by The Pandemic Century- One Hundred Years of Panic, Hysteria
Not long after, Fraser decided it was also time for him to close the book on the EIS investigation and begin the laborious process of drafting his final report, known as an EPI-2. Despite going over the hotel with a fine-tooth comb and many hours of interviews, he was still no closer to identifying the pathogen or means of transmission. Privately, he was dismissive of the nickel carbonyl theory as the metal usually has an incubation period of less than thirty-six hours and rarely causes fevers above 101°F. Nor did he think it was food poisoning: the Legionnaires had purchased food from many different sources, and EIS officers had been unable to implicate a common meal at the convention. Similarly, although the cross-connection between the AHU and the Bellevue’s potable water supply suggested a waterborne illness, more than one-third of the ill delegates insisted they had never drunk water at the hotel. By contrast, almost all the hotel employees, none of whom had suffered Legionnaires’ disease, said they frequently drank from a fountain in the lobby.
In October, Fraser discussed the case with his boss, John V. Bennett. Bennett had been the lead investigator of the 1965 outbreak at St. Elizabeths Hospital. Based on the epidemiology of the outbreak (the cases were associated with proximity to open windows), Bennett had suspected airborne transmission. However, all attempts to identify the etiological agent had failed, so at the end of the investigation Bennett had filed the blood from St. Elizabeths patients in the CDC’s serum banks in the hope that it might prove useful at a later date. “When you solve the Legionnaires’ outbreak you will solve my outbreak at St. Elizabeths,” he told Fraser.
Mulling over Bennett’s words, Fraser thought an airborne agent might explain the sickness of both the Legionnaires and Broad Street pneumonia cases—individuals who had passed by the hotel without entering it. Fraser also noted the strong association between illness and time spent in the lobby, and the fact that after the convention a fault had been discovered in the AHU serving the lobby area, prompting the hotel’s management to have the filters cleaned. This cleaning operation had taken place on August 6 and “may have inadvertently limited the ability of investigators to identify a toxic or microbiologic agent in the air handling system,” Fraser wrote in his report. On the other hand, the relatively low attack rate “may rule against an airborne agent to some degree.” All that Fraser could say with confidence is that the illness resembled an infectious disease and there had been no secondary spread. Unfortunately, despite exhaustive microbiological studies, all the tests had been negative. It was possible that as new tests and technologies became available, new toxins capable of causing pneumonitis might be discovered, but that lay in the future. At present, “no toxin is known that causes just this pattern of disease,” he concluded, “and toxicological studies are also negative so far.”
In all his years as an outbreak investigator Fraser had never come across a case quite like it; it stuck in the craw, but he had to admit defeat. So did Langmuir. The Philadelphia outbreak, he informed the press, constituted “the greatest epidemiological puzzle of the century.”
* Rickettsia is the name for a family of bacteria transmitted by the bites of chiggers, ticks, fleas, and lice. The best known rickettsial diseases are typhus and Rocky Mountain Spotted Fever.
† Many bacteria will continue to grow in tissue postmortem, hence the importance of embalming and cold storage to prevent putrefaction. However, most bacteria that cause disease cannot survive more than a few hours in a dead body.
CHAPTER V
LEGIONNAIRES’ REDUX
“The discovery of the etiologic agent of Legionnaires’ disease was accomplished in the face of overwhelming odds. All of the combined bacteriology and pathology experience accumulated since the beginning of the century pointed away from this agent being a bacterium.”
—WILLIAM H. FOEGE, director, CDC, Senate Subcommittee
on Health and Scientific Research, November 9, 1977
In a period when antibiotics and vaccines seemed to have closed the book on infectious disease, the Legionnaires’ outbreak challenged medical confidence that America was on the brink of a germ-free era. Little wonder then that the CDC’s failure to solve the puzzle of the century produced a lingering sense of insecurity and anxiety. However, outside of medical and public health circles, the same sense of anxiety did not attach to swine flu. This is odd when you consider that CDC officials had been warning of an epidemic since February. Indeed, in late March, President Ford had gone on television to drive home the concerns that an outbreak of swine flu was imminent. Flanked by the two godfathers of the polio vaccine, the scientists Albert Sabin and Jonas Salk, Ford told the American public that he had been advised that “there is a very real possibility that unless we take effective counteractions, there could be an epidemic of this dangerous disease next fall and winter.” Accordingly, he was seeking a $135 million appropriation from Congress for sufficient vaccine “to inoculate every man, woman, and child in the United States.”
Congress approved the appropriation bill in April, and by the middle of August had also passed legislation waiving corporate liability for the immunization campaign. Ironically, Congress’s willingness to idemnify vaccine manufacturers had little to do with its enthusiasm for the insurance business and everything to do with the fears raised by the Philadelphia outbreak. Even though on August 5 Sencer had told senators he did not think the outbreak was due to swine flu, politicians were petrified he might be wrong and that they would end up being branded obstructionists who had impeded the delivery of a life-saving vaccine. Nonetheless, scientists’ and politicians’ enthusiasm for the flu immunization campaign was not shared by the public, with a Gallup poll in September indicating that only about half of all Americans were willing to be immunized. In other words, faced with a repeat of an epidemic on the scale of 1918, the public’s response was to shrug its shoulders. This indifference turned to resistance when at the beginning of October the campaign finally got under way. Within ten days, one million Americans had rolled up their sleeves and received the jab, but on October 11 the campaign suffered a disastrous blow when it was reported that three elderly people in Pittsburgh, Pennsylvania, had died hours after having the inoculation. The deaths prompted a media scare that led to nine states suspending their vaccination programs. To calm the public’s nerves and restore confidence, President Ford and his family were photographed receiving jabs at the White House. CDC scientists, meanwhile, attempted to educate the public, explaining that the risk of temporally associated deaths within forty-eight hours of inoculation occurred at a rate of 5/100,000 a day. By comparison, the anticipated death rate per day for all causes among citizens in Pennsylvania was 17/100,000. In other words, it was to be expected that some people would die after receiving the flu shot, but that did not mean there was a causal connection.
Unfortunately, by 1976 the public’s unquestioning acceptance of scientific authority was beginning to wane, as were memories of life before vaccines against polio, measles, and other debilitating childhood diseases. Moreover, by now influenza experts in other countries were beginning to question the American scientific consensus that the swine flu isolated at Fort Dix was the harbinger of a new pandemic strain, a skepticism shared by WHO officials in Geneva who advocated a “wait and see” policy. As October turned to November with no signs of the feared pandemic, skepticism hardened. Still, the campaign might have been saved were it not for reports of cases of Guillain-Barré syndrome (GBS). A rare and occasionally lethal neurological syndrome, GBS occurs at a steady rate in the general population and, if a pandemic had been occurring, would have been deemed an acceptable risk. However, in the absence of a pandemic, the reports in December that as many as thirty people had developed the syndrome within a month of receiving the flu jab sparked widespread alarm, prompting the government to suspend the campaign so that the association with the vaccine could be investigated. The program was never restarted, and as cases of the syndrome soared—by the end of December, 526 cases were being reported, of which 257 had
received the flu jab—the press and politicians in Washington began looking for a fall guy. The New York Times was especially harsh, labeling the campaign a “fiasco” and, in view of the fact that the pandemic had never materialized, a waste of time and effort. The result was that when Jimmy Carter moved into the White House in January, Joseph Califano Jr., the incoming secretary of Health, Education, and Welfare (HEW), demanded Sencer’s resignation. Shamefully, Califano informed Sencer of his dismissal minutes before the pair were due to appear together at a meeting in Washington on the moratorium of the swine flu program. Worse, the whispered conversation in an HEW hallway was captured by the TV cameras, deepening Sencer’s humiliation. According to public health historian George Dehner, this was “shabby treatment” for someone who had given sixteen years of service to the CDC, eleven as its director. On the other hand, Dehner writes, in his efforts to convince administration officials of the vaccination campaign’s necessity, Sencer had deliberately downplayed the scientific uncertainty so as to give “a distorted vision of the new virus.” The result was that “only the most dire vision of a swine flu pandemic remained.”
Ironically, just three weeks before Sencer’s very public firing, Shepard had rushed into his office to announce that he and McDade had solved the puzzle of the century. The culprit was a hitherto unknown Gram-negative bacterium. Other researchers had missed it because the bacterium was difficult to see using a conventional Gram stain. However, McDade had solved the problem using a different staining technique. According to Garrett, after all the pressure and frustration of the past year, Sencer was reluctant to accept what Shepard was telling him. “Shep, how sure are you?” “Better than ninety-five percent,” he replied, “but I’d like to run a few more experiments before we go public.”
There is an old saying in medical research: “fortune favours the prepared mind.” The saying is usually attributed to Louis Pasteur, who famously stumbled on a vaccine against chicken cholera in 1880 when a colleague inoculated some chickens with an old culture of chicken cholera germs. In McDade’s case, however, it was because he was a novice to public health microbiology and therefore not schooled in the same thought processes as his colleagues that a chance observation led him to the answer that had eluded them. It also helped that McDade was a worrier and a perfectionist. To his way of thinking, the peculiar rod-shaped bacteria he had weakly glimpsed through his microscope in August represented a loose end, and he didn’t like it. However, it was not until late December that he thought to return to the problem. The impetus was a conversation with a man who had cornered him at a party shortly before Christmas. “I don’t know how he knew I was CDC but he did,” recalled McDade. “He said, ‘We know you scientists are a little weird but we count on you and we’re very disappointed.’ I stuttered because I didn’t know what to say. But it bothered me and stuck in my mind.”
It had always been McDade’s habit to use the week between Christmas and New Year’s to resolve any outstanding paperwork in preparation for the new working year. While tidying his office he spotted the glass slides with the smears he had taken from the guinea pigs in a box on the shelf and decided to have another look. The exercise, he recalled, was like “searching for a missing contact lens on a basketball court with your eyes four inches away from the floor.” Eventually, however, McDade spotted a cluster of organisms in the corner of one of the microscopic fields. To McDade’s way of thinking, the fact that the organisms were clustered together “suggested that it wasn’t just organisms that happened to be there but were actually growing there inside the guinea pig.” It was at this point that McDade decided to have another go at culturing the organism. His thinking was “if I can rule out that it has nothing to do with the disease my conscience will be salved and I can go about my business.” It was at this point that McDade’s expertise as a rickettsial specialist and his willingness to depart from conventional patterns of thought came into play. Going to the freezer drawers containing spleen tissue from suspect guinea pigs that had been put on ice in August, McDade thawed the samples and inoculated some of the tissue into embryonated eggs. However, this time he withheld antibiotics in order to allow whatever organisms were present in the guinea pig tissue to grow freely. Five to seven days later the eggs died and McDade took new smears. As before, he applied a Gimenez stain, a technique which had been developed specifically for rickettsial organisms, and, once again, spotted the same rod-shaped bacteria growing in clusters. Could these bacteria be responsible for the guinea pigs’ deaths, and could the same bacteria be responsible for the Legionnaires’ disease outbreak? To answer the question, McDade retrieved some preserved serum from the Legionnaires’ cases and mixed it with the organism he had found in the eggs. If a patient’s serum contained antibodies specific to the organism, an observable reaction would take place. It did. “They just lit up dramatically,” he said. “My neck hair bristled. I wasn’t sure what I’d got there but I knew it was something.”
McDade immediately shared his findings with Shepard, and together they ran further tests using paired serum samples taken from Legionnaires two or more weeks apart. If it could be shown that the reaction took place at much higher dilution in the second serum sample than in the first, this would be strong evidence that the patient had recently recovered from the disease caused by the organism. At the same time, McDade and Shepard repeated the test using blinded samples from both Legionnaires’ and non-Legionnaires’ patients, some of whom had had other pneumonias or were healthy. Nearly fifty years later, McDade vividly recounted the moment of discovery:
When we’d finished all the tests, later that evening, they brought down the paper and we broke the code. All the normal specimens from healthy people were negative, specimens from patients with other pneumonias were all negative. Then we looked at the Legionnaires’ disease specimens. Specimens taken from Legionnaires early in the illness had little or no antibodies, and specimens taken later in illness had very high levels of antibody, which suggested they had been infected with this bacteria. So that was the moment when we knew we’d found the etiologic agent.
When Shepard informed Sencer of the breakthrough he could hardly contain his excitement and insisted that they issue an announcement in the next edition of the Morbidity and Mortality Weekly Report, the CDC’s house journal, and schedule a press conference for the same day, January 18, 1977. This was earlier than Shepard and McDade had been anticipating—normally, scientific discoveries take several months to be written up before being submitted to a scientific journal. Because of the political pressure on Sencer, however, he could not wait for the usual peer review process. Worried that they would be laughing stocks if their methodology was subsequently found to be faulty, Shepard and McDade double-checked their results. Then, out of curiosity, McDade decided to look in the CDC’s stores for serum from other unsolved outbreaks. That’s when he came across the stored blood from the patients at St. Elizabeths Hospital. McDade injected the blood into chicken eggs, then added the organism he had isolated in Philadelphia. The eggs lit up immediately, indicating that there was an antibody reaction and that the St. Elizabeths patients had been infected with the same organism. Bennett’s intuition had been correct: in solving the Philadelphia outbreak, Fraser and his team had also solved the mystery of the earlier outbreak in Washington, DC.
News of Shepard and McDade’s discovery traveled around the world, prompting scientists at other research establishments in Europe and elsewhere to duplicate the CDC’s results. As scientists exchanged information and examined old case files, it became apparent that St. Elizabeths was not the only prior outbreak of Legionnaires’ disease. Blood specimens from patients at the Oakland County Health Department in Pontiac, Michigan, in 1968 also tested positive for antibodies to Legionella pneumophila, as the organism was now known, suggesting that they had been infected by the same agent, though why there had been no pneumonia in the “Pontiac Fever” cases and why the outbreak had not resulted in any fatalities was unclear. That was not all:
in May 1977, Marilyn Bozeman, a rickettsia specialist at the Walter Reed Army Institute of Research, in Bethesda, Maryland, informed McDade that she had seen very similar organisms in guinea pigs while investigating specimens taken from an outbreak in 1959. Like McDade, she had assumed these were contaminants and described them as “rickettsia-like.” It was only later, when she ran new tests, that she found they were actually two new species of Legionella, Legionella bozemanii and Legionella micdadei. It was subsequently found that L. micdadei had also been responsible for an outbreak of “Fort Bragg fever” in 1943 and that Walter Reed also had an isolate of L. pneumophila dating from 1947.
Then, in early summer, came news of an outbreak at a medical center in Burlington, Vermont. EIS officers rushed to the scene, and by September they had documented sixty-nine cases of Legionnaires’ disease. However, once again, the source of the exposure eluded them. Soon, there were reports of other outbreaks in hospitals across the United States. The most notable was an outbreak that began at the Wadsworth Medical Center, a veterans’ hospital in Los Angeles, in the summer and which by the end of the year had claimed sixteen lives. At around the same time, a smaller epidemic broke out at a hospital in Nottingham, England, sickening fifteen people. Once again, no common source was found, but two of the sera from patients sent to the CDC for analysis tested positive for antibodies to Legionella. That was not all: in 1978 CDC scientists confirmed that Legionella had been responsible for a mysterious outbreak of pneumonia at the Rio Park hotel in Benidorm, Spain, that had been blamed for the deaths of three Scottish holidaymakers five years earlier. The result was that when, in 1980, another outbreak occurred at the same hotel, epidemiologists took water samples and found the bacterium lurking in the showerheads. Apparently, an old water well had been brought back into use five days before the start of the outbreak and had fed water infected with L. pneumophila directly into the hotel. Investigators concluded that those who showered and washed first thing each morning were at most risk because the bacteria multiplied overnight in water standing in peripheral pipe work. In all, a total of fifty-eight people were sickened, and one woman died. Like the outbreak at the Bellevue, the Rio Park outbreak sparked considerable press interest and inspired the thriller writer Desmond Bagley to pen a novel, Bahama Crisis (1980), in which a Caribbean holiday resort’s water system is deliberately seeded with Legionella bacteria in an act of industrial espionage.