The present study shows that elastic fibres in the airway wall can be visualized by FCFM, a novel bronchoscopic imaging modality, and that a laminar pattern of these fibres is associated with reduced lung function. The elastic fibres in histology and FCFM images exhibited 3 distinct patterns. There was good agreement in semi-quantitative pattern score between histology and FCFM, but there were no differences in such patterns between asthma patients and controls. These findings indicate that FCFM can be used to capture structural changes in the airway wall in humans in vivo, and might become a real-time imaging tool to estimate the type and degree of airway remodelling in chronic airway diseases such as asthma in the near future. Since FCFM has the ability to visualize the airway wall of the whole bronchial tree in vivo during bronchoscopy, it has complementary advantages as compared to taking snapshot biopsies at several locations.
There are some study data and case reports in literature examining the association between FCFM and histology of e.g. preinvasive bronchial lesions and sarcoidosis [13, 25]. However, to our knowledge this is the first study to investigate the histological substrate of autofluorescence in 488 nm FCFM images using in vivo human endobronchial biopsy specimens obtained from asthmatic patients and healthy control subjects with histology as 'gold standard'. The agreement of elastic fibre patterns between histology and FCFM is a novel finding and extends previous in vitro observations showing that bronchial mucosa and elastin powder extracted from human lung had similar autofluorescence spectra, suggesting that the autofluorescence in FCFM mainly originates from the elastic fibres present in the airway wall . We did not observe differences in elastic fibre patterns between asthmatics and controls. This contrasts previous studies using histology [11, 12], which is likely explained by differences in disease severity of the asthmatic patients, which in our study included mild disease.
The association between the elastic fibre patterns and lung function is a novel finding and adds to the validity of our histological and FCFM scoring. A plausible explanation for the lower FEV1 %predicted in the 'lamellar' group as compared to the other groups is that the parallel organisation of the thickened elastic fibres in a layer just beneath the epithelium changes airway wall mechanics and thereby FEV1. Airway wall mechanics are different in asthma as compared to controls , but additional ECM components e.g. collagen, proteoglycans, and glycoproteins are likely to contribute to this as well. Even though ECM may thicken the airway wall and thereby promoting luminal narrowing, any accompanying stiffening can stabilize the airways from collapse . It is still unknown how elastic fibre patterns could influence either of the above mechanisms. Hence, our current structure-function observations should be considered as hypothesis-generating.
The present results suggest that FCFM is an adequate method to examine bronchial elastic fibre morphology in vivo and might be an important tool to detect asthmatic patients who are prone to loss of lung function, at an early stage enabling timely intervention. Based on data found in literature we would expect that the degree of airway remodelling differs with asthma severity, including a varying amount or organization of elastic fibres [[10, 12, 28–30]]. Therefore, subsequent studies including larger numbers of subjects with varying asthma severity and FCFM images of multiple locations in the bronchial tree are needed. These will give a more detailed insight into the association between histology, FCFM, and airway function.
The strength of our study is that we applied strict patient selection criteria and that histology and FCFM were obtained from the same endobronchial sites. The bronchial main carina was chosen as the location for FCFM and biopsy to minimize imaging artefacts resulting from inadequate positioning of the miniprobe superimposed on the movement of the airways due to tidal breathing. However, there are potential limitations that need to be addressed. First, the number of 16 subjects was relatively low when using 3 semi-quantitative scores. Although this study was powered on kappa, power estimation based on this value is not firmly developed yet. This is due to the fact that this estimation is dependent on the kappa expected to be found and the marginal frequencies, which are the proportions of test subjects in each semi-quantitative category of histology and FCFM. Second, FCFM images during bronchoscopy were captured by placing the miniprobe perpendicularly to the surface of the airway main carina followed by a biopsy from the same location. It was technically impossible to orientate the small biopsy specimens in such a way that the cutting plane was identical to the plane of view during the FCFM recordings. This may have introduced bias in the semi-quantitative scoring of the histological sections. However, this bias seems to be limited as only the superficially located elastic fibres in the subepithelial layer were graded.
Our findings show a good agreement between pattern scores by histology and FCFM. The FCFM miniprobe has a fixed depth of view of 50 μm and therefore captures images at the level of the subepithelial layer. Accordingly, elastic fibre patterns in the subepithelial layer were scored in the histological sections. Pattern scores by histology and FCFM proved to be close in resemblance. By analysing both the autofluorescence patterns and the surrounding darker areas in FCFM images, this imaging modality also has the potential to visualize airway remodelling in general, which today is only possibly by histology of biopsy specimens. Other bronchoscopic real-time imaging modalities visualizing airway wall structures have recently been introduced including anatomical optical coherence tomography (aOCT) and endobronchial ultrasonography (EBUS) [[31–34]]. While aOCT and EBUS may visualize the different layers of the airway wall, FCFM can image a specific airway structural component in microscopic detail. Nevertheless, all three imaging techniques are not suitable to replace histology in the clinical setting yet. The technical part has to be further improved to acquire even higher quality images with minimal imaging artefacts.