H1N1 LESSONS LEARNED Pandemic underscored influenza's unpredictability

Editor's Note: This is the first of a series of articles reviewing the world's experience with pandemic H1N1 influenza and what we've learned in the past year. Look for further installments in the days to come.

Apr 23, 2010 (CIDRAP News) – A year's experience with the 2009 H1N1 influenza virus has underscored the endless unpredictability of flu.

The virus and the resulting illness defied expectations on many levels. It was first detected in North America, not Southeast Asia. Unlike the pathogens that caused previous pandemics, the virus was not a new subtype, but rather an H1N1 strain, making it a cousin of a seasonal flu strain that's been around for decades.

Unlike seasonal flu, though, the illness hit children and nonelderly adults much more than people over 65. It was relatively mild for most people, unlike the 1918 flu or the devastating disease associated with H5N1 avian flu, which had heavily shaped pandemic preparations. But it killed far more children and young people than seasonal flu typically does, with the vast majority of deaths involving people younger than 60.

The pandemic also brought the first hint that obesity is a risk factor for severe flu. And it reinforced an old and somewhat underappreciated lesson: that pregnant women who contract flu also can get seriously ill and even die.

In one of the biggest surprises, the pandemic virus made seasonal viruses do a vanishing act this past winter, at least in the United States. Although type B and H3N2 viruses circulated to some degree in certain parts of the world, such as Asia, they failed to show up or appeared in only tiny numbers in the US.

This week, a year after the pandemic virus first burst into the headlines, several experts contacted by CIDRAP News reflected on the past year and the lessons it has taught about the virus and the illness. Future stories will look at treatment issues, the H1N1 vaccine, the vaccination program, and other public health responses.

New, but not new
A big surprise to many—and a source of terminological confusion—was that the H1N1 virus is not a new subtype. As an H1N1 virus, it is the same subtype as the virus that caused the 1918 pandemic and whose descendants have circulated in humans ever since, except for one 20-year interruption.

As virologist Vincent Racaniello, PhD, of Columbia University describes it, the H1N1 virus of 1918 went into both people and pigs at the time and evolved on mostly separate tracks in the two different hosts. In humans the subtype circulated until 1957, when it was replaced by the H2N2 virus, which triggered the pandemic of 1957-58. H1N1 then vanished from humans until 1977, when it mysteriously reappeared, probably as the result of a lab accident. It then continued to evolve as one of three seasonal flu strains, with H3N2 and type B.

Meanwhile, the H1N1 in pigs continued to circulate, but evolved much more slowly than its human counterpart, according to Racaniello, who authors "Virology Blog." Because pigs, as food animals, don't live very long, their viruses don't face very much selection pressure, he explained. "So that virus that went into them in 1918 didn't change very much. It did reaasort with a variety of human and avian viruses, but the key proteins, the HA and NA, are direct descendants and haven't changed very much. To me that's the most amazing thing, that pigs have almost been like a freezer for this virus."

While the H1N1 in pigs changed little, human strains of H1N1 changed a lot, under pressure from the immune system of their long-lived hosts, Racaniello said. By 2009, the human and pig versions were so different that when the swine version jumped into humans, many people had little protection against it.

The fact that seasonal H1N1 and the pandemic H1N1 are the same subtype obscures how different they are, Racaniello said. "Subtypes are kind of artificial definitions," he commented. "The H1N1 pandemic virus is as different from the previous seasonal H1N1 as it is from an H3N2. So the fact that we're calling it H1N1 doesn't really mean anything. We should get used to the idea that a pandemic could be caused by the same subtype [as a seasonal virus]."

Dr. Daniel B. Jernigan of the Influenza Division at the Centers for Disease Control and Prevention (CDC) expressed a similar view. "[The pandemic H1N1] is not a new subtype. But the distance from the nearest neighbor causing routine human disease is pretty far, so in that sense it's not unexpected that you'd have significant numbers of people who have not had exposure to anything like this," he said.

As it turned out, of course, those who had some protection against the pandemic virus were older people, who benefited from their encounters with H1N1 viruses before 1957. Those who had some exposure in the 1930s, 1940s, and to some degree the 1950s probably have some immunity, said Jernigan.

Geographic expectations
Another surprise to many, given that the last two pandemics and H5N1 avian flu had their roots in Southeast Asia, was that the pandemic virus was discovered in North America. The first known cases of the 2009 H1N1 occurred in Mexico, with cases popping up in the United States shortly afterward.

But Racaniello said it's not known where the virus really emerged. "We had the first reports in North America, in Mexico and the US. That doesn't mean it went from pigs to people there; it doesn't mean it originated in those areas." But even if it did, he added, "I don’t think it's surprising, because we have lots of pigs in North America, and [pigs are] where this virus came from."

The virus's real birthplace may never be pinpointed, because its ancestors are missing from the surveillance records for at least 10 years, he said. "The parents haven't been detected for a long time. I don’t know if we're going to figure that out. The moment may have passed."

Wave pattern
In the United States, the 2009 virus first spread widely in late April and May, slowed but hung around through midsummer, heated up again in late summer as children returned to school, and peaked in late October and early November. Then it waned fairly quickly, and a feared third wave never materialized in the winter.

In this pattern of waves the pandemic differed sharply from seasonal flu but bore some resemblance to past pandemics, especially that of 1957. "If you look at pandemic epi curves, what we saw in 1957 was very similar to what we saw this year, in the shape of the curve and the time it appeared," said Jernigan. He said no significant winter wave was noticed in 1957-58, but mortality data later suggested that one did occur.

The absence of a winter wave of cases this time around may be explained by a shrinking number of susceptible people. With the combination of immunity in older people, the virus's presence through the summer, and a major wave of cases in the fall, "What we may have done was burn through a significant amount of the population, so there weren't enough susceptibles left to have a third wave," said Jernigan.

In addition, by winter a growing number of people had gained immunity through vaccination, further limiting the number of easy targets for the virus, noted William Schaffner, MD, chair of the Department of Preventive Medicine at the Vanderbilt University School of Medicine and a board member of the Infectious Diseases Society of America. The CDC's current estimate of the number of people vaccinated is between 72 million and 81 million.

Impact on the population
The true impact of the pandemic on the population won't be known for some time to come, given that pandemic viruses tend to dominate flu seasons for several years and that it takes 2 to 3 years to develop accurate flu mortality data for a given season. The number of cases can only be estimated, because costly confirmatory testing was sharply reduced within weeks of the outbreak's start.

But it's safe to say that the public generally regards the pandemic as mild, or at least milder than news reports led people to expect. The perception is based on the fact that for most people the illness resembled seasonal flu and also that the number of cases dropped steeply after the fall peak. In addition, the CDC's latest estimate of deaths in the pandemic is 12,270, well below the oft-cited CDC estimate of 36,000 in a typical flu season—though the two estimates are based on different methods.

But the perception of the pandemic as mild overlooks the reality that cases and deaths have targeted mainly children and young adults, in sharp contrast to seasonal flu. About 90% of the deaths in a typical flu season are in elderly people, whereas most of the pandemic deaths have been in younger adults and children.

"Normally severe flu strikes children under 4 and people over 65 and spares healthy people from 5 to 65. What we see is it did not spare older children and younger adults," said Andrew Pavia, MD, chief of pediatric infectious diseases at the University of Utah Health Sciences Center and chair of the pandemic flu task force for the Infectious Diseases Society of America (IDSA).

The CDC estimates that 35 million cases have occurred in adults aged 18 to 64, 19 million in children through age 17, and just 6 million in those older than 64. The proportions are similar for hospital cases, estimated at a total 270,000. And out of an estimated 12,270 deaths, there were 9,420 in the younger adult group, 1,580 in the elderly, and 1,270 in children.

The H1N1 burden on younger people was also underscored by a recent report in BMC Infectious Diseases with data from the early months of the pandemic in 11 countries: It showed that more than 75% of cases occurred in people age 30 and under, with the highest incidence in those 10 to 19. Only 3% of cases were in people 65 and older.

"What we had in terms of overall severity doesn't fit easily into a headline," said Pavia. "We had a disease that caused largely moderate illness in most people," but there was a risk of severe disease for young children, pregnant women, people with underlying diseases. About 278 children died of confirmed H1N1, far more than the previous record of about 88 in a flu season, he added.

In view of the disease burden on younger people, a team of researchers led by Cecile Viboud of the National Institutes of Health recently tried to refine our understanding of the pandemic's impact by estimating the total years of life lost. They estimated the number of deaths to be between 7,500 and 44,100, with the lower number based on deaths officially attributed to pneumonia and influenza and the higher number representing all-cause deaths, including those from diseases associated with flu but not necessarily reported as such.

Their main finding was that, in life-years lost, the pandemic was at least as severe as a nasty regular flu season and possibly worse than the pandemic of 1968-69.

Health system stressed
As for the pandemic's impact on the healthcare system, the general verdict is that there were signs of stress, but the system was not overwhelmed.

"Our clinical capacities were stretched but not broken by this pandemic," said Schaffner. He credits the built-in protection enjoyed by older people for sparing the system. Without that, he said, "We would've been in much hotter soup just taking care of patients. Because in this country we've pared down our excess medical capacity, just because it's so expensive."

Pavia said hospitals were not too hard hit overall, but some emergency departments, particularly pediatric emergency units, were "extremely stressed." Some had to set up satellite units in their parking lots during the peak period. Pediatric emergency departments have probably done more disaster training than other departments because of the heavy load of respiratory illnesses they face each winter, he commented.

Intensive care units (ICUs) also were put to the test, Pavia said. One colleague reported that his hospital's ICU at one point ran short on dialysis machines, not on ventilators, which are often cited as likely to be scarce in a pandemic. "In planning we didn't think about dialysis machines," he said.

But from the reports he has heard, ICUs had more problems with staffing than with equipment, Pavia said. Workers became exhausted from caring for young adults who required long periods of intensive care. "It may be that the greatest vulnerability in our intensive care capacity is personnel, the ability to sustain effort over a long period of time," he commented.

The clinical picture
The clinical illness caused by the novel H1N1 virus was generally similar to seasonal flu, experts say. But the intense scrutiny of the pandemic brought some new aspects to light and underlined previous lessons.

"Clinically I'd have to say the illness is fairly similar to what had been reported before," said Kathleen M. Neuzil, MD, MPH, an associate professor of medicine in the division of allergy and infectious diseases at the University of Washington in Seattle and chair of the influenza working group for the CDC's Advisory Committee on Immunization Practices (ACIP).

"I think we have to be careful in interpreting some of [the findings], because this would have to be the most studied virus, if not in history, then for a very long time," she said. "There was a lot more testing and a lot more capturing of different types of disease."

For example, there have been reports that quite a few H1N1 patients had no fever, seeming to suggest that this could be peculiar to the pandemic virus. But in the past, said Neuzil, the CDC definition of influenza-like illness included fever, so afebrile flu cases would not have been counted. "The pandemic made us look a little harder at the illness; it made us look more broadly," she said.

In fact, Jernigan said the pandemic and seasonal viruses probably differ little in this respect. "There were some reports that up to a third or even half of individuals didn't have fever, which is not so different from seasonal flu, but we don't have a lot of good data."

Another tentative finding is that gastrointestinal (GI) symptoms are more common in H1N1 cases than with seasonal flu, which, if true, would have implications for transmission.

"The GI symptoms do appear to be somewhat more prominent than with seasonal," said Pavia. "We've seen that in H5N1 in Southeast Asia. It's interesting in terms of clinical presentation, but it also raises the possibility that some patients will be able to transmit infection by shedding virus in their stool."

Jernigan said the CDC has observed a higher-than-expected rate of GI problems in H1N1 patients, but he's not sure if it's real. "I don't know that we have enough data to say that this virus truly causes more GI illness," he said. "It's not unheard of for you to get more GI problems when you get flu."

A feature that has turned out to be less common than expected in H1N1 cases is bacterial coinfections, Jernigan reported. "We saw them, but they were not as prominent as we expected, probably due to a lot of antibiotic use and vaccination against strep pneumo," he said.

Streptococcus pneumoniae and Staphylococcus aureus are the typical causes of pneumonia secondary to flu, he noted. "We've been vaccinating against strep pneumo for the past 10 years or so, and we may be seeing a benefit during the pandemic of that vaccination program."

Community acquired methicillin-resistant S aureus (MRSA) has emerged in recent years and makes a "bad combination" with flu, Jernigan added. "We saw that, but we didn't see it in the amount we thought we would—maybe because people are using more of the right antibiotics."

In terms of transmissibility, the experts agreed that the H1N1 virus seems to differ little from seasonal flu. However, Schaffner said it appears that children can shed the virus considerably longer than is the case with seasonal flu—for a week or a week and a half after recovery. In seasonal flu cases, it gets much harder to recover virus by 3 or 4 days after the onset of symptoms, he said.

Obesity emerges as risk factor
One thing that clinicians noticed about adult flu patients with serious H1N1 infections is that many were obese. "People were struck by the number of morbidly obese patients in their units," said Pavia.

It's an association that has not been prospectively studied, he said, but added, "I think it's real. I think it's probably a moderate-strength risk factor, and we were not in a position to notice it before."

"That is something we did find that is unique," Jernigan affirmed. "It came up early because of the kinds of individuals that were getting sick. They were otherwise health individuals, 30s to middle-aged, that were in ICUs and needed long-term respiratory support. It does look like obesity, especially morbid obesity, is a perhaps a significant risk factor."

A recent study by a CDC team using data from the early months of the pandemic showed that patients who were morbidly obese were nearly five times as likely to be hospitalized as were normal-weight patients. Their risk of death was also higher, but the difference was less dramatic.

The reason for the finding is unclear, but it may be simply that obesity is more common now than it once was, Jernigan commented. He also cited a recent study in which obese mice seemed to have poorer immunologic memory compared with lean mice.

Schaffner commented that the identification of obesity as a risk factor for complications "comes as no surprise to clinicians. They know that taking care of large people with lung infections is much more difficult than people who are lean," because of physical factors.

Danger for pregnant women
Another risk factor, albeit not a new one, that the pandemic highlighted was pregnancy. "One lesson was that flu is a dangerous disease for pregnant women. We knew it but it wasn't that prominent in our consciousness," said Pavia.

Reports from the first months of the pandemic suggested that pregnant women accounted for about 5% of H1N1 deaths, though they make up only about 1% of the US population. This finding echoed, though on a much smaller scale, what happened in the 1918 pandemic, which took a heavy toll on pregnant women.

Jernigan said there were indications early in the pandemic that obstetricians and gynecologists "weren't really thinking about flu as something that might harm their patients." The pandemic experience has raised their awareness of the risk, in his view. And messages about the risk to pregnant women pushed H1N1 vaccination rates for pregnant women well above the typical levels for seasonal flu shots, he added.

Other risk factors for serious H1N1 cases include asthma and, particularly in children, neurologic disorders, both of which are also seen in seasonal flu, Jernigan reported.

The pandemic brought a number of reports from various countries that racial and ethnic minorities seemed more susceptible to severe H1N1 illness. Jernigan said that pattern holds for both pandemic and seasonal flu: Majority and minority groups have about the same risk of contracting flu, but minorities are more likely to get seriously sick. The reasons are various and include access to care and vaccination, care-seeking behavior, and reluctance to be vaccinated.

Among patients who got severely sick, some showed signs of suffering a "cytokine storm," an overly intense immune response that floods the bloodstream with pro-inflammatory chemicals and causes the lungs to fill up with fluid, according to Pavia. The phenomenon is believed to have been a common feature in fatal cases in the 1918 pandemic.

"There were certainly patients who appeared to have the kind of catastrophic shock-like illness that's been blamed on cytokine storm," he said. "We don't as of today have a lot of evidence that those severe multi-organ system dysfunction patients had cytokine storm; it just resembles that and fits with the hypothesis."

Other lessons learned
One of the important lessons of the pandemic was the value of early antiviral treatment for patients with severe illness or at risk for severe illness, according to Pavia. He said it was hard to get this point across to clinicians. (A later story will look at the use of antivirals in the pandemic.)

Another lesson of the past year, in Pavia's mind, is the need for networks of clinical investigators prepared to swing into action when an outbreak starts.

"We did a pretty good job of getting clinical information quickly—the CDC, NIH, and other investigators stepped up to the plate, but we don't have an organized system to do that," he said. Clinical investigations "were cobbled together on the fly." They worked reasonably well, but systems set up in advance would be more effective, he added.

A more general lesson, according to Neuzil, is the need to avoid "tunnel vision" in pandemic preparedness, as exemplified by focusing heavily on the H5N1 threat. "One lesson for preparedness is for us to think very broadly and just look at multiple different scenarios," she said.

That was done to some extent, particularly regarding increased surveillance and the stockpiling and distribution of antivirals, Neuzil added.

Outlook for next season
One abiding mystery of the pandemic, experts agree, is why the 2009 H1N1 virus pushed the seasonal flu strains almost completely off the stage this past winter. That makes the question of what will happen next winter all the more interesting.

The disappearance of the seasonal strains "remains a fascinating mystery," said Pavia. "Everything we'd say would be pure speculation. We expected that H3N2 and B would've resurfaced this year, and it's hard to explain their absence based on cross-immunity related to [2009] H1N1." Some level of cross-immunity between the novel virus and the seasonal H1N1 probably accounts for the latter's exit, he added.

The experts expect the pandemic virus will be around again come winter, but they offer a range of views on which other flu viruses will be competing with it. The approaching flu season in the Southern Hemisphere will offer the first clues.

Racaniello expects that the pandemic strain will follow the pattern of past pandemics and dominate flu seasons for at least 10 years. "It'll probably be just H1N1 and B viruses; the H3N2 and the previous seasonal H1N1 will be gone," he said.

Pavia agreed that the pandemic H1N1, with or without drift mutations, will circulate next winter. "And I'd guess B will be back. But what will happen to H3N2 or H1N1 Brisbane [seasonal H1N1], I have no idea."

Schaffner said he, too, expects the 2009 H1N1 virus to be back. But because of the likely level of population immunity, he said, "My expectation is it'll be a minority player in 2010-11. I'm sticking my neck out here."

And Jernigan commented, "If we had to guess, I think we'd say seasonal H1N1 is not likely to hang around." He expects the pandemic strain to be the dominant player and to become the new seasonal H1N1. As for H3N2, which displaced H2N2 back in 1968, "I don't know exactly what will happen this fall, but I would guess that the H3N2 would actually persist. This is a guess."

See also:

CDC's latest estimates of cases, hospitalizations, and deaths
http://www.cdc.gov/h1n1flu/estimates_2009_h1n1.htm

CDC's Flu View surveillance update
http://www.cdc.gov/flu/weekly/

Mar 23 CIDRAP News story "Study: In life-years lost, H1N1 pandemic had sizable impact"

This week's top reads