Pale Rider: The Spanish Flu of 1918 and How It Changed the World

by Laura Spinney

  Iijima’s estimate is problematic, however. One of the assumptions he made was that the flu arrived at the ports, and that poor communications prevented it from penetrating the interior. Yet Taiyuan, the ‘capital’ of Shansi and very much inside that interior, was already connected to Peking by rail in 1918, and anecdotal evidence suggests that the epidemic was anything but mild in Shansi. In 1919, a man who had first-hand experience of fatal epidemics, Percy Watson, described the outbreak there–by which he meant the illness that raged over three weeks in October 1918–as ‘one of the most fatal epidemics reported in the medical literature this past year’.6 On 2 November 1918, describing the same outbreak, the North China Herald mentioned thousands of dead in Taigu, a town in Shansi. And contemporary reports kept by the Chinese Post Office referred to many victims in two neighbouring provinces, Hopei to the east and the confusingly named Shensi (Shaanxi) to the west. In Hopei, the flu was reported to have killed more postal workers than had a visitation of pneumonic plague early in 1918. It seems at least possible, therefore, that flu was widespread in China in 1918 and 1919, that it followed a similar pattern as elsewhere in the world–of mild spring wave, severe autumn wave and possible recrudescence in early 1919–and that in parts of the country, at least, the death toll was very high indeed. In the case of China, Patterson and Pyle may have been closer to the mark.

  In 1998, on the eightieth anniversary of the pandemic, Australian historian and geographer Niall Johnson and German flu historian Jürgen Müller revised the global death toll upwards again. Their justification was that the earlier estimates represented tips of a largely unreported iceberg, that the under-reporting affected rural populations and ethnic minorities disproportionately, and that there were indications that some of those populations–partly for reasons of historical isolation–had suffered very heavy losses. By then, the death toll in India alone had been estimated to be as high as 18 million–three times what Indians believed it to have been in 1919–making Jordan’s 21.6 million seem ‘ludicrously low’ by comparison. Johnson and Müller came up with a figure of 50 million, of which Asia accounted for 30 million. But, they stressed, ‘even this vast figure may be substantially lower than the real toll, perhaps as much as 100 per cent understated’.7

  An understatement of 100 per cent means that the number of dead could have been as high as 100 million–a number so big and so round that it seems to glide past any notion of human suffering without even snagging on it. It’s not possible to imagine the misery contained within that train of zeroes. All we can do is compare it to other trains of zeroes–notably, the death tolls of the First and Second World Wars–and by reducing the problem to one of maths, conclude that it might have been the greatest demographic disaster of the twentieth century, possibly of any century.

  In the annals of flu pandemics, the Spanish flu was therefore unique. Most scientists now agree that the event that triggered it–the spillover of the pandemic strain from birds to humans–would have happened whether or not the world had been at war, but that the war contributed to its exceptional virulence, while at the same time helping to spread the virus around the world. It would be hard to think of a more effective dissemination mechanism than the demobilisation of large numbers of troops in the thick of the autumn wave, who then travelled to the four corners of the globe where they were greeted by ecstatic homecoming parties. What the Spanish flu taught us, in essence, is that another flu pandemic is inevitable, but whether it kills 10 million or 100 million will be determined by the world into which it emerges.

  PART SIX: Science Redeemed

  René Dujarric de la Rivière in an army laboratory, Calais, 1915

  13

  Aenigmoplasma influenzae

  In the dog days of August 1914, an ageing Ilya Mechnikov–Russian exile, Nobel laureate, ‘lieutenant’ of Louis Pasteur and mentor of Yakov Bardakh, Wu Lien-teh and others–battled his way across a Paris in the grip of mobilisation to reach the Pasteur Institute, one of the world’s leading centres for the study of infectious diseases and the production of vaccines. When he arrived, he found it under military command. Most of the younger scientists had left for active service and all of the experimental animals had been killed. The man who had renounced God at the age of eight, who believed fervently that the progress of civilisation depended on the advancement of science, surveyed his deserted empire and quaked.

  In his novel Journey to the End of the Night, Louis-Ferdinand Céline immortalised Mechnikov as Serge Parapine, an eccentric and demented genius who ‘always had enough hair on his cheeks to make him look like an escaped convict’, and who raged and muttered through the smelly corridors of the renowned Parisian institute where he worked. The institute’s other inhabitants were ‘grey-haired, umbrella-carrying schoolboys, stupefied by the pedantic routine and intensely revolting experiments, riveted by starvation wages for their whole adult lives to these little microbe kitchens, there to spend interminable days warming up mixtures of vegetable scrapings, asphyxiated guinea pigs, and other nondescript garbage’. But as Mechnikov intuited on that summer’s day, the era that Céline described so scathingly–his era, in which important battles had nevertheless been won against crowd diseases, and faith in science rode high–was about to end.

  First, though, there was a war to be fought–and diseases to be kept at bay. One of the young scientists who had left the Pasteur Institute at the outbreak of war was René Dujarric de la Rivière, a twenty-nine-year-old aristocrat from the Périgord who, like others of his contemporaries, had been swallowed up by the army’s network of laboratories. Four years later, when the second wave of the Spanish flu broke out, he was working in the central army laboratory in the city of Troyes. ‘I was there in the Champagne region when an artillery troop came through on its way to the front. They never left. All of them, men and officers alike, were suddenly struck down and had to be hospitalised urgently.’1 The army launched a vaccination campaign, using a vaccine against pneumonia-causing bacteria that had been developed at the Pasteur Institute before the pandemic. Dujarric had spent time in Richard Pfeiffer’s lab in Breslau–where Pfeiffer, known to his colleagues as the ‘Geheimrat’ or privy counsellor, was treated with profound respect–but he had begun to doubt that Pfeiffer’s bacillus was really the cause of flu.

  He wasn’t alone. Pfeiffer’s bacillus–Haemophilus influenzae, to give it its scientific name–is a real bacterium that lodges in the nose and throat and causes infections, some of them severe, but while it had been found in many of the flu cases analysed, it hadn’t been found in them all. In New York, bacteriologists William Park and Anna Williams of the city health department had collected lung tissue from dozens of flu victims post-mortem, then grown the bacteria colonising it on agar gel in order to identify the species present. Even when Pfeiffer’s bacillus was among them, they found, it seemed to exist in different strains. That was odd: in a pandemic, you’d expect to find the same strain consistently. And it certainly wasn’t the only bacterium in the mix: streptococci, staphylococci and pneumococci were there too, in legions, and they could also cause respiratory disease. Alexander Fleming, a captain in the British Army’s medical corps at the time, had confirmed Park’s and Williams’ results using tissue from, among other places, Étaples. Some had gone a step further. As early as 1916, Milton Rosenau, a doctor in Boston, had voiced his suspicion that the causative agent of flu was a virus–an organism small enough to pass through the pores of the porcelain Chamberland filters that were routinely used to trap bacteria out of liquid at that time, and hence commonly referred to as a ‘filterable virus’.

  Dujarric probably knew of Fleming’s work, possibly even of Park’s and Williams’, and of Rosenau’s suspicions. In 1915, before moving to Troyes, he had run the army laboratory for the northern region, in Calais, and while there had crossed paths with Sir Almroth Wright, the British inventor of the typhoid vaccine. Wright had requisitioned the casino in nearby Boulogne for a laboratory–beds replaced gaming tables, chandel
iers were swathed in linen sheets–and put his junior colleague Fleming and others to work in it. They shared the space with an American hospital set up by Harvard University. Wright was well known by then, and the casino received a constant stream of visitors. He got on well with the French, according to Fleming’s French biographer, André Maurois (who acted as an interpreter and liaison officer with the British Army), despite differences in the British and French attitudes to the war. For the French it was a quasi-religious ceremony to be treated with great solemnity, while the British did their duty and took what opportunities they could for relaxation. Maurois recounts how Fleming and another man, probably Wright, were enjoying a wrestling match one day, when a door opened and in came a delegation of senior French Army doctors. The wrestlers leapt to their feet and immediately engaged the visitors in a scientific discussion, but, recalled a witness, ‘I will never forget the expression on the French doctors’ faces on discovering that scene.’

  They may not have agreed about the place of contact sports in the theatre of war, but they were converging on the notion that the cause of flu might not be Pfeiffer’s bacillus. That idea was therefore in the air when, walking in the streets of Troyes one day, early in October 1918, Dujarric bumped into his old friend and fellow Pasteurian, Antoine Lacassagne. The two had not seen each other since before the war, but Lacassagne had been sent to Troyes to help vaccinate the troops. ‘After chatting a moment, he made me a curious proposition,’ Lacassagne recalled years later. ‘Dujarric asked me to do him the favour of injecting him with the filtered [blood] of a flu patient, the experiment that he felt would confirm his hypothesis. I pointed out the moral dilemma he was placing me in, but he finally convinced me that it was better that I do it, in the best conditions, than that he inject himself–something he was otherwise determined to do. I administered the injection on the morning of Tuesday 8 October, in his army laboratory.’2

  Lacassagne had to leave for Paris the next day, and he didn’t discover the outcome of the experiment until months later. For two days, Dujarric remained well, then he noticed the first symptoms. He managed to describe the course of the disease: ‘Third and fourth day, after an abrupt onset, intense and persistent frontal headache, pain all over… temperature between 37.8˚ and 38.2˚… Fourth night agitated, nightmares, sweats. On the fifth day the pain disappeared; very pleasant euphoria after the indefinable sense of malaise that had marked the previous two days… In the following days everything returned to normal, except for a lingering fatigue, then on the seventh day cardiac symptoms emerged, and these persist: intermittent but very disagreeable chest pains, irregular pulse, breathlessness at the slightest effort.’

  In a second experiment performed a few days later, he painted his own throat with a filtered emulsion of flu patients’ sputum and waited, but experiencing no further symptoms, concluded that the first experiment had immunised him against the second. Miraculously, given his own state of health and the chaos around him, he managed to write up his findings and transmit them to the Pasteur Institute’s director, Émile Roux, within a matter of days. It was only a preliminary study, he admitted in the report that Roux presented on his behalf on 21 October, to the French Academy of Sciences, but the key point was that the blood with which he had been injected had been filtered, hence free of bacteria. It raised the possibility that the flu was caused by a virus.3

  What did Dujarric mean by a virus? He probably wasn’t quite sure himself. All he could really say was that it was something smaller than a bacterium, that was capable of transmitting disease. He probably would have hesitated before describing it as a living organism, however (and indeed, the debate over whether a virus is dead or alive continues today: can an organism be described as alive if it is incapable of reproducing on its own?), and he may have at least allowed the possibility that what he had infected himself with was something more like a venom.

  Coincidentally, in the same proceedings of the academy, two other Pasteurians, Charles Nicolle and Charles Lebailly, reported the same conclusion. They were working in the Pasteur Institute’s outstation in Tunis, and in the first days of September they had inoculated a monkey and two human volunteers with the sputum of a Spanish-flu patient–unfiltered in the case of the monkey, filtered in the case of the humans. The monkey, which had received the inoculum via the inner lining of its eyelids and nostrils (considered a part of the airborne route), showed signs of a flu-like disease a few days later–high temperature, loss of appetite, lassitude. The human who had received the filtrate under his skin fell sick on the same day, but the one who received it into his blood remained well. Nicolle and Lebailly concluded that the cause of the disease was a filterable virus that could not be transmitted by the blood.

  Dujarric de la Rivière and Nicolle and Lebailly were the first to publish, independently but simultaneously, the finding that the flu was probably caused by a virus. Before 1918 was out, German, Japanese and British scientists had performed similar experiments and arrived at similar conclusions. Like Dujarric, the German, Hugo Selter of the University of Königsberg, had experimented on himself. The first half of the twentieth century was an era of self-experimentation (Mechnikov had deliberately given himself cholera, among other potentially lethal diseases) but perhaps it was easier to risk one’s life when all around you were risking theirs–that is, in time of war. Members of the British team, who published an initial account of their findings in December 1918, did not experiment on themselves. But one of them, Graeme Gibson, was preparing a follow-up report when, worn down by long hours in the army lab at Abbeville, near Étaples, he caught the flu. He died before it was published the following March.

  For all their bravery, the credibility of these scientists’ findings is tainted. The experiments were conducted during the pandemic, at a time when it would have been impossible for them to ensure that their laboratories were free of contamination by the ubiquitous flu virus, so it is hard to know by which route their experimental subjects received the infection. Anyone paying attention will have noticed that Dujarric’s and Nicolle’s and Lebailly’s results contradict each other: Dujarric thought that he had given himself flu via an injection of filtrate into his blood, while the pair in Tunis ruled out the blood as a transmission route. Nicolle and Lebailly were right, in fact: influenza is not transmissible by the blood, so Dujarric cannot have caught it from the injection that Lacassagne gave him. He probably caught it via the usual route–the air–while bending over the four gravely ill soldiers whose blood he took in preparation for the experiment, developing symptoms after the usual incubation period of two or three days. As happens so often in science, in other words, Dujarric was right for the wrong reasons.

  Rosenau and his colleague John Keegan in Boston also tried, in the thick of the autumn wave, to demonstrate that the causative agent of the flu was filterable, but they were unable to transmit the disease. Others failed too, but their results are as unreliable as those of their French counterparts. One reason their human volunteers may have failed to get sick, for example, was that they had been exposed to the virus during the spring wave, and acquired some immunity. Within the scientific community at the time, however, people interpreted the results according to their preferred theory. The Geheimrat himself, Richard Pfeiffer, remained convinced that ‘his’ bacillus was the most likely candidate. His supporters felt that if Rosenau had found no virus, it was because there was no virus to be found (putting his trust in the data, Rosenau agreed with them–a case of being wrong for the right reasons). When it came to explaining the troubling finding that Pfeiffer’s bacillus was absent from the lungs of some flu victims, on the other hand, the Pfeiffer camp blamed bad tools and methods. It didn’t help, in terms of penetrating the shadows around the disease, that antibacterial vaccines had shown some efficacy against it–because they had worked against those lethal secondary infections.

  Only in the 1930s did the shadows begin to lift. One of the peculiar aspects of the 1918 pandemic was that it coincided wit
h an epidemic of a very similar disease in pigs–so similar, in fact, that the pig disease was dubbed ‘swine flu’. At the time, veterinarians regarded it as a new disease in swine, but from then on it erupted periodically in herds. In 1931, following one such outbreak, an American virologist, Richard Shope, confirmed what Dujarric, Selter and the others had tried to demonstrate earlier, in far more difficult circumstances: that flu was caused by a filterable virus. Two years later, a team of British scientists working at the National Institute for Medical Research in London, did the same in humans. After a ferret sneezed in the face of one of them, Wilson Smith, he came down with flu. They went on to show that a filterable agent could transmit flu from a ferret to a human and back again (whether that agent was an organism or a toxin was still an open question, though by 1950 the London team had come to believe, correctly, that they were dealing with an organism).

  From the humble beginnings of a ferret’s sneeze, the vast and complex biology of influenza began to unfold. When a virus infects a person, his or her immune cells secrete tiny morsels of protein called antibodies that attach themselves to the virus, disabling it. Antibodies can linger in the blood for years after the infection has passed, providing a record of past infections, and by the 1930s, scientists already had tests for detecting them in serum (the clear liquid in which all the other components of blood float). When they saw that antibodies produced during one flu outbreak did not necessarily protect people against another, they realised that flu came in different varieties. Three types of flu were eventually identified (a fourth has been added very recently): A, B and C. A and B cause epidemics, but only A causes pandemics. C is altogether milder and less contagious than the other two. The virus that caused the Spanish flu was, needless to say, an A.