When Science Goes Wrong (28 page)

When they knocked at the address given for Fagim Dayanov on Escadronaya Street, for example, they found his widow, Rema. She told them that Fagim, like several other victims, had been employed at the ceramics factory, which was situated three kilometres from Compound 19 in a south-easterly direction. Fagim had remained healthy during the first few weeks of the outbreak. When the cause was identified as anthrax, vaccinations were made available; Rema agreed to be vaccinated but Fagim refused. On May 3, probably as part of the disinfection campaign, Fagim was told to clean the roof of the factory. He became ill on May 4 and was taken to the hospital on May 6. Rema and their son were not allowed to visit him until three days later, by which time he was desperately ill. The couple was kept separate by a glass partition and had to communicate by exchanging notes. On the evening of May 10 Fagim wrote:
Rema – don’t go home. We may never look into each other’s eyes again
. Rema eventually did go home, and Fagim died that night. When Rema returned to the hospital, not knowing that her husband was dead, she was directed to the morgue, where her husband’s body lay on a gurney, naked, with his chest, abdomen, and braincase sliced open. The autopsy had already been completed.

Besides these often-painful interviews with the victims’ relatives, Meselson’s team had extensive discussions with the two pathologists, Faina Abramova and Lev Grinberg, who had conducted most of the autopsies. It was Abramova who made the initial diagnosis of anthrax, based on the observation of the anthrax bacillus in tissue samples. She related how the KGB had come to her hospital to confiscate all the medical and autopsy records and tissue samples, but she successfully hid much of the material: she placed jars containing tissue samples among other items in her pathology museum, and she concealed paperwork in filing cabinets devoted to other matters. Thus, Abramova and Grinberg were able to give Meselson’s team a comprehensive description of 42 cases. All of them had shown signs of infection in the thoracic cavity, with bleeding from thoracic lymph nodes and from the mediastinum (the area between the left and right lungs). Some victims also had involvement of the gut, but this involvement was probably secondary to a systemic infection, because the mesenteric lymph nodes (the nodes that receive lymph from the intestines) were affected in only a few cases. In short, the pathological findings were consistent with inhalation anthrax in most or all cases, just as Peter Gumbel had concluded in his
Wall Street Journal
articles the previous year.

The two weeks that Meselson’s team had allotted themselves for the trip to Russia came to an end before they were able to carry out most of their interviews with the victims’ relatives, and they resolved to make a return trip the following year. Nevertheless, even with the results that they obtained on the 1992 visit, two things were clear: the outbreak was caused by the airborne release of anthrax spores and, second, most of the victims either lived or worked in the Chkalovskiy district of Sverdlovsk at the time they became ill. The researchers’ findings were thus pointing strongly toward the conclusion that Meselson had long resisted, namely that the outbreak was caused by an accident at Compound 19, the suspected germ-warfare institute.

In the autumn of 1992, an event occurred that to some degree made Meselson’s investigations irrelevant. The second-in-command at Biopreparat, the network of Soviet facilities that Vladimir Pasechnik had fingered as a germ warfare agency, defected to the United States. His name was Kanatjan Alibekov, but he later anglicised it to Ken Alibek. When interviewed by the CIA, Alibek poured out a truly horrifying account of the Soviet germ warfare programme, a description that went far beyond what Pasechnik had revealed. Among other things, he said that the Soviets had produced hundreds of tons of anthrax. The lethal dose of anthrax when inhaled is about 10,000 spores, or ten billionths of a gram, so the Soviets had manufactured enough of the spores – in theory – to kill the entire population of the planet millions of times over. Alibek identified Compound 19 as being heavily involved in research, production and weaponisation, and he said that he himself had developed a strain of anthrax that was far more lethal than those that occurred in nature.

Meselson was still determined to complete his own study, and in August of 1993 he and Guillemin returned to Sverdlovsk. They interviewed more relatives of the deceased victims, as well as some of the few victims who had recovered, and they obtained detailed information about the victims’ whereabouts during the outbreak. In addition, they visited several villages outside Sverdlovsk where farm animals had come down with anthrax during the same period. One place they were not allowed to visit was Compound 19, the suspected site of anthrax release; although anthrax production had probably long ceased it was still a military facility, and Meselson could not obtain the required authorisation to enter.

Back in the United States, Meselson and Guillemin put all the information together in an attempt to pin down the time and place of the anthrax release. The first person to fall ill, a 48-year-old ceramics factory employee named Vera Kozlova, did so on Wednesday April 4, so there must have been some release of anthrax before then – probably a day or two before, given what was known about the minimum incubation period for inhalation anthrax. So many victims fell ill in the few days after Kozlova was struck down that it seemed likely that a single release event very early in that week had caused most of the infections.

Looking at their records, Meselson and Guillemin saw that nearly all the victims either lived in the Chkalovskiy district to the southeast of Compound 19, or had worked in or visited that district in the relevant time span. Six victims lived or worked in Compound 19 itself. As already mentioned, several of the victims worked at the ceramics factory southeast of Compound 19. Another group of victims were five military reservists; these men all lived and worked outside of Chkalovskiy, but during the week starting Monday April 2, all five took a course at Compound 32, a military facility immediately to the south of Compound 19. The reservists came in each morning and left the district in the late afternoon. This strongly suggested that the anthrax release occurred during daytime hours on one of the weekdays from April 2 to April 6, most likely on April 2 or April 3, given that the earliest disease onset was on April 4.

When the researchers took a map of Sverdlovsk and plotted the daytime locations of all the victims during this time span, a striking pattern emerged. Nearly all the victims were located in an extremely narrow, cigar-shaped zone that began in Compound 19 and extended in a south-easterly direction. The closest victims – six of them – had been within Compound 19 itself. The most distant had been four kilometres away on the south-eastern outskirts of the city. Although no humans were affected beyond that distance, animals were: the veterinary cases occurred in six villages that lay precisely along a continuation of that south-easterly line. The most distant of these villages lay more than 50 kilometres from Compound 19.

Five victims were ‘outliers’ – they neither lived nor worked in the high-risk zone. Three of them had occupations, such as truck driver, that might have taken them into the zone. Information about the other cases was lacking.

Thus it was clear to the researchers that the anthrax cloud had originated at Compound 19 and the wind had blown it toward the southeast. This was the final clue that allowed Meselson to pin down the date of the accident. Sverdlovsk airport, like all major airports around the world, reports weather conditions every three hours to a United Nations agency, and this information is archived by the US National Center for Atmospheric Research in Boulder, Colorado. When he obtained the records for the week in question, Meselson found that the wind had blown steadily to the southeast or south-southeast throughout the daylight hours on Monday, but never blew in that direction during the following two days. The wind did blow to the southeast for part of the day on Sunday April 1, but this day was ruled out because the military reservists were not in the area until Monday. Thus Monday, April 2, was pinned down with near certainty as the date of the incident. As icing on the analytical cake, it turned out that one of the reservists had attended classes at Compound 19 on only a single day – Monday.

Meselson now turned to another question: how much anthrax was released? To answer this, he had to make a lot of ‘guesstimates’ about factors such as the height of the release above ground, atmospheric conditions and the number of spores that would be required to cause a fatal infection. In the end, he concluded that the total release was quite small – less than a gram (four hundredths of an ounce), and perhaps as little as a few milligrams.

The researchers published their findings in two papers. A 1993 paper in the
Proceedings of the National Academy of Sciences
presented the evidence from the pathological investigations, which pointed to the conclusion that most of the anthrax infections were acquired by inhalation and not by food consumption. In the following year, Meselson’s group published a paper in
Science
in which they presented the findings of their epidemiological studies, including the time, place and estimated amount of anthrax release.

Together, the two studies presented far more detailed and convincing evidence that the anthrax outbreak originated in the biological warfare facility than had any previous investigation. No one reading the
Science
paper could fail to be impressed by how weeks of plodding epidemiological footwork had allowed the researchers to plot a cigar-shaped ‘arrow’ that pointed unmistakably and accusingly at Compound 19.

Still, the study did have its detractors. For one thing, there were some who felt the whole topic was moot, given that the Russians had already admitted that their germ warfare institute caused the outbreak. Also, Meselson had not been able to answer the question that people were now asking, which was what exactly went wrong in Compound 19 that led to the anthrax release. Finally, there were some critics, such as germ-dispersal expert Bill Patrick, of Fort Detrick (home to the US Army Medical Research and Materiel Command), who felt that Meselson’s estimate of the amount of anthrax released was far too low. ‘We hooted [when we heard his numbers],’ is what Patrick told the authors of Germs.

The significance of the dispute over the amount of anthrax released is this: if the amount was very small, it was conceivable that it happened not as a result of weapons production but in the course of legitimate research or vaccine development – activities that were permitted under the Biological and Toxin Weapons Convention. Tom Mangold and Jeff Goldberg, in their 1999 book
Plague Wars
, suggested that Meselson gravitated to a low estimate because he was clinging to his long-held hope or belief that the Soviets were abiding by the convention. They dug up some statements made by Meselson in the mid-90s that were consistent with this interpretation. Still, Meselson sticks by his estimate, and it has recently been confirmed in a re-analysis by physicist Dean Wilkening of Stanford University.

One still-unresolved mystery has to do with the ‘tail’ of anthrax cases which extended for six weeks beyond the initial outbreak. Most of these victims clearly had inhalation anthrax, not the gastrointestinal form, so it doesn’t seem likely that they got sick from eating tainted meat, as the CIA analysts had originally proposed. How then had they acquired their infections?

One theoretical possibility is that the anthrax release was not a single event but continued for several weeks, albeit at a lower rate. If this were the explanation, however, one would expect the later victims to be located in other places than in the cigar-shaped zone where the early victims lived or worked, because the wind blew in other directions on later days. In reality, they were located in exactly the same high-risk zone as the early victims.

A second possibility would be that the later victims acquired their infections from anthrax that had been deposited on surfaces during the initial release and then re-entered the atmosphere at some later time. Such an explanation seems particularly appropriate for victims like Fagim Dayanov who, as described earlier, fell ill in early May, a day after he was sent to clean the roof of the ceramics factory. Did he stir up a secondary aerosol of anthrax during that cleaning operation and thus breathe in enough spores to develop an infection?

When I asked Matt Meselson about this, he acknowledged that secondary infections were a possibility, but he considered this explanation unlikely. Anthrax spores are so tiny, he told me, that once they bind to surfaces they are held there by surprisingly strong electrostatic forces and are very difficult to pull back into the atmosphere. Larger clumps of spores can be re-aerosolised, but such clumps are not effective infectious agents because they do not penetrate deep into the air sacs of the lungs, which they need to reach order to trigger an infection.

Meselson favours a third explanation for the late cases, which is that those victims inhaled the anthrax spores on Monday April 2, just like the early victims, but simply took longer to fall ill – up to six weeks in some cases. This contradicts conventional medical wisdom, which says that the incubation period for inhalation anthrax is just a very few days – just one day in some cases. Meselson dug up some old studies, however, in which monkeys were exposed to anthrax by inhalation. Some of these monkeys took weeks to fall ill. Such long incubation periods were not observed in more recent studies, but according to Meselson that was because researchers simply didn’t wait that long – the animals were sacrificed after a few days whether they were sick or well.