Researchers air new theories about 1918 flu mortality

Jan 13, 2012 (CIDRAP News) – Experts who have studied the 1918 pandemic have long puzzled over why the mortality rate in young adults was so high, a feature that might be explained by pathologic immune responses related to previous flu exposure, according to a new theory from a research team.

The extreme immunologic response to the 1918 pandemic virus could have made patients more susceptible to deadly secondary bacterial infections, according to two researchers who explained the rationale behind their hypotheses yesterday in a perspective piece in an early online edition of Emerging Infectious Diseases (EID).

The authors who detailed the hypotheses are Dr G. Dennis Shanks, director of the Australian Army Malaria Research Institute, and Dr John Brundage, a physician and epidemiologist with the Armed Forces Health Surveillance Center in Silver Spring, Md.

In September 2011, the same two authors published an epidemiologic analysis of naval medical records that suggested the first wave of the 1918 pandemic was caused by an antigenically distinct virus that left its hosts vulnerable to getting sick again in the second wave with infections that weren't lethal.

However, another study published about the same time suggested that cases in the first and second waves of the 1918 pandemic looked much the same, and it found no evidence of viral mutations that would explain why the fall wave was so much more severe. The study was based on tissue samples from pandemic victims.

During the 1918 pandemic, illness rates were highest among children, but mortality was highest in young adults, a highly unusual pattern for influenza.

Shanks and Brundage wrote that an analysis of who got sick and when they became ill provides some useful clues that might help explain the unusual mortality curve. Their assertions hinge on their view that the first of the three pandemic waves was caused by a different virus, given that clinical patterns varied between the first two waves.

A review of 1918 pandemic flu mortality rates in different populations suggests that exposure to bacterial pathogens played an important role, they wrote. For example, case-fatality rates from secondary pneumonia were highest in Australian and American soldiers with pandemic flu who were relatively new to their settings.

The authors hypothesized that many people in 1918 had their second exposures to an immunodominant T-cell epitope that was present on both the internal protein of the 1918 pandemic strain and on a similar previous strain, such as the 1889 pandemic virus. The second exposure could have led to a pathologic immune response that made patients vulnerable to deadly secondary bacterial infections, which they said may have contributed to the highly lethal nature of the second 1918 pandemic wave.

They also suggested that the pathologic responses made people vulnerable to bacteria that they had not been exposed to before, noting military recruits in new locations or soldiers crowded on troop ships would've had such exposures. Mortality rates were lower in people who had previous exposure in their occupational settings, such as military clinics, the article says.

Factors that led to high mortality during the 1918 pandemic, such as commercial and social isolation, no longer exist on a global scale, and most young adults are likely to be exposed to numerous viral and bacterial pathogens, according to Shanks and Brundage, who said their hypothesis may explain why the mortality rate for the 2009 H1N1 pandemic was relatively low.

The hypothesis about the T-cell epitope may have some implications for the development of universal flu vaccines, since it would be important "to ensure that T-cell-mediated responses to future seasonal and pandemic strains are protective but not pathogenic," they wrote.

In an editorial that accompanied the report, Dr David Morens and Dr Jeffery Taubenberger wrote that Shanks and Brundage's hypotheses were thought-provoking, but they said that a higher percentage of people of all ages, not just young adults, experienced flu that led to secondary bacterial pneumonia in 1918. Both were coauthors of the recent study on tissue samples from 1918 pandemic victims and are with the National Institute of Allergy and Infectious Diseases.

They also pointed out that case-fatality rates for 1918 flu-related pneumonia were similar to nonflu pneumonia caused by the same bacterial organisms.

Pathogenesis theories about severe or fatal 1918 flu must explain why the virus predisposed more people to secondary bacterial pneumonia and must address bacterial cofactors, Morens and Taubenberger wrote. They suggested that a similar phenomenon happened a year earlier in US military training camps with measles and bacterial pneumonia. "Such simple exposure variables might explain some of the mortality rate differences pointed out by Shanks and Brundage," they wrote.

The possible effects in 1918 of exposure to earlier circulating viruses are difficult to assess, according to the authors. How long the 1889 flu virus circulated and what age-groups may have been exposed are all speculative, they noted. Some 1918 data suggest older people who were alive during pandemics in the 1830s and 1840s may have had partial protection.

"Taken together, the information produces more questions than it answers," the two wrote, noting that only virologic or serologic evidence could clarify the picture.

Although the hypotheses draw attention to some possible clues, such as inconsistent wave-to-wave protection data, and suggest avenues for more experimental research, key questions about the 1918 pandemic still persist, Morens and Taubenberger wrote.

Dr Lone Simonsen, an epidemiologist in the Department of Global Health at George Washington University who has studied the 1918 pandemic, told CIDRAP News that Shanks and Brundage raise a novel hypothesis that the editorial writers question.

Neither of the papers incorporates two critical clues about the 1918 H1N1 pandemic, she said. An analysis of viruses from specimens obtained during the summer pandemic wave, published in September 2011 by Morens and Taubenberger's group, isolated the H1N1 virus, which strongly argues that it was already circulating then and caused the milder summer wave. She added that epidemiologists had long argued that this was the case, after demonstrating that the unusual pattern of nearly all deaths occurring in young adults also occurred during the summer wave.

"Taken together, these clues argue strongly that the summer wave was caused by a 1918-like H1N1 virus that had not yet acquired the extreme virulence seen in the devastating autumn wave," she said. "Or else, the bacterial pathogens that caused deadly co-infections linked to the devastating mortality in autumn were not present during the summer wave."

Simonsen said another important clue to consider is how adults over age 45 were completely spared in nearly all countries studied so far, except for remote locations. She pointed to what she called a common misinterpretation regarding the so-called W-shaped curve, noting that those who consider the data should be aware that high year-round background mortality rates in young children and elderly can give a false impression of unusual pandemic mortality in these age-groups.

Elderly people in the United States and Europe were completely spared in the 1918 pandemic, and thus must have been protected by previous exposure to a related virus that circulated many years earlier, she said. "But, the fact that all age-groups were dying at high rates in remote populations that may have been isolated from earlier influenza importantly demonstrates that extreme 1918 mortality shows that previous exposure is a critical clue to explaining at least the sparing of the elderly in the US and Europe."

One thing that students of the 1918 pandemic can agree on is that many unanswered questions remain, such as how likely it would be that an extremely deadly influenza pandemic like the devastating 1918 event will happen again, and why the age pattern was so different from seasonal influenza and other pandemics, she said.

"Searching for answers to such questions is what makes this historic pandemic research so fascinating," said Simonsen, who added that an interdisciplinary approach that integrates epidemiology, virology, immunology, anthropology, sociology, and history will be needed to make the many clues from the past fall into place.

Shanks DG, Brundage JF. Pathogenic responses among young adults during the 1918 influenza pandemic. (Perspective) Emerg Infect Dis 2012 Feb;19(2) [Full text]

Morens DM, Taubenberger JK. 1918 influenza, a puzzle with missing pieces. (Commentary) Emerg Infect Dis 2012 Feb;19(2):[Full text]

See also:

Sep 20, 2011, CIDRAP News story "Autopsy study leaves severity of fall 1918 pandemic wave a mystery"

Sep 30, 2011, CIDRAP News story "Study suggests 1918 flu waves were caused by 'distinct' viruses"

Newsletter Sign-up

Get CIDRAP news and other free newsletters.

Sign up now»


Unrestricted financial support provided by

Bentson Foundation 3MAccelerate DiagnosticsGilead 
Grant support for ASP provided by

  Become an underwriter»