Bacterial superinfection may be tied to pneumonia hospitalizations in African children

May 2005

NEW YORK — Many children who are admitted to the hospital for viral pneumonia may have a bacterial superinfection, Keith P. Klugman, MD, PhD, said here at the 17th Annual Infectious Diseases in Children Symposium.

This would have important treatment implications, since antibiotics are not normally prescribed for viral pneumonias, and would be especially important during a pandemic or epidemic of influenza.

In a study in South Africa of clinical and radiologically confirmed pneumonia in children with confirmed viral infections, a much larger portion of pneumonia was found to be due to Streptococcus pneumoniae than researchers had expected, according to Klugman, who is with the respiratory and meningeal pathogens research unit in Johannesburg, South Africa and also with the department of international health and division of infectious disease at Emory University in Atlanta.

“Now, I’m not saying that every child with influenza needs antibiotics to stop the bacterial pneumonia. It may be 1 in 1,000 kids who have influenza who end up in the hospital with pneumonia. But what I am saying is that if a child ends up in hospital with pneumonia, even if you’re convinced it’s viral, that kid needs antibiotics because the risk of a bacterial superinfection in the severely ill child with pneumonia is overwhelming according to our data.”

In a very large trial of more than 40,000 children, Klugman is testing the nine-valent conjugate pneumococcal vaccine in South Africa. The vaccine includes serotypes 1 and 5, which are important causes of invasive disease outside the United States.

They looked at invasive pneumococcal disease in both HIV-negative and -positive children because HIV in Africa is associated with a great burden of invasive disease, Klugman explained.

Although the vaccine was efficacious in children who did not have HIV, it also had efficacy of around 65% in children with HIV.

Elusive cause

The cause of pneumonia can be difficult to determine, he said. “To this day, we still have a large fraction of pneumonia where we don’t know the etiology,” he said. In studies, the conjugate vaccine can be used to extrapolate the etiology of pneumonia because the vaccine can serve as a probe to define the fraction of pneumonia caused by the pneumococcus, according to Klugman.

“The background to all of this is Haemophilus influenzae pneumonias. The Hib vaccine trial against pneumonia was conducted in the Gambia in Africa and it showed a 21% reduction in all radiologically proven pneumonia. Now, based on blood cultures, a small fraction of blood cultures in kids with pneumonia would have grown Hib in the pre-vaccine era, but this study suggested that a large fraction of the pneumonia that was killing kids was due to Hib because, in a double-blind trial, the kids who got the vaccine had 21% less x-ray confirmed pneumonia,” he explained.

Researchers are taking a similar approach with the pneumococcal conjugate vaccine. In a study of the seven-valent vaccine approved in the United States (Prevnar, Wyeth), pneumonia was not a primary endpoint of the pivotal study by Steve Black, MD, and colleagues, but they reviewed the Kaiser database to see how many kids in the trial had a diagnosis of pneumonia. “They found a small, non-significant drop in pneumonia, when there was any clinical pneumonia that was simply a diagnosis stated on the chart, but when a radiograph was obtained, there was a 10% reduction in the kids who got the conjugate vaccine. And if there was consolidation on x-ray, there was a 20% reduction in that endpoint in vaccinated children compared to controls,” Klugman said. The study by Klugman and colleagues in South Africa did define pneumonia as a primary endpoint.

Using a WHO definition of alveolar consolidation on x-ray, they found between a 20% and 25% reduction in children who did not have co-infection with HIV. “We were concerned that in HIV-infected kids the etiology of pneumonia is so broad that that kind of x-ray endpoint would not be specific enough.”

They saw about 10% to 12% reduction in children with HIV.

“We wanted to see whether we could use the vaccine to define a larger burden of disease that could be prevented by a conjugate vaccine,” he said.

“You can use an x-ray as a definition of pneumococcal pneumonia, but there’s a lot of pneumococcal pneumonia that is not defined by consolidation on x-ray,” he said.

So, they chose a wider definition: cough, fast breathing and chest wall indrawing plus either consolidation on x-ray, or plus bronchial breath sounds and/or rhonchi. Wheeze in the absence of x-ray consolidation was considered bronchiolitis and was excluded, he said.

Bacterial superinfection

Klugman said there is strong evidence to support the idea that much of what is considered respiratory syncytial virus (RSV) or influenza-associated pneumonia may be due to bacterial superinfection.

His vaccine trial found that the vaccine prevented 44% of pneumonias related to influenza and parainfluenza and a 21% reduction of pneumonia associated with RSV. There was an overall 31% reduction in pneumonia with any viral etiology. “So, here we have a pneumococcal conjugate vaccine preventing between 30% and 45% of what is assumed to be viral pneumonia,” he said.

“We don’t think it’s prevention of the viral pneumonia at all. The virus is there, and the virus is clearly playing a critical role in getting this kid into hospital with pneumonia, but it’s not the only thing there. There’s clearly a pneumococcal superinfection that is landing the kid in the hospital with pneumonia,” Klugman explained.

Klugman said he thinks it may work this way: a child who is colonized in the nasopharynx with a pneumococcus, probably newly colonized, gets influenza or RSV. That viral infection disturbs the nasal flora and allows the pneumococcus to invade the respiratory tract, giving rise to pneumonia.

“These data suggest that a significant fraction of influenza and other viral morbidity may be due to the pneumococcus and the good news of course is that it is also preventable by this conjugate pneumococcal vaccine,” Klugman said.

For more information:

• Klugman KP. Prevention of pneumococcal pneumonia. Presented at the 17th Annual Infectious Diseases in Children Symposium. Nov. 20-21, 2004. New York.
• Madhi SA, Klugman KP. Vaccine trialist group. A role for Streptococcus pneumoniae in virus-associated pneumonia. Nat Med. 2004;10(8):811-813. Epub 2004 Jul 11.
• Klugman KP, Madhi SA, Huebner RE, et al. For the vaccine trialists group A trial of a 9-valent pneumococcal conjugate vaccine in children with and those without HIV infection. New Engl J Med. 2003;349(14):1341-1348.
• Madhi SA, Heera JR, Kuwanda L, et al. Use of procalcitonin and C-reactive protein to evaluate vaccine efficacy against pneumonia. PLOS Med. 2005;2(2):147-151.
• Black SB, Shinefield HR, Ling S, et al. Effectiveness of heptavalent pneumococcal conjugate vaccine in children younger than five years of age for prevention or pneumonia. Pediatri Infect Dis J. 2002;21:(9)810-815.