Optical Diagnosis for Preneoplasia, the Search Continues

Optical Diagnosis for Preneoplasia, the Search Continues In the study by Nguyen et al1 published in this issue of the Journal of Bronchology & Interventional Pulmonology, the investigators sought to determine if preneoplastic lesions of the bronchus and larynx could be better discriminated by combining autofluorescence with narrow band imaging (NBI), and if real-time computerized analysis of NBI images could allow for optical diagnosis in the future. Three cases were shown that compared carcinoma in situ (CIS) with normal tissue, CIS with postradiation tissue change, and severe dysplasia with an earlier biopsy scar. Airway inflammation, postradiation, and postelectrosurgery tissue changes and earlier biopsy scars are examples that contribute to the false-positive rate of autofluorescence bronchoscopy (AFB). NBI technology uses 2 wavelengths of light; blue light (390 to 440 nm), which is absorbed by superficial capillaries, and green light (530 to 550 nm), which is absorbed by the blood vessels beneath the mucosa. These improve visualization of abnormal angiogenesis that occurs in preneoplastic and cancerous lesions. Shibuya et al2 showed that dotted vessels detected on NBI correlated with angiogenic squamous dysplasia, which is a recognized precursor of early squamous carcinoma. Herth et al3 also reported that NBI was more specific than AFB in the detection of airway preneoplasia, without compromising its sensitivity. Although NBI aided the bronchoscopists in the choice of site for biopsy and demarcated the lesion for local bronchoscopic therapy, the investigators failed to highlight that a change of bronchoscopes from AFB to NBI was necessary for combined imaging. Moreover, careful documentation of airway sites with abnormal fluorescence was necessary, which could lead to longer procedural time, greater patient discomfort, and the need for additional sedation that might compromise patient safety. Our study showed that dual imaging with video and AFB (SAFE 3000, Pentax, Japan) was not only sensitive for preneoplasia (0.86) but also specific (0.94) and discriminatory for airway inflammation, fibrosis, and earlier biopsy scars, as simultaneous display of video and AF images of the lesion allows precise visual assessment by providing both functional and anatomic information. The time taken for airway inspection and biopsy was 9 minutes, comparable to white light bronchoscopy, whereas sequential white light bronchoscopy and AFB with the same bronchoscope added 5 to 13.8 minutes to the standard procedural time.4 Computerized analysis of NBI images using color plots to further aid the bronchoscopist in discriminating high-grade dysplasia from normal airway mucosa is novel, but an overlap between normal and dysplastic/CIS lesions was observed in the cases studied. Notwithstanding these represented preliminary results, clinicopathologic correlation and validation are required before clinical application. Moreover, color plots representative of normal airway, inflammation, fibrosis, and varying grades of dyplasia have to be analyzed and made available in real time during bronchoscopy for optical diagnosis to become a reality. My colleagues and I have recently published our experience of using the Onco-LIFE device (Novadaq Technologies Corp, Canada), which allows composite quantification of red reflectance and green fluorescence intensity signals by expressing numerically the red to green ratio (R/G ratio) of the area of interest. A receiver-operating curve based on 3362 adequate biopsies with their corresponding R/G ratios from 738 patients was generated. A derived R/G ratio of 0.54 or more correlated with moderate dysplasia or worse, which was validated by a prospective study (Fig. 1). We concluded that color fluorescence ratio could serve as an objective method to guide biopsy but was premature for optical diagnosis.5 The search continues.FIGURE 1.: Color fluorescence ratio (R/G ratio) of carncinoma in situ. White light image shows suspicious focal mucosal thickening of LB6. Autofluorescence image shows abnormal fluorescence with distinct margin of LB6. R/G ratio of target within the brackets was 1.5, derived by dividing the average red reflectance with green fluorescence signals.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.