Many types of bacteria and viruses can cause pneumonia, but there is no easy way to determine which microbe causes a particular disease in a particular patient. This uncertainty makes it difficult for physicians to choose effective treatment, as antibiotics widely used to treat bacterial pneumonia do not help patients with viral pneumonia. In addition, limiting the use of antibiotics is an important step in curbing antibiotic resistance.
MIT researchers have now developed a sensor that can distinguish between viral and bacterial pneumonia infections, which they hope will help doctors choose the right treatment.
“The problem is that there are so many different pathogens that can cause different types of pneumonia, and even the most extensive and advanced testing cannot identify the specific pathogen that causes someone’s disease in half of the patients. If you treat viral pneumonia with antibiotics, you can contribute to antibiotic resistance, which is a big problem and the patient will not get better, ”said Sangita Bhatia, John and Dorothy Wilson, professor of health sciences and technology at the Massachusetts Institute of Computer Science and Technology. specialist and a member of MIT’s Koch Integrated Oncology Research Institute and the Institute of Medical Engineering and Science.
In a study of mice, the researchers showed that their sensors could accurately distinguish between bacterial and viral pneumonia within two hours, using a simple urine test to read the results.
Bhatia is the senior author of this week’s study Proceedings of the National Academy of Sciences. Melody Anahtar ’16, PhD ’22 is the main author of the article.
Signs of infection
One of the reasons why it is difficult to distinguish between viral and bacterial pneumonia is that there are many microbes that can cause pneumonia, including bacteria. Streptococcus pneumoniae the and Haemophilus influenzaeand viruses such as influenza and respiratory syncytial virus (RSV).
In developing the sensor, the research team decided to focus on measuring the host’s response to infection, rather than trying to identify the pathogen itself. Viral and bacterial infections trigger different types of immune responses, including the activation of enzymes called proteases, which break down proteins. The MIT team found that a pattern of enzyme activity could serve as a sign of a bacterial or viral infection.
The human genome encodes more than 500 proteases, many of which are used by cells that respond to infection, including T cells, neutrophils, and natural killer (NK) cells. A team led by Purvesh Khatri, an associate professor of medicine and biomedical information at Stanford University and one of the authors of the paper, collected 33 publicly available data sets of genes identified during respiratory infections. By analyzing those data, Khatri was able to identify 39 proteases that respond differently to different types of infection.
Bhatia and his students then used the data to create 20 different sensors that could interact with those proteases. Sensors are made up of nanoparticles coated with peptides, which are broken down by specific proteases. Each peptide is identified by a reporter molecule, which is released when the peptides are separated by proteases raised in the infection. Those messengers eventually pass urine. Urine mass spectrometry can be used to determine which proteases are most active in the lungs.
The researchers tested the sensors on five different mouse models of pneumonia caused by infections. Streptococcus pneumoniae, Klebsiella pneumoniae, Haemophilus influenzaeinfluenza virus and mouse pneumonia virus.
After reading the results of urine tests, the researchers used machine learning to analyze the data. Using this technique, they were able to teach algorithms to differentiate healthy people from pneumonia, as well as to distinguish whether the infection was viral or bacterial based on those 20 sensors.
The researchers also found that their sensors could detect five pathogens tested, but with less accuracy than tests to distinguish between viruses and bacteria. One of the possibilities for researchers is to develop algorithms that can not only differentiate bacterial infections from viral infections, but also identify the class of microbes that cause bacterial infections, which will help doctors choose the best antibiotic to fight this type of bacteria.
Urine-based reading can also be identified in the future with a strip of paper similar to a pregnancy test, which allows you to make a diagnosis at the point of care. To this end, the researchers identified some of the five sensors that will bring home testing closer to reach. However, more work is needed to determine whether the abbreviated panel has better genetic and clinical variability in humans than in mice.
In their study, the researchers also identified some patterns of response to different types of infection. Mice with bacterial infections were more likely to have proteases secreted by neutrophils, which is to be expected because neutrophils are more likely to respond to bacterial infections than to viral ones.
Viral infections, on the other hand, produced protease activity from T cells and NK cells, which are usually more responsive to viral infections. One of the sensors that generates the strongest signals is associated with a protease called granism B, which triggers programmed cell death. The researchers found that this sensor was highly activated in the lungs of mice with viral infections, and that NK and T cells were involved in the response.
To deliver the sensors to the mice, the researchers injected them directly into the trachea, but they are now developing versions for human use, which can be controlled or a nebulizer similar to an asthma inhaler can be used. They are also working on a way to determine the results with the help of breathalyzers instead of urine tests, which can give even faster results.
The study was funded in part by the Bill and Melinda Gates Foundation, the Janssen Research and Development Foundation, the Koch Institute Support (major) grant from the National Cancer Institute, and the National Institute of Environmental Sciences.