In Pneumonia’s Tug-of-War, Lung Microbiome Could Tip the Balance

Northwestern University scientists have potentially uncovered a previously unknown, hidden player in pneumonia.

In a new study, scientists found the lungs’ own microbial community, or microbiome, appears to influence how the illness evolves, who responds well to treatment and whether a patient will recover successfully or continue to deteriorate.

Using lung samples from pneumonia patients, the team tracked how microbial ecosystems and immune responses evolved over time. Among their findings, they discovered patients most likely to recover shared two characteristics: Their lung microbiomes resembled oral microbiomes, and their microbial communities were dynamic rather than stable.

The findings eventually could help physicians predict patient outcomes, tailor personalized antibiotic treatment plans and develop therapies that nurture beneficial microbes in the lungs. It also could improve understanding of the elusive disease, which claims tens of thousands of lives annually in the United States, according to the US Centers for Disease Control and Prevention (CDC).

The study was published Dec. 10 in the journal Cell Host & Microbe.

“Most people are familiar with the gut microbiome or skin microbiome but are surprised to learn the respiratory tract also has a microbiome,” said Northwestern Engineering’s Erica Hartmann, who led the study. “For a long time, people actually thought the lungs were sterile and microbes were present only during an infection. It turns out that’s not the case. We wondered if the microbiome might help explain why some pneumonia patients respond to treatment and others do not. Ultimately, we hope this leads to better diagnostics and improved patient outcomes.”

An expert in microbiomes, Hartmann is an associate professor of civil and environmental engineering at the McCormick School of Engineering. The study was conducted in collaboration with Northwestern’s Successful Clinical Response in Pneumonia Therapy (SCRIPT) project. Study coauthor Richard Wunderink is a professor of medicine at Northwestern University Feinberg School of Medicine and principal investigator of the SCRIPT project.

Pneumonia’s innate unpredictability

Each year, pneumonia sends roughly 1.2 million people to emergency departments in the US, according to the CDC. Despite its prevalence, pneumonia remains surprisingly difficult to predict and treat. Even if two patients have the same diagnosis and receive the same antibiotic, they can have vastly different outcomes.

“Pneumonia is defined by its symptoms, not by its cause,” Hartmann said. “There is a huge proportion of pneumonia patients for which doctors can’t tell if it’s bacterial, viral, or fungal. Hospital-acquired pneumonia and community-acquired pneumonia also are quite different. Depending on the type of bacterial infection the patient has, the antibiotics may or may not be effective.” 

“This unpredictability has significantly hampered research efforts to understand pneumonia pathogenesis,” Wunderink said. “For way too long, we have used the 19th-century tool of bacterial cultures to study an important 21st-century problem. Sequencing data like this will allow greater understanding of how patients get pneumonia, what microbes are actually causing pneumonia, and ultimately, what pathogen is causing the pneumonia in the patient I am caring for right now.”

Identifying four distinct ‘pneumotypes’

To better understand the illness, Hartmann and her collaborators aimed to identify the microbes present in pneumonia patients. Working with Wunderink at Northwestern Medicine, the scientists collected multiple lung samples from more than 200 critically ill pneumonia patients in hospitals’ intensive care units. Then, they identified the microbes within the samples and measured how many bacteria were present.

After tracking the microbiomes over time, the team identified four distinct microbial patterns, or “pneumotypes,” associated with different types of pneumonia, including community-acquired, hospital-acquired, and ventilator-acquired. Patients’ lungs were either dominated by microbes typically found in the mouth, on the skin or a mix of both. The fourth pneumotype was dominated by common pathogen Staphylococcus aureus.

Hartmann and her team discovered that the lung microbiomes and host’s immune response were intertwined and changed together. They also found that patients with oral-like pneumotypes were more likely to recover successfully. Skin-like and mixed pneumotypes were not clearly associated with recovery but also not clearly associated with decline. And the patients with Staphylococcus-dominated pneumotypes tended to have the worst outcomes.

“We’re still trying to understand what this means,” Hartmann said. “One speculative hypothesis is that the lungs already have constant exposure to oral-like microbes. The upper respiratory tract includes the mouth and throat, so saliva is constantly moving down and getting coughed back up. The immune system might already be adapted to those oral-like microbes, so it knows how to respond when it encounters them.”

Shifts linked to success

The scientists also found that the worst outcomes were associated with the most stable lung microbiomes.

“Lungs are like any other ecosystem,” Hartmann said. “When an ecosystem is perturbed, it shifts. Those shifts might give it the potential to kick out a pathogen. But if the community is too stable, then it might not be flexible enough to defend itself. Again, though, we don’t really know, so this is all highly speculative.”

To help confirm these speculations, Hartmann and her collaborators plan to conduct experiments in cellular cultures. The team could only obtain lung samples through bronchoscopy, which requires patients to already use ventilators. Although the experiments included ventilated patients without pneumonia as controls, it could not include healthy controls.

“Going forward, we want to culture these organisms and put them in a flask together to see how they interact,” Hartmann said. “But it does seem that the microbial communities and different pneumotypes do matter. And whether or not that pneumotype remains stable also matters. And that’s fascinating.”