Resolving uncertainty in acute respiratory illness using optical molecular imaging

Ventilator associated pneumonia (VAP) and acute respiratory distress syndrome (ARDS) are two respiratory conditions unique to mechanically ventilated patients. The diagnosis of these conditions, and therefore any subsequent treatment, are befuddled by uncertainty. VAP rates vary considerably according to the diagnostic or surveillance criteria used. The pathogenesis of ARDS is well understood but when the internationally agreed consensus criteria are employed, the histological hallmarks are absent about half the time, indicating a disconnection between the clinical diagnosis and what is known about the biology of this condition. It is argued that tests of biological function should be considered in addition to clinical characteristics in order to improve the utility of diagnosis. Given that the pathological sequelae of both VAP and ARDS are driven by an over exuberant host neutrophil response, the activated neutrophil was selected as a potential biological imaging target. Optical molecular imaging uses visible and near visible wavelengths from the electromagnetic spectrum to derive or visualize information based on the optical properties of the target tissue. Optical wavelengths are safe and cheap to work with, producing much higher resolution images than those relying on x-rays or gamma radiation. The imaging modality can be coupled with exogenously applied chemistry to identify specific biological targets or processes. The hypothesis that optical molecular imaging could be used to detect activated neutrophils in real time in the alveolar region of patients was tested. A bespoke optical molecular imaging agent called Neutrophil Activation Probe (NAP), designed in-house, was used to test the hypothesis. NAP is a dendrimeric compound delivered to the alveolar region of a patient in microdoses (≤100 micrograms), becoming fluorescent only on contact with activated neutrophils, and can be detected by optical endomicroscopy. Both the imaging agent and the endomicroscope are delivered to the distal lung via routine bronchoscopy. The agent was tested extensively in the laboratory to demonstrate function, specificity, and safety. Ex vivo testing took place using human and ovine lungs. A regulated dose escalation Phase I clinical trial of investigational medicinal product (CTIMP) in healthy volunteers, patients with bronchiectasis, and mechanically ventilated patients with a pulmonary infiltrate on chest radiography (NCT01532024) was designed and conducted. The aim of the Phase I study was to demonstrate the safety of the technique and to confirm proof of concept. In order to support the requirement for a technique that interrogates alveolar neutrophils two supplementary clinical studies were performed. Firstly, two VAP surveillance techniques (CDC surveillance and HELICS European VAP surveillance) were compared with clinically diagnosed VAP across consecutive admissions in two large tertiary centres for one year. Secondly, the utility of circulating neutrophils to permit discrimination between acute respiratory illnesses was examined. Blood samples from mechanically ventilated patients with and without ARDS underwent flow cytometric assessment using eight clusters of differentiation and internal markers of activation to determine neutrophil phenotype. All clinical studies received the appropriate regulatory, ethical, and/or Caldicott guardian approval prior to commencement. NAP became fluorescent only in the presence of three processes specific to neutrophil activation: active pinocytosis, progressive alkalinization of the phagolysosome, and the activity of human neutrophil elastase. High optical signal was detected following the application of NAP in the alveolar regions of explanted lungs from patients with cystic fibrosis, known to be rich in activated neutrophils. Using an ex vivo ovine lung ventilation and perfusion model optical signal was demonstrated following segmental lung injury. The safety and specificity of the technique in a small cohort of healthy volunteers and mechanically ventilated patients was demonstrated. The technique was tested on a small cohort of patients with bronchiectasis, which provided the first opportunity to obtain broncho-alveolar lavage samples for laboratory correlation. Fluorescent signal was shown in the lavaged neutrophils, labeling that could only have taken place in the alveolar region. The supportive clinical studies found the concordance between actual VAP events was virtually zero even though the reported VAP rates were similar. Furthermore, the rate at which clinicians initiate antibiotics for VAP was approximately five times higher than either surveillance VAP rate. The study of circulating neutrophils from the blood of healthy volunteers and mechanically ventilated patients with and without ARDS indicated circulating neutrophil activation phenotype was not capable of discriminating between clinically diagnosed ARDS and other acute respiratory illnesses. In summary, an ambitious programme of work was completed to develop and support an optical molecular imaging technique that meets the rigorous requirements for human application and can be applied at the bedside to yield immediate visual results. The spatiotemporal relationship of neutrophil activation in real time both in the laboratory and in volunteers and patients was visualized. The visualization of neutrophil activation at such a resolution has never been achieved before in humans, healthy or unhealthy. The Phase I study was not powered to determine utility but recruitment has begun to a Phase II CTIMP (NCT02804854) to investigate the utility, accuracy, and precision of the imaging technique in a large cohort of mechanically ventilated patients. Ultimately, it is proposed that the technique will facilitate diagnosis, stratify patients for treatment and monitor treatment response using this technique