Of pattern-recognition receptors and „machine learning“

CAPNETZ STIFTUNG supports research on community-acquired pneumonia (CAP) and other lower respiratory tract infections through the CAPNETZ database and CAPNETZ biomaterials. As examples of research projects conducted based on the CAPNETZ cohort the following recent publications are presented:

Pattern-recognition receptors with reduced sensitivity to Legionella infection in CAPNETZ patients

Legionella pneumophila is among the common CAP-causing pathogens in the clinical sector and has a mortality rate of up to 34 %. Ruiz-Moreno et al¹ investigated the underlying molecular mechanisms and found a genetic marker in the pattern-recognition receptor cGAS-STING. 

Pattern-recognition receptors are essential for the cellular immune response as they detect molecular patterns of bacteria, viruses, or parasites and elicit the secretion of signaling molecules of the immune response such as cytokines and interferons.

Interestingly, many CAPNETZ patients that were infected with Legionella pneumophila had a specific genetic variant of cGAS-STING called HAQ which exhibits lower sensitivity for molecular patterns of Legionella pneumophila as compared to more common variants of cGAS-STING. As a result, the immune reaction of carriers of HAQ cGAS-STING, which is found in around 2-3% of Europeans, can be impaired.

While the study investigated cases of Legionella pneumophila infection, the authors note that HAQ cGAS-STING could also be relevant in the response to other pathogens.

Macrolides in CAP treatment: yes or no? Machine learning guides decision making

Moderate and severe cases of CAP have a high mortality rate. To reduce the risk of death or complications, a prompt medication is absolutely crucial. Therefore, treatment with antibiotics is initiated before the pathogen is determined, and a combination of antibiotics is administered to cover a broad range of potential pathogens. The combination of beta-lactam antibiotics with macrolides is often the therapy of choice due to broad effectivity. However, macrolides have been reported to interact with other drugs and can have side effects which are likely associated to pre-existing health conditions.

Koenig et al² screened clinical parameters recorded for CAPNETZ patients which received or did not receive macrolides and developed an algorithm, which connects patient data to clinical outcomes. Using machine learning certain parameters could be identified that were predominantly associated to successful application of macrolide therapy. The researchers concluded that in absence of chronic cardiovascular and respiratory conditions combined with the lack of increase in leukocyte counts in respiratory secretions, macrolide treatment decreased the mortality rate by 27%, as compared to standard treatment. Conversely, patients with cardiovascular and pulmonary comorbidities combined with aberrant leukocyte counts should not receive macrolide treatment. 

Ultimately, the approach identified 3 simple anamnestic questions which could guide clinicians in patient stratification. However, the findings still require evidence via a controlled clinical study.

CAPNETZ STIFTUNG: Data exploitation in over 200 publications

Research projects such as the above would not have been possible without CAPNETZ STIFTUNG. By granting access to the vast databank and biomaterial bank from over 20 years of CAPNETZ study, more than 200 scientific papers, reviews, and other articles were published and complemented by additional German-language medical and scientific communications such as a designated column in the magazine Pneumologie

Researchers in Germany, Europe, and beyond can apply for access to CAPNETZ data and biomaterials to contribute to our growing network of CAP research.

¹ Quelle: The common HAQ STING variant impairs cGAS-dependent antibacterial responses and is associated with susceptibility to Legionnaires’ disease in humans – PubMed (nih.gov)

² https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1006829