The Lost Benefits of Antiretroviral Drug Use in South Africa: Fact or Fiction?

In their Perspectives article entitled ‘Estimating the Lost Benefits of Antiretroviral Drug Use in South Africa’ [Chigwedere et al (J AIDS 2008; 00:000-000)], a team of researchers from Harvard compared the number of persons who received antiretroviral (ARV) drugs for treatment or prevention of mother-to-child- HIV transmission (PMTCT) between 2000 and 2005 with an alternative, hypothetical scenario of those persons for whom it was reasonably feasible to use ARVs for treatment or PMTCT during the same period. Their conclusion: “More than 330,000 lives or approximately 2.2 million person-years were lost because a feasible and timely ARV treatment program was not implemented in South Africa ” and “Thirty-five thousand babies were born with HIV, resulting in 1.6 million person-years lost by not implementing a PMCT prophylaxis program using nevirapine”.

Although these hypothetical numbers sound startling, many more lives could have been possibly saved had nutritional supplementation been factored into their comparison, especially since persons living with HIV/AIDS in South Africa are known to be malnourished and nutritional supplementation has been shown in other well-controlled studies to prolong lives in patients with this disease. The report, however, completely ignores the role of nutrition in this patient population and fails to acknowledge the severe toxicity of ARV drugs such as zidovudine (ZDV or AZT) and nevirapine, which have been shown to damage the immune system and the liver, posing a risk for drug-induced immune deficiency. The report’s findings are thus misleading and covered with several inaccurate statements and inferences.

An analysis of key distortions of this misleading report is presented below:

THE J AIDS PERSPECTIVE SAYS: Except among very few scientists, such as Peter Duesberg, the scientific community has accepted HIV as the cause of AIDS for more than 20 years.

OUR ANALYSIS SAYS: There is no doubt that HIV exists. However, based on the time it takes from HIV infection to AIDS and the manifestation of AIDS-related symptoms, HIV cannot be the sole cause of AIDS. Thus, the period it takes for the development of disease symptoms following infection with HIV, a persistent virus, is 9 to 10 years – a remarkable fact, to say the least. In marked contrast, the incubation period for manifestation of symptoms for acutely infecting viruses such as measles is 8-10 days and for influenza 1 to 10 days. Additionally, HIV-infected persons, free of AIDS, with a long-term history of disease non-progression (“long-term non-progressors”) have been reported. These observations are consistent with the conclusion that HIV infection is not the sole cause of AIDS and that other factors, such as nutritional deficiencies linked to viral immune deficiency, must play a role in the AIDS development. This fact, however, has been ignored by the HIV community and pharmaceutical companies, which, in the hopes of promoting and exclusively marketing toxic ARV drugs, purposely undermine the role of nutrient status and other factors in disease development.

THE J AIDS PERSPECTIVE SAYS: ZDV was tested for AIDS treatment in controlled randomized clinical trials and its side effects were clearly documented and disclosed.

OUR ANALYSIS SAYS: The controlled randomized trials referred to here pertain to the licensing study used to obtain approval of zidovudine (AZT) use in AIDS (Fischl et al., (a) New England Journal of Medicine (1987); 317: 185-191); Richman et al., New England Journal of Medicine (1987); 317: 192-197). This was one of the very few controlled studies, mostly short-term, that looked at the effects of ARVs on clinical parameters such as survival or disease progression. Although the study provided the only evidence for clinical benefit of ARVs, the trial was problematic from start to finish. The benefits were short-term in duration and were attended by serious adverse reactions. Approximately 24 % of AZT recipients (4-month survivors) developed anaemia (reduced number of red blood cells or haemoglobin concentration), compared to 4% of those who received a placebo. Anaemia is the result of the toxic effect of ARVs on the bone marrow, the production site of red and white blood cells. Most importantly, 21% of the AZT recipients had to be kept alive by red-cell transfusions compared to 4% of the placebo (Richman et al., 1987).

Moreover, the short-term benefit of this ARV study was not maintained in a follow-up 21-month study (Fischl et al., Journal of the American Medical Association (1989) 262: 2405-2410) , which found a rapid decline in the survival benefit of AZT reported for the first 4 months. No long-term clinical proof for survival exists from any other controlled studies. A few years after the US AZT licensing study, the British-French Concorde study was published (Seligman et al., Lancet (1994); 343: 871-878) showing unambiguously that AZT was unable to prevent AIDS in HIV-asymptomatic individuals.

THE J AIDS PERSPECTIVE SAYS: Later studies showed that in combination with other drugs, therapy was very efficacious, resulting in the name highly active antiretroviral therapy for triple-drug cocktails.

OUR ANALYSIS SAYS: The ‘later studies’ referred to here are not placebo-controlled studies but rather comparative trials in which one ARV drug regimen was compared with another without inclusion of an ARV-free control group. In the specific reference cited (Hammer et al (1997) New Engl J Med 337 (11): 725-733), two nucleoside analogues [zidovudine (or stavudine) and lamivudine] plus a protease inhibitor (indinavir) were compared with two nucleoside analogues alone in persons with HIV infection and CD4 cell counts of 200 per cubic millimeter or less. The study found that the proportion of patients whose disease progressed to AIDS or death was lowered from 11% with 2 nucleoside analogues to 6% with the 3-drug combination. Since no placebo group was studied, it is difficult to assess the efficacy of the triple combination therapy with respect to a drug-free control. Hence the efficacy of the triple drug therapy is only relative with respect to nucleoside analogues which by themselves were found to have poor efficacy in placebo-controlled trial. Thus, the efficacy of the drug combination evalauted was based on comparison with a low therapeutic standard.

Although the doses of NRTIs (nucleoside reverse transcriptase inhibitors) have been lowered in the triple drug combination cocktails such as Highly Active Antiretroviral Therapy (HAART), the toxicity is still evident as documented by the reports of thrombocytopenia and anaemia in the following studies: Carbonara et al., J Infect (2001) 42(4): 251-256; and Mocroft et al., AIDS (1999); 13: 943-950.

Additionally, both NRTIs and non-nucleoside reverse transcriptase inhibitors (NNRTIs) contained in triple drug cocktails (e.g. HAART) are hepatotoxic, i.e. cause damage to the liver (Abrescia et al., Current Pharmaceutical Design (2005); 11: 3697-3710). Further, NRTIs in HAART have been linked to mitochondrial toxicity (Fleischer et al., Clin Infect Dis (2004); 38(8): e79-80) and the protease inhibitors in HAART have been associated with a lipodystrophy syndrome, characterised by dyslipidemia, i.e. a severe impairment of the metabolism and storage of fats in the body (Carr et al., Lancet (2001); 357: 1412-1414).

Many ARVs used in HAART also have other serious adverse side effects. A number of metabolic complications are associated with the use of some ARVs, including derangement in glucose and lipid metabolism, bone metabolism and lactic acidaemia (sustained accumulation of lactic acid in the blood), which have been documented in industrialised countries. Thus, there are attendant risks of drug-induced toxicity in triple combination cocktails.

THE J AIDS PERSPECTIVE SAYS: ZDV was tested for PMCT of HIV in a randomized clinical trial that showed much benefit and little risk. The consensus is that ZDV’s benefits outweigh its side effects………..

OUR ANALYSIS SAYS: The authors of the J AIDS report cite a study by Connor et al [N Engl J Med (1997); 337: 725-733) showing reduction of maternal-infant transmission of HIV-type 1 with zidovudiine (ZDV) treatment. Although in the cited report a regimen consisting of ZDV reduced the risk of maternal-infant HIV transmission by approximately two thirds, it is questionable whether zidovudine’s benefits outweigh its side effects. In other studies damaging prenatal/perinatal effects of ZDV have been reported in the PMCT setting. These studies include the following:

Given the poor efficacy of ZDV in adult AIDS patients (reviewed above) together with the toxic side effects of this DNA-synthesis inhibitor to all rapidly dividing cells in the body, including those of the bone marrow and blood-cell system, its highly unlikely that the benefits of ZDV outweigh its risks.

THE J AIDS PERSPECTIVE SAYS: More than 330,000 lives or approximately 2.2 million person-years were lost because a feasible and timely ARV treatment program was not implemented in South Africa .

OUR ANALYSIS SAYS: To estimate the average life-years that ARV therapy adds to patients with AIDS in Africa , the authors of the J AIDS article used “the very conservative estimate of an average ARV treatment benefit of 6.7 years per patient”. This applies to ARV use in late disease treatment compared to 9.8 years if ARV treatment is provided earlier. It essentially translates to average benefits of 6.7 to 9.8 years for ARV treatment in South Africa . These numbers are not impressive compared to global rate of AIDS mortality based on WHO estimates. Thus, for the year 2000, which represents the beginning of the period evaluated in the J AIDS report, it can be calculated based on WHO reports [WHO, Weekly Epidemiological Record (2001); 76: 381-388] that the average annual global mortality rate for all HIV-positives (including the minority on ARV drugs) is ~1.4% (assuming that all newly diagnosed cases died in the same year). According to this estimate, less than 14% of the global population would die from AIDS in 9.8 years, the projected maximum lifespan for persons on ARV treatment. This suggests that HIV positive persons on ARV drugs in the South African population estimated in the Perspectives article would have a 7 times higher chance of dying from AIDS compared to the average annual AIDS mortality of all HIV positives on earth (including the minority on ARV drugs).

The high death rate from AIDS in South Africa , despite ARV treatment, underlies the fact that the population is primarily malnourished. In an analogous malnourished cohort of HIV-infected persons in Singapore, Paton et al (HIV Medicine (2006); 7(5): 323-330) reported that for patients who commenced HAART, the hazard ratio for death for those with moderate to severe malnutrition was 6-fold higher than for those with normal nutrition status. A significant association between nutritional status and survival was also observed in the pre-HAART era. Thus, if nutritional supplementation had been provided to the South African HIV-infected population, then by extrapolation, it can be estimated that the subjects would have survived 6 times longer (approx 40.2 years) than the 6.7 life years on ARV therapy alone. Given these estimates, the emphasis of the missed ARV therapy on patient life years is misplaced compared to the lost benefits that could have accrued from nutritional supplementation therapy.

While all deaths from AIDS are tragic, the fact is that a great many of them could have been avoided through nutrient supplementation. No mention is made of this nor of the fact that ARV therapies can be toxic with limited effects on survival. As long as the myth persists that toxic ARVs are safe and effective in prolonging lives of HIV/AIDS patients, many will continue to die when their lives otherwise could have been prolonged through the use of nutrient supplementation.