Detection of Bacterial Colonization in Lung Transplant Recipients Using an Electronic Nose
Aim: To assess the diagnostic accuracy of electronic nose (eNose) technology to detect bacterial colonization in lung transplant recipients (LTRs).
Take home message: eNose technology shows high accuracy in non-invasively detecting bacterial colonization in lung transplant patients, offering a potential alternative to invasive diagnostics and enabling earlier intervention to improve patient outcomes.
Introduction
This study examines the use of electronic nose (eNose) technology to detect bacterial colonization (BC) in lung transplant recipients (LTRs). Bacterial colonization can lead to complications such as chronic lung allograft dysfunction (CLAD) and often requires invasive procedures for diagnosis. eNose technology analyzes volatile organic compounds (VOCs) in exhaled breath, offering a non-invasive alternative for early detection.
Methods
A cross-sectional analysis was conducted on 81 LTRs at Erasmus University Medical Centre, Netherlands. Patients provided exhaled breath samples analyzed by the SpiroNose®, which transmits sensor responses to the BreathBase® platform. Patients were classified as BC-positive if bacterial cultures showed consistent pathogen presence without acute infection symptoms. Sensor data were processed and assessed using machine learning algorithms, with performance evaluated by the area under the curve (AUC), sensitivity, and specificity.
Results
eNose analysis distinguished BC-positive from BC-negative patients with an AUC of 0.82 in the training set and 0.97 in the validation set. In the validation set, the model achieved 67% sensitivity, 100% specificity, and an accuracy of 93%, significantly exceeding the baseline classification rate. These results demonstrate high diagnostic accuracy and a strong potential for real-time detection of BC.
Discussion
This study highlights eNose technology as a promising tool for non-invasive BC detection in LTRs. By identifying distinct breathprint changes associated with BC, eNose technology could reduce the need for invasive diagnostics, enabling earlier intervention and potentially extending CLAD-free survival. Further validation in diverse cohorts will support the clinical application of eNose as a routine monitoring tool for LTRs.
