Top Menu

Under suspicion: the HIV drug that held out hope for millions

Fresh cause for concern over the side-effects of nevirapine

Doubts emerge over clinical studies of crucial HIV drug

By Neville Hodgkinson – The Business , 30/31 January 2005


A tragedy of global proportions is unfolding over a toxic anti-HIV drug given to hundreds of thousands of women and babies in the developing world in the belief that it can help prevent the spread of Aids.

The drug, nevirapine, has become so central to Aids agencies’ efforts to support African and other developing nations that they are defending its use in dozens of poor countries, despite evidence that flaws in claims for its safety and effectiveness were covered up at the highest level by government scientists in the United States.

Nevirapine is acknowledged by Boeringer Ingelheim, its German manufacturer, to be capable of causing severe liver damage and life-threatening skin reactions soon after patients start taking regular doses . This month a new warning about its dangers was issued by US health officials. Deaths have been reported from several countries.

But a study published by The Lancet in 1999 purported to show that when given as a single dose to HIV-positive mothers at delivery, and to the babies within three days of birth, the drug safely reduces transmission of HIV from mother to child.

The study was initiated and funded by the powerful Division of Aids of the National Institute of Allergy and Infectious Disease (NIAID), part of the American Government’s massive National Institutes of Health (NIH) complex at Bethesda, Maryland.  The research, conducted on maternity wards in Kampala, Uganda, was led by investigators from Johns Hopkins University, Baltimore, Maryland.  The findings were hailed by Dr Anthony Fauci, NIAID’s director, as opening up “an entire new avenue” towards prevention of transmission of HIV in countries that could not afford more expensive drugs.

To help get nevirapine established for this purpose, Boehringer Ingelheim offered to provide it free for five years to government hospitals in developing countries.  Along with the drug’s endorsement by the World Health Organisation and the Joint United Nations Programme on HIV/AIDS (UNAIDS ), this gave Aids organisations a powerful tool for pressing for its rapid introduction and there are now 122 programmes administering the drug in 57 developing countries.  The product, which is also used as part of some anti-viral “cocktail” treatments for Aids patients, has today become the company’s sixth best-selling drug with sales totalling E310m ($412m, ₤217m) in 2003.

On 19 January, the US Food and Drug Administration (FDA) warned that cases of liver damage were more common with nevirapine, especially in women, than with other anti-HIV drugs. Some instances have been fatal, including in pregnant women. The FDA said doctors should weigh benefits and risks before prescribing the drug, adding that no serious toxicity or deaths have been reported with the single-dose treatment. The drug is not licensed for this use in the US or Europe.

In July 1999, even before the Uganda study results had been published, Boehringer Ingelheim asked South Africa’s Medicines Control Council (MCC) to fast-track approval of the single-dose regime in mothers and babies.  When South African authorities insisted it was not proven safe and that caution was needed, massive criticism followed from within and outside the country.

Worldwide media derision was directed particularly against South Africa’s president, Thabo Mbeki, who had questioned the relevance of Western approaches to Aids, particularly in the African setting, arguing that poverty and malnutrition were the real causes of immune deficiency there

The push behind nevirapine was one of the most extensive promotional drives by the world’s media for a pharmaceutical product, perhaps surpassed only by the way NIAID catapulted AZT, the first purported anti-Aids drug, onto world markets through studies that were also shown subsequently to be deeply flawed.  Manufactured by Burroughs Wellcome (a company now subsumed in GlaxoSmithKlin e), AZT was said to be the “gold standard” of Aids treatment until the biggest and longest trial, conducted jointly by French and UK government researchers, showed more deaths in patients given the drug early than in “controls” who received it later.

That finding was also downplayed and AZT survived, becoming used in much smaller doses as part of the “cocktail” treatments.  AZT is also given to pregnant mothers and their babies for the same purpose as nevirapine; it was in comparison with AZT that nevirapine was declared to halve the risk of transmission of HIV in the Uganda study, from 25% to 13%.

Yet some scientists argue that AZT is useless and dangerous.  They say that although nevirapine and AZT can reduce the proportion of babies who test HIV-positive, this may simply be a result of general suppression of the immune system by the drugs.  Scores of conditions, including pregnancy itself, as well as infections such as TB that are especially prevalent in Africa, have been shown to affect the immune system in ways that cause positive reactions with test kits used to diagnose “HIV” infection, but have nothing to do with HIV.

This interpretation is supported by a follow-up study in the Uganda trial showing that despite the halving of HIV-positivity attributed to nevirapine, 18 months later the overall death rates among the babies did not differ significantly between the AZT and nevirapine groups.  The death rates were also extremely high, averaging 12%.

Furthermore, the Uganda trial had no drug-free group of subjects to compare with the treated groups, even though other studies have shown “HIV” transmission rates below 13% when untreated.

In January 2001, South Africa’s MCC yielded to pressure from Aids campaigners and granted Boehringer Ingelheim a provisional licence for mother and baby treatment in South Africa, based on the single Lancet study.

Problems began to surface however in 2002, after the manufacturer applied to the FDA for a licence to sell the drug for the same purpose in America.  The Uganda study was central to the application; but an audit of the findings showed problems, including differences of opinion between the American researchers and Ugandan hospital staff over what constituted “serious adverse events” in the trial subjects.  When investigators asked for the original case files, trial overseers were unable to produce most of them.

In March 2002 the company withdrew its application, citing “questions regarding reporting and documentation” in the study.  Dr John LaMontagne, deputy director of the NIAID, reassured the press that “there is no question about the validity of the Lancet study … the problems are in the rather arcane requirements in record keeping.”

Last month, Associated Press, the American news service, reported that top NIH officials, including Fauci, had been warned by an audit team that the study may have under-reported severe adverse events among the trial subjects, including at least 14 deaths.  Researchers had also acknowledged that “thousands” of adverse events were not reported.

Boehringer Ingelheim, in a pre-audit check of the trial, found incomplete safety data and arbitrary definitions of severity of adverse events in patients, compounded by the fact that the investigators were “not actually seeing the patients whose events they are evaluating.”  A 16-page report on these findings, faxed to the NIAID’s Division of Aids, was marked by one of the division’s officials “Sensitive information.  Asked for it to be destroyed when audit is upon us.”

Other documents showed how NIH chiefs downplayed and delayed reporting the problems, fearing the potential impact on the drive to extend the use of nevirapine.  In particular, Fauci and Dr Edmund Tramont, head of the Division of Aids, were concerned not to jeopardise a $500m mother and child HIV prevention initiative for Africa, announced a few weeks later by President Bush, specifically to support the nevirapine roll-out.

In March 2003, Tramont released a report on a Division of Aids re-evaluation of the Uganda study, again asserting that the study’s overall conclusions on safety and effectiveness were reliable.  But Tramont excluded from this report the findings of an expert panel that had specifically reviewed the data on safety and which concluded that the safety claims in the original study could not be validated.  The panel drew attention to lack of adherence to serious adverse event reporting requirements, poor quality of subject records, a failure of study staff to use proper toxicity grading scales, and absence of staff supervision and quality control.

Tramont’s report also appears hard to reconcile with the findings of an independent audit of the study contracted previously by NIAID which found a “large number of infants with apparent failure to thrive past six months of age”; and that some and perhaps many infants had serious health problems at 12 months, suggesting “more pathology than had previously been reported.”

In July 2003, NIH appointed Dr Jonathan Fishbein, an expert on drug safety and development, to develop and oversee standards of conduct in clinical research in the Department of Aids through a newly-created post, Director of the Office for Policy in Clinical Research Operations.  Before his appointment, Fishbein was vice president of North American Medical Services at Parexel International, one of the largest contract research organisations in the world.

Four months into the job, Fauci presented him with a certificate expressing appreciation of his “outstanding contributions and efforts in support of the NIAID mission”.  But thereafter, things began to sour.  “It soon became apparent that I was hired not so much to change things at DAids [Division of Aids] but to give the appearance that I was,” Fishbein told Anthony Brink, a South African lawyer who has for several years been chronicling the nevirapine story (see ).

After persistently trying to see the irregularities he had uncovered addressed within the Division of Aids, Fishbein was sacked.  Unable to interest other NIH departments, or the Department of Health and Human Services that is responsible for NIH, he took his concerns to the US Congress.  Officials of the Subcommittee on Oversight and Investigations of the Energy and Commerce Committee took him seriously; they agreed with NIH a new review of the Uganda study, to be conducted by the Institute of Medicine.  But the review is charged only with assessing the integrity of the data and will not address issues of scientific misconduct, cover-up, or reprisal.

In testimony submitted to the inquiry earlier this month, Fishbein says it quickly bec ame apparent to him “that something did indeed go terribly wrong” with the Uganda study.  His statement, posted at, tells of “an atmosphere of intimidation” perpetrated by the Division of Aids leadership, especially towards its regulatory affairs branch; and of how NIAID officials “have conspired, and continue to conspire, to hide from the public serious deficiencies” in the Uganda trial that would otherwise invalidate its results.

Those deficiencies included poor record-keeping and lack of follow-up on adverse events, suggesting that “the care of study subjects was not sufficiently overseen by the study staff and jeopardised the safety of the subjects given this potent drug’s known serious and potentially life-threatening side effects.”

To date, Fishbein’s whistle-blowing efforts appear to have failed to shift opinion among leading Aids scientists, drug activists and their media supporters, who are so deeply convinced drugs are the way to prevent the spread of Aids that they regard the controversy as threatening lives.

One American journal, Science, said in its December 24 issue that the revelations had dismayed Aids researchers and clinicians around the world.  It quoted Dr Clifford Lane, NIAID’s deputy director, as worrying that the story “may cause people to stop using nevirapine”, so that “infants could be infected and die needlessly.”

A report from Washington in the UK science journal Nature, headlined “Nevirapine trial was not flawed, say researchers”, quoted the doctor-president of Global Strategies for HIV Prevention as declaring: “There are already mothers who are refusing to take nevirapine.  This is the most successful therapy in the entire Aids epidemic.  It should not be attacked.”

Nature said scientists and patient advocates had united to defend the treatment, pointing out that trials in South Africa, Malawi and Thailand have confirmed nevirapine’s safety and effectiveness. But in July 2003 South African drug regulators withdrew nevirapine’s provisional licence for perinatal use, citing “data integrity” problems in the Uganda study.

The Malawi trial cited is not comparable with the Uganda one, and shows quite different results: a reported transmission rate of 20.9% in babies who received nevirapine, and 15.3% in babies who received nevirapine and AZT together.

Furthermore, the conduct of this study, also led by Johns Hopkins researchers, has been challenged by staff working at the Zomba Central Hospital, one of the trial sites.  Dr Peter Safar, head of the department of Obstetrics and Gynaecology, and Dr Christian Fiala, an Austrian gynaecologist and long-standing critic of the theory that immune deficiency in Africa is caused by HIV, wrote to the South African Medical Journal in 2002 protesting that the organisers had “ignored the most basic principles of research in medicine”.

They said the deficiencies included a lack of proper information and counselling over HIV testing, failure to inform the women of possible side-effects and refusal to tell doctors and nurses on the wards which drugs their patients were receiving.


Top of page