Clinical Optical Coherence Tomography In Head And Neck .

Page 3 of 7Figure 3: A normal human lip. BM, basement membrane; EP, epithelium; GW, glass window of probe tip; LP, lamina propria; MP, muscularis propria. Bar 1 mm.ously. Being independent of mechanically moving parts within the system,images are therefore created at amuch faster rate, allowing for ‘videorate’ display of B-scans as well asthe generation of three-dimensionaland four-dimensional sequences inreasonable time6,7.Last but not the least, additionalinformation from the acquired signalcan be used to derive more information from the tissue investigated. Forexample, intravascular speed or vocalcord vibrations can be determinedvia measurement of Doppler shift(Doppler-OCT)8, or certain tissuestructures such as fibrous proteinscan be highlighted by detectingthe polarisation state of the backreflected light (polarisation sensitiveor PS-OCT)9.Clinical applications in headand neckOCT has recently been investigated interms of its applicability in otolaryngology. The current review is focusedon its clinical application in head andneck cancers, mostly for the differentiation of premalignant and earlyinvasive lesions of the UADT. Thisreview also discusses its use in aidingthyroid cancer diagnosis as wellas in judging the severity of UADTmucositis as a side-effect of radiationtherapy. Other interesting applicationfields of OCT in otorhinolaryngologywhich are not discussed in this paperinclude its use in otology for visualisation of the tympanic membrane,functional assessments of vocal cordmobility and function and neonatology for three-dimensional reconstruction of the upper airway.The Medline database wassearched for appropriate and relevant publications in February 2013via PubMed using the followingsearch strings: ‘optical coherencetomography’ and ‘optical imaging’in combination with ‘oral cavity’,‘larynx’, ‘thyroid’ and ‘mucositis’.Early tumour diagnosis in theUADTAs tissues have different optical backscattering properties, OCT allows oneto visualise and to differentiate themost superficial tissue layers of theUADT. As demonstrated in Figure 3, itthus provides microanatomical information on the integrity or disruption of the epithelium, basementmembrane (BM) and supportinglamina propria.The earliest studies on clinicalapplications of OCT in the UADTwere reported from Russia. In 1997,Sergeev et al. were the first to reportimaging of the larynx as part of theUADT with OCT in vivo10. Normaltissue and cancerous tissue wasvisualised with OCT and the authorsobserved ‘a loss of normal tissuestratification in tumours’ and thoughtthat OCT would be ‘an interestingtool for early diagnosis of tumoursand for guidance of biopsies’. In2001, Shakhov et al. published adescriptive study on TD-OCT examinations in 26 patients with smalllaryngeal squamous cell carcinomas(SCC)11. The authors concluded that a‘stratification seen in OCT images isa criterion for a healthy larynx’ andthat ‘disappearance of such stratification is a sign of pathologic tissuealterations’. In addition to ratherrecently published, site-specific‘normal values’ for epithelial thicknesses within the larynx12 and theoral cavity13, their findings thus formthe basis for our current interpretation of OCT images in the UADT withregard to diagnosis of early invasiveSCC versus premalignancy. This canbe illustrated by Figures 4 (dysplasiaof the tongue) and 5 (early SCC ofthe floor of the mouth). Other clinically challenging diagnoses, such asdifferentiating hyperplasia versusdysplasia, might also be possibleusing real-time OCT14,15 as dysplasticlesions seem to show a persistentlyhigher decrease in signal intensityover the axial run of the epitheliallayer than hyperplastic lesions.LarynxThe larynx is the most frequentlyinvestigated location with OCT inthe UADT, as its sturdy structureas well as its relatively thin epithelium in healthy conditions makesit an optimal location for opticalimaging1,11,12,14,16. OCT has beenstudied in 500 patients in singlecentre trials as an adjunct diagnostictool for its value in assessing and diagnosing early stage laryngeal pathologies7,9,11,12,14,16-25, but with consistentlyand significantly positive resultswith regard to sensitivity, specificityand correlation with biopsy results.In most cases, OCT imaging probeswere placed directly onto the tissueduring microlaryngoscopy. However,non-contact OCT image acquisitionvia 90 rigid laryngoscopy or flexibleLicensee OA Publishing London 2013. Creative Commons Attribution Licence (CC-BY)F: Betz CS, Volgger V, Silverman SM, Rubinstein M, Kraft M, Arens C, Wong BJF. Clinical optical coherencetomography in head and neck oncology: overview and outlook. Head Neck Oncol. 2013 Mar 06;5(3):35.CompeƟng interests: none declared. Conflict of Interests: none declared.All authors contributed to the concepƟon, design, and preparaƟon of the manuscript, as well as read and approved the final manuscript.All authors abide by the AssociaƟon for Medical Ethics (AME) ethical rules of disclosure.Critical review

Page 4 of 7Critical reviewFigure 5: A) An OCT image of the ‘normal’ floor of the mouth shows contrast betweenEP and LP. B) An OCT image of the adjacent muscle invasive tumour shows total loss ofmicrostructure, without visible EP or LP. EP, epithelium; LP, lamina propria.transnasal laryngoscopy in patientswho were awake or intraoperativelyvia surgical microscopes has alsobeen described7,21-24.In 2005, Wong et al.16 observedlaryngeal mucosa in 82 patientswith various pathologies duringendoscopy. The authors describedthat OCT enabled them to obtain‘microanatomical information aboutthe epithelial layer, the integrity ofthe basement membrane and laminapropria’. The study’s most important qualitative observation was ‘theclear border between epitheliumand lamina propria in all benign andpremalignant lesions’. The authorsstated that the main benefit fromOCT was its ability to illustratesmall foci of destruction in the BM,because they define invasive tumourgrowth.In a study published by Armstronget al. in 200619, the authors performeda total of 26 TD-OCT scans duringmicrolaryngoscopy and comparedthese images to the histopathologicaldiagnosis (n 24) in patients withsuspicious lesions of the larynx. Inthis series, ‘a loss of the demarcation between the epithelium and thesubmucosal tissue was demonstratedin all (n 18) cases of invasive cancer’.In the remaining 6 patients withnon-malignant and premalignantchanges, an intact basal membranecould only be demonstrated in asubgroup of 3 patients. Similar to ourown study, the authors attribute thislack in specificity to the fact that thefalse-positive results were all foundin ‘bulky lesions’; they also state that‘thick, hypercellular tissues increasebackscattering at the surface andlimit the propagation of light intodeeper layers of the specimen. Thisresults in a limited capability to bothidentify the basement membraneand image deeper tissue structures’.Apart from this fact, an imprecisecorrelation of a histopathologicalsection and OCT image might also be(at least in part) responsible for theirfalse-positive findings.In another important study, Kraftet al. prospectively investigated 193patients (217 laryngeal lesions)undergoingmicrolaryngoscopyalone and in conjunction with OCT14.Microlaryngoscopy with OCT led tothe correct diagnosis in 89% of cases,while microlaryngoscopy alone wasonly able to correctly diagnose 80% ofthe cases (p 0.006). The exact gradeof dysplasia was correctly predictedin 71% of patients with precancerousdisease when OCT was utilised.Sensitivity for identifying dysplasiaincreased when OCT was added tomicrolaryngoscopy (66% vs. 78%),as did sensitivity for predicting invasion (87% to 93%). Apart from theirencouraging results, the authorsmentioned three commonly encountered problems with OCT-based diagnosis: 1) due to strong light absorption, hyperkeratosis may preventdeeper layers from being adequatelyevaluated; 2) ulcers may be confusedwith invasive carcinomas due to theabsence of a BM and 3) microinvasivecancer often cannot be safely delineated from high grade dysplasia orLicensee OA Publishing London 2013. Creative Commons Attribution Licence (CC-BY)F: Betz CS, Volgger V, Silverman SM, Rubinstein M, Kraft M, Arens C, Wong BJF. Clinical optical coherencetomography in head and neck oncology: overview and outlook. Head Neck Oncol. 2013 Mar 06;5(3):35.CompeƟng interests: none declared. Conflict of interests: none declared.All authors contributed to the concepƟon, design, and preparaƟon of the manuscript, as well as read and approved the final manuscript.All authors abide by the AssociaƟon for Medical Ethics (AME) ethical rules of disclosure.Figure 4: A) ‘Normal’ tongue OCT with high contrast between EP and LP. B) White tongue plaque OCT with irregular surface, decreasedcontrast and intact BM. C) Photograph of the tongue. EP, epithelium; LP, lamina propria; MP, muscularis propria. Black arrows indicatethe BM. Bar 1 mm.

Page 5 of 7Figure 6: A) An OCT image of the ‘normal’ thyroid tissue with uniform thyroid follicles(arrows). B) An OCT image of the thyroid nodule with a thickened thyroid capsule andC) thyroid follicle enlargement (arrows).carcinoma in situ due to the currentspatial resolution of OCT. The authorsconcluded that OCT is a simple, rapidand reliable aid in the early diagnosisof laryngeal disease.In 2011, Burns et al. used PS-OCTand conventional OCT to imagelaryngeal disease. The authors foundPS-OCT to be useful especially whendifferentiating between normal vocalcord tissue and scar tissue, becausethe increased collagen content in thescar tissue led to different birefringence patterns9. Also, PS-OCT wasfound to be useful for determiningthe exact extent of malignant invasionand submucosal spread into adjacentnormal structures. However, largerstudies are needed to determine thetrue value of PS-OCT9,26.In conclusion, OCT appears to bean ideal imaging tool for evaluatingdiscrete and superficial laryngealabnormalities. To date, its application helps in locating the best sitefor decisive biopsy in widespreadlesions and to determine the lateralextensions of those mucosal pathologies. The results from the studiesperformed so far suggest that OCTmight serve as an ‘optical biopsy’for live tissue, which would help todecrease costs, time to treat (‘onestop shop’), patient anxiety and therisk of permanent harm to the vocalcord function from unnecessarybiopsies. On the downside and apartfrom the aforementioned generallimitations of OCT, some anatomicalsites within the larynx (such as theanterior commissure) can be difficultto image15, and certain conditionssuch as hyperkeratosis and ulcerations can be hard to interpret. Withupcoming technical advances andfollowing more thorough investigations, however, the method has thepotential to gain a much more widespread acceptance in laryngology.Oral cavityFewer but not less interesting singlecentre studies have been publishedon the clinical use of OCT in the oralcavity. As early as 2004, Fomina etal.27,28 investigated 43 patients with56 intraoral lesions. The authors werea

Clinical optical coherence tomography in head and neck oncology: overview and outlook CS Betz1*, V Volgger1, SM Silverman2, M Rubinstein3, M Kraft4, C Arens5, BJF Wong3 Abstract Objective Optical coherence tomography is a high-resolution and minimally inva-sive optical imaging method, which provides in vivo cross-sectional