' HIV Clinic recorded an incidence of 1.3 per 1,000 person-years of severe cases of hepatomegaly with steatosis and lactic acidosis. (The number of persons multiplied by the number of years equals person-years; e.g., one person followed for ten years equals ten person-years and ten persons followed for one year also equals ten person-years.) A more recent French study found a frequency of 8.4 per 1,000 person-years for symptomatic high lactic acid levels.

In the case of ddI, the incidence of severe pancreatitis has decreased in recent years as people's overall health has improved with the success of potent anti-HIV regimens. Frequent monitoring should be done to detect the presence of pancreatic enzymes in the blood -- an early sign of possibly developing pancreatitis. Sometimes, however, this condition develops rapidly and without forewarning; doctors therefore should not let their guard down. Somewhat alarmingly, the precise risk factors predisposing an individual on ddI to pancreatitis, as with other mitochondrial toxicities, remain unexplored. The effects of combining ddI with other drugs such as d4T and hydroxyurea also have come as a surprise. Even more surprising is that this situation emerged after a decade of testing and marketing ddI. Overall, more adverse effects are seen in the setting of alcohol abuse.

NRTIs rarely inflict severe mitochondrial damage -- with a few exceptions, notably adefovir's impairment of the kidneys; ddI, d4T and ddC's neuropathy; and high-dose AZT's anemia. Frequent monitoring of organ function through lab tests (specifically, tests of liver function, phosphorus, and carbon dioxide, as well as complete blood counts) usually can prevent serious outcomes. Monitoring serum lactate, while somewhat difficult because the test must be conducted immediately after blood is drawn, may also be useful, as indicated by one recent, preliminary report by Marianne Harris, MD, from the British Columbia Centre for Excellence in HIV/AIDS. Also, stopping the drugs commonly results in gradual recovery, probably because mitochondrial replication restores energy production. However, the FDA continues to receive reports of deaths from lactic acidosis, although mild increases in lactic acid levels that result in fatigue may be much more common than life-threatening cases.

New Long-Term Mitochondrial Toxicities

After a decade's experience with one NRTI or another, there are now three or four potent, well-established drug combinations that include protease inhibitors (PIs) or non-nucleoside reverse transcriptase inhibitors (NNRTIs). Combining these drugs also combines their toxicities in new and unfamiliar ways. More importantly, antiviral drugs are now being taken for much longer periods of time than in the past. With durable suppression of HIV caused by highly active antiretroviral therapy (HAART), the health of the individual has stabilized but residual HIV survives (see "HIV Persists despite HAART," Spring 2000 BETA). Not only is this a new era in HIV therapy; it is a new era with regard to long-term toxicities.

Chronic, unsuppressed HIV infection has long been known to trigger a wasting syndrome in which people with HIV lose lean tissue mass (mostly muscle), sometimes to a life-threatening extent. Some researchers have noted that subcutaneous (under the skin) fat loss accompanies the loss of lean tissue, especially in women and overweight men. The worst deficit, it appears, has been an inability to regain lean mass lost during severe opportunistic infections (OIs).

Today, some people taking HAART have trouble maintaining or restoring proper fat distribution. Fat tends to accumulate over time around the central organs and sometimes around the back or front of the neck. Meanwhile, lipoatrophy -- the depletion of subcutaneous fat stores, especially in the limbs and cheeks -- continues.

This widespread body fat rearrangement, affecting up to two-thirds of participants in various studies, has been a curious phenomenon (see "Body Fat Changes: More than Lipodystrophy," January 1999 BETA, and "HAART Attack: Metabolic Disorders during Long-Term Antiretroviral Therapy," April 1999 BETA). It was assumed at first that the new PIs were responsible for this "lipodystrophy" (now sometimes also referred to as HARS, HIV-Associated Adipose Redistribution Syndrome).

After the first flood of reports, physicians began to recall that they had seen some cases of fat accumulation at the base of the neck and between the shoulder blades (dorsocervical fat pads, commonly called "buffalo humps") in the pre-HAART days. This buildup sometimes was accompanied by lipoatrophy. In addition, there is a marked similarity between HAART-associated lipodystrophy and a form of multiple symmetrical lipomatosis (abnormal fat accumulations), also known as MSL or Madelungs disease. Both syndromes involve peripheral fat loss and buffalo humps. Persons with MSL do not have fat deposits in the torso, but they do exhibit neuropathy. Several reports indicate that mutations at a particular point in the mitochondrial DNA accompany MSL.

A recent survey of body fat abnormalities in Australia's HIV positive population recorded an overall prevalence of 54%, including 63% of those who had taken PIs and 32% of those with exposure to treatments that did not include PIs. The same research group, headed by Andrew Carr, MD, and David Cooper, MD, of St. Vincent's Hospital in Sydney, went on to compare 14 men receiving NRTIs only who were experiencing peripheral fat loss with or without fat accumulation with three different control groups without lipodystrophy. These three control groups were comprised of 32 treatment-naive men, 28 men on NRTIs only, and 44 men receiving NRTIs plus PIs. A fourth comparison group contained 102 men with lipodystrophy who were taking both NRTIs and PIs.

The NRTI-only cases of lipodystrophy were distinguished by higher lactic acid levels, which correlated with losses in body weight and signs of liver dysfunction. These signs included swollen livers and elevated liver enzymes in the blood, all of which are suggestive of mitochondrial toxicity.

Liver toxicity has particular implications for aggravating mitochondrial toxicity, since one of the roles of the liver is to remove lactic acid from the blood. Liver function is critical for fat metabolism, too, given that organ's central role in converting excess blood lipids (fats) into glucose (sugar) or vice versa as required by the body's activity level and dietary intake. PIs can lead to high lipid and insulin levels.

Recent weight loss, fatigue, and nausea were significantly more common in the Australians' NRTI lipodystrophy group than in the combined therapy lipodystrophy group. The lipodystrophic men on PIs had higher blood levels of lipids (cholesterol and triglycerides), glucose, and insulin than did their counterparts on NRTIs only.

Body shape alterations were similar in the two groups. However, in a study done in France by L. Maulin, MD, and colleagues, this was not the case. The French researchers found that those on PI-containing regimens had markedly more visceral (abdominal) fat than those either on NRTIs only or taking no anti-HIV drugs. Both the protease and nucleoside groups had about the same amount of subcutaneous fat, which was substantially less than in the treatment-naive group.

Risk Factors

Significant risk factors for lipodystrophy in the Australian study included current use of d4T and the number of years on any NRTI. Each year of NRTI use increased the risk of lipoatrophy 1.26-fold. In contrast, each year of PI use trebled the risk of any form of lipodystrophy. Discontinuing NRTIs led to reductions in blood lactic acid and liver enzymes, nausea, and fatigue, but lost weight was not recovered (over an average four months of follow-up).

The French study also showed a strong association between lipoatrophy and d4T. In 1999, the same researchers reported on what happened when 29 persons with peripheral fat wasting were taken off d4T and switched to another drug(s). Fourteen who were also receiving a PI changed from d4T to AZT or abacavir (Ziagen). The rest replaced their d4T-containing NRTI regimen with AZT/3TC/nevirapine (Viramune). After six months, blood triglyceride levels decreased by approximately 40% in both groups while subcutaneous fat rose 40%.

Problems Ranking the NRTIs

Glaxo Wellcome, the maker of AZT, 3TC, and abacavir, is quick to point to such studies. Whether d4T is especially culpable in lipoatrophy has yet to be decisively documented, however. Most persons on d4T have taken AZT previously and have lived longer with HIV infection.

Likewise, laboratory studies that attempt to rank the different NRTIs' respective mitochondrial toxicity are open to question since they may not reflect what happens in living cells in the body. There, toxicity is determined not just by the drugs' chemical inhibition of DNA polymerase gamma, but also by the concentration of each compound within mitochondria and the ease with which it is activated via phosphorylation (addition of a phosphate group) before DNA polymerase can process it.

The energy demands of cells in the body, for which they rely on their mitochondria, are yet another decisive factor in eliciting the NRTIs' inhibition of mitochondrial activity. Gender and HIV infection itself are major factors that influence the metabolic and, hence, the energy profile of cells throughout the body.

Beyond DNA

NRTIs also may have other effects on mitochondrial metabolism besides interference with DNA replication. For example, while lodenosine seemed to have little effect on mitochondrial DNA in lab tests, it causes serious heart toxicity in mice and fatal liver toxicity in humans in a way that resembles mitochondrial failure. This NRTI in fact does disrupt mitochondrial functioning, as evidenced by the heightened lactic acid levels and death in laboratory cell lines exposed to it. According to Yung-chi Cheng, a Yale University researcher who did the original mitochondrial assays on lodenosine, its toxic effect probably results from inhibition of lactic acid dehydrogenase, an enzyme that converts lactic acid into a compound that the mitochondria can oxidize.

A parallel situation exists for AZT, claims a recent study conducted in a laboratory liver cell line by Jean-Pierre Sommadossi, PharmD, and colleagues from the University of Alabama. Unlike ddC and ddI, AZT did not inhibit mitochondrial DNA synthesis in this case, but AZT increased lactic acid production in cell cultures whereas other NRTIs did not. The researchers found that exposure to AZT directly reduced the activity of several of the mitochondria's major energy-producing enzymes. This was thought to be the reason for the contradictory results.

Contributing Toxicities

At this past spring's Keystone Symposium on New Biological Approaches to HIV-1 Infection, Scott Raidel of the Department of Pathology at Emory University showed striking photos of mitochondria in the heart muscle of mice bioengineered to contain a gene for the Tat protein of HIV; the mitochondria were extremely elongated and their internal structures disrupted. That is, mice that were given genetic implants so that they developed higher levels of the Tat protein were then observed to have significant mitochondrial damage. Their DNA content was about 30% lower than in normal mice.

When NRTIs enter the body, they encounter an environment that may be already prone to the type of mitochondrial damage that these drugs promote. Tat is a protein that helps convert the HIV DNA genome within cells back into RNA for packaging in a new generation of virions (complete virus particles). Among other things, it suppresses the cells' antioxidant protective system, rendering them more sensitive to inflammatory activation by such cytokines as tumor necrosis factor (TNF). (Cytokines are hormones or messenger proteins that coordinate immune system responses.) Uninfected cells in lymph nodes are chronically exposed to Tat and suffer the effects of oxidative stress. Many die with swollen, effects.

One widely advertised agent that has some effect in countering the anemia associated with certain NRTIs is erythropoietin (Procrit), a growth factor that stimulates red blood cell production. Erythropoietin helps reverse anemia caused by AZT but has no effect on other sources of fatigue, including muscle weakness or high lactic acid levels associated with mitochondrial dysfunction. People taking erythropoietin must continue to be monitored for such effects, and in some cases the offending drug still must be stopped or changed.

Another agent that might reduce mitochondrial toxicity is replacement L-acetyl carnitine, which is itself reduced by mitochondrial toxicity. L-carnitine is a product of protein breakdown that is used to transport fat components into the mitochondria for oxidation. L-acetyl carnitine, a variant of L-carnitine, shares this function and also helps damaged nerve cells regenerate. One four-person study has now indicated that L-acetyl carnitine promotes nerve growth and reduces symptoms in people with drug-related neuropathy. (Note that L-acetyl carnitine is not the L-carnitine sold in health food stores.)

A previous Italian study, led by Andrea Cossarizza of the University of Modena, of T-cells taken from people with primary (early) HIV infection found electric charge alterations in mitochondrial membranes and noted a strong tendency for these cells to undergo spontaneous cell death. In the test tube, L-acetyl carnitine, as well as N-acetyl cysteine and nicotinamide (niacin, a B vitamin), was able to rescue the cells, which correlated with the subjects' blood HIV and TNF levels. This association with TNF suggested that the cell rescue involved a process that reversed oxidative stress.

A Spanish study, led by José Garcia de la Asunción and colleagues at the University of Valencia, has tried a similar strategy using high doses of the antioxidant vitamins C (1 g per day) and E (0.6 g per day). Chemical markers of muscle damage and oxidative stress decreased in eight HIV positive persons on AZT plus the vitamins compared with a control group on AZT alone. The same was true in a group of study mice. Examination of the mitochondria in the mice's muscle cells further showed that the mice receiving AZT plus the vitamins retained normally organized mitochondria whereas the mice on AZT alone had swollen, disrupted mitochondria.

The Big Picture

Mitochondria play a vital role in the body's metabolism; protecting mitochondria is essential to sustaining life. Although an abundance of research suggests that certain NRTIs severely disrupt specific mitochondria, the exact significance and nature of that toxicity requires further elucidation. Development of nontoxic treatments to support energy production and of safer NRTIs is critically important. However, that research is still in rudimentary stages and lacks substantial industry or government support. Worse yet, the commercial implications of present NRTIs' mitochondrial toxicity threaten to disrupt the objective scientific process.

Still, it should be noted that mitochondrial toxicity results at least in part from the use of antiretroviral drugs -- and there are risks and benefits involved with the use of most if not all drugs, not only anti-HIV medications. Simply put, when the risks outweigh the benefits, the drug in question is not used; when the benefits outweigh the risks, the drug is used. While further research and data may change the risk-benefit ratio, the benefits for the vast majority of people with HIV whose lives have been improved and extended through the use of antiretroviral therapy still far outweigh the risks.

Dave Gilden is Director of Treatment Information Services at amfAR.

Blanche, S. and others. Persistent mitochondrial dysfunction and perinatal exposure to antiretroviral nucleoside analogues. The Lancet 354(9184): 1084-1089. September 25, 1999.

Bridges, E.G. and others. Favorable interaction of _-L(-) nucleoside analogues with clinically approved anti-HIV nucleoside analogues for the treatment of human immunodeficiency virus. Biochemical Pharmacology 51(6): 731-736. March 22, 1996.

Brinkman, K. and others. Mitochondrial toxicity induced by nucleoside-analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral-therapy-related lipodystrophy. The Lancet 354(9184): 1112-1115. September 25, 1999.

Carbonari, M. and others. Death of bystander cells by a novel pathway involving early mitochondrial damage in human immunodeficiency virus-related lymphadenopathy. Blood 90(1): 209-216. July 1, 1997.

Carr, A. and others. A syndrome of lipoatrophy, lactic acidæmia and liver dysfunction associated with HIV nucleoside analogue therapy: contribution to protease inhibitor-related lipodystrophy syndrome. AIDS 14(3): F25-32. February 18, 2000.

Chen, C.H. and others. Effect of anti-human immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Molecular Pharmacology 39(5): 625-628. May 1991.

Cossarizza, A. and others. Mitochondria alterations and dramatic tendency to undergo apoptosis in peripheral blood lymphocytes during acute HIV syndrome. AIDS 11(1): 19-26. January 11, 1997.

Dalaka, M.C. and others. Mitochondrial myopathy caused by long-term zidovudine toxicity. New England Journal of Medicine 322: 1098-1105. 1990.

Dutschman, G.E. and others. Metabolism of 2',3'-dideoxy-2',3'-didehydro-beta-L(-)-5-fluorocytidine and its activity in combination with clinically approved anti-human immunodeficiency virus beta-D(+) nucleoside analogs in vitro. Antimicrobial Agents and Chemotherapy 42(7): 1799-1804. July 1998.

Fortgang, I.S. and others. Hepatomegaly and steatosis in HIV-infected patients receiving nucleoside analog antiretroviral therapy. American Journal of Gastroenterology 90(9): 1433-1436. September 1995.

Fouty, B. and others. Riboflavin to treat nucleoside analogue-induced lactic acidosis. The Lancet 352(9124): 291-292. July 25, 1998.

Garcia de la Asunción, J. and others. AZT treatment induces molecular and ultrastructural oxidative damage to muscle mitochondria: prevention by antioxidant vitamins. Journal of Clinical Investigation 102(1): 4-9. July 1998.

Ibdah, J.A. and others. A fetal fatty-acid oxidation disorder as a cause of liver disease in pregnant women. New England Journal of Medicine 340(22): 1723-1731. June 3, 1999.

Kotler, D.P. and others. Studies of body composition and fat distribution in HIV-infected and control subjects. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 20(3): 228-237. March 1, 1999.

Kulkarni, P.M. and others. Lactic acidosis and hepatic steatosis in HIV seropositive patients treated with nucleoside analogs. Digestive Disease Week 2000. San Diego, CA. May 21-24, 2000. Abstract and poster 95.

Lewis, W. and others. Mitochondrial ultra-structural and molecular changes induced by zidovudine in rat hearts. Laboratory Investigation 65: 228-236. 1991.

Lewis, W. and Dalaka, M.C. Mitochondrial toxicity of antiviral drugs. Nature Medicine 5(1): 417-422. January 1999.

Lo, J.C. and others. "Buffalo hump" in men with HIV-1 infection. The Lancet 351(9106): 867-870. March 21, 1998.

Luzzati, R. and others. Riboflavin and severe lactic acidosis. The Lancet 353(9156): 901-902. March 13, 1999.

Mulligan, K. and others. Cross-sectional and longitudinal evaluation of body composition in men with HIV infection. Journal of Acquired Immune Deficiency Syndromes 15(1): 43-48. May 1, 1997.

Pan-Zhou, X.R. and others. Differential effects of antiretroviral nucleoside analogs on mitochondrial function in HepG2 cells. Antimicrobial Agents and Chemotherapy 44(3): 496-503. March 2000.

Parker, W.B. and Cheng, Y.C. Mitochondrial toxicity of antiviral nucleoside analogs. Journal of NIH Research 6: 57-61. March 1994.

Purnell, J.Q. and others. Effect of ritonavir on lipids and post-heparin lipase activities in normal subjects. AIDS 14(1): 51-57. January 7, 2000.

Raidel, S. and others. HIV-1 tat construct targeted to adult cardiac ventricular myocytes alters heart structure and function. Keystone Symposium on Novel Biological Approaches to HIV-1 Infection Based on New Insights into HIV Biology. Keystone, CO. April 4-10, 2000. Abstract 457.

Saint-Marc, T. and others. The effects of discontinuing stavudine therapy on clinical and metabolic abnormalities in patients suffering from lipodystrophy. AIDS 13(15): 2188-2189. October 22, 1999.

Saint-Marc, T. and others. Fat distribution evaluated by computed tomography and metabolic abnormalities in patients undergoing antiretroviral therapy: preliminary results of the LIPOCO study. AIDS 14(1): 37-49. January 7, 2000.

Savès, M. and others. Severe hepatic cytolysis: incidence and risk factors in patients treated by antiretroviral combinations: Aquitaine cohort, France, 1996-1998. AIDS 13(17): F115-121. December 3, 1999.

Tsai, C.H. and others. Effect of anti-HIV 2'-beta-fluoro-2',3'-dideoxynucleoside analogs on the cellular content of mitochondrial DNA and on lactate production. Biochemical Pharmacology 48(7): 1477-1481. October 7, 1994.

Wallace, D.C. Mitochondrial diseases in man and mouse. Science 283(5407): 1482-1488. March 5, 1999.

Page last updated 16 August 2000