During the 2009 H1N1 influenza pandemic, many previously healthy children became critically ill, developing severe pneumonia and respiratory failure, sometimes fatal. The largest nationwide investigation to date of influenza in critically ill children, led by Children's Hospital Boston, found one key risk factor: Simultaneous infection with methicillin-resistant Staphylococcus aureus (MRSA) increased the risk for flu-related mortality 8-fold among previously healthy children.
Moreover, almost all of these co-infected children were rapidly treated with vancomycin, considered to be appropriate treatment for MRSA. The fact that they died despite this treatment is especially alarming given the rising rates of MRSA carriage among children in the community.
"There's more risk for MRSA to become invasive in the presence of flu or other viruses," says study leader Adrienne Randolph, MD, MsC, of the Division of Critical Care Medicine at Children's Hospital Boston. "These deaths in co-infected children are a warning sign."
The researchers hope their findings, published November 7 by the journal Pediatrics, (eFirst pages) will promote flu vaccination among all children aged 6 months and older. (No flu vaccine is currently available for children younger than 6 months.)
"The 2009 H1N1 virus has not changed significantly to date," notes Tim Uyeki, M.D., MPH, of the Influenza Division of the National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention (CDC), a senior investigator on the study. "Infections of children in the U.S. with 2009 H1N1 virus are expected this season and need to be prevented and treated appropriately. Influenza vaccination protects against 2009 H1N1 illness."
With emergency funding from the National Institutes of Health, Randolph and her colleagues in the Pediatric Acute Lung Injury and Sepsis Investigator's Network tracked 838 children admitted to 35 pediatric ICUs around the country with probable 2009 H1N1 influenza from April 2009 to April 2010. Their vaccination status wasn't consistently known, but H1N1 vaccine did not become available until September 2009 or later.
The median age of the children critically ill with H1N1 was 6 years. Most had respiratory failure, two thirds required mechanical ventilation, and some required extracorporeal membrane oxygenation (ECMO) for advanced cardiac and respiratory support. Their disease progressed rapidly, and 75 children (9 percent) died, two thirds of them within two weeks of ICU admission.
"Some children were quickly overwhelmed, and many died despite centers doing everything to save them," says Randolph. "Early in the pandemic, centers were worried that they would run out of ventilators, that they would run out of ICU beds."
While most of the children critically ill with H1N1 had one or more chronic health conditions that increased their risk, such as asthma, neurologic disorders or compromised immune systems, 251 children (30 percent) were previously healthy.
Among these otherwise healthy children, the only risk factor that was identified for death from influenza was a presumed diagnosis of MRSA co-infection in the lung – which increased the risk for mortality 8-fold (P<0.0001).
"It is not common in the U.S. to lose a previously healthy child to pneumonia," says Randolph. "Unfortunately, these children had necrotizing pneumonia – eating away at their tissue and killing off whole areas of the lung. They looked like immunocompromised patients in the way MRSA went through their body. It's not that flu alone can't kill – it can – but in most cases children with flu alone survived."
Treatment challenges
Use of antiviral agents for critically ill patients with influenza is now routinely recommended. In this study, 88 percent of the children admitted to the ICU received Tamiflu (oseltamivir) during their stay, but only 6 percent had received it prior to hospital admission. Randolph believes it's possible that earlier antiviral treatment might have saved some of these children, especially those who died of influenza without bacterial co-infection. (Influenza antiviral medications work best when given within the first two days of symptom onset.)
The study also found that most of the MRSA co-infected children who died had received vancomycin promptly at or before ICU admission. The researchers aren't sure why this antibiotic of choice didn't save these children – perhaps it couldn't penetrate the lung, or perhaps the disease moved too rapidly.
Physicians seeing children with serious lower-respiratory-tract disease during flu season are urged to give early antiviral treatment (Tamiflu or zanamivir [Relenza]) and antibiotics covering MRSA and other flu-associated bacteria, even before suspected infections are confirmed in the lab, the researchers say.
But other approaches are urgently needed. "MRSA is hard to develop a vaccine against – researchers have been trying since the 1960s and have been unsuccessful," says Randolph. "So the only way to prevent these severe complications is to get everyone vaccinated against the flu, and do more studies of MRSA colonization so we can prevent it in the community and in kids."
A growing threat
Recent studies point to a rise in the number of children carrying MRSA. A study in Pediatrics in 2010 found that the number of children hospitalized for MRSA infection increased from 2 in 1,000 admissions in 1999 to 21 in 1,000 admissions in 2008. This rise was attributed mostly to community-acquired cases – not cases acquired in the hospital.
Many observers attribute the rise to increased use of antibiotics in both people and animals. "The more antibiotics we take, the more we colonize ourselves with antibiotic-resistant organisms such as MRSA," says Randolph.
Influenza appears to suppress the immune response, making children who are already colonized more susceptible to invasive bacterial disease. "Previously, MRSA has not been considered a common cause of pneumonia in kids but this may be changing," Randolph says. "It's likely that flu and other viral infections let MRSA invade and that there's some synergistic reaction between flu and these bacteria."
While recent data show an increase in MRSA co-infection in children dying from seasonal influenza, this is the first study to collect data on a large number of children with no risk factors for severe flu.
Severe pregnancy complications reduce likelihood of having more children