Never a dull moment in Washington. We had just returned from Iraq. My family, who had been on pins and needles for the entire time I was away, was enormously relieved to have me home, but there was little time to absorb all that I had seen. Something was brewing in my area of infectious disease that had bubbled over into the public consciousness and was causing anxiety. It was related to an influenza virus that had emerged among chickens and was beginning to infect humans in China. The fact that humans became infected with a bird virus immediately generated concerns throughout the world, including in the United States, about the possibility of a pandemic influenza outbreak.
People have feared this kind of natural disaster since the historic global pandemic of influenza in 1918, often referred to erroneously as the Spanish flu. In reality, that global pandemic might have actually started in the United States toward the end of World War I, and it claimed the lives of between 50 and 100 million people worldwide. The population of the world then was 1.8 billion, making this equivalent to 200 to 400 million deaths today. There were more deaths due to influenza among military personnel on both sides of the conflict than deaths due to bullets and bombs. At that time there were no vaccines for influenza, no antibiotics for the secondary bacterial infections that killed many influenza-infected people, and no intensive care units that might have saved many lives and thus might have mitigated the impact of the pandemic.
Of note, in 1918 the influenza virus had not yet been identified. Today we know an enormous amount about influenza. There are two major types of influenza: influenza A and influenza B. The most common type is influenza A, which has been the cause of all known influenza pandemics. Different influenza A viruses are designated by an international classification according to the types of protein on the surface of the virus. Two of the most important influenza surface proteins are hemagglutinin and neuraminidase, represented by the letters H and N, respectively, together with a number next to each letter. There are eighteen separate H proteins and eleven separate N proteins among influenza A viruses. Each virus has an H with a given number and an N with a given number. The pandemic of 1918 was caused by an influenza virus with the designation of H1N1. Influenza viruses did not start off evolutionarily as human viruses, and only a small proportion of these have infected humans. The remainder are animal influenzas, primarily viruses of wild waterfowl such as ducks.
Over centuries, certain of the influenza viruses have adapted very well to humans such that now we generally consider them human viruses, although they remain prevalent in many species of birds including chickens, as well as in mammals such as pigs, hence the term “swine flu” for some viruses. The pandemic of 1918 is thought to have originated from a bird influenza that jumped species to humans.
Each year human outbreaks of influenza occur almost invariably during the winter season in a given region of the world. In the United States and other Northern Hemisphere countries, this seasonal outbreak usually takes place from December through March. In the Southern Hemisphere in countries such as Australia, the winter flu season is June through September. The occurrence of seasonal influenza is predictable in that we are almost guaranteed to have an outbreak every winter. However, the precise influenza virus that will dominate during any given winter is much less predictable.
From one year to another there is usually a slight change in the circulating influenza virus caused mostly by mutations in the genetic makeup of the virus. This slight change is referred to as a drift. Because of exposures to influenza viruses over previous years as well as the protection afforded by vaccinations, most of the world has some degree of what is known as background immunity to influenza. This at least partially protects us from actual infection or from serious disease caused by the infection. This is the reason we do not have a global catastrophic outbreak of influenza every year.
In contrast, when an influenza virus emerges that is markedly different from previous years, the phenomenon is referred to as a shift in the virus. Because this very different virus is new to most people, there is little background protection against the novel strain. This can lead to a high rate of global infection and an increased level of suffering and death. Such outbreaks are referred to as pandemics.
After the pandemic of 1918, the next pandemic occurred in 1957 with a brand-new influenza virus referred to as H2N2. During that pandemic there were 1 to 2 million deaths globally. A subsequent pandemic occurred in 1968 with a new influenza virus referred to as H3N2. This pandemic resulted in roughly 700,000 deaths worldwide. Since then, H3N2 influenza viruses have transitioned and evolved over the years to become the typical seasonal influenza virus that we experience with varying degrees of drifts each year. The most recent influenza pandemic, which occurred in 2009, originated somewhere near the border of California and Mexico. This was a relatively mild pandemic referred to as H1N1 swine flu because it jumped species from a pig to the human population. The global mortality estimates from that pandemic ranged from 151,000 to 575,000.
Influenza viruses have long been an intermittent and sometimes major problem for chickens—and chicken farmers. Entire flocks of chickens can get infected and either die or need to be culled to prevent further spread to other flocks. Note, not all chicken influenza viruses are especially harmful to the chickens and do not require dramatic measures such as massive culling. While chicken influenzas occur worldwide, they are most common in the Far East, particularly in China. Occasionally, one of these chicken influenza viruses jumps species to infect one or two people, but these occurrences are rare and do not attract much attention. This was not the case in the spring of 1997, when an influenza virus designated H5N1 infected and caused high mortality in three chicken farms in Hong Kong. Soon thereafter there were reports from Hong Kong of avian influenza H5N1 infections that had jumped from chickens to eighteen humans. The more troubling problem was that six of these people died, a case fatality rate of 33 percent. This was astronomical compared with the usual death rate of seasonal influenza, which is a fraction of 1 percent. Even in the catastrophic 1918 pandemic, the mortality rate was only 1 to 2 percent.
The fear that this virus might develop the capability of spreading easily from human to human, together with the economic considerations associated with a highly pathogenic chicken virus spreading to other chicken farms, led the health authorities in Hong Kong to take drastic measures. They did a mass culling, killing essentially all of the chickens in Hong Kong. This draconian course of action ended the 1997 H5N1 outbreak among chickens, and thus there were no additional human infections.
For the next six years H5N1 fell off the radar screen. That changed in 2003 when it reemerged in China and spread to countries in Asia and Southeast Asia. Of concern were the sporadic human cases that occurred. In one situation, health officials could not rule out the possibility of human-to-human transmission within a family unit. This is what we were facing at the end of February 2004, when I returned from Iraq.
Over the twenty years that I had been director of NIAID, threats of pandemics almost never turned into actual pandemics, and there was no solid evidence at that point of efficient or sustained human-to-human transmission of H5N1 influenza, which would be the first indication of a looming pandemic. Nonetheless, we began the public health and scientific process of preparing for the worst. In the broad sense, the CDC was responsible for the public health aspects of the preparedness and response to a potential outbreak. They carefully tracked the cases in collaboration with their Chinese and other Asian counterparts and were prepared to identify, isolate, and contact trace infected individuals if the virus ever entered our country. At NIAID my colleagues and I began the process of doing what we do best—developing and testing vaccines against this new influenza strain.
As part of this effort, in the early fall of 2004, we launched the NIAID Influenza Genome Sequencing Project in collaboration with other HHS agencies and other non-HHS organizations to determine the complete genetic sequences of H5N1 viruses from thousands of chickens and from the occasional human who was infected. We rapidly provided the sequences to the broader scientific community funded predominantly by NIAID. This was information they required both to do the nuts-and-bolts research on the testing of sensitivity to known anti-influenza drugs such as Tamiflu and to begin the important process of developing a pre-pandemic vaccine. All this was proceeding very well. But then we hit a trouble spot on another front that diverted our attention.
The supply of influenza vaccines administered in the United States each year for seasonal influenza generally becomes available in the early fall preceding the winter influenza season. The United States usually obtains its seasonal influenza vaccines from several pharmaceutical companies based throughout the world. For the 2004–5 influenza season we were expecting between forty-six and forty-eight million doses from the Chiron Corporation and about fifty-two million doses from other companies. During the summer of 2004 at the peak time of vaccine preparation and production, the FDA and the CDC were informed that there was a “minor problem” with about four million doses being produced in Chiron’s Liverpool, U.K., plant due to bacterial contamination. The FDA and the CDC were led to believe that this problem was “being taken care of.” But on October 5, 2004, the British regulatory agency abruptly suspended Chiron’s license to ship the influenza vaccine to the United States because the Liverpool plant failed to pass inspection. This came as a big surprise to the FDA and was a huge issue for the United States because this would result in a severe shortage of seasonal influenza vaccine doses in the fall and winter of 2004–5. So the rising concern for a potential H5N1 bird flu pandemic was compounded by the reality of a vaccine shortage for seasonal flu. And what had started off as a purely public health issue quickly evolved into a political hot potato as we happened to be in the final months of the 2004 election campaign.
Almost immediately there was an explosion of press interest and an array of congressional hearings. On October 8, 2004, just three days after the British regulatory agency suspended the Chiron license, the CDC director, Julie Gerberding, the acting FDA commissioner, Lester Crawford, and I testified before the House Government Reform Committee regarding the broader issue of the fragility of the influenza vaccine supply. I totally believed that my FDA colleagues were telling the truth when they said they were not told about the seriousness of the situation in the Liverpool plant until the British announced the suspension of the Chiron license.
On October 12, I gave a White House briefing in which I reiterated my confidence that the FDA was correct because I was hearing from other sources that their story matched with the way others had seen the events unfold. But the damage had already been done from a public perception standpoint, and we had to deal with it. The president’s aides were concerned that Senator John Kerry, the Democratic nominee for president, would try to embarrass President Bush in the presidential debate to be held the next night by claiming that the FDA knew about the problem and covered it up. The last thing that I wanted to do was to get involved in a fight between political parties, something I had successfully avoided for so many years as director of NIAID. But the political game was on. I offered my humble suggestion that they just come out with the plain truth, and that the president state outright that a contamination in the vaccine plant was the cause of our potential vaccine shortage.
During the debate, that is how it played out. President Bush admitted right up front in response to a question from one of the moderators, Bob Schieffer of CBS, that “we relied upon a company out of England to provide about half of the flu vaccines for the United States citizens, and it turned out that the vaccine they were producing was contaminated.” The president then took the high ground: “And so we took the right action and didn’t allow contaminated medicine into our country.” When asked to comment on President Bush’s answer, Senator Kerry changed the subject and talked about the weaknesses in our health-care system and how they had gotten worse over the past four years of the Bush administration. The debate went on to other issues.
George W. Bush was reelected on November 2, 2004. I doubt if the issue of the Chiron vaccine debacle had any effect one way or another on the voting. And the fact that the election was over seemed to defuse the tension around this issue. The 2004–5 influenza season came and went without incident. Relatively speaking, the public health impact of the vaccine shortage was insignificant. We were lucky.
No sooner had the second term of the Bush administration begun than the situation with H5N1 influenza started heating up again. In January 2005 we learned of the first documented human-to-human transmission of the H5N1 bird flu in Thailand. The bird flu issue again loomed large, and we accelerated the process of developing a pre-pandemic vaccine against the circulating H5N1 bird flu.
The development of a vaccine against an emerging influenza virus such as H5N1 is a complicated process that starts off with the isolation of the virus in question. Usually the CDC as well as the WHO obtains the virus specimens and creates what is referred to as seed viruses that are manipulated to grow readily in chicken eggs. Growing the virus in eggs is the standard method of obtaining enough virus to produce an influenza vaccine. A consensus is developed as to what the best seed virus is—that is, the one that grows most efficiently in eggs—and then this virus is distributed to various pharmaceutical companies so that they can manufacture test doses of vaccine. Some of these doses are then given to us at NIAID to test in various clinical trials.
When a new vaccine is being made, you need to know if it is safe and if it induces a response (immunogenicity) in the person being injected that would predict that when that person gets exposed to the actual virus, they will be protected. In other words, if I vaccinate you, I can measure antibodies in your blood that tell me even before you get exposed to or infected with the virus whether those antibodies will protect you against infection. That is called “predictive of being protective against exposure to the virus,” and it is an important aspect in NIAID’s early clinical trials.
At the same time the NIAID vaccine trials are being conducted, the various pharmaceutical companies begin accelerated production of vaccine that the government ultimately purchases for the stockpile. The clinical trials tell us what the optimal dosage and number of doses are for the various populations who might receive the vaccine, such as adults, the elderly, and children. This entire process takes several months at a minimum.
On April 27, 2005, Julie Gerberding and I briefed President Bush and his senior staff in the White House about the current situation with the H5N1 bird flu. The epicenter of bird-to-human transmissions had shifted to Vietnam, where several individual cases were reported.
The president was in good spirits as he entered the Roosevelt Room from the Oval Office, joking with people in the room including the newly appointed HHS secretary, Mike Leavitt, former governor of Utah, before taking his seat at the center of the large wooden table. Julie gave a quick summary on the increasing number of bird-to-human cases and human-to-human cases being reported from Southeast Asia and Indonesia. The president turned to me: “Okay, Fauci, your turn.” “Mr. President, thus far the safety and potency of the candidate H5N1 vaccine look good,” I said. Discussion turned to the need to stockpile H5N1 influenza vaccine once it was adequately tested. This was of particular interest to Vice President Cheney and Scooter Libby because both of them had been major driving forces in establishing Project BioShield and supporting the Strategic National Stockpile.
By August, we had successfully demonstrated in normal volunteers that with the proper dose we could induce responses that we could predict would be protective against the H5N1 influenza were it to develop into an outbreak. The remaining challenge was to make enough vaccine in the immediate future to have a sufficient stockpile of tens of millions of doses and to be able to scale up to hundreds of millions of doses in the event of a pandemic.
The good news regarding the potential efficacy of the vaccine was tempered by the recent sobering news that migratory birds that likely were infected by the chickens in China and the Far East had brought the H5N1 infection to chicken flocks in Russia, Mongolia, and Kazakhstan, and birds using established flyways would likely bring the infection to India and Europe. Given the international implications of the virus’s spread, I was asked to brief Condoleezza Rice, now secretary of state, on August 17. Although the tone of the briefing was serious, Condi joked that every time we met it was about some real or potential global health crisis. She said she was glad to see me, but we both knew the truth.
The possibility of an influenza pandemic was completely overshadowed by a real catastrophe that took place on August 23, 2005, when Hurricane Katrina struck Louisiana and Mississippi. In the aftermath of this tragedy—one of the most horrific natural disasters in the history of the United States—the Bush administration, particularly the Federal Emergency Management Agency, and even the president himself were sharply criticized for their inadequate response to the hurricane. This criticism generated an increased sensitivity on the part of the administration to never again be unprepared for natural disasters. When I again briefed the president in the Roosevelt Room on our progress in developing a pre-pandemic vaccine about ten days after Katrina made landfall, he was emphatic: “Fauci, failure is not an option here.”
On October 6, Mike Leavitt and I briefed President Bush and his senior staff in the Oval Office for a meeting the next day with the senior executives of major pharmaceutical companies. The president sat in his usual chair in front of the fireplace. “Mr. President,” I said, “our capacity to manufacture influenza vaccines is rather fragile. Remember the disaster with the Liverpool contamination.” I went on to tell him that our ability to scale up rapidly in response to a pandemic outbreak was weak, particularly when we had to rely on the rate of growth of viruses in eggs to produce vaccines. For this reason I said we needed to get more companies to invest in newer technologies and that it would be helpful if he emphasized this point.
The next day, the president walked into his meeting and got right to the point. He emphasized the importance of revitalizing our vaccine production capacity not only for the immediate issue of H5N1 influenza but also for every seasonal influenza. He told the pharmaceutical executives that we needed to ramp up capacity and create a sustainable market for seasonal influenza vaccine that would provide a major head start in the event of a pandemic flu.
The pharmaceutical executives left that meeting with a clear message from the president. How they would respond was less clear.
I left with clarity on a different issue. We did not just need to increase our vaccine production capacity. That would be only an incremental improvement. What we needed was a game changer. We just had to do better than continually chasing after each new emerging influenza strain. We needed a “universal” influenza vaccine—one vaccine for all strains of influenza, be they seasonal or pandemic. In 2010, we began in earnest a concentrated research effort at NIAID to develop one.
For the next year, the interest in H5N1 influenza among the media, Congress, and the general public waxed and waned depending on the occurrence or not of new infections in chickens in various parts of the Far East, eastern Europe, and North Africa. Despite this interest the number of actual cases in humans remained relatively low. By the end of May 2006 there had been 218 confirmed cases of bird flu in humans associated with 124 deaths.
By late 2008, from a countermeasures standpoint we were as prepared as we could be. We had successfully developed an H5N1 vaccine and had stockpiled forty million doses to be able to vaccinate twenty million people with the required two doses apiece. In addition, we had stockpiled a considerable amount of antiflu medications, particularly Tamiflu. While the virus continued to smolder among chickens, it never evolved the capacity to spread efficiently from human to human, there were no human infections of bird flu in the United States, and we never experienced a pandemic outbreak of H5N1.
But once again, we were lucky. I believed then that as long as there were viruses or other pathogens that infect animals and that could evolve molecularly to infect humans, there was a risk of spillover, similar to what we had seen with SARS in 2002–3. This was especially true in places such as Southeast Asia where animals live in proximity to humans—on farms or in the wild—known as the animal-human interface.
One might imagine that it would be terribly frustrating to put in enormous efforts of preparation for events that never happen. However, that is not how I feel. At NIAID, our job was to conduct and support the basic and clinical research that would allow us to develop the best possible vaccines, treatments, and diagnostics for any future pandemics. I knew that sooner or later the day could come when our efforts at preparedness might save millions of lives.