The history of AIDS drugs and reasonable pricing

The Biden administration is moving to correct a three-decade-old policy that has cost taxpayers and patients billions of dollars through exorbitant drug pricing

I am devoting my last post before the Thanksgiving holiday to a historical footnote. It involves:

• The central role government plays in developing new drugs;

• The forced disconnect of the government’s role from drug pricing; and

• The Biden administration’s admirable effort to correct a flawed policy that dates from the AIDS crisis of the 1980s and early 1990s.

I am motivated by a recent exchange of letters (subscription required) in the December 7, 2023 New York Review of Books between myself and Harvard historian Daniel J. Kevles. I questioned one of the claims in his critical review of a new book by journalist Alexander Zaitchik: Owning the Sun: A People’s History of Monopoly Medicine from Aspirin to Covid-19 Vaccines. The exchange revolved around his assertion that a private company, Burroughs Wellcome, was the first to identify the first anti-AIDS drug (azidothymidine or AZT) in the mid-1980s.

Biden’s bombshell

But let’s start with the news. In early September, the Health and Human Services Department signed a $326 million contract with Regeneron Pharmaceuticals to develop new COVID-19 drugs. The government hopes the next generation of COVID drugs, which use monoclonal antibodies, will work better than Pfizer’s Paxlovid, which has a long list of side effects and has failed to protect many patients from rebounding disease. The agreement included a clause that requires Regeneron to sell the new products, should they win Food and Drug Administration approval, at a price that is equal to or less than its retail price in other high-income countries.

This is the second time in recent years that the government has placed price controls on potential pharmaceuticals that it had a hand in developing. The collaborative agreement between the U.S. and Novavax to develop a COVID-19 vaccine also contained a clause requiring the company sell its product at a price at or below the price in other advanced economies.

There is no disputing the fact that government-funded research was crucial to the rapid development of COVID vaccines. Several federal agencies spent $337 million before the pandemic on the development of the mRNA technology behind the most effective vaccines. Once the vaccine was invented, the feds spent $2.2 billion on clinical trials. And, it offered $29.2 billion in guaranteed purchase contracts after it was approved by the FDA, according to a recent study.

No price controls on the first AIDS drugs

These two COVID-era pricing clauses stand in stark contrast to the policy established in 1995 at the height of the AIDS crisis.  Fearing industry would refuse to collaborate with the government in developing new drugs to combat AIDS, Harold Varmus, then head of the National Institutes of Health, rejected using government “march in” rights under the 1980 Bayh-Dole Act, which enabled the government to pursue fair pricing for drugs that relied on government-funded research during their development.

At that juncture, at least three drug companies were racing to develop protease inhibitors, which would become the final piece of a three-drug regimen that turned AIDS from a death sentence into a manageable disease. At least one of the companies had relied heavily on government funding, and all of them were dependent on NIH-funded pre-clinical research.

Varmus did not want a repeat of the greeting angry AIDS activists gave AZT, which, though only marginally useful on its own, had come to market in 1987 at the then unprecedented price of $10,000 a year. The activists argued government scientists had discovered AZT, and demanded its price reflect that contribution. (This history is recounted in chapters 4-6 in my 2004 book, The $800 Million Pill.)

Still in print!

The controversy over who actually discovered AZT would rage for the next seven years, eventually forcing Burroughs Wellcome (the remnants of BW are now owned by GlaxoSmithKline) to slash its price to more manageable levels. But the company won in patent court in 1994 after it sued generic manufacturer Barr Pharmaceuticals (now owned by Teva) for patent infringement. Barr claimed it had the right to manufacture AZT as a generic because BW failed to name two scientists at the National Cancer Institute as co-inventors, arguing they were the ones who first identified the drug as active against HIV. The court rejected that claim. The generic was delayed for another decade.

Was the patent court’s judgment dispositive?

In his review of Owning the Sun, Kevles wrote “Zaitchik misguidedly presents … NCI had ‘invented and rediscovered the drug,’ and so could have claimed ownership and licensed it for production and sale at a much lower price. But … it was not the NCI that recognized AZT as a likely agent against HIV but Burroughs Wellcome, whose scientists had observed that the drug adversely affected mouse viruses similar to HIV. At the company’s request, the NCI tested AZT and was the first to find that it displayed effectiveness against live HIV.”

Like the AIDS activists in the late 1980s, I took issue with that characterization of the events in late 1984 and early 1985 when AZT was first identified as active against HIV. In my letter to the NYRB, I drew on my own research in the NIH archives to show that Samuel Broder, then head of NCI, had asked numerous companies for drug candidates that could inhibit HIV. NIH-funded scientists had already discovered the retrovirus and characterized its genome (Robert Gallo), and developed the world’s only assay for testing potential drug candidates (Hideki Mitsuya).

Broder wanted to cast as wide a net as possible to look for something that might inhibit reproduction of this new and dangerous pathogen. He asked numerous drug companies to submit candidates that might be active against the retrovirus. BW was just one of several companies that responded, sending in multiple candidates, including AZT. Only NCI could determine which ones, if any, were active against HIV. “The NCI mass screening project identified AZT, a failed cancer drug that had been developed in 1964 by an NIH-funded scientist,” I wrote in my letter to the NYRB.

Kevles took issue with that chronology of events by referencing the patent litigation, which began six years later. “Burroughs Wellcome began testing various compounds for activity against mouse viruses similar to HIV and by early December (1984) had observed that AZT prevented the mouse viruses from replicating,” he wrote in his rejoinder to my letter. “The company was sufficiently confident of AZT’s likely efficacy against HIV that by February 6, 1985, it had completed the draft of an application for a patent on AZT that included a range of dosages for the treatment of HIV infection.” That was two days after it sent a sample of AZT, plus four dozen other potential drugs to NCI.

BW filed its first patent application for AZT as an anti-AIDS drug in Great Britain in late March, about a month after receiving the NCI results that AZT was the most promising candidate. It wouldn’t file in the U.S. until September. Its patent applications did not list either Broder or Mitsuya as co-inventors. That exclusion became the basis for Barr’s claim that the patent was invalid. After it applied to the FDA to begin making a generic version, BW filed suit claiming infringement of its intellectual property.

What did Burroughs Wellcome really know?

Did the fact presented in the patent litigation corroborate Kevles’ claim that BW identified AZT? The litigation’s outcome hinged on the arcane matter of conception. When did BW conceive that AZT would be effective against HIV? The company’s draft patent application, written two days after sending the candidate to NCI, proved it came before getting the test results, the three-judge Appellate Court panel ruled.

Yet the judges admit both at the top and near the end of their decision that:

“Broder and Mitsuya received from Burroughs Wellcome a group of compounds, known to Broder and Mitsuya only by code names, selected for testing by the Burroughs Wellcome inventors. They then tested those compounds for activity against HIV in their patented cell line. The test results revealed for the first time that one of the compounds, later revealed to be AZT, was exceptionally active against the virus.”

My reading of the Appellate Court’s ruling leaves open the possibility that BW’s decision to begin drafting patent applications was merely a prudent business decision. In those days, U.S. patent law granted patents based on first to invent, not first to file (it was changed to the latter in 2013). Getting something down on paper for AZT would have established the paper trail needed to prove it had already conceived of using that particular chemical against HIV/AIDS.

But what would have happened if a different drug candidate had popped up as superior in the NCI screen? As cancer researchers are well aware, drugs that cure cancer in mice rarely work in humans, and some that don’t work in mice do work in humans.

I don’t have access to the trial record, and my calls to the former CEO at Barr and its legal team (it was represented by Dan Webb, the former U.S. attorney who prosecuted John Poindexter in the Iran-Contra scandal and is now co-director of Winston and Strawn) have not been returned. I would like to know if they produced evidence at trial that showed BW had similar draft applications for any of the other drug candidates it sent along to NCI, which, as I noted earlier, totaled four dozen, according to this contemporaneous account in Nature magazine.

As I wrote above, this could be portrayed as merely a historical footnote. But it has relevance in an era when pharmaceutical innovation is more dependent than ever on publicly-funded research and the price of drugs has become a major barrier to access, not to mention an unnecessary drain on public and private purses. At the least, publicly-supported agencies that fund research should stop making the same mistake that Broder, who later went to work for a pharmaceutical company, and NIH made in the 1980s and 1990s.

When the government fails to insist that government-funded inventors be included on patent applications by the companies that eventually bring new drugs to market, it sacrifices a valuable tool for negotiating reasonable pricing.

Have a happy Thanksgiving weekend everyone!