Head & neck optical diagnostics: vision of the future of surgery
Review paper and Proceedings of the Inaugural Meeting of the Head and Neck Optical Diagnostics Society (HNODS) on March 14th 2009 at University College London.
The aim of our research must be to provide breakthrough translational research which can be applied clinically in the immediate rather than the near future. We are fortunate that this is indeed a possibility and may fundamentally change current clinical and surgical practice to improve our patients' lives.
KeywordsOptical Coherence Tomography Raman Spectroscopy Optical Coherence Tomography Imaging Negative Resection Margin Optical Coherence Tomography System
Upper aero-digestive tract (UADT) carcinomas continue to be the 6th most common cancer worldwide with approximately 270,000 new oral cavity tumours per year . Unfortunately, the majority of these tumours present in late stage with the attendant functional, psychological and economic costs to their victims. Early diagnosis is often delayed as tumour precursors or early cancers are hardly visible and not picked up by common imaging methods. It's clearly evident that screening and early detection of the cancer and its early precursors have the potential to reduce the morbidity and mortality of this disease. In that context, current oral examination methods including incandescent light or toluidine blue, reflectance visualization and illumination with chemi-luminescent light source, are largely subjective, dependent on the experience of the examiner and are considered in-effective tools in primary care settings .
The true or important surgical margin of these lesions is still not defined . In the treatment of cancer the fundamental surgical goal is to remove all local malignant disease and leave no residual malignant cells. Studies have demonstrated the benefit of achieving negative resection margins in terms of disease free local recurrence and overall survival. The surgical margins for head & neck cancer may vary widely depending on the site of disease. This variation reflects the biological and anatomical environment of the tumour site at macroscopic and microscopic levels. There is no accepted standard for the quantity of normal tissue to be removed and the effect of positive margins on recurrence rate appears to be considerably dependent on the site of the tumour. The extent of tumour volume resection is determined by the need for cancer control and the peri-operative, functional and aesthetic morbidity of the surgery.
Resection margins are currently assessed intra-operatively by frozen section and retrospectively after definitive histological analysis of the resection specimen. There are limitations to this assessment. The margin may not be consistent in three dimensions and may be susceptible to errors in sampling and histological interpretation. Assigning the true excision margin may be difficult due to post-excision changes secondary to shrinkage and fixation .
Local recurrence occurs even among tumours with extensive histological demonstration of adequate resection margins. Sites with significant recurrence rates after negative resection margins are oral cavity, sub-mandibular region, tonsil and pharynx. Therefore, it is accepted that cancers at these sites require larger margins of excision than tumours elsewhere in the head and neck .
Tumour margin detection in real-time without the use of additional molecular stain would be desirable. Surgical use of a microscopy tool would be ideal, but most systems focus on microscopic disease evaluation, whereas tools for macroscopic scanning of tissue such as enhanced endoscopy imaging are less well developed.
Screening for second primaries in patients with a history of cancer. This requires imaging techniques or an approach where a larger area can be scanned quickly.
Distinguishing potentially malignant visible primary lesions from benign ones. Here fibre-optic point measurements can be used as the location of the lesion is known.
Elastic Scattering Spectroscopy
Optical spectroscopy mediated by fibre-optic probes can be used to perform non-invasive, or minimally-invasive, real-time assessment of tissue pathology in situ. The method of elastic scattering spectroscopy (ESS) is sensitive to the sub-cellular architectural changes, such as nuclear grade and nuclear to cytoplasm ratio, mitochondrial size and density...etc., which correlates with features often used by pathologists when performing histological assessment. ESS is also sensitive to ultrastructural changes that are beyond the Abbe resolution limit for optical microscopy, and thus conveys information that may not be available from conventional histology. ESS has proved to be a promising method for detecting premalignant and malignant changes in a variety of organ areas, including oral tissues, with high sensitivity and specificity. Several head and neck tissues, including lymph nodes, archival bones, skin and resection margins, have been interrogated using ESS with very promising results [4, 5].
Differential path-length spectroscopy (DPS)
The first measurements in biopsy proven squamous cell carcinoma showed significant changes in both vascular and morphological parameters. Measurements on thick keratinized lesions however failed to generate any vascular signatures. This is related to the sampling depth of the standard optical fibres used. Recently the group developed a fibre-optic probe with a ~1 mm sampling depth. Measurements on several leukoplakias have shown that with this new probe one may now sample just below the keratin layer to obtain vascular signatures. This enables clinically significant diagnostic measurements .
Spectral scatter scanning system
A novel technique showing great translational utility is a recent raster scanning scatter spectroscopy system that has been evaluated for imaging the spectral signature remitted from tissue, with real-time classification algorithm, which maximizes the ability to identify regions of tumour from regions of normal tissue. The system uses a wide band of wavelengths from 400 nm up to 700 nm, and recovers the scatter power, scatter amplitude, and absorption species, from the reflectance from a 100 micron spot, allowing imaging of tissue a high frame rate. The system uses dark field illumination and spectrometer detection in the emission channel together with a scanning mirror. The early prototypes of the system were tested on pancreas tumours and prostate tumour margin detection, and current work is ongoing in breast cancer margin delineation .
The applications of Raman spectroscopy in the life sciences are still in the early stages of development. Raman spectroscopy is being investigated in a broad spectrum of biological and toxicological sciences. In oncology Raman is being investigated as a diagnostic tool for characterising cancer cells and early malignant changes and distinguishing these from normal cells. Raman spectroscopy has the distinct advantage over other optical techniques that it provides information on molecular composition and structure of living tissue. There is a strong rationale for using Raman spectroscopy in epithelial cancer. Although Raman spectroscopy has been investigated for several decades, clinical head & neck studies are scarce .
A significant problem associated with Raman applications arises from inherently weak signal produced by the Raman Effect. Biomedical samples are extremely intricate systems which reflect complex Raman spectra. Raman bands due to biological constitutes are generally overlapped, making it difficult to identify individual components correctly. Furthermore, due to the minimal sample preparation encountered in the clinical environment, biomedical sample samples usually produce a strong fluorescent background which may completely obscure the true Raman signals .
In an attempt to improve intra-operative tissue diagnosis, tools and methods for an "optical biopsy" have been proposed, some of them exploiting fluorescent properties of endogenous or exogenous fluorochromes [9, 10].
Fluorescence spectroscopy tries to capture characteristic spectral features of fluorochromes and correlate these with the disease state. Several mathematical methods have been proposed to evaluate recorded spectra to maximize the discrimination between "normal" and "malignant".
Fluorescence imaging aims at highlighting malignant tissue, especially where it is not evident under white light in a large field of view. Autofluorescence as well as drug-induced fluorescence can be detected and displayed with commercial equipment. They usually rely on capturing fluorescence in one or two colour channels and remission in another channel. Sophisticated image processing to quantify fluorescence or eliminate disturbing signal is only slowly becoming available.
The fluorescence contrast is even slightly enhanced by using exogenously applied fluorescent markers or their precursors (e.g., 5-aminolevulinic acid induced Protoporphyrin IX). Recent advances include the possibility to extract true spectra of single fluorophores ("intrinsic spectra") by mathematically eliminating the undesired influences of scattering and absorption. As well, tumour-specific enzymes are about to be specifically targeted by fluorescent markers "smart probes" in order to improve both sensitivity and specificity.
Optical Coherence Tomography
Despite the recent development of Fourier domain OCT that significantly increases imaging speed and sensitivity, the OCT system that achieves both high speed and high sensitivity simultaneously at 1.3 μm is not currently available. The recent development of a Fourier-domain-mode-lock (FDML) swept source based OCT system can now achieve high speed (>100 kHz A-scan rate) and high spatial resolution (<4 μm) simultaneously. In addition, the development of various miniature scanning probes that allow high-speed 3-D OCT imaging will be reported. This has been augmented by the development of a non-iterative digital focusing method to alleviate the compromise between lateral resolution and depth measurement range, which allows high lateral resolution over the full depth measurement range .
The major clinical applications of OCT in head and neck surgery that have been explored recently are: examination of the true vocal folds with the aim of identifying and characterizing pre-cancerous and early stage malignancy and examination of the paediatric/neonatal sub-glottis. OCT imaging can discern subtle differences in the sub-glottic mucosa and hopefully provide a means to identify patients at risk for extubation failure, and ideally in the future be used in the neonatal ICU to optimize endotracheal tube management .
This technology still as yet gives the unfulfilled promise of the basement membrane recognition which is all important in defining the extent of malignant spread and disease prognosis. Its resolution is still clearly just below the tissue level but rapidly approaching the cellular scale. Further developments in hardware (i.e. optics, probe design and laser scanning) and software processing (including pattern recognition and rendering) continue to improve resolution and clinical utility.
Several other optically based systems for tissue interrogation are described as well as their current ex-vivo and practical applications
The potential role of infrared spectroscopy in biomedical science has been described to distinguish different biomolecules by probing chemical bond vibrations and using these molecular and sub-molecular patterns to define and differentiate pathological from healthy samples. Several protocols have been described to exploit the potential of infrared spectroscopy in defining spectral profiles in salivary gland disease attributable to various kinds of cancer and the corresponding healthy tissues. Researchers suggest the potential of infrared micro-spectroscopy imaging, in combination with multivariate data analysis, to highlight even subtle biochemical and morphological changes, distinguishing various kinds and grades of neoplasia in human tissues .
Confocal Endomicroscopy (CEM) is a non invasive imaging tool enabling "optical biopsies" of tissues at cellular level. Clinical studies have successfully reported the accuracy of CEM for the characterization of gastrointestinal, dermatologic and ocular diseases. Researchers have assessesed the potential use of endomicroscopy in combination with clinically approved fluorophores to characterize premalignant and malignant lesions in human larynx. Imaging of squamous cell carcinoma provided clear information on the heterogeneous distribution of tumour cells surrounded by stroma. Cellular anomalies and disorders of keratinisation such as dyskeratosis and keratin pearls were also discerned by CEM and the images corroborated with histological data .
Optical diagnosis of the head and neck is a rapidly developing area of clinical research that can be readily translated to inform patient treatment and overall quality of life. Much still needs to be achieved and granting organisations are directed to pay attention to this specialty where relatively small investments may lead to enormous dividends in terms of improvements in treatments throughout the fields of medicine and surgery.
To the organisers, speakers and delegates who provided a powerful and rich forum to critically discuss the latest developments in their fields.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.