Journal of Cancer Research and Therapeutics

REVIEW ARTICLE
Year
: 2017  |  Volume : 13  |  Issue : 6  |  Page : 908--915

Optical diagnostics in oral cancer: An update on Raman spectroscopic applications


Aditi Sahu, C Murali Krishna 
 Advanced Center for Treatment Research and Education in Cancer, Tata Memorial Center, Chilakapati Laboratory, Kharghar, Navi Mumbai, Maharashtra, India

Correspondence Address:
Dr. C Murali Krishna
Advanced Center for Treatment Research and Education in Cancer, Tata Memorial Center, Chilakapati Laboratory, Kharghar, Sector 22, Navi Mumbai - 410 210, Maharashtra
India

Abstract

Raman spectroscopy (RS) is a sensitive vibrational spectroscopic method that can detect even subtle biochemical changes during the onset of disease. Consequently, RS has been extensively investigated for disease diagnosis, including cancers. Oral cancers are known to suffer from dismal survival rates, which have not improved for several decades. As delayed diagnosis contributes to the low disease-free survival rate observed in oral cancers, RS has also been explored for the early diagnosis of oral cancers. This review summarizes the major developments in the field, including diagnosis, surgical margin assessment and prediction of treatment response, and in the overall management of oral cancers. The article comprises an overview of epidemiology, diagnosis, treatment, and recently introduced diagnostic adjuncts for oral cancers, the basic principle, instrumentation of RS, multivariate analysis that impart objectivity to the approach, and finally a discussion on the recent applications in oral cancers. PubMed and Google Scholar database have been used to compile information available online till December 2015.



How to cite this article:
Sahu A, Krishna C M. Optical diagnostics in oral cancer: An update on Raman spectroscopic applications.J Can Res Ther 2017;13:908-915


How to cite this URL:
Sahu A, Krishna C M. Optical diagnostics in oral cancer: An update on Raman spectroscopic applications. J Can Res Ther [serial online] 2017 [cited 2018 Aug 17 ];13:908-915
Available from: http://www.cancerjournal.net/text.asp?2017/13/6/908/191032


Full Text

 Introduction



Oral cancers, a subtype of head and neck cancers, are cancers of the oral cavity. Oral cavity represents the first structure of the aerodigestive tract and is composed of distinct anatomic subsites. Lips, buccal mucosa, the upper and lower alveolar ridges with their attached gingiva, the retromolar trigone, the hard palate, the floor of the mouth, and the anterior two-thirds of the tongue majorly constitute the oral cavity as shown in [Figure 1]. Oral cancers can arise in any of these subsites. In Western countries, tongue, floor of the mouth (FOM), and lip account for ~70% of all cancers while in the Indian subcontinent, buccal mucosa and tongue along with the lip are the most commonly affected subsites.[1],[2] This disparity in most commonly affected subsites can be attributed to differences in ethnic, social, and lifestyle-related factors. Oral cancers are a major health problem worldwide, with an annual incidence estimate of approximately 275,000 cases.[3] Oral cancers form a significant health burden in developing countries like India, where they account for over 30% of all cancers with 80,000 new cases reported each year.[4] Tobacco (both smoking and smokeless) and alcohol are major etiological factors.[5] Visual inspection, followed by biopsy and histopathology of suspicious lesions found during clinical examination, is the gold standard for diagnosis.{Figure 1}

Oral cancers are often preceded by clinically visible mucosal alterations termed “precancer stages.” These precancer stages may refer to the presence of a benign lesion or morphologically altered tissue that has a greater than normal risk of malignant transformation. Leukoplakia, erythroplakia, oral submucous fibrosis (OSMF), tobacco pouch keratosis, and lichen planus are some forms of oral precancer or premalignant lesions. These lesions may or may not be dysplastic on histopathological assessment. The removal of lesions with moderate or severe dysplasia is advocated while mild dysplasia is followed up for reversal or progression. Treatment for oral cancer includes surgery, radiotherapy, and chemotherapy; surgery combined with chemotherapy and radiotherapy improves overall survival. In spite of the advancement in surgical and treatment modalities, low disease-free survival rates have been observed for several decades. The main reasons for the dismal survival rates include - diagnosis mainly in advanced stages, recurrence, inadequate access to health services, and lack of primary knowledge about causative factors.

It is known that early detection of oral cancer and recurrence can enhance survival rates and improve overall quality of life. Adjunct techniques such as tissue staining using toluidine blue, oral cytology, tissue fluorescence (VELscope), and chemiluminescence (ViziLite) based methods are being explored as complementary techniques for early diagnosis.[6],[7] While tissue staining involves the use of metachromatic dyes such as toluidine blue which have high affinity for DNA and can differentiate normal and abnormal tissues based on DNA content, oral cytology involves the collection of transepithelial samples from oral mucosa and subsequent cytomorphometry to identify abnormal cells. Light-based methods such as VELscope and ViziLite employ native tissue characteristics such as fluorescence and reflectance to identify abnormal regions. Serum- and saliva-based molecular diagnostic markers are also being investigated. In recent times, optical spectroscopic approaches have also been explored for oral cancer diagnosis. Optical spectroscopy involves the study of light-tissue interaction. The optical spectrum derived from any tissue contains information about the histological and biochemical makeup of that tissue. Because of the accessibility of the oral cavity, there has been increasing interest in the use of fiber-optic probe coupled optical spectroscopy systems to provide tissue diagnosis in real-time, noninvasively, and objectively (with the use of multivariate data analysis). Techniques such as fluorescence spectroscopy, reflectance spectroscopy, elastic scattering spectroscopy, infrared spectroscopy, and Raman spectroscopy (RS) are increasingly being investigated for oral cancer applications.[8]

 Raman Spectroscopy



RS is a vibrational spectroscopy method based on the inelastic scattering of light. Inelastic scattering of light, also known as Raman effect, was discovered by Sir C. V. Raman after seminal experiments on scattering.[9] This effect was discovered in the year 1928, for which Raman received the Nobel Prize in 1930. When a sample is irradiated with intense monochromatic light, phenomena such as absorption, scattering, and reflection occur. Most of the scattered photons have the same frequency of the incident light (Rayleigh scattering) while a small proportion (one in ten million) are inelastically scattered, i.e. with a frequency different from the incident photons; this phenomenon is termed as Raman effect. When the frequency of the scattered light is lower than the frequency of incident photon, the process is called Stokes shift. If the frequency of scattered photon is higher than incident photon, the process is called anti-stokes shift. The different types of scattering are illustrated in the simplified energy diagram shown in [Figure 2]. The energy difference between the incident and scattered photon (Raman shift) is represented as wavenumber (/cm).{Figure 2}

Instrumentation

RS is an inherently weak process: Only 1 in 10 million photons are Raman scattered. Thus, sophisticated instrumentation, i.e. powerful excitation source, high-throughput spectrograph, and sensitive detection systems are a prerequisite. The introduction of low-noise charged coupled device (CCD) detector technology, highly efficient imaging spectrographs, and compact semiconductor laser excitation sources enabled extensive Raman spectroscopic applications in diverse areas. Typically, Raman spectrometer is made up of (i) excitation source, (ii) optical system, (iii) spectrograph, and (iv) detection and computer control/processing systems. For most applications, a continuous-wave laser is employed as the excitation source in Raman spectrometers. The optical system consists of light steering mechanisms which direct and select laser wavelength for sample excitation. Rayleigh rejection system, which prevents elastically scattered light to be incident on the spectrograph and detector, is the most crucial component of Raman system as it facilitates the rejection of Rayleigh scattered light and detection of the comparatively weaker Raman scattered light. The main function of the spectrograph is to disperse light into its component wavelengths. CCDs are the most commonly employed detectors for RS in recent times. A schematic of the major components of typical Raman spectrometer for in vivo applications is shown in [Figure 3]. Specialized variants of RS are employed for specific applications, some of these include-Raman microscopy and imaging, resonance RS, surface-enhanced RS (SERS),[10] drop-coating deposition Raman (DCDR),[11] coherent anti-Stokes RS (CARS).[12]{Figure 3}

Spectral preprocessing and data analysis

RS can serve as an objective approach for both qualitative and quantitative analysis as the Raman spectral data are amenable to computational tools. These methods derive spectral information and use it to obtain classification between the groups. Data analysis can be either univariate or multivariate. Multivariate methods are more commonly employed to extract maximum features from the data. Important multivariate tools include principal component analysis (PCA), hierarchical cluster analysis (HCA), linear discriminant analysis (LDA), partial least-square (PLS)-based methods, artificial neural networks, and genetic algorithms. Spectra are subjected to multivariate analysis post-spectral preprocessing such as CCD response correction, removal of background signals by first derivatization or baseline correction, interpolation, and normalization.

Applications

Due to attributes such as sensitivity, high information content, and nondestructive nature, RS has been extensively applied in the fields of chemistry, biology, geology, pharmacology, forensics, pharmaceuticals, and material sciences. It is known that disease is accompanied with a concomitant change in native tissue biochemistry. RS can detect these changes and facilitate disease diagnosis. This is the basis for Raman spectroscopic diagnosis of diseases, including cancers. RS has shown that potential in diagnosis of several diseases including cancers, both ex vivo and in vivo. RS has extensively been employed for diagnosis of oral cancers.

 Raman Spectroscopy and Oral Cancer Applications



The first Raman spectroscopic applications in oral cancers were investigated by Bakker Schut et al.[13] in 2000. This group explored the in vivo classification of normal and dysplastic tissue in rat palate after cancer was induced by application of 4-nitroquinoline 1-oxide. Since then, several studies have investigated the potential of RS in the management of oral cancers. Different approaches - ex vivo, in vivo, biofluids, cell based, and imaging have been explored. Raman spectroscopic applications in oral cancer from the year 2009 onwards have been summarized in this review. PubMed and Google Scholar have been used for accessing publications in this field. The applications are summarized under the three categories - (a) diagnosis, (b) surgical margin detection, and (c) prediction of treatment response.

Diagnosis

Ex vivo studies

The animal study by Bakker Schut et al. was followed by a study on human oral frozen cancer biopsies by Venkatakrishna et al. in 2001.[14] Raman spectroscopic measurements from formalin-fixed tissues were consequently demonstrated, and significant difference between the normal and malignant epithelial regions was observed.[15] Malini et al.,[16] in the year 2006, carried out an extended study to discriminate normal, cancerous, precancerous, and inflammatory conditions. Lipid-rich features and predominant protein features were observed in normal and tumor conditions, respectively. Classification between different groups was explored using PCA coupled with multiparametric “limit test”, and high sensitivity and specificity were achieved.[16] Hu et al. acquired spectra of 66 human oral mucosa tissues (43 normal and 23 malignant) using confocal Raman microspectroscopy in 2008. After preprocessing spectra using wavelet-based analysis, PCA along with the calculation of areas under bands 1004, 1156, 1360, 1587, and 1660/cm was used as a classification method.[17] Another study by Sunder et al.[18] in 2011 was carried out to evaluate the applicability of near-infrared (NIR) RS in differentiating normal epithelium and different grades of oral cancer. Findings demonstrated that oral carcinomas of different pathological grades can also be identified with RS. Shifted-excitation Raman difference spectroscopy study (SERDS)[19] on 12 oral squamous cell carcinoma (OSCC) tissues could differentiate between malignant and benign areas with sensitivity of 86% and specificity of 94%. Keratin as a marker for OSCC identification using RS was recently demonstrated on 24 tissues samples with a sensitivity and specificity of 77–92% and 100%, respectively.[20] Rapid detection of oral cancer using 24 normal and 32 oral tumor sections on Ag-TiO2 nanostructured SERS substrate has been recently shown, achieving 100% sensitivity and 95.83% specificity.[21] Following these successful ex vivo studies on both fixed and frozen tissues, in vivo oral cancer studies on humans were concomitantly initiated.

In vivo studies

The first in vivo Raman spectroscopic study on humans was carried out by Guze et al.[22] for identifying site wise variations in the human oral cavity. In this study, the feasibility of spectral acquisition from oral cavity, reproducibility of Raman spectroscopic signature of normal oral mucosa among different anatomical oral sites was evaluated on 51 subjects of different races (Asian and Caucasian) and genders. This study, carried out on high-wavenumber region, suggested that spectra for different oral sites within the same ethnic group are significantly different, and the Raman signal was not influenced by gender or ethnicity. The differences between anatomical subsites could be due to varying degrees of keratinization. Therefore, the study suggested clustering of sites based on anatomical and spectral similarities. The consequent study by Bergholt et al.[23] aimed to characterize the in vivo Raman spectroscopic properties of different normal oral tissues in the fingerprint region (800–1800/cm) and to understand the biochemical basis for differentiation. Fitting of reference biochemicals that constitute oral cavity and PLS-discriminant analysis of 402 high-quality spectra from twenty subjects found that histological characteristics influence Raman spectra. Based on the findings, anatomical regions in the oral cavity were divided into three different clusters based on their histological and spectroscopic characteristics. The three groups included - (a) buccal mucosa, inner lip and soft palate, (b) dorsal, ventral tongue, and FOM, and (c) gingiva and hard palate. Another study by Krishna et al. suggested clustering of subsites within four major anatomical clusters based on spectral patterns - (a) outer lip, and lip vermillion, (b) buccal mucosa, (c) hard palate, and (d) dorsal, lateral and ventral tongue, and soft palate. Further, the authors also suggest the use of anatomy-matched algorithms to increase discrimination between healthy and abnormal conditions.[24]

Our group reported the first in vivo spectral acquisition from oral cancer patients in clinically implementable time.[25],[26] Using fifty subjects, malignant lesions on buccal mucosa could be differentiated from contralateral normal area with high sensitivity and specificity. Subsequently, the studies to detect oral premalignant conditions were carried out by the same group on 104 subjects. Spectra were acquired from premalignant patches, contralateral normal, tumor sites, and healthy subjects with or without tobacco habits. Data analysis using PCA and principal component-LDA (PC-LDA) suggest potential of RS in differentiating premalignant lesions from healthy, normal, and tumor conditions.[27] In another study, the origin of Raman signals in tissues was also investigated.[28] The effect of age-related physiological changes and their possible influence in classification between normal and pathological conditions was also explored by our group.[29] Next, the feasibility of detection of malignancy-associated-changes or cancer-field-effects (CFEs) in oral cancer patients was evaluated.[30] Spectra were acquired from 84 subjects and subjected to PC-LDA. Findings indicate differences between contralateral normal sites of tobacco and nontobacco habitués. Contralateral normal of nonhabitue oral cancer subjects was distinct from healthy controls, suggesting microarchitectural changes that characterize CFE could be detected using in vivo RS. All these studies explored the detection of cancer and precancer lesions on buccal mucosa subsite. Krishna et al. investigated the classification between spectra acquired from multiple normal sites and histopathologically verified precancer (OSMF and leukoplakia) and cancer lesions. Using probability-based multiclass diagnostic algorithm, sensitivity and specificity of 94.2% and 94.4% were obtained for a binary normal versus abnormal model.[31] In a recent study by Guze et al.,[32] Raman spectral discrimination of premalignant and malignant lesions from contralateral normal and benign lesions was explored. Spectral differences were apparent between groups; the premalignant and malignant lesions could be classified with 100% sensitivity and 77% specificity. Further, the study also states that the anatomic origin of the tumor may not have a bearing on normal versus tumor classification.[32] Another recent study by our group [33] has shown major spectral differences between buccal mucosa and tongue while the lip was shown to have an intermediate position. Individual and pooled subsite diagnostic algorithms were explored with respect to spectra and classification. Although the individual models gave better classification efficiency, the pooled subsite gave reasonable sensitivity and specificity and can therefore be employed for preliminary screening for normal versus abnormal changes in the oral cavity. Further, it may be possible that the detection of premalignant and early cancer states could benefit from subclassification of sites. The last two studies suggested the use of single diagnostic model for the detection of oral cancers.

The utility of the pooled diagnostic algorithm for normal versus abnormal classification was tested on 157 healthy and oral cancer subjects by our group. Spectra were acquired from healthy mucosa, contralateral normal mucosa, premalignant and malignant lesions on 7 different subsites, buccal mucosa, lip, tongue, FOM, retromolar trigone, gingival, and tongue in the oral cavity. A total of 1128 healthy spectra, 1107 contralateral spectra, 106 premalignant, and 277 tumor spectra were acquired from the different subsites of the recruited subjects. Spectra were recorded with an HE-785 commercial Raman spectrometer, and data analysis was carried out using PC-LDA. PC-LDA was carried out using 4 factors; the scatter plot plotted using scores of factor 2 and 3 is shown in [Figure 4]. Leave-one-out cross validation (LOOCV) was employed to validate the PC-LDA results. LOOCV yielded an overall classification of 98%, 54%, 29%, and 67% for healthy, contralateral normal, premalignant and malignant conditions. As no abnormal spectra misclassified with healthy, a sensitivity of 100% and a specificity of 98% were obtained. The results indicate that RS can potentially classify healthy against all cancer-related abnormal conditions of the oral cavity with high sensitivity and specificity. Thus, in vivo RS can be used as noninvasive, rapid, and objective preliminary test during screening for oral precancer and cancers. The “abnormal” sites/subjects can then be subjected to further confirmatory procedures.{Figure 4}

Less invasive samples

Apart from the obvious advantages of the in vivo approach, extensive applications using these methods are restricted due to the need of a dedicated instrument and stringent experimental conditions at every screening center.

Serum, saliva, urine

Less invasive samples such as saliva, blood, urine, and exfoliated cells have therefore been investigated for oral cancer diagnosis to enable distance diagnosis. Saliva is the most accessible and pain-free sample. Because of the dilute concentrations of biomolecules, extremely sensitive detection mechanisms are crucial. SERS of saliva could differentiate between healthy and oral cancer samples in the first study conducted in 2007.[34] Another study has shown differences in SERS spectra from squamous cell carcinoma (SCC) positive and healthy control samples. Differential peaks were evident at wavelengths 400, 811, 884, 955, 1130, 1270, and 1610/cm.[35] Biofluids such as blood reflect metabolic disturbances due to the onset of disease in the body and are known to be one of the ideal samples for disease diagnosis in clinics. Blood plasma and serum have also been explored for oral cancer diagnosis. Harris et al.[36] demonstrated the feasibility of head and neck cancer detection using RS of a peripheral blood sample. Blood samples were collected from twenty head and neck cancer patients (tumors at tongue, larynx, skin, and salivary glands), and plasma was obtained. Conventional LDA approaches showed an accuracy of around 65% while genetic evolutionary algorithm obtained an efficiency of 83%. A study carried out by our group [37] demonstrated the potential of serum RS in differentiating serum from healthy and oral cancer subjects. The pilot study exploiting resonance Raman effects of beta-carotene on seventy serum samples and data analysis using PC-LDA yielded classification efficiency of ~78%. Cancers at different subsites, buccal mucosa and tongue, could also be differentiated. Blood plasma was explored for oral cancer diagnosis by Rekha et al. in 2013 using 28 plasma samples. Spectra were acquired from LabRam HR 800 in the spectral range 800–1800/cm. Marked spectral differences were observed between cancer and normal plasma corresponding to protein, amino acid, glucose, and lipids. Successful classification between the groups was observed using PCA and LDA.[38] The utility of NIR excitation wavelength in classifying normal and oral cancer samples was next verified by our group using 86 samples. PC-LDA yielded similar classification efficiency of ~ 78% in distinguishing normal and cancer groups.[39] A recent study by our group has further demonstrated the potential of serum RS in oral cancer screening. Using ~ 340 samples, the feasibility of differentiating normal from all abnormal (oral cancer, precancer, and cancer control) was observed. A comparable sensitivity and specificity (64% and 80%, respectively) with regard to existing screening methods such as mammography, Pap staining, and fecal occult blood test was achieved.[40] Urine has also been explored for oral cancer diagnosis. A recent study for oral cancer diagnosis using urine was carried out by Elumalai et al. Characterization of the metabolites of human urine of normal subjects and oral cancer patients in the fingerprint region (500–1800/cm) was explored. PC-LDA findings yield sensitivity and specificity of 98.6% and 87.1%, respectively to discriminate healthy and cancer patients.[41]

Less invasive samples-exfoliated cells

Detection of malignant changes at the cellular level can help in early cancer diagnosis. Several studies have explored the detection of such changes in both cultured cell lines and exfoliated cells. Exfoliative cytology is a simple, rapid, and less invasive technique:[42] it is thus well accepted by the patients and is suitable for routine application in population screening programs. In the proof of concept study carried out by our group, feasibility of cell collection and spectral acquisition and differences between healthy and precancer [43] and tumor [44] were investigated. Post-spectral acquisition, Pap staining of the same cell pellet was also carried out and used for correlation. Data analysis was carried out using both spectra- and patient-wise approaches. Healthy, premalignant, and tumor cell pellets could be differentiated by PCA and PC-LDA. In a study by Carvalho et al., spectral differences in the cytoplasm, nucleus and nucleolus regions of SCC cell line (SCC-4), dysplastic cell line (dysplastic oral keratinocyte) and normal epithelial, primary cultures were investigated. For each cell line, spectra were acquired from twenty cells, and data analysis was carried out using PCA. Findings indicate successful discrimination between cancer and healthy cell lines; dysplastic and cancer cell lines could also be differentiated based on cytoplasmic content. Thus, early precancerous conditions and noninvasive oral cancer harboring dysplastic regions may be identified using RS.

Raman imaging

Raman mapping experiments on oral mucosal tissues have also been reported. As oral mucosa is not homogenous and comprises different layers and histological characteristics, signal contributions from individual layers have to be understood. In the first study by Cals et al.,[45] the method of RS-based histopathology was developed and standardized. The study revolved around Raman microspectroscopic mapping of unstained frozen sections, followed by histopathological annotation of features in Raman images. Twenty experiments were conducted on different tissue sections obtained from tongue SCC; K-means cluster analysis (KCA) and HCA were used for data analysis. Findings indicated Raman mapping followed by KCA and HCA, can be used as a reproducible method to effectively define the spectral characteristics of individual histopathological structures for oral mucosa. In a subsequent study by Daniel et al.,[46] Raman mapping was explored for oral cancer diagnosis. Normal and oral cancer tissue sections could be distinguished based on the spectral parameters. PCA and KCA were employed to construct pseudocolor images. Raman maps could clearly delineate tumor margins. Mapping carried out on a blind sample also yielded correct identification of the sample. A similar study by our group [47] aimed to understand biochemical variations in normal and malignant oral buccal mucosa. Data were acquired from 10 normal and SCC tissues. Raman maps of normal sections could resolve the layers of epithelium, i.e. basal, intermediate, and superficial while inflammatory, tumor, and stromal regions were identified in tumor maps. PCA could successfully classify epithelium and stromal regions of normal cells. The classification between cellular components of normal and tumor sections was also observed.

Surgical margin assessment

Tumor-positive resection margins lead to recurrence in oral cancer patients and consequently lower disease-free survival rates. The sensitivity of RS could be exploited for the detection of surgical margins in oral cancer tissues. Potential of RS in surgical demarcation was investigated by two recent studies. In the first study by Barroso et al.,[48] differential water content in malignant and surrounding normal tissue was used as a basis for identifying surgical margins using Raman bands of OH- and CH-stretching vibrations in high-wavenumber region. The water content in SCC was significantly higher than surrounding healthy tissue. Thus, tumor tissue could be detected with a sensitivity of 99% and a specificity of 92% after using a cutoff water content value of 69%. In another study by Cals et al.,[49] Raman imaging of normal and tissue sections from ten oral cancer patients was carried out and 127 pseudo-color Raman images were generated. These images were linked to the histopathological evaluation of same sections, and spectra were annotated based on histopathological findings. LDA was used to build models for tumor and surrounding healthy tissue. Thus, RS could successfully differentiate tumor and surrounding healthy tissues. As Raman measurements are fast and can be carried out on freshly excised tissue without any preparation, the development of an intraoperative tool for guiding tumor resection may improve patient outcomes.

Prediction of treatment response

Raman spectra have been shown to correlate with molecular and cellular changes associated with disease, including cancer. RS can therefore be used for monitoring treatment response in oral cancers. In a recent retrospective study by our group, serum RS could differentiate between cohorts of patients with or without recurrence. DNA and protein content were major spectral features to differentiate the recurrence and nonrecurrence groups.[50] Prediction of radio response by RS was investigated in vitro using radioresistant oral cancer cell lines established by fractionated radiation.[51] Raman spectral differences were observed between parental tongue cancer cell line, 50 Gy, and 70 Gy cell lines; PCA also yielded three distinct clusters. The change in molecular profile acquired by radioresistant sublines was successfully detected by RS.

 Conclusions



Oral cancers are associated with poor disease-free survival rates. Improvements in screening, diagnostic, and monitoring approaches can lead to improved treatment outcomes. Raman spectroscopic applications in oral cancer have been extensively investigated. These studies have demonstrated the potential of RS in being an objective; real-time screening, diagnostic, and therapeutic monitoring adjunct for oral cancer diagnosis.In vivo approaches have demonstrated the potential of RS in screening and early diagnosis of abnormal mucosal conditions at all oral subsites. The studies on minimally invasive samples, including serum, have shown potential in preliminary screening of oral cancers; urine-based approaches have shown promising results which need to be validated on a large sample size. RS of exfoliated cells has shown promise in differentiating normal and tumor samples and in identifying early precancerous changes in cell lines. Raman imaging studies have helped in understanding the spectral contributions from the different layers of the epithelium; further studies on rapid scanning methods can help in real-time surgical demarcation. The studies on margin assessment have shown feasibility of clearly differentiating tumor from surrounding normal using high-wavenumber region and Raman imaging. The studies on prediction of treatment response have successfully identified recurrence-prone patients and changes associated with the acquisition of radio-resistance in a cell-line model. Overall, these studies have strongly demonstrated the potential of RS and preparedness of this instrument for noninvasive and less-invasive diagnosis of oral cancers. Translation of this approach to clinics may help in improved preliminary oral cancer screening, early diagnosis, and enhance disease-free survival rates.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1Kirita T, Omura K. Oral Cancer: Diagnosis and Therapy. Tokyo: Springer; 2015.
2Carnelio S, Rodrigues G. Oral cancer at a glance. Internet J Dent Sci 2004;1:1-12.
3Warnakulasuriya S. Global epidemiology of oral and oropharyngeal cancer. Oral Oncol 2009;45:309-16.
4Coelho KR. Challenges of the oral cancer burden in India. J Cancer Epidemiol 2012;2012:701932.
5Khandekar S, Bagdey P, Tiwari R. Oral cancer and some epidemiological factors: A hospital based study. Indian J Community Med 2006;31:157-9.
6Mehrotra R, Gupta DK. Exciting new advances in oral cancer diagnosis: Avenues to early detection. Head Neck Oncol 2011;3:33.
7Messadi DV. Diagnostic aids for detection of oral precancerous conditions. Int J Oral Sci 2013;5:59-65.
8Swinson B, Jerjes W, El-Maaytah M, Norris P, Hopper C. Optical techniques in diagnosis of head and neck malignancy. Oral Oncol 2006;42:221-8.
9Raman CV, Krishnan KS. A new type of secondary radiation. Nature 1928;121:501-2.
10Haynes CL, McFarland AD, Duyne RPV. Surface-enhanced Raman spectroscopy. Anal Chem 2005;77:338A-46A.
11Zhang D, Xie Y, Mrozek MF, Ortiz C, Davisson VJ, Ben-Amotz D. Raman detection of proteomic analytes. Anal Chem 2003;75:5703-9.
12Begley R, Harvey A, Byer RL. Coherent anti-stokes Raman spectroscopy. Appl Phys Lett 1974;25:387-90.
13Bakker Schut TC, Witjes MJ, Sterenborg HJ, Speelman OC, Roodenburg JL, Marple ET, et al. In vivo detection of dysplastic tissue by Raman spectroscopy. Anal Chem 2000;72:6010-8.
14Venkatakrishna K, Kurien J, Pai KM, Valiathan M, Kumar NN, Murali Krishna C, et al. Optical pathology of oral tissue: A Raman spectroscopy diagnostic method. Curr Sci 2001;80:665-9.
15Krishna CM, Sockalingum GD, Kurien J, Rao L, Venteo L, Pluot M, et al. Micro-Raman spectroscopy for optical pathology of oral squamous cell carcinoma. Appl Spectrosc 2004;58:1128-35.
16Malini R, Venkatakrishna K, Kurien J, Pai KM, Rao L, Kartha VB, et al. Discrimination of normal, inflammatory, premalignant, and malignant oral tissue: A Raman spectroscopy study. Biopolymers 2006;81:179-93.
17Hu Y, Jiang T, Zhao Z, editors. Discrimination of Squamous Cell Carcinoma of the Oral Cavity Using Raman Spectroscopy and Chemometric Analysis. Intelligent Networks and Intelligent Systems, 2008 ICINIS'08First International Conference on; 2008, IEEE.
18Sunder N, Rao N, Kartha V, Ullas G, Kurien J. Laser raman spectroscopy: A novel diagnostic tool for oral cancer. J Orofac Sci 2011;3:15.
19Christian K, Johanna M, Werner A, Kathrin B, Tesfay GM, Robert H, et al. Raman difference spectroscopy: A non-invasive method for identification of oral squamous cell carcinoma. Biomed Opt Express 2014;5:3252-65.
20Chen PH, Shimada R, Yabumoto S, Okajima H, Ando M, Chang CT, et al. Automatic and objective oral cancer diagnosis by Raman spectroscopic detection of keratin with multivariate curve resolution analysis. Sci Rep 2016;6:20097.
21Girish CM, Iyer S, Thankappan K, Rani VD, Gowd GS, Menon D, et al. Rapid detection of oral cancer using Ag-TiO 2 nanostructured surface-enhanced Raman spectroscopic substrates. J Mater Chem B 2014;2:989-98.
22Guze K, Short M, Sonis S, Karimbux N, Chan J, Zeng H. Parameters defining the potential applicability of Raman spectroscopy as a diagnostic tool for oral disease. J Biomed Opt 2009;14:014016.
23Bergholt MS, Zheng W, Lin K, Ho KY, Teh M, Yeoh KG, et al. Characterizing variability in in vivo Raman spectra of different anatomical locations in the upper gastrointestinal tract toward cancer detection. J Biomed Opt 2011;16:037003.
24Krishna H, Majumder SK, Chaturvedi P, Gupta PK. Anatomical variability of in vivo Raman spectra of normal oral cavity and its effect on oral tissue classification. Biomed Spectrosc Imaging 2013;2:199-217.
25Singh SP, Deshmukh A, Chaturvedi P, Krishna CM. Raman spectroscopy in head and neck cancers: Toward oncological applications. J Cancer Res Ther 2012;8 Suppl 1:S126-32.
26Singh S, Deshmukh A, Chaturvedi P, Krishna CM, editors.In Vivo Raman Spectroscopy for Oral Cancers Diagnosis. Bellingham, Washington, USA: SPIE BiOS, International Society for Optics and Photonics; 2012.
27Singh SP, Deshmukh A, Chaturvedi P, Murali Krishna C.In vivo Raman spectroscopic identification of premalignant lesions in oral buccal mucosa. J Biomed Opt 2012;17:105002.
28Deshmukh A, Singh SP, Chaturvedi P, Krishna CM. Raman spectroscopy of normal oral buccal mucosa tissues: Study on intact and incised biopsies. J Biomed Opt 2011;16:127004.
29Sahu A, Deshmukh A, Ghanate AD, Singh SP, Chaturvedi P, Krishna CM. Raman spectroscopy of oral buccal mucosa: A study on age-related physiological changes and tobacco-related pathological changes. Technol Cancer Res Treat 2012;11:529-41.
30Singh SP, Sahu A, Deshmukh A, Chaturvedi P, Krishna CM.In vivo Raman spectroscopy of oral buccal mucosa: A study on malignancy associated changes (MAC)/cancer field effects (CFE). Analyst 2013;138:4175-82.
31Krishna H, Majumder SK, Chaturvedi P, Sidramesh M, Gupta PK.In vivo Raman spectroscopy for detection of oral neoplasia: A pilot clinical study. J Biophotonics 2014;7:690-702.
32Guze K, Pawluk HC, Short M, Zeng H, Lorch J, Norris C, et al. Pilot study: Raman spectroscopy in differentiating premalignant and malignant oral lesions from normal mucosa and benign lesions in humans. Head Neck 2015;37:511-7.
33Sahu A, Deshmukh A, Hole AR, Chaturvedi P, Krishna CM.In vivo subsite classification and diagnosis of oral cancers using Raman spectroscopy. J Innov Opt Health Sci 2016;9:165007. doi: 10.1142/S1793545816500176.
34Kah JC, Kho KW, Lee CG, James C, Sheppard R, Shen ZX, et al. Early diagnosis of oral cancer based on the surface plasmon resonance of gold nanoparticles. Int J Nanomedicine 2007;2:785-98.
35Davies K, Connolly JM, Dockery P, Wheatley AM, Olivo M, Keogh I. Point of care optical diagnostic technologies for the detection of oral and oropharyngeal squamous cell carcinoma. Surgeon 2015;13:321-9.
36Harris AT, Lungari A, Needham CJ, Smith SL, Lones MA, Fisher SE, et al. Potential for Raman spectroscopy to provide cancer screening using a peripheral blood sample. Head Neck Oncol 2009;1:34.
37Sahu A, Sawant S, Mamgain H, Krishna CM. Raman spectroscopy of serum: An exploratory study for detection of oral cancers. Analyst 2013;138:4161-74.
38Rekha P, Aruna P, Daniel A, Prasanna SW, Udayakumar K, Ganesan S, et al., editors. Raman Spectroscopic Characterization of Blood Plasma of Oral Cancer. Photonics (ICP), 2013 IEEE 4th International Conference on; 2013, IEEE.
39Sahu A, Sawant S, Talathi-Desai S, Murali Krishna C. Raman spectroscopy of serum: A study on oral cancers. Biomed Spectrosc Imaging 2015;4:171-87.
40Sahu AK, Dhoot S, Singh A, Sawant SS, Nandakumar N, Talathi-Desai S, et al. Oral cancer screening: Serum Raman spectroscopic approach. J Biomed Opt 2015;20:115006.
41Elumalai B, Prakasarao A, Ganesan B, Dornadula K, Ganesan S. Raman spectroscopic characterization of urine of normal and oral cancer subjects. J Raman Spectrosc 2015;46:84-93.
42Diniz-Freitas M, García-García A, Crespo-Abelleira A, Martins-Carneiro J, Gándara-Rey J. Applications of exfoliative cytology in the diagnosis of oral cancer. Med Oral 2004;9:355.
43Sahu A, Shah N, Mahimkar M, Garud M, Pagare S, Krishna CM, editors. Raman Spectroscopy and Oral Exfoliative Cytology. Bellingham, Washington, USA: SPIE BiOS, International Society for Optics and Photonics; 2014.
44Sahu A, Tawde S, Pai V, Gera P, Chaturvedi P, Nair S, et al. Raman spectroscopy and cytopathology of oral exfoliated cells for oral cancer diagnosis. Anal Methods 2015;7:7548-59.
45Cals FL, Bakker Schut TC, Koljenović S, Puppels GJ, Jong RJ. Method development: Raman spectroscopy-based histopathology of oral mucosa. J Raman Spectrosc 2013;44:963-72.
46Daniel A, Prakasarao A, David B, Joseph L, Murali Krishna C, Ganesan S. Raman mapping of oral tissues for cancer diagnosis. J Raman Spectrosc 2014;45:541-9.
47Behl I, Kukreja L, Deshmukh A, Singh SP, Mamgain H, Hole AR, et al. Raman mapping of oral buccal mucosa: A spectral histopathology approach. J Biomed Opt 2014;19:126005.
48Barroso EM, Smits RW, Bakker Schut TC, ten Hove I, Hardillo JA, Wolvius EB, et al. Discrimination between oral cancer and healthy tissue based on water content determined by Raman spectroscopy. Anal Chem 2015;87:2419-26.
49Cals FL, Bakker Schut TC, Hardillo JA, Baatenburg de Jong RJ, Cals FL, Bakker Schut TC, et al. Investigation of the potential of Raman spectroscopy for oral cancer detection in surgical margins. Lab Invest 2015;95:1186-96.
50Sahu A, Nandakumar N, Sawant S, Krishna CM. Recurrence prediction in oral cancers: A serum Raman spectroscopy study. Analyst 2015;140:2294-301.
51Yasser M, Shaikh R, Chilakapati MK, Teni T. Raman spectroscopic study of radioresistant oral cancer sublines established by fractionated ionizing radiation. PLoS One 2014;9:e97777.