There is no easy way to determine which type of bug is causing a patient's illness. It's hard for doctors to choose effective treatments because the antibiotics commonly used to treatbacterial pneumonia won't help patients with viral pneumonia. Limits on the use of antibiotics is a step in the right direction.
MIT researchers have designed a sensor that can distinguish between the two types of pneumonia.
Even with the most extensive and advanced testing, the specific pathogen causing someone's disease can't be identified in about half of patients. If you treat a viral pneumonia with antibiotics, you could be contributing to antibiotic resistance, which is a big problem, and the patient won't get better.
The researchers used a urine test to read the results of their study on mice.
The study will be published this week in the Proceedings of the National Academy of Sciences. The lead author is a graduate student.
There are signatures of infections.
It has been difficult to distinguish between the two types of pneumonia due to the fact that there are so many different types of infections.
The researchers decided to focus on measuring the host's response to infections rather than trying to detect the pathogen. There are different types of immune responses caused by viral andbacterial infections. According to the MIT team, the pattern of activity of those enzymes can be a sign of a disease.
Many of the proteases in the human genome are used by cells that respond to infections. A team led by Purvesh Khatri, an associate professor of medicine and biomedical data science, collected 33 publicly available datasets of genes that are expressed during respiratory infections. 39 proteases were identified that appear to respond differently to different types of infections.
The data was used to create 20 different sensors. The sensors have a coating on them that can be cleaved. When the peptides are cleaved, they are labeled with a reporter molecule that is freed. The reporters excrete in the urine. Mass spectrometry can be used to determine which proteases are most active in the lung.
Five different mouse models of pneumonia were tested by the researchers with their sensors.
The researchers used machine learning to analyze the results from the urine tests. Using this approach, they were able to train a program that could distinguish between pneumonia and healthy controls, and also distinguish between an infectious disease and a benign one.
The researchers found that their sensors were able to distinguish between the five pathogens they tested, but with less accuracy than the test to distinguish between the two. One possibility the researchers may pursue is developing a system that can distinguish between the two types of infections and help doctors choose the best antibiotic to fight them.
Future detection with a paper strip is similar to a pregnancy test and would allow for point-of-care diagnoses. The researchers were able to identify a subset of five sensors that could be used to test at home. More work needs to be done to determine if the reduced panel would work in humans with more genetic and clinical variability than mice.
There are patterns of responses.
The researchers were able to identify some patterns of host response to different types of infections. It was expected that the proteases produced by the neutrophils would be more visible in mice withbacterial infections.
T cells and NK cells respond to viral infections in a different way. The strongest signal came from one of the sensors that was linked to the granzyme B. The researchers found that the sensor was activated in the lungs of the mice with the infections.
To deliver the sensors in mice, the researchers injected them directly into the trachea, but they are now developing versions for human use that would be similar to an asthma inhaler. A way to detect the results using a breathalyzer instead of a urine test is being worked on.
More information: Host protease activity classifies pneumonia etiology, Proceedings of the National Academy of Sciences (2022). DOI: 10.1073/pnas.2121778119. Journal information: Proceedings of the National Academy of Sciences
All Rights Reserved © 2024
