Narrow Band Imaging in the identification and monitoring of oral potentially malignant diseases and oral cancer

Abstract

Oral cancer ranks as the sixth most common cancer in the world, and despite advances in surgical techniques over the past 30 years, has a five year survival rate of only 50%. The poor survival rate is partly attributed to the fact that the majority of patients are diagnosed at advanced stages of the disease. Early detection and intervention is therefore important for improving patient prognosis as invasive and disfiguring treatment that has high morbidity and mortality can be avoided. Ideally, oral potentially malignant disorders (OPMDs) are detected and monitored before they can progress to cancer. However, OPMDs can be difficult to detect using standard operatory lights, as some lesions may have only very subtle changes in colour and texture. This has led to the development and use of various visualisation adjuncts that claim to improve the clinician’s ability to detect lesions. One visualisation adjunct that can be used intraorally is Narrow Band Imaging (NBI), which is an endoscopic technique that utilises filtered blue and green light to highlight the mucosal surface texture and underlying vasculature. The ability to enhance vasculature makes NBI a useful technology for highlighting neoplastic lesions that may not be clinically evident with white light alone, as angiogenesis is an early occurrence in carcinogenesis. Use of NBI in the oral cavity has been a fairly recent development, and thus the literature regarding its efficacy and usefulness is relatively limited. The first experimental chapter aimed to investigate the efficacy of NBI for aiding the detection of OPMDs in a prospective series of patients with at least one white, red or red-white oral mucosal lesion. All patients underwent a conventional oral examination, followed by an exam using the white light mode then the NBI mode of a NBI system. A total of 272 lesions from 95 patients were observed. When using conventional oral examination as the gold standard, NBI had 100%, 74.63%, 92.38%, 100% and 93.77% sensitivity, specificity, positive predictive value, negative predictive value and accuracy respectively for the detection of OPMDs. NBI aided the detection of 24 lesions undetected by conventional oral examination and 13 lesions undetected by white light endoscopy. Unfortunately, surgical intervention for the treatment of OPMDs and oral cancer cannot be completely avoided despite the emphasis of early detection and monitoring. When surgery is indicated, a key factor for having a good prognosis is to have clean resection margins. Current methods for delineating the margin between healthy and diseased tissue may not necessarily result in a clean margin at a molecular level, even though they may appear healthy at a histological level. The second experimental chapter aimed to correlate mRNA and microRNA expression profiles at the primary tumour margins of oral cavity squamous cell carcinoma defined by NBI and white light. Three samples were taken from each of the eighteen resected oral squamous cell carcinomas – one from the NBI margin, white light margin and centre of the tumour. RNA was isolated from each sample and hybridised to mRNA and miRNA gene expression microarrays. A total of 4 794 genes and 137miRNAs were found to be differentially expressed. The type of miRNAs and genes expressed at the tumour was different compared to at the NBI margin – specifically, the NBI margin was more likely to contain molecularly normal cells. When integrating differentially expressed miRNA with differentially expressed mRNA, there were 91 potential miRNA-mRNA pairs (28 miRNA with 87 genes) identified. Several of the miRNAs and genes have been implicated in anti-apoptosis, tumorigenesis and metastasis. These studies demonstrate that NBI has great utility as a visualisation adjunct for detecting OPMDs and OSCCs, as it can aid the delineation of healthy tissue from diseased at both tissue and molecular levels. The use of NBI for monitoring OPMDs and determining tumour resection margins may improve the overall survival rate of OSCCs.