Anthony Brink: The trouble with nevirapine

PART ONE

 Take nothing on its looks; take everything on the evidence, Pip; there is no better rule.

Jaggers the solicitor in Great Expectations

Charles Dickens

Nevirapine, the manufacturer tells us, is a “non-nucleoside analog reverse transcriptase inhibitor.” It prevents “viral replication” by binding “directly to the HIV RT enzyme”, thereby disrupting its ability to foster the formation of viral DNA. In other words it clogs reverse transcriptase a bit like spilt crude oil does penguins. But at a specific site. Like the bird’s head only.

Or think of the enzyme being like a bricklayer laying DNA building blocks. Nevirapine handcuffs his hands. Drugs in the AZT class, similarly described, but without the “non-” prefix, are also said to inhibit reverse transcriptase. But indirectly, like useless straw bricks slipping into the place of clay ones, to prevent the wall going up. So the theories go.

Let’s close our eyes and pretend just for now that reverse transcriptase has unambiguously been shown to exist as a distinct enzyme unique to retroviruses and not also a component of uninfected human cells, and that retroviruses exist outside textbooks, like Coffin’s, and children’s imaginations, as in the Superman cartoon movie Cold Vengeance: “a Roscoe’s retrovirus…100 per cent fatal”, and that retroviruses can be malevolent and have sufficient genetic complexity for the execution of their nefarious intentions, and that HIV is one of them. Killing off our CD4 immune cells. No evidence for that? OK, not attacking them directly, but hypnotising them into committing suicide. Even those they haven’t been anywhere near. Like telepathically. They call it “programmed cell-death.” For ‘AIDS sufferers’ with sky-high CD4 cell counts, we’ll think up an explanation another day. Likewise one for folks in peak health with cell counts at rock bottom. Who should be gasping in hospices. Overcome by opportunistic infections. Except that we’re talking about some US Olympic athletes. HIV sits dormant in our cells, a lurking lentivirus, poised to jump out and attack us after about a decade or so. No, no, we’ve dumped that theory; the new one is that it replicates from the start, incredibly rapidly but being neutralised by antibodies, although not quite efficiently enough to avoid being eventually overrun. Actually that model is now in the toilet too, so we ‘AIDS experts’ are not too sure what to say anymore, but who wants to get bogged down by details when we’ve got to get out there and save lives? I mean let’s keep our eye on the bigger picture. Because like hey, people are dying.

Nevirapine was first mooted as a potential new hi-tech anti-HIV agent at the start of the ‘90’s. In Impure Science: AIDS, Activism and the Politics of Knowledge, sociology Professor Steven Epstein tells us that, “The second generation of antiviral AIDS drugs – the non-nucleoside reverse transcriptase inhibitors that had looked so promising in vitro – performed poorly in clinical trials.” This news was “nothing short of shattering” to guys like Theo Smart of ACT UP New York, sister to our own Treatment Action Campaign – massive expectations for it having been pumped up by manufacturer Boehringer Ingelheim and its surrogates in AIDS activist organisations across the country. Former history professor Elinor Burkett picks up the trail in The Gravest Show on Earth: America in the Age of AIDS: In February 1993, a first year medical student at Harvard, Yung-Kang Chow, tooled around with the flop drug, mixing it with the older ones, AZT and ddI. The test tube action he claimed to have seen was packaged in a twelve-page press release by the Harvard Medical School as the next thing: a likely cure for AIDS. The New York Times, the television stations, and everybody else fell for it. Never mind that Chow was light on proof. The nevirapine combo idea got its next big boost from a splash in Nature later that month. With all the hype and hullabaloo, a large-scale clinical trial of this novel drug combination approach was set up in the US over sixteen centres. But after the 7^th International AIDS Conference in Berlin was over and English and American researchers got back to look at Chow’s magical findings, and did a few experiments of their own, they couldn’t replicate his results. So they started complaining. Chow’s supervisor, top ‘AIDS expert’ Martin Hirsch, took another look at Chow’s original research. It was a cock-up, he announced. Chow had made a fundamental blunder vitiating his conclusions. Dreadfully sorry, everyone. All the newspapers that had written about nevirapine with such excitement in February and March now rained tomatoes on it. Along with Nature,which published a disavowal in July.

Now you might imagine that a formal concession that the root study on the basis of which the drug went to clinical trials was invalid would be cause to call off the trials. Because mirth had taken the place of the model for the novel treatment approach. After all, humans were being treated with a very poisonous chemical (we’ll see below) for which there was no in vitro warrant, no evidence of any efficacy – no matter how ingeniously this smart drug had been engineered. According to a design brief like all good engineering projects. Assumed sound. One of its parameters being that reverse transcriptase is that distinct stuff mentioned earlier – and part of the tiny foe perpetrating the crimes, all likewise described. So it would be a big mistake to think we’re just going to dump it after all that time and trouble. After all that money invested in producing it, and all those high hopes of making so much more. No way.

On completion of the clinical trials, this is what the investigators reported: (i) Combined with AZT and ddI in patients who’d taken antiretrovirals before, “nevirapine produced a sustained improvement in CD4 count when compared with ZDV [AZT] plus ddI.” (The fact that CD4 cell counting as a surrogate marker for clinical health had been discredited two years earlier in the biggest best longest AIDS drug trial yet conducted, the Concorde trial in England, Ireland and France, didn’t dampen the party.) (ii) The CD4 cell count boost was most pronounced among folk who’d been on AIDS drugs previously, with cell counts of between 50 and 200 mm³. (iii) In the case of antiretroviral drug-naïve patients (first-timers) with CD4 cell counts “between 200 and 600 cells/mm³, nevirapine plus AZT and ddI resulted in a 140-cell absolute change from baseline at 52 weeks, compared with a 26-cell increase with ZDV/ddI, and a 2-cell decrease with ZDV/nevirapine.” Looking at the last figure, couldn’t they see the futility of it all? You take AZT and nevirapine together and your cell count actually goes down by the end of the trial. We note that the trial overseers reported no effect on ‘viral load’, that is the measure, according to ‘AIDS experts’, of HIV infection levels, and their reduction by a given drug. (Subsequent investigations turned up reductions in ‘viral load’, but transient only, returning to baseline levels within weeks.)

On the basis of this crap, Boeringer Ingelheim duly made a pitch for FDA approval. Obviously nevirapine wouldn’t have made it out of the starting blocks in any ordinary drug evaluation process.  But this was no ordinary procedure. It was a quickie.

On 21 June 1996 after a fast-track review that lasted just 119 days, nevirapine got its ticket. But it was a qualified one. It was not to be used on its own. That’s because even on the worthless measure of efficacy used by ‘AIDS experts’ (CD4 cell counting) it was ineffective solo. Get that? It doesn’t work as an ‘anti-HIV’ drug on its own. Moreover, there was no evidence to show that the drug afforded any clinical benefits: made you feel better, and improved your health. In terms of the Accelerated Approval Regulations, Boehringer Ingelheim was required to go home, do some more studies on humans, and come back to describe and verify the clinical benefit that it proposed the drug might have. Cool new system. For drug companies. These times being pro-business ones.

A press release on the same day stated, “studies also showed that the virus rapidly becomes resistant when nevirapine is used alone.” Quite how, no one has ever ventured to propose. Let alone suggest more frankly that on a plainer interpretation of the data, the drug is simply ineffective. It’s something like saying: ‘We lost Vietnam because all those frigging gooks became resistant to napalm and saturation bombing and the systematic selective assassination of their intellectuals by our special operatives.’

“Therefore, nevirapine is only recommended for use in combination with at least one other antiretroviral agent” in the nucleoside analogue class: AZT, ddC, ddI and d4T. One that the patient hasn’t taken before. To modulate the laboratory marker – CD4 cell counts – a bit better than AZT and like drugs on their own. Not make the patient feel any better. The fact that CD4 cell counts rise for a while in reaction to exposure to metabolic poisons like AZT even among HIV-negative people (AIDS 1996; 10:1444-1445) evidently passed the FDA by. As did the fact that AZT and nevirapine combined was no good according to the cell-count data.

What’s more, the effects of nevirapine on “surrogate endpoints” were only noticeable among patients “with HIV infection who have experienced clinical and/or immunological deterioration.” Not for people appearing well, and whose lab test results were considered normal, despite having a virus ravaging their immune systems, according to their doctors.

That nevirapine is extremely poisonous was admitted on the drug’s label – advising discontinuation “in patients who develop a severe rash or a rash accompanied by fever, blistering, oral lesions, conjunctivitis, swelling, muscle or joint aches, or general malaise.” And make no mistake, when the manufacturer talks of rash, it isn’t referring to a brush with stinging nettles. It means a generalised symptom of drug intoxication so severe in some cases that it shows up with thick layers of your skin dying off and peeling in great chunks. An ad for the drug – featuring the endorsement of “the dosing convenience of VIRAMUNE” by transatlantic sailor Mike Schmidt – “WHEN THINGS GOT ROUGH VIRAMUNE DIDN’T GET IN MY WAY” – explains: “Severe and life-threatening skin reactions have occurred in patients treated with VIRAMUNE, including Stevens-Johnson syndrome and toxic epidermal necrolysis. Fatal cases of toxic epidermal necrolysis have been reported.” The ad also warned of “severe…liver toxicity, including fatal cases”, apart from the bother of “fever, nausea, headache, and abnormal liver function tests.”

It stands to reason that a drug with this kind of appalling toxicity profile – even worse than AZT – must have some clinical benefit, shown, if not by the stage it was licensed under the business-friendly new approval regime, surely by the time it was offered to the public. But au contraire, as the ad spelt out: “…indicated for use in combination with other antiretroviral agents for the treatment of HIV-1 infection. This indication is based on analyses of changes in surrogate endpoints. At present, there are no results from controlled clinical trials evaluating the effect of VIRAMUNE in combination with other antiretroviral agents on the clinical progression of HIV-1 infection, such as the incidence of opportunistic infections or survival.” Can you credit this? That a drug so toxic without any proven health benefits should even be on the market? But this is the land of the free. And after AZT, anything.

A sucker for every new offering from the drug industry, Project Inform in the US jumped on it. It promptly released an “ACTION ALERT [to] URGE CALIFORNIA STATE OFFICE OF AIDS TO APPROVE THE INCLUSION OF NEVIRAPINE INTO AIDS DRUG ASSISTANCE PROGRAM FORMULARY. …Please urge the California Department of Health Services to move quickly to make this promising treatment available! Action Needed: Write or fax Kim Belshé, Director of the California Department of Health Services. Urge her to add Nevirapine to the California ADAP formulary immediately. Stress that including this therapy is expected to be cost neutral or a cost saving for the state. You can use the following information from Project Inform’s statement on Nevirapine to help craft your message [‘etc’].”

Project Inform went on: “Nevirapine may be useful in preventing mother-to-child transmission of HIV.” About which, the sailor’s ad had some interesting things to sayunder Pregnancy and Nursing Mothers. For pregnancy, the drug is “Category C”, in view of the absence of  “adequate and well-controlled studies in pregnant women” so the drug “should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus.” A risk indicated by rodent studies finding “a significant decrease in fetal body weight occurred at doses providing systemic exposure approximately 50% higher…than that seen at the recommended human clinical dose.” As far as nursing mothers go, “a single oral dose” given about six hours before delivery “readily crosses the placenta and is found in breast milk” so “mothers should discontinue nursing if they are receiving VIRAMUNE.” So as not to expose the baby to more than it got in the womb. Because it’s so poisonous. We’ll return to nevirapine to prevent mother to child transmission in Part Four. You won’t know whether to laugh or cry.

It is popularly believed by AIDS activists and their friends in journalism that at recommended doses, nevirapine is safer and less toxic than AZT. In fact the contrary is the case. Boehringer Ingelheim’s glossy 86-page Viramune Product Monograph version 3.0 states: “Nevirapine is generally very well tolerated.” After letting us know that people have died of its toxicities. And revealing that in one major study cited, “The overall incidence of nevirapine related serious adverse events was 4.3% for patients who received combination therapy.” But in an analysis posted on aidsmyth.com, Fintan Dunne in Ireland draws our attention to different numbers in the company’s Nevirapine data sheet posted on the New Zealand Medicines Safety Authority website: “The major clinical toxicity of VIRAMUNE is rash…occurring in 16% of patients in combination regimens in Phase II/III controlled studies. … 35% of patients treated with VIRAMUNE experienced rash…severe or life-threatening…in 6.6 % of [cases] … Overall 7% of patients discontinued VIRAMUNE due to rash. Rashes are usually mild to moderate, maculopapular [pimply blemish] erythematous [red] cutaneous [skin] eruptions with/without pruritus [itching], on trunk, face and extremities. Severe and life-threatening skin reactions have occurred in patients treated with VIRAMUNE, including SJS and TEN (toxic epidermal necrosis). Fatal cases of SJS, TEN and hypersensitivity reactions have been reported. … Severe and life-threatening hepatoxicity, including fatal fulminant hepatitis, has occurred. …” Dunne rightly deplores the false description of nevirapine toxicity manifest on the skin as rash, because it is “not cutaneous” nor “local” as the word suggests, but is “systemically driven…signalling a very serious illness.”

But Boehringer Ingelheim’s claim to the New Zealand government that ‘rash’ occurs in 35 per cent of cases isn’t true either. It’s an average figure from all the trials combined. Dunne points out that in the clinical trial coded B1 1046: “In treatment-naïve patients, the overall incidence of adverse side-effects is doubled. Serious gastrointestinal side-effects appear.” Concerning the latter he plausibly contends that these are the result of toxic tissue manifestations similar to external ones, having quoted a pair of clinical experts defining SJS/TEN’s “definitive characteristics” as being “massive epidermal sloughing at the dermo-epidermal junction… Gastrointestinal involvement may occur because of mucosal sloughing of the mouth, esophagus, stomach, and rectum [ranging in effect] from anorexia to development of a necrotic [dead] bowel.” And finally, he points out the finding in B1 1046 that: “Rash affects 50% of subjects – not 35%. Nevirapine and AZT used together produce side-effects at a rate that diverges strongly [upward] from the average of all trials.”

On 5 January 2001 the US Centers for Disease Control’s Morbidity and Mortality Weekly Report published a report by MedWatch, a voluntary drug toxicity reporting system set up the FDA, entitled Serious Adverse Events Attributed to Nevirapine Regimens for Postexposure Prophylaxis After HIV Exposures – Worldwide, 1997-2000, reporting severe toxicity after an average of just two weeks of treatment given to healthy medical workers. The New York Times summed up on the same day: “Federal health officials advised doctors yesterday not to prescribe a standard H.I.V. prevention drug to healthy health care workers stuck by needles. The drug, nevirapine, can produce liver damage severe enough to require liver transplants, and has caused death in such use, the Centers for Disease Control and Prevention said in its weekly report. But nevirapine should still be used for two other groups, the centers said. One is in treating people infected with H.I.V., the AIDS virus. The second is to prevent transmission of H.I.V. from mothers to their infants during childbirth. … The agency said it and the federal Food and Drug Administration had identified 22 cases of severe liver, skin and muscle damage related to nevirapine taken after possible exposure to H.I.V. from March 1997 through September 2000.” Too poisonous for doctors. But great for the rest of us. Excuding rape victims: “The agency does not recommend nevirapine to prevent H.I.V. infection among people recently exposed to the virus through unsafe sex, [the CDC’s Dr Julie] Gerberding said.” But including pregnant women. That’s because the “benefit outweighs the risk, largely because the mother and child take only one pill,“ said the CDC’s Dr Helen Gayle. “No serious adverse effects” had been noted in pregnancy studies. Gayle was wrong about that, as we’ll discover in Part Four. The UN AIDS programme accordingly supported the CDC’s recommendation that the drug be given to pregnant women and was “working with Boehringer Ingelheim to start programs in developing countries.” Just in case you never knew that when it comes to pharmaceutical drugs, we’re all one big happy family.

After securing the US market by getting it past the FDA, the rest was a breeze. As it had been with AZT. Boehringer Ingelheim turned next to the European Agency for the Evaluation of Medicinal Products, claiming in its submission: “The pharmacology safety programme addressed among others, effects on central and autonomous nervous system, cardiovascular, and renal respiratory systems and did not reveal any severe side effects at relevant dose levels and/or concentrations.” These “toxicokinetic data” were derived from studies on “both rats and dogs”, on the basis of which the company proposed that “there seemed to be acceptable safety margins” – notwithstanding that clinical trials with humans as test subjects had already revealed a shocking toxicity profile. The Safety discussion reads like an excerpt from Orwell’s 1984. Suffice it to say for present purposes that it was obvious even then that nevirapine was exceedingly poisonous. As revealed by its effect on the liver, and ‘rashes’ of such severity that some patients died. Both of which toxic manifestations had been predictable from the reactions of test animals, discussed earlier in the dossier. But don’t hassle. Boehringer Ingelheim smoothed things over with this public-spirited precaution: “Therefore having reassessed the risk/benefit profile of nevirapine, the warnings concerning the occurrence of these events have been reinforced. In addition, new recommendations for the liver and cutaneous monitoring of patients…have been introduced through an urgent procedure.”

The EMEA approved the drug on 5 February 1998.On the advice of an expert panel however, the EMEA recommended that nevirapine be categorised in its register of approved drugs for prescription “under exceptional circumstances” only. Being a very dangerous drug. And the EMEA required commitments from Boehringer Ingelheim to conduct further clinical studies, on the basis of which it would make a final risk/benefit evaluation.

The trouble is that new bits of fine print on the paper in the box didn’t make any difference. Inasmuch as strychnine is strychnine. Reports continued to flow in of people suffering serious skin and liver damage, some fatally. On 12 April 2000 the EMEA got nervous, and issued from London an urgent EMEA PUBLIC STATEMENT ON VIRAMUNE (nevirapine): SEVERE AND LIFE-THREATENING CUTANEOUS AND HEPATIC REACTIONS, which it “thought…necessary to provide…to the public.” The “prescribing and patient information” contained in the package insert was to be substantially revised and amplified further, essentially boiling down to: ‘If you get very sick very quick as other people taking this stuff have done, with your liver packing in, and your skin starting to rot, especially mucosal surfaces like eyes, mouths and the lower orifices, then chat to your quack. Even think about ditching it.’

In November 2000, the FDA in the US caught up with the Europeans. It issued an alert of its own concerning nevirapine liver toxicity, after a study found that it caused a third of patients to quit taking the drug – as against the picture painted in Boehringer Ingelheim’s glossy Viramune Product Monograph 3.0 dished out at the 13^th International AIDS Conference in Durban a couple of months earlier: “Nevirapine related hepatitis was reported infrequently in clinical trials. The incidence…was 1.0%…in BI clinical trials (Pollard et al. 1998) … In the four controlled trials of combination therapy…treatment related hepatitis was observed in…1.1%… The occurrence of hepatitis and other liver related events with double and triple therapy regimens including nevirapine from 13 recently completed and currently ongoing clinical trials has also been reported (Pollard et al. 1998). In these 13 studies, the overall incidence of hepatitis possibly related to nevirapine (0.5%) was similar to those in the earlier studies discussed above. … Fatal cases of hepatitis have been reported.”

 Is this the same drug we’re talking about?