Across the United States there are a handful of anonymous, cavernous warehouses. They exist partly because we don’t want a repeat of 1918. Packed inside are the weapons we’d need to fight another global pandemic. The bunkers contain our Strategic National Stockpile of emergency medicines, and they symbolize both the strides and the limits of our medical prowess and logistical readiness.
Apparently, each warehouse is the size of a couple of Walmarts joined together. The stockpile was created by the CDC in 1999 and now has an inventory of medicines and emergency materials that cost more than $7 billion. The stockpile, which is maintained by the CDC, contains supplies for a public health scenario that overwhelms state and local government. It has antibiotics and vaccines and antiviral medications, surgical equipment and life support machines, and protective suits like the kind we saw during the Ebola outbreak. Its inventory is massive, and the details are classified. Originally, the goal of the stockpile was to provide medical supplies in the event of a chemical or nuclear attack, but recently the inventory has expanded to an “all hazards approach” that includes hurricanes and earthquakes. And, of course, a flu pandemic.
Maintaining the stockpile is an enormous undertaking. Medicines, just like cartons of milk, have expiration dates. While your local supermarket makes sure to bring its soon-to-expire products to the front of the shelves, medicines in the stockpile just sit around waiting for a massive crisis. Once they expire, they must be replaced. Keeping the whole thing going costs at least $500 million per year—which you might consider a small price to pay for peace of mind.
The Supply Service Center, a warehouse in Perry Point, Maryland, doesn’t have a name as grandiose as Strategic National Stockpile, but it plays an important part in the year-round federal response to influenza. Perry Point looks pretty, even in the late fall as the grass turns brown and the trees become branchy skeletons. The center sits on a peninsula minutes from the Delaware border, where the Susquehanna funnels into the Chesapeake Bay, a couple of hours’ drive north of Washington, D.C. Perry Point is like a quaint little town all to itself, with a baseball diamond, a community center, and several cul-de-sacs, all bordered by forest and fresh water. The veterans’ hospital that shares the peninsula has its own police and fire department, and even its own post office. Mark Burchess is the deputy director of Supply Chain Management Services, which is part of the Department of Health and Human Services.
“A cross between Amazon and a local pharmacy,” Burchess calls his workplace, which is a warehouse that serves two purposes. The first is to distribute medical supplies to those who work in the federal government. Burchess has everything, or so it seems: vaccines for an embassy, disposable gloves and masks for first responders, gas sensor “sniffers” to detect biological threats, and blankets for hurricane victims. If he doesn’t have it in stock, he knows where it can be purchased. The second purpose of the warehouse is to maintain “stockpiles of national significance” like those that would be needed during a pandemic flu.
Clients from across the federal government store supplies in the warehouse the way you might store stuff in your basement. There are boxes for the Department of Defense and pallets for the Department of Homeland Security, as well as material stored by BARDA, the Biomedical Advanced Research and Development Authority. BARDA is a government office that develops and purchases drugs and equipment that would be needed in a public health emergency.
Burchess was excited that the center, working with BARDA, was about to double the size of its refrigeration storage unit. He couldn’t say what it would hold, other than “stuff” that was vital to an influenza response. Precautions were plenty. The refrigeration unit came with two standard compressors, a third compressor for backup, and enough spare parts on-site to build a fourth compressor, should the need arise. The center also had enough gasoline on-site to run the electrical generators for more than a week. The warehouse would soon acquire medications for multidrug-resistant tuberculosis that it would store on behalf of the CDC. This place was one huge doomsday prepper campus.
On the soaring shelves that seem to stretch on forever are medicines, antidotes, and colossal boxes containing vials of flu vaccine. Everything here seems to come in wholesale sizes, as if we were at Costco. Vaccines are stored in handy five-gallon drums, each containing thousands of individual doses. The strain is printed clearly on the side of each drum. Many names would be familiar to most physicians, but others are mysterious or classified.
From Perry Point, the center sends supplies across the world every day, from autoclaves that sterilize instruments to massive electrical generators. Although owned and operated by the U.S. government, it does not receive government funding. Instead, it works as a business, making money on the items that it ships. If the CDC needs ten thousand syringes for a vaccine trial in Africa, it turns to Burchess, who gives a price that includes a markup to run the business. Like other businesses, he has competition. Several other government agencies, like NASA and the Department of the Interior, have similar supply centers, so prices remain competitive.
Every year the Supply Service Center ships out the influenza vaccine, as well as the paraphernalia needed to run a successful vaccination program: needles and syringes and gloves and inoculation cards. Burchess is especially proud of the vaccines he shipped to Customs and Border Protection during the 2009 H1N1 influenza outbreak, helping protect those who protect us. During that outbreak, almost 120 million doses of the H1N1 vaccine were shipped across the United States. The center not only distributes influenza vaccines but, rather surprisingly, collects them too. Many of those 120 million doses were given, but millions were left unused in hospitals, doctors’ offices, and warehouses around the country. Each state dealt with the unused doses differently, and some declared them medical waste. What was delivered by FedEx one day might need to be picked up by a team in white protective suits the next, and the center provided a reverse logistics operation to bring many of the doses back to a central location for destruction.
In 2009 Perry Point also held most of the U.S. stock of the anti-influenza drug peramivir, which at the time did not have approval from the FDA. As a result, it was being provided under an emergency-use authorization. The team at Perry Point made sure that it was delivered to patients within twenty-four hours of the CDC’s request. Peramivir, sold under the trade name Rapivab, is one of three medications that belongs to a class of drugs called neuraminidase inhibitors. The other two are zanamivir, whose brand name is Relenza, and oseltamivir, branded as Tamiflu.
“Neuraminidase” is a clunky word, but what this viral enzyme does is quite elegant. Once the flu virus gets into your cells through the cell membrane, it takes advantage of the cozy conditions to replicate. Then the baby flu virus particles have to escape the cell. They make their way to the surface and then push through the cell membrane. At first they remain tethered to the cell membrane like dinghies attached by ropes to their mother ship. Neuraminidase is the enzyme that acts like a knife. It allows the baby flu viruses to cut the rope and make their own way out and into the world. Without neuraminidase, baby flu particles could not spread the infection and replicate. The neuraminidase inhibitor drugs prevent viral neuraminidase from working. No knives, no free-ranging influenza viruses.
Neuraminidase inhibitors were first discovered in the 1960s by a group of researchers from Scotland, but it wasn’t until thirty years later that scientists began to test them. The drug was first manufactured as a powder that needed to be inhaled, and after a few clever tweaks an oral formula was developed. That first drug was oseltamivir, marketed as Tamiflu. It was promoted as a silver bullet against influenza and a powerful weapon in our strategic stockpile. Rapivab is only available intravenously, so it is prescribed far less often than the other two drugs, and only to very sick patients. Relenza, the third drug, needs to be inhaled, and it captured much less of the market than Tamiflu, the best-known neuraminidase inhibitor.
In 2014 a TV ad for Tamiflu ran more than eleven thousand times. “Prescription Tamiflu attacks the flu virus at its source,” said the voice-over. “Sometimes, what we suffer from is bigger than we think. . . . The flu is a really big deal, so why treat it like it’s a little cold? Treat it with Tamiflu.”
Another ad targeted mothers whose children had the flu. “The flu virus. It’s a really big deal,” a narrator said as a worried mom watched over her coughing daughter. “Mom knows it needs a big solution: an antiviral.”
But given the way the medicine actually performs in patients, it turns out that it is not much of a solution at all. If you read the small print accompanying the ad carefully, it claims that on average, Tamiflu shortens a child’s flu symptoms by about a day—but only if taken within the first forty-eight hours after the onset of symptoms. After that, it’s even less effective. The helpful narration—mom and her healthy daughter are now shopping for fresh fruit—tells us that children and adolescents taking the drug “may be at an increased risk of seizures, confusion, or abnormal behavior,” but you shouldn’t worry because “the most common side effects are mild to moderate nausea and vomiting.” The drug package itself notes that vomiting occurs twice as often in children taking Tamiflu compared to those not treated. In fact, Tamiflu often gives you some of the very symptoms you are trying to relieve, and at best will shorten your misery from influenza by a day.
But Tamiflu is part of the Strategic National Stockpile. There are clearly experts who are convinced of its benefits. The fascinating story of Tamiflu played out over several years and across multiple continents, and it shows how decisions made more than two decades ago still influence our approach to the treatment of influenza today.
* * *
The story begins with the outbreak of H5N1 avian flu in Hong Kong in 1997. The virus infected only eighteen people, but six of them died. This got the attention of the World Health Organization. Two years after the outbreak the WHO published a report that mentioned that “two closely related compounds have been developed that bind to the active site in a minor protein found on the surface of influenza viruses, the enzyme neuraminidase.” Trials of these compounds were under way, and if they were approved for use, they might be useful in the treatment of influenza, regardless of the specific strain. The WHO report disclosed that it had been written “in collaboration with the European Scientific Working Group on Influenza (ESWI).” But it turned out that the ESWI was funded by at least seven pharmaceutical companies that stood to profit from an outbreak, or at least from the fear of one. These manufacturers joined together to promote “a favorable climate” in Europe for vaccines and related research. Several years after the report came out, it was revealed that one of its authors was a paid consultant for one of the drug manufacturers. Both the ESWI and the WHO were supposed to provide objective scientific advice, but there were clear conflicts of interest. The push toward stockpiling Tamiflu began in earnest on the basis of a specious and probably biased recommendation.
Around the time of the 1997 avian influenza outbreak, neuraminidase inhibitors were in the early phase of clinical trials, but there was still enough data to make some recommendations regarding their use. In 1999 the Cochrane Collaboration, an international group of thirty-seven thousand medical contributors, issued the first of three reports about neuraminidase inhibitors. When it comes to how well drugs work, Cochrane reviewers pledge to issue reports that are “free from commercial sponsorship and other conflicts of interest.” They don’t just take the manufacturer’s word. They look for any published and unpublished papers, and all the trials that were reported or not reported. Using all this data they make recommendations about everything from vitamin E during pregnancy (it doesn’t work to prevent early labor) to yoga for back pain (seems to be better than no regular exercise). I’ve used their work to help me make decisions about which drugs to recommend to my patients. Cochrane found that as a treatment the neuraminidase inhibitors only shortened the duration of flu symptoms by about one day, although they were somewhat more effective when used to prevent influenza.
In late 1999 the FDA approved Tamiflu as a treatment for influenza, but within a few months it sent the manufacturer, Hoffman–La Roche, a warning letter about its advertising campaign. The FDA found that the Roche campaign lacked balance, contained misleading information about how the drug worked, and overstated claims about the drug’s effectiveness. Roche had claimed that Tamiflu “reduces the duration of flu so you can feel better faster,” but this blurred and exaggerated the evidence from clinical trials. Despite these concerns, the FDA had approved the use of Tamiflu to prevent influenza, and to treat influenza in children under four years of age. In 2002 the European Union also approved the drug, in time for another outbreak of avian flu in Asia two years later.
In November 2005 the avian influenza scares prompted President George W. Bush to make the short trip from the White House to the National Institutes of Health in Bethesda, Maryland. His speech there was ominous. “At this moment,” he said, “there is no pandemic influenza in the United States or the world. But if history is our guide, there is reason to be concerned.”
Those were chilling words, and the president outlined how his administration planned to tackle those concerns. He described the small outbreaks of avian flu in 1997 and 2003. He told the audience he’d gotten his flu shot. And he urged vigilance in a manner similar to his speeches on terrorism.
“If the virus were to develop the capacity for sustained human-to-human transmission, it could spread quickly across the globe,” Bush said. “Our country has been given fair warning of this danger to our homeland—and time to prepare.”
The president presented a three-pronged approach: First there would be an effort to detect flu outbreaks earlier. Second, the government would stockpile vaccines and antiviral drugs, and request $1.2 billion from Congress to purchase enough avian influenza vaccine to immunize 20 million people. Third, the president asked the country to develop emergency pandemic plans in all fifty states and every local community. This three-part approach, said the president, would “give our citizens some peace of mind knowing that our nation is ready to act at the first sign of danger.” Overall, the president requested $7.1 billion to fund his strategy, although a year later Congress approved only half of that amount. Now avian influenza was on everyone’s mind, even though the chances of catching it in the United States were very low.
The 2004–2005 influenza season in the U.S. was not particularly bad, and yet there was a rapid rise in the sales of Tamiflu and other antiviral medications. In the fall of 2005 five times as much Tamiflu was being prescribed in the U.S. than had been the year before. This increase was even higher in those who had no chronic medical conditions, as well as in children. This suggests that healthy people were buying up Tamiflu, preparing for a potential outbreak. The increase had nothing to do with the amount of flu that was actually out there, since there was no more than usual. What was unusual was the presidential address on the matter, and the media’s coverage of and hand-wringing over avian influenza. In Canada there was an even greater hoarding of Tamiflu by worried citizens. Prescriptions for the drug there rose tenfold, and threats of a shortage prompted Roche Canada to restrict distribution of the drug.
In the midst of the avian flu scare, Britain’s minister of health made a somber assessment. “We must assume we will be unable to prevent it reaching the UK,” said Chief Medical Officer Sir Liam Donaldson. “When it does, its impact will be severe in the number of illnesses and the disruption to everyday life.”
The BBC, not known for its sensationalism, reported that unless millions of doses of Tamiflu were stockpiled, an outbreak of avian influenza in the United Kingdom could kill more than 50,000 people. And so the British government made plans to purchase and stockpile over 14 million doses of Tamiflu. In the U.S., Tamiflu was already part of the national stockpile, but the supply was more modest—only 2.3 million doses were on hand, though more was on order.
Barely two months after President Bush’s speech and the news that the U.S. would stockpile Tamiflu, the Cochrane group published another analysis of antiflu medications. The authors reviewed more than thirty clinical trials of the older antiflu drugs and nine trials of the newer Tamiflu. The older drugs were no longer effective, and Tamiflu, the new kid on the block, had some serious limitations. There was no role for the drug in patients who had an influenza-like illness that was not specifically caused by the influenza virus. More startling was the lack of evidence that it could even fight avian influenza, the very disease for which it had been stockpiled. Tamiflu didn’t reduce deaths from avian flu when it was used in Southeast Asia, and now the virus was showing resistance to the medication. Over time Tamiflu succeeded in one respect: it made some types of flu resistant to it. In Europe at least 14 percent of the circulating influenza viruses were resistant to Tamiflu by 2008. In short, the drug worked only to strengthen the virus that it claimed to combat.
The drug was next tested in 2009, when there was an outbreak of swine flu in the U.S. This was similar to the swine flu that jumped from pigs to humans in 1976. This flu, you may recall, was identified as an H1N1 type, a combination of swine flu viruses that usually infected pigs in America and Europe. By June 2009 there were more than 30,000 cases of swine flu in seventy-four countries. The World Health Organization declared a pandemic. The CDC hosted dramatic press conferences in Atlanta, but this H1N1 swine flu, though new, was not as virulent as experts feared. Swine flu viruses generally cause only mild disease, and were no more harmful than older, standard strains of influenza.
But the swine flu was getting a lot of media attention. Just as the nickname “Spanish flu” had made the 1918 epidemic seem exotic and alien in newspaper headlines, “swine flu” sounded threatening and feral, which contributed to another run on Tamiflu. In Boston, the law firm of Ropes & Gray went the extra mile for its 1,900 employees and their families. It gave them prescriptions for Tamiflu, no doctor’s visit required. The firm reminded its staff to take the medication only at the onset of flu symptoms, but it did not mention that the antiviral was unlikely to be of any significant help. In an editorial, the Boston Globe admonished Ropes & Gray for being part of the problem. “While there has been very little resistance to Tamiflu among swine flu patients so far,” the editorial said, “that could change as cases mount up and more physicians prescribe the drug.” The CDC issued a curt statement: “This is something we would not like to see widely practiced by employers.”
Leaving aside the question of whether Tamiflu even worked, there was another issue: fairness. News of the Ropes & Gray decision suggested that the medicine was not being distributed justly. A privileged minority—lawyers for global firms and those with the right connections—were at the front of the line. They had access to antiviral medications before they fell ill. The have-nots would have to wait and see. Dr. Karen Victor, an internist at a Boston hospital, said the main issue was access. “[The firm believes] Ropes & Gray’s employees’ appearance at work is so important,” she said, “that they will put that above fairness to society.”
Many countries reported that the flu virus was 100 percent resistant to Tamiflu. But by 2009, despite all the evidence that it worked poorly, Tamiflu and other neuraminidase inhibitors had become part of the national stockpile of the U.S., Britain, and at least ninety-four other countries around the world. The medication was a huge financial success. Governments ordered more than $3 billion worth of antivirals. Gilead, the pharmaceutical company that initially developed Tamiflu, reported royalties of over $52 million in the first quarter of 2009. And Roche, the Swiss giant that licensed the drug from Gilead, made even more: $590 million in sales in one quarter alone. And then, rather suddenly, academics struck another blow against Tamiflu.
Back in 2003 a group of researchers led by Dr. Laurent Kaiser in Geneva had looked for any papers that examined how Tamiflu performed, and then pooled the results. They’d found ten studies, though only two had been published. The other eight had been reported on to some degree, or not reported on at all, and were otherwise languishing in the drawers (or on the computers) of their authors. The review by Dr. Kaiser and her colleagues was funded by Roche, the manufacturer of Tamiflu, and it concluded that the medication reduced lung complications, antibiotic use, and hospitalizations in otherwise healthy patients. It was a very important paper for two reasons. First, it provided evidence that Tamiflu worked, right around the time when there was discussion about placing it in the national stockpile. And second, it was used as evidence in later reports by the prestigious Cochrane Collaboration.
But in July 2009 Dr. Keiji Hayashi, an astute Japanese pediatrician, contacted the Cochrane group with a concern. He wondered what was in those eight unpublished studies, and why their findings had not been shared. Without knowing the results of the unpublished Tamiflu studies, the Cochrane reviewers might find only studies that cast the drug in a good light. The Cochrane group had indeed relied upon the Kaiser paper, which itself had relied on those eight unpublished studies. They had erred in doing so. They acknowledged this error and spent the next several years trying to fix the mistake.
Why were eight of those ten trials about Tamiflu never published? Perhaps they were poorly conducted or too small in scope, and would therefore not be rigorous enough for a scientific journal. But knowing the details of a study is a critical part of the scientific process. For example, if the effects of Tamiflu had been studied only in young and otherwise healthy patients, we could not reach a conclusion about how the drug might work in those who were elderly, or those with other medical conditions. These details are critical in evaluating any clinical trial. Kaiser’s group reached its conclusions based on evidence that it alone had seen, and that no one else could properly corroborate. That’s bad form, to say the least.
It is also possible that the unpublished trials actually found that Tamiflu had no benefit. Medical journals are a business and, like drug manufacturers, they have a bottom line. They’re more likely to publish a positive and exciting study that will make headlines. Negative trials (in which it turned out that a new drug didn’t work as claimed) are rarely submitted for publication (that’s the fault of the researchers), and if they are, they’re often rejected by journal editors as not exciting enough to be published (that’s the fault of the editors). And negative trials usually end up at the bottom of an electronic filing cabinet. But negative drug studies are crucial if the goal is to properly judge the safety and effectiveness of a treatment. When there’s a bias inherent in the publication process, we can’t be sure that all studies, both positive and negative, have been aired publicly. If doctors don’t have access to all the data, they can’t be sure that they are making the best scientific regulations.
After the problem was raised by Dr. Hayashi, the Cochrane group, led by Dr. Tom Jefferson, got to work. Jefferson contacted Roche and asked the company to provide the missing data. At first, the company refused to release the data because, it said, another team was already performing a review of it. When Jefferson asked why that should prevent the Cochrane group from also reviewing the data, Roche agreed to release it—but only if Jefferson signed a confidentiality agreement. That would prohibit him not only from sharing the data but also from acknowledging that he had even signed the confidentiality agreement itself. Jefferson never signed, but Roche did agree to release some of its Tamiflu data. When Jefferson reviewed it, there were too many missing details to reach any conclusions. In December 2009 Jefferson released an updated review without including any of the unpublished data. They summarized that neuraminidase inhibitors, including Tamiflu, have “modest benefit—reduction of illness by about one day.” The review concluded that “they should not be used in routine control of seasonal influenza.” In addition, they questioned the effects of Tamiflu on lower respiratory tract infections. And because these drugs did not prevent infection or stop nasal viral excretion, “they may be a suboptimal means of interrupting viral spread in a pandemic.”
The Tamiflu controversy then broke from the ivory tower of academia and swept into the British Houses of Parliament. There, a member named Paul Flynn sponsored a motion that was polite in phrasing but excoriating in message. It used words that would make any drug manufacturer blanch: “surprised,” “uncertain,” “concerned” and, most pointedly, “fatal side-effects including heart attacks.” “It is unwise to continue with . . . the programme,” the motion concluded. Outside of Parliament, Flynn went further by suggesting on his blog that the leftover stockpiles of Tamiflu could actually be used for something practical: salting Britain’s snowy roads.
Flynn authored a report on behalf of the European Parliamentary Assembly, which found a general lack of transparency in the way the swine flu pandemic had been handled by the World Health Organization. These criticisms were echoed in the British Medical Journal, which published a series of articles questioning the industry ties of several experts at the WHO. Dr. Fiona Godlee, the editor in chief of the journal, noted that the WHO’s guidance on the use of antivirals in a pandemic was written by a flu expert who was getting paid by the manufacturer of Tamiflu. That hardly inspired confidence in its findings.
Over the years, industry and the medical community continued their tug-of-war. Roche refused to release its internal data to investigators. Then, feeling the heat, it commissioned an independent review of Tamiflu, in which neither the authors nor their institution received any funding for their work. Roche statisticians cooperated with the review, answering any data-related questions that arose. When it was published in 2011, the review showed that Tamiflu reduced some of the complications from influenza that would have required antibiotics. This might be useful in a pandemic and could, if verified, provide a limited use for the drug. In order to shake loose the truth, the British Medical Journal began to publish correspondence between the Cochrane group, Roche, the CDC, and the WHO. It was all part of the journal’s “open data” campaign. Sunlight was proving to be a better treatment than Tamiflu.
Within a few months, Roche released all the trials requested by the Cochrane group, as did GlaxoSmithKline, which produced another neuraminidase inhibitor, zanamivir. This allowed the Cochrane group to complete an updated review, which it released in April 2014. Finally Cochrane could produce an analysis based on all the published and unpublished clinical trials of Tamiflu (and its cousin zanamivir, marketed as Relenza). Cochrane found that, when taken as prevention, these drugs might reduce the risk of developing influenza; once a patient was sick, they reduced the duration of symptoms of the flu by less than one day. However, Tamiflu was most efficient at triggering side effects: nausea and vomiting, and sometimes psychiatric effects like hallucinations, anxiety, and even seizures. It could also damage your kidneys. The most damning finding was that Tamiflu did not reduce the risk of hospitalization, or of pneumonia, which was the very reason the United States and other countries had put Tamiflu in their strategic stockpiles. The Cochrane report, and several others that followed, blew a hole in the notion that Tamiflu should be stocked at all.
The debate over Tamiflu continues today, although perhaps with less intensity. New academic papers have been published, but they seem to have done little to change anyone’s mind. In January 2015 the Lancet released an analysis that included all the published and unpublished Roche-sponsored trials, and any other relevant trials reviewers could find. They found that Tamiflu decreased the risk of hospitalization and confirmed the already well-documented finding that Tamiflu shortened sickness by about a day. The study was not supported by Roche directly but by a foundation called the Multiparty Group for Advice on Science, which is (you guessed it) supported by Roche.
Despite all the controversy, and the corruption of the study pool by pharmaceutical interests, Tamiflu continues to be tested in clinical trials. As of 2017 there were at least eight open trials in the U.S. and Canada focused only on groups that are at high risk for the flu: the elderly, those who have underlying lung or heart conditions, or those whose immune systems are not working properly. For the rest of us, taking Tamiflu to prevent or treat influenza is a waste of time. Even the CDC implies this. Its most recent guidelines recommended antiviral drugs only for those high-risk patients.
But Tom Frieden, who was director of the CDC in 2014, had strongly supported the use of antiviral medications. At a telephone news briefing held late that year, he told listeners the opposite of what the data showed: that antiviral drugs can soften and shorten illness and reduce the risk of dying from the flu. “Antiviral treatment is particularly important this year,” he said before name-checking Tamiflu, adding: “If you are sick, talk to your doctor promptly about getting antiviral treatment.”
A reporter from Reuters then asked Frieden how he could square his recommendations with the evidence of Tamiflu’s ineffectiveness. Frieden replied that the CDC had looked at all the data, both published and unpublished, and that there was “strong” evidence that Tamiflu was impactful. It was not, he admitted, a “miracle drug,” and he wished there were better options. But there weren’t. The logic, then, is to take the drug not because it’s effective—it isn’t—but merely because it exists.
* * *
Peter Doshi has been following the Tamiflu saga for years. He is an editor focusing on drug regulation and marketing at the British Medical Journal, which campaigned for greater transparency about Tamiflu. Doshi is also a member of the Cochrane group that reviewed the neuraminidase inhibitors, which makes him an authority on the issue. Doshi had studied medical history and East Asian studies at Harvard and earned his doctorate from MIT, where his thesis was on the medical politics of influenza. After his postdoctoral work at Johns Hopkins he moved across town to the University of Maryland, where he teaches in the School of Pharmacy.
We met there, where he has a commanding view of downtown Baltimore from the twelfth floor. In his uncluttered office there was a heavy copy of Wittgenstein’s Philosophical Investigations on the table where we sat. It is a book that identifies the relationship between language and reality, and so mirrors the debate about Tamiflu and influenza treatments as a whole.
Doshi is deeply troubled by how governmental agencies had conflicting assumptions of flu treatments. Inside Tamiflu packaging is an insert that contains, in very fine print, all the information that the manufacturer must legally disclose, like dosing and side effects. All medicines have this insert, but almost no one reads them. The Tamiflu insert notes that although influenza may be complicated by serious bacterial infections, the drug “has not been shown to prevent such complications.” This is what the Cochrane review found, and what the FDA required the manufacturer to disclose.
Doshi points out that when the federal government was making its pandemic influenza plans, it seems to have come to an entirely different conclusion. Like the CDC, the Department of Health and Human Services believed that Tamiflu “will be effective in decreasing risk of pneumonia, will decrease hospitalization by about half . . . and will also decrease mortality.” The positive conclusions drove the “stockpile assumption,” as Doshi called it, and helped get these antivirals into more widespread use. The CDC held sway over what flu medications made it into the stockpile, and the CDC thought Tamiflu worked. And so a drug whose packaging included written doubts about its efficacy was squirreled away in large quantities for a pandemic it wouldn’t be able to slow.
Doshi believed that the level of evidence required for approving neuraminidase inhibitors was too low. This started in the late 1990s, when the first of these drugs was approved, and all Tamiflu had to do, Doshi said, was to “walk over that low bar.” He noted that there are, even today, questions about how the drug works. In addition to inhibiting neuraminidase, it also seems to have a direct effect on the central nervous system, lowering the fever associated with infections. If this finding is correct, Tamiflu’s effect on influenza-like illnesses may be no better than aspirin’s.
The controversy will continue, because the conflicting data allowed for diverging conclusions. Doshi holds out for a definitive trial whose costs, while high, would be “a drop in the bucket compared to the costs of stockpiling these drugs.”
Doshi’s doctoral thesis was titled “Influenza: A Study of Contemporary Medical Politics,” and he understands the bureaucratic realities of influenza. If there was a catastrophic flu outbreak, we would turn to the federal government to provide us with the very best medicines available. Who, precisely, would ever take the responsibility to declare that the antiviral medications at our disposal are not useful, and that the money to keep them in the stockpile might be better spent elsewhere?
Peter Palese is a longtime professor and chair of the Department of Microbiology at the Icahn School of Medicine at Mount Sinai (formerly the Mount Sinai School of Medicine) in New York. He has been involved in virtually every aspect of research that touches the influenza virus. His lab was the first to develop the technology that allowed the 1918 virus to be rebuilt from scratch. He tested the resurrected 1918 virus on mice to determine how dangerous it was. And he was part of the group that studied why the influenza virus spreads in the winter. Palese is an author of more than four hundred research papers studying the virus, and for him there is no debate. He disagrees with the FDA-required package information that says Tamiflu hasn’t been shown to reduce flu-related bacterial infections. It was all a question of timing; once there were symptoms of the flu, it was too late for the drug to work. But it is, he believes, a very good drug if given early on.
Palese actually used the resurrected 1918 virus in a study of Tamiflu. Along with others, he infected mice with the virus and then, six hours later, started treating them with Tamiflu. He found that Tamiflu was able to protect 90 percent of the mice from a lethal infection. One problem in reading the report of this experiment, however, is that it isn’t clear how many mice were used. That’s a really important detail because if only a few mice were used, we might not be confident of the result. It seems like a crucial experiment to verify, but because of the difficulties of experimenting with the live 1918 influenza virus, no one has done so. So all we know from this experiment is that antivirals seem to protect some mice, when given early enough. That’s something to cling to, some hope amid the depressing news about Tamiflu. But it’s not much.
Palese is a fervent believer in the science and benefits of Tamiflu and equates its critics with discredited antivaxxers. To be clear, the two issues are in no way comparable. There is no link between autism and vaccines. Study after study has shown this. Meanwhile, the evidence shows that taking Tamiflu when you have influenza only reduces your symptoms by about one day. To conflate the two issues is unfair. But Palese’s opinion reveals the very deep feelings that drive the issue.
A godsend, or a grift? Or something in between? The debates over the neuraminidase inhibitors is perfectly illustrated in two pieces of mail I received during the same week while writing this book. One was an information card, sent by the manufacturer to emergency physicians like me, encouraging us to prescribe peramivir, the intravenous drug for influenza. The other was my monthly copy of Emergency Medicine News (“the most trusted news source in emergency medicine”). On the front page was an article that announced, in a large and bold caption, that Tamiflu “belongs in the dust bin of clinical practice.”
If an emergency physician can feel confused about flu medications, then patients surely may feel the same. I’m duly skeptical about a drug like Tamiflu, but imagining a repeat of a 1918-type pandemic makes me think: if an influenza virus like 1918 leaped from birds or pigs, what else do we have to fight it? Noxious fumes from a factory in Falmouth? A century after the 1918 pandemic we still don’t have the miracle drug for influenza. We still cling to imperfect treatments, and seek peace of mind by stockpiling something like Tamiflu, you know, just in case. Our predicament and the limits of our power are best summarized by a former CDC field officer who told me: “Tamiflu doesn’t work. Now hurry up and take it.”