Journal of Gastroenterology and Hepatology (2002) 17 (Suppl.) S85–S104

QUADRENNIAL REVIEW

New optical technologies for earlier endoscopic diagnosis of premalignant gastrointestinal lesions RALPH S DACOSTA,* BRIAN C WILSON* AND NORMAN E MARCON † *

*Department of Medical Biophysics, University of Toronto, Ontario Cancer Institute/University Health Network, Toronto and †St Michael’s Hospital, Center for Therapeutic Endoscopy & Endoscopic Oncology, Toronto, Ontario, Canada

Abstract Gastrointestinal malignancies continue to be the second leading cause of cancer-related deaths in the developed world. The early detection and treatment of gastrointestinal preneoplasms has been demonstrated to significantly improve patient survival. Conventional screening tools include standard white light endoscopy (WLE) and frequent surveillance with biopsy. Well-defined endoscopic surveillance biopsy protocols aimed at early detection of dysplasia and malignancy have been undertaken for groups at high risk. Unfortunately, the poor sensitivity associated with WLE is a significant limitation. In this regard, major efforts continue in the development and evaluation of alternative diagnostic techniques. This review will focus on notable developments made at the forefront of research in modern gastrointestinal endoscopy based on novel optical endoscopic modalities, which rely on the interactions of light with tissues. Here we present the ‘state-of-the-art’ in fluorescence endoscopic imaging and spectroscopy, Raman spectroscopy, optical coherence tomography, light scattering spectroscopy, chromoendoscopy, confocal fluorescence endoscopy, and immunofluorescence endoscopy. These new developments may offer significant improvements in the diagnosis of early lesions by allowing for targeted mucosal excisional biopsies, and perhaps may even provide ‘optical biopsies’ of equivalent histological accuracy. This enhancement of the endoscopist’s ability to detect subtle preneoplastic changes in the gastrointestional mucosa in real time and improved staging of lesions could lead to curative endoscopic ablation of these lesions and, in the long term, improve patient survival and quality of life. © 2002 Blackwell Publishing Asia Pty Ltd Key words: ALA, diagnosis, dysplasia, endoscopy, fluorescence, immunophotodetection, light scattering spectroscopy, monoclonal antibodies, optical coherence tomography, premalignant, Raman.

INTRODUCTION At present, when patients present with symptoms of obstruction, pain or bleeding due to cancer, the lesion is usually large, easily identified by endoscopy or radiology, and most likely to be at an advanced stage with reduced chance for cure. The ideal scenario would be to have a reliable marker of early neoplastic disease (i.e. fecal, serological, or urinary) to better select asymptomatic patients for endoscopic examination. The ability to detect subtle lesions still confined to the mucosa would then permit their cure by either endoscopic ablation or minimally invasive surgery.

It is well recognized that conventional endoscopy using white light does not detect dysplasia and that subtle lesions (i.e. flat adenomas) may be missed. Furthermore, the ability to detect dysplasia within fields of transformed mucosa, such as Barrett’s esophagus and long-standing chronic ulcerative colitis, is a major clinical challenge and remains a strong motivation to develop new endoscopy systems to complement white light endoscopy. Optically based endoscopic technologies detect relative changes in the way light interacts with tissue along the disease transformation pathway (i.e. from dysplasia through to invasive cancer). Whereas conventional

Correspondence: Dr NE Marcon, St Michael’s Hospital, Center for Therapeutic Endoscopy & Endoscopic Oncology, 16-062 Victoria Wing, 30 Bond Street, Toronto, Ontario, M5B 1W8, Canada. Email: [email protected]

S86

RS DaCosta et al.

endoscopy is limited to detecting lesions based on gross morphological changes, these new optically based devices offer the potential of detecting the very earliest mucosal changes at the microstructural, biochemical and molecular levels.

LIGHT-INDUCED FLUORESCENCE Principles Tissue illuminated with ultraviolet (UV) or short wavelength visible (VIS) light from a laser source or a filtered lamp, emits fluorescence light of a longer wavelength. This tissue ‘autofluorescence’ arises from endogenous molecules within the tissue, called fluorophores, such as connective tissues (collagen, elastin), cellular metabolism related coenzymes (reduced nicotinamide adenine dinucleotide (NADH), flavin adenine dinucleotide (FAD), and flavin mononucleotide (FMN)), aromatic amino acids (tryptophan, tyrosine, phenylalanine), by-products of heme biosynthesis (porphyrins) and lipopigments (lipofuscin, ceroids). Different excitation wavelengths activate different groups of fluorophores, each of which emits at a range of different wavelengths (Table 1). Tissues also contain molecules called chromophores that absorb light without re-emission of fluorescence. Tissue absorption is strongly wavelength dependent and may significantly modify the in vivo fluorescence signal observed at the tissue surface due to a ‘filtering’ of both the excitation and fluorescence emission light. The main chromophore in gastrointestinal tissues in the visible wavelength (400–700 nm) range is hemoglobin. Tissues are composed of a complex milieu of several fluorophores that occur in different concentrations and, in the gastrointestinal (GI) tract, at different depths. Thus, the mucosa, submucosa and muscularis propria have distinct fluorophore compositions, so that even though each fluorophore has a distinct fluorescence

Table 1

spectrum, the fluorescence measured at the luminal surface comprises contributions from the fluorophores in the various layers.1,2 Excitation and fluorescence emission wavelength bands are often broad, relatively featureless and overlap with one another, so that identifying individual fluorophores in a given tissue spectrum is difficult. Changes in the intrinsic fluorescence of the tissue layers that occur during disease transformation are due to alterations in their biochemical composition (i.e. metabolic state, microenvironment).1–4 In addition, changes in the mucosal layer thicknesses (resulting from preneoplastic cellular growth) or in the blood concentration collectively contribute to the differences seen between normal and diseased tissues. Thus, the use of autofluorescence to detect early cancers or premalignant GI lesions is dependent on changes in one or more of the following factors: (i) the tissue architecture (mucosal thickening or loss of layered structures); (ii) the light absorption and scattering properties of each layer, particularly hemoglobin in the capillary networks; (iii) the distribution and concentration of fluorophores in the different layers; (iv) the biochemical microenvironment of the tissue that may alter the fluorescence yield or spectral shape, and (v) the metabolic status of the tissue (e.g. NADH is only fluorescent in the reduced form). Hence, although complex, tissue autofluorescence is sensitive to alterations in tissue morphology and biochemistry, resulting from malignant transformation. There is a strong wavelength dependence in the degree to which these changes alter the in vivo fluorescence measurements, as the excitation and emission wavelengths used determine the dominant fluorophores involved. For example, tissue proteins, composed of amino acids, are autofluorescent only when excited by UV wavelengths, while all other fluorophores, such as those above, are also excited by visible wavelengths. Additionally, in general, as the excitation wavelength increases, so too does the penetration depth of the excitation light. Thus, there may be one or more optimal excitation or emission wavelength bands depending on the anatomical site or application. Unfor-

Some known tissue endogenous fluorophores and wavelengths of their excitation and emission maxima

Endogenous fluorophore Tryptophan Phenylalanine Tyrosine Collagen elastin NADH FAD, flavins Pyridoxine Pyridoxal-5¢-phosphate Porphyrins Ceroid, lipofuscin

Biological source Amino acids

Structural proteins Metabolic cofactors Vitamin B6 compounds By-product of heme biosynthesis; bacterial fauna Lipopigment granules; age related; lipid oxidation products

Wavelength of max. fluorescence excitation (nm)

Wavelength of max. fluorescence emission (nm)

280 260 275 330 350 340 450 330, 340 330 400–450

350 280 300 390 420 450 515 400 400 635, 690

340–395

430–460, 540–640

New technologies for endoscopic diagnosis tunately, for each type of tissue the optimal excitation and emission wavelengths are not known a priori, and must be determined from ex vivo tissue samples. This has the advantage of providing detailed fluorescence spectroscopy and microscopy data, but the accuracy of the results must be considered carefully as the measured spectra may be affected by altered microvasculature (hemoglobin) or metabolic changes.1 The measurement of excitation and emission matrices (EEM) of known and suspected biological fluorophores in solu has also helped identify the optimal excitation and emission wavelength bands for tissue autofluorescence spectroscopy.5 The other option for fluorescence diagnosis is to use exogenous fluorescent drugs. Clinical studies to date have largely exploited the selective localization in neoplastic tissue of photosensitizers used in photodynamic therapy (PDT), many of which are also fluorescent. Examples include hematoporphyrin derivative (HpD), tetra(m-hydroxyphenyl)chlorin (mTHPC), phthalocyanines (PC), benzoporphyrin derivative (BPD) and tin etiopurpurin (SnET2). The advantages of using druginduced fluorescence are that the signal is strong compared to autofluorescence, so that the instrumentation may be simpler and cheaper, and that the optimum excitation and emission wavelengths are known a priori, which allows the tissue autofluorescence ‘background’ to be subtracted. However, the diagnostic efficacy of such agents is dependent on the degree of selective localization of the drug within (premalignant) lesions. Additionally, the procedures must be performed at the optimum time for the drug pharmacokinetics, which may vary from patient to patient and tissue to tissue. Different drugs, methods of delivery (e.g. antibodies and liposomes) or modes of administration (e.g. topical or systemic) may be necessary for different anatomical sites or disease states. Use of fluorescent drugs also involves extra costs and regulatory issues. Although much lower drug doses are required for diagnosis than are typically used therapeutically, another concern may be cutaneous photosensitization. A compound that has shown promise to obtain good fluorescence contrast between normal and neoplastic tissues is 5-aminolevulinic acid (ALA). ALA is a ratelimiting precursor in the heme biosynthesis pathway, so that when it is administered to the patient it increases heme synthesis, the penultimate step of which is the porphyrin compound protoporphyrin IX (PpIX), which has a distinct red fluorescence emission between 625 and 725 nm. In general, PpIX appears to accumulate preferentially in malignant tissues. The reasons for this are not well understood. In vitro observations of the differential PpIX synthesis induced by ALA in tumor and normal cells are contradictory,6 while considerable variation in PpIX fluorescence in vivo has been reported from lesion to lesion in the same patient7 and strong dependence on the anatomical site has also been observed.8 Suggested mechanisms of preferential accumulation for PpIX accumulation in tumors include deficiencies in iron or in ferrochelatase, an enzyme required for conversion of PpIX to heme.9 Enhancement of PpIX accumulation may be accomplished by the administration of exogenous compounds such as the

S87 iron chelator desferrioxamine10 that inhibit PpIX conversion to heme by ferrochelatase, and therefore may allow for a lower ALA dose. ALA has shown rapid metabolic breakdown,11 and few side-effects, with a short duration of skin photosensitivity,12 and chemical modification of ALA (e.g. ALA esters) has shown some promise in increasing tissue penetration.13 However, selecting the best drug dose and time delay before fluorescence detection remains problematic,14 as with other exogenous fluorophores.

Clinical experience Two primary endoscopic methods (point spectroscopy and imaging) have been developed to investigate tissue fluorescence in vivo. Fluorescence point spectroscopy involves the use of a contact optical fiber probe (single or multiple fibers, typically several meters in length) that delivers excitation light (from a laser or wavelength filtered lamp) to the tissue surface and collects the resulting fluorescent light. Typically, a central delivery fiber illuminates the tissue, while a surrounding circular array of fibers collects the emitted fluorescence. Optical filters block the detection of scattered excitation light. The fiber bundle probe, typically 0.5–1 mm in diameter, is delivered endoscopically via the biopsy channel and placed in contact with the tissue surface of suspicious lesions and visibly normal mucosa seen under white light endoscopy (WLE) control. Fluorescence light is separated into component colors by a spectrograph, and displayed as a fluorescence intensity versus wavelength curve (Fig. 1). Various mathematical procedures can be applied to these spectra to extract diagnostic information, such as taking the ratio of two or more fluorescence emission wavelength bands. The effectiveness of some diagnostic algorithms to date has been shown to depend on anatomic site, however, so that different algorithms may be required for different segments of the GI tract. During endoscopy, point fluorescence measurements must be guided by white light viewing of the luminal surface. The white light background overwhelms the fluorescence signal, so the measurements must be performed either while the white light is temporarily turned off, by using a combination of pulsed excitation sources and fast-gated detectors, or by measuring the alternating-current fluorescence signal using an intensity-modulated excitation light source. Laser-induced fluorescence point spectroscopy was the first approach used for tissue autofluorescence diagnosis in the GI tract. The majority of the initial studies involved tissue analysis ex vivo, particularly in the colon. In 1990, Kapadia et al. used 325 nm excitation on excised human colon tissues, applying multivariate linear regression analysis of the fluorescence emission band (350–600 nm), and demonstrated accuracies of 100%, 94%, and 100% for identifying normal, hyperplastic, and adenomatous tissues, respectively.15 In 1991, Richards-Kortum et al. measured excitation and emission matrices in ex vivo colon polyps, using a range of excitation and emission wavelengths.16 The optimal excitation wavelengths for discrimination of normal

S88

RS DaCosta et al.

(a)

(b) 3500

3500

3000

Fluorescence/Intensity (a.u.)

3000 2500

2500

2000

2000

1500

1500

1000

1000

500

500

0 400

450

500

550

600

650

700

750

800

0 400

500

550

600

650

700

750

800

(d)

(c) 160

600 Fluorescence/Intensity (a.u.)

450

140

500 120

400

100

300

80 60

200

40

100

20

0 400

0 450

500

550

600

650

700

750

800

400

450

Wavelength (nm)

colon from adenomatous polyps were around 330, 370, and 430 nm. Subsequently, they showed that adenomas could be identified accurately against normal colon using 370 nm excitation and emission wavelengths of 404, 480 and 680 nm. In 1992, Schomacker et al. confirmed that fluorescence spectroscopy could be used to differentiate between non-neoplastic and neoplastic colon tissues with a sensitivity of 80% and a specificity of 92%.17 In the first in vivo study, reported by Cothren et al. (1996) 370 nm excitation was used and a 2-D scatter plot of fluorescence intensities at two emission wavelengths (460 and 680 nm) allowed differentiation of adenomas and non-adenomatous colon tissue in 97% of cases. The first blinded in vivo fluorescence spectroscopy study of the colon reported by this group identified the correct tissue type in 88% of cases, with a sensitivity of 90% and a specificity of 95%.18 The first in vivo fluorescence spectroscopy study of the esophagus was reported in 32 patients in 1995 by Panjehpour et al. Using 410 nm excitation, a stepwise multivariate discriminate analysis was used to select the best fluorescence emission wavelengths to differentiate normal from malignant tissues.19 Classification of tissue spectra as either normal or malignant was achieved with a sensitivity of 100% and a specificity of 98%. In a later study, in 36 patients with Barrett’s esophagus, the spectra were analyzed using a method based on the fluorescence intensity at 480 nm compared with the fluorescence integrated over all wavelengths. Specifically, a model based on a so-called differential normalized fluorescence (DNF) index was developed. Briefly, a set of fluorescence spectra (430–716 nm) from 15 patients

500

550

600

650

Wavelength (nm)

700

750

800

Figure 1 Examples of point fluorescence spectra collected in vivo from normal colon, hyperplastic polyp, adenomatous polyp and colon adenocarcinoma, using 437 nm excitation. (a), Normal colon; (b), hyperplastic polyp; (c), adenomatous polyp; (d), colon adenocarcinoma.

with normal squamous esophagus were normalized to the same total integrated intensity, and used to determine an average normalized fluorescence spectrum to be used as a baseline. The DNF spectrum for each new fluorescence spectrum was then determined by subtracting this baseline spectrum. DNF index values at 480 nm for normal and cancerous tissues were selected and graphed in a scatter plot. The two populations representing normal and malignant DNF indices could be separated on this scatter plot, with an accuracy of 96% for non-dysplastic Barrett’s esophagus; 100% for low-grade dysplasia (classified as benign); and 90% for high-grade dysplasia (classified as premalignant). Seventy-two percent of low-grade dysplastic-containing focal high-grade dysplasia lesions, however, were incorrectly identified as benign. In the same study, seven of seven patients with high-grade dysplasia, but none of six patients with low-grade dysplasia, were identified. Hence, this system could not detect low-grade lesions or focal high-grade dysplasia. To date, the use of the DNF algorithm has been limited to the esophagus. Despite shortcomings, this important study demonstrated the potential to identify endoscopically occult premalignant lesions by autofluorescence point spectroscopy to guide sites for targeted biopsy. Most of the in vivo point spectroscopy studies to date have involved steady-state fluorescence measurements. An alternative approach based on time-resolved autofluorescence spectroscopy was demonstrated recently when Mycek et al. reported the in vivo results of 17 patients with 24 polyps (13 adenomatous, 11 nonadenomatous).20 Time-resolved fluorescence spectra represent the decay of fluorescence intensity at a given

New technologies for endoscopic diagnosis

S89

emission wavelength as a function of time after a brief pulse of excitation light (~10 ns). The decay time of colonic adenomas was shorter than that of nonadenomas, yielding an 85% differential sensitivity, 91% specificity, 92% positive predictive value, and 83% negative predictive value. Further evaluations are required to confirm the clinical effectiveness of this technique in other anatomic sites. Figure 2 shows our own preliminary time-resolved data for normal squamous, Barrett’s and esophageal carcinomas, which indicate a significant diagnostic difference with disease state. Georgakoudi et al. have described another point spectroscopy system where they combine multiple fluorescence emission spectra and reflectance in assessing 16 patients with known Barrett’s esophagus undergoing standard WLE surveillance.21 This protype produced a fast EEM of tissue, in vivo, provided by 11 different excitation wavelengths (337–620 nm) generated by a 337 nm nitrogen-dye laser which pumped 10 different dye cuvettes on a rapidly rotating wheel. Reflectance spectra were also collected from the same tissue area using white light (350–700 nm) coupled into the same delivery fiber which was touched to the mucosal surface. The reflectance spectra were used to correct the measured fluorescence spectra for artefacts induced by changes in tissue scattering and absorption, rather than by tissue biochemistry, as reflectance is affected by the same scattering and absorption processes. Differences in fluorescence emission spectra identified by statistical algorithms (principle component analysis) allowed differentiation of high-grade dysplasia (HGD) from low-grade dysplasia (LGD) and non-dysplastic Barrett’s sites (NDB). Intrinsic fluorescence spectroscopy had a sensitivity of 100% and a specificity of 97% in distinguishing HGD versus (LGD and NDB), and a sensitivity of 79% and a specificity of 88% in distinguishing (LGD and HGD) versus NDB. However, despite these impressive values, the present paper did not discuss differentiation between LGD versus HGD, and LGD and NDB. Despite the potential benefits and relative technical simplicity of fluorescence point spectroscopy, a major drawback is the sampling of only a small volume of tissue (~1–3 mm3) immediately beneath the probe tip. Although capable of taking multiple readings quickly, this technique is dependent on placing the probe in the right spot (as for biopsy).Targeted sampling is currently only possible with lesions that are visible. This inherently limits the sensitivity. In addition, unlike imaging, the spectral information lacks contextual information, which is often needed to distinguish abnormal from surrounding normal tissues.

ARB INT

RESID

ACORR

ARB INT

RESID

ACORR

ARB INT



Figure 2 Fluorescence decay curves of human normal squamous esophagus, Barrett’s esophagus, and esophageal adenocarcinoma following a brief (10 nsec) light pulse, showing differences in the decay times of each tissue type. In each panel, the inner curve is the excitation light pulse, and the outer curve is the fluorescence emission profile. Excitation wavelength = 340 nm; Emission wavelength = 350–500 nm. Note the steeper slope for the Barrett’s esophagus and esophageal carcinoma.

RESID

ACORR

S90 Instruments that can produce fluorescence images of high resolution and in real time, to complement WLE, have recently become available, enabling in vivo imaging to interrogate large areas of the mucosal surface in parallel with WLE examinations.WLE images are made from primarily diffuse reflectance (with some contribution from specular reflectance), when the different wavelength components of a broadband (white) light source are multiply scattered and absorbed in tissue. In WLE the illuminating and reflected light photons have the same wavelength, as opposed to fluorescence where the emitted fluorescence photons are longer in wavelength compared with the (near) monochromatic excitation light. As the light scattering and, particularly, absorption in the tissue are wavelength dependent, the different wavelengths effectively interrogate the tissue to different depths (red being the most and blue being the least penetrating). The resultant image provides the endoscopist with visual clues to the tissue surface topography and underlying patterns of vasculature. Fluorescent images are formed using only selected emission wavelength bands. There are various ways in which this can be achieved. For example, the emission bands can be selected using special optical filters, and then detected by separate cameras to form the final displayed fluorescence image. Thereby, real-time, falsecolor fluorescence images of the tissue can be viewed, switching rapidly between this and WLE. Such a system for fluorescence imaging was first demonstrated for the bronchus as a screening tool for dysplasia and carcinoma in high-risk patients. This led to a commercial system (LIFE-Lung; Xillix Technologies Corp., Richmond, BC, Canada), which used blue-light excitation, with separate red and green fluorescence imaging channels. Using LIFE in combination with WLE, the detection of moderate to high-grade bronchial dysplasia was increased by 171%, compared with white-light bronchoscopy alone, with only a 22% decrease in specificity. Currently, prototype imaging systems derived from the original lung device are being evaluated for GI endoscopy, and initial clinical feasibility studies have been reported by ourselves and other collaborating groups.22–25 The most recent prototype uses a detachable camera module, which connects to the optical head of a conventional fiberoptic endoscope. The module contains two individual high sensitivity cameras, one for green (490–530 nm) and one for red (590–700 nm) fluorescence. The digital images are combined to produce a real-time, false-color image, where normal tissue generally appears green and abnormal tissue appears red. The ratio of red to green fluorescence is standardized over normal mucosa. The current system can be switched rapidly (~4 s) between WLE and LIFE allowing rapid surveying of wide areas of the mucosa. Sensitivity and specificity values are determined by correlating the positive (red fluorescence) and negative (green fluorescence) images with histological diagnosis of corresponding biopsy samples. In a preliminary study, we demonstrated detection of atypia or higher grade lesions in the colon with a sensitivity and specificity of 87% and 79%, respectively. In the non-inflamed colon, we had no false negatives.

RS DaCosta et al. False-positive lesions have been seen with acute colitis, both infective and idiopathic. Diminutive polyps are extremely common and because the endoscopist cannot usually determine whether these are hyperplastic or adenomatous, these polyps are removed or fulgurated. This can be labor intensive. The LIFE imaging system can differentiate between hyperplastic polyps, which have no known malignant potential, and adenomatous polyps, which should be removed (Fig. 3).Table 2 shows our results of a preliminary comparison between WLE and WLE combined with LIFE in 26 patients with nondysplastic and dysplastic colonic lesions. Although most of these lesions were detected as nodules or polyps by WLE, the ideal is to detect flat dysplastic lesions not recognized by obvious topographic irregularity. Detection of lesions of this type is the ultimate goal of any effective screening program and would have a significant impact on survival. Villous adenomas are often sessile and can cover several square centimeters of colonic mucosa. Their removal usually involves the submucosal injection of saline prior to polypectomy. Complete removal is sometimes problematic and recurrence is sometimes difficult to identify with WLE in the area of the scar. Fluorescence endoscopy, however, facilitates this nicely. Although cancer of the esophagus in most Western countries is relatively uncommon in comparison with cancer of the colon, the 5-year survival rate is only 5%, primarily because most patients at the time of diagnosis have symptomatic dysplasia, advanced disease and therefore reduced survival despite intensive surgical and radiotherapy protocols. Considerable interest is being focused on Barrett’s and the rising incidence of adenocarcinoma in the distal esophagus. The main selection factor to identify patients with Barrett’s is those patients with symptoms of reflux, in whom the screening protocol involves multiple random biopsies. Panjehpour et al. first reported on fluorescence spectroscopy to detect high-grade dysplasia within Barrett’s.5 We have used the LIFE imaging system in a number of patients with Barrett’s (Fig. 4). Although we have detected areas of dysplasia occult to previous WLE in a number of Barrett’s patients, further development is required to reduce the false-positive rate that is due to the Barrett’s epithelium having abnormally high red fluorescence.23

Table 2 Comparison of WLE and WLE + LIFE in differentiating between hyperplastic and adenomatous colonic polyps using the Xillix LIFE-GI imaging system in 64 patients*

Sensitivity Specificity Positive predictive value Negative predictive value Accuracy

WLE (%)

WLE + LIFE (%)

80 69 59 87 71

95 80 71 97 85

WLE, white light endoscopy; GI, gastrointestinal. *Taken from Song WK et al. Am. J. Gastroenterol. 2001; 96 (Suppl.): S1–S376.

New technologies for endoscopic diagnosis To date, several key reports have been published on the clinical effectiveness of prodrugs, such as ALA, to detect preneoplastic GI lesions. Messmann et al. (1998) reported an experimental ulcerative colitis (UC) rat model in which ALA-PpIX fluorescence (UV excitation; i.v. administration of ALA at 25–200 mg/kg bodyweight; visualization 1–4 h later) was used to enhance the detection of low-grade dysplastic lesions.26 The best results were obtained with 75 mg/kg ALA, for which dysplasia (LGD and HGD) was detected with a sensitivity of 92%, although the specificity was disappointing at 35%. Focal lesions as small as five abnormal crypts were observed in ex vivo tissue samples, which is a significant finding, as aberrant crypts are thought to be the earliest precursors to dysplastic lesions in the colon.27,28 The major limitation of this study was that UC also showed appreciable concentrations of PpIX, resulting in high false-positive rates, although identification of low-grade dysplasia against an active colitis background may be difficult even for the skilled pathologist using biopsied tissues.29 Lowering the ALA dose increased the specificity, with a notable decrease in the false-positive rate, although at the price of reduced sensitivity, which may have been due, in part, to the use of the unaided eye for detecting the fluorescence. Messmann et al. (1999) suggested improving the technique by topical administration of the ALA, by optimizing the time period between drug administration and fluorescence observation, and by attempting to detect highgrade dysplasia in quiescent colitis. An alternative may be to keep the ALA dose low and increase the detection sensitivity, for example, using a high sensitivity imaging system (see below).14 Another recent article by Messmann et al. (1997) reports six patients with dysplastic lesions in the GI tract, two of whom had Barrett’s, one with low-grade and the other with high-grade dysplasia.30 The patients drank 10–20 mg/kg ALA and the endoscopic examination was performed 1–6 h later using blue light excitation and compared with conventional WLE. In this limited and selected group of patients, Messmann et al. claimed that low-dose ALA facilitated the identification/differentiation of dysplastic lesions in a background of Barrett’s, and showed histological proof that the red fluorescent area was dysplastic. Surprisingly, an area of adjacent ulceration did not fluoresce red, despite the fact that inflammation is a confounding feature that is thought to be responsible for high false-positive rates in a variety of imaging systems. In this study, a 40% falsepositive biopsy rate was attributed to inflammation. Low-dose oral ALA resulted in higher background red fluorescence in normal duodenal and squamous epithelia compared with normal stomach mucosa and Barrett’s epithelium. However, this is contradicted by recent findings of van den Boogert et al. in a rat model of Barrett’s using intravenous ALA, where it was concluded that there was no selectivity in ALA-induced fluorescence between normal squamous and adjacent Barrett’s mucosa.31 Studies are also currently underway at our institutions to assess the combined role of ALA in diagnostic fluorescence imaging (and ablation using PDT) in Barrett’s esophagus. Interestingly, preliminary data

S91 on LIFE imaging using very low-dose oral ALA (i.e. ~2 mg/kg) have shown selectivity of ALA in Barrett’s mucosa relative to squamous mucosa that was not observed at higher doses (Fig. 5). ALA-PpIX fluorescence has also been studied in a variety of other organs, such as the bladder,32 oral cavity,33 and brain.34 Peng et al. recently reviewed the use of ALA for PDT, while Marcus et al. summarized both the clinical and preclinical development of ALA as a fluorescence diagnostic agent.6,35 Further studies will continue to investigate the complementary aspects of exogenous fluorescent compounds and fluorescent imaging systems with conventional endoscopy. Their use in everyday endoscopic practice will be influenced by extra costs, regulatory approval, possible drugrelated toxicities, and overall cost-effectiveness as part of a screening program.

RAMAN SCATTERING SPECTROSCOPY Principles Among the other spectroscopic techniques under investigation, Raman spectroscopy provides the most detailed information about the molecular composition of tissue. Sampling tissues up to a depth of about 500 mm, the spectral features are much narrower than those observed in fluorescence and reflectance spectroscopy, and they are molecule specific. However, the Raman effect is more difficult to implement than fluorescence, mainly because tissue Raman signals are much weaker than autofluorescence, and they can be masked by the broad-band fluorescence background. In addition, special optical fiber probes are required to minimize the fluorescence and Raman signals generated in the probe itself. The Raman effect is an inelastic light-scattering process. Most light photons are scattered in tissue without a change in frequency (elastic scattering that is responsible for reflectance). However, a very small fraction of the incident light is scattered inelastically such that the photon energy is thereby changed. As photon energy is proportional to frequency, this Raman scattered light is shifted to a lower frequency (i.e. longer wavelengths). Using a spectral analyzer the molecular information contained in the Raman emission spectrum can be extracted. As every molecule possesses a unique pattern of Raman spectral peaks, the molecular composition of a tissue sample can be accurately determined. More exactly, the population of common molecular bonds can be identified. Considering that the onset of cancer is accompanied by changes in chemical composition, Raman scattering is a potentially powerful diagnostic technique. The targets of Raman scattering include a wide range of specific tissue molecules. Proteins, lipids and nucleic acids all exhibit distinct Raman signatures. The use of UV light to induce Raman spectra is possible. This produces tissue autofluorescence that is well separated from the detected Raman spectral bands, which allows filtering to remove the unwanted

S92

RS DaCosta et al.

Figure 3 In vivo examples of WLE and autofluorescence endoscopic imaging of normal colon, a hyperplastic polyp, an adenomatous polyp, and corresponding histopathology. With fluorescence, normal and hyperplastic mucosa appear in false-color as cyan, while the adenomatous polyp appears red. Images were taken using Xillix LIFE-GI imaging system (437 nm excitation, green (490–540 mn) and red (> 590 mn) emission).

fluorescence. Also, in UV resonance Raman (UVRR) spectroscopy, certain Raman spectral bands can be appreciably amplified using excitation light corresponding to a particular absorption band. However, UV excitation light is potentially mutagenic, and compared with near-infrared (NIR) light, UV light does not penetrate deeply into tissues. Near-infrared light (~700–1300 nm) penetrates deeply into tissue, making this technique well suited to probing early GI lesions (depth of 500 mm), and NIR minimizes tissue fluorescence compared with visible light. In a paper by Shim and Wilson, potential artefacts in Raman spectroscopy due to handling of ex vivo tissue samples collected from biospy were determined.36 It was concluded that tissue samples should be immediately frozen and prior to Raman spectroscopy, the tissue should then be acclimatized at room temperature in phosphate-buffered saline (PBS) and immersed in PBS during spectroscopic examination. This study demonstrated that Raman spectroscopy and microscopy could be performed on biopsied tissues. Using a specially designed optical probe, ex vivo spectra were able to differentiate dysplasia from intestinalized dysplasia in esophageal biopsies with a sensitivity of 77% and a specificity of 93%.

Clinical experience Several investigators have used different configurations to acquire in vivo Raman spectra, but there are few reports for the GI tract.37 Recently, Shim et al. designed and built a near-infrared (NIR) fiberoptic device for in vivo Raman spectroscopy (IVRS) measurements and reported the first in vivo Raman spectra of human GI tissues measured during routine clinical endoscopy with acceptable signal-to-noise ratio (SNR) and short collection times.38 This was achieved by using the system with an optically filtered fiberoptic probe (~2 mm diameter) capable of ‘beam-steering’ that was passed through the endoscope instrument channel and placed in contact with the tissue surface. Background signals were suppressed and the light collection efficiency was optimized.39 Spectra were obtained with good SNR in ~5 s. The effects on the spectra of varying the pressure of the probe tip on the tissue and of the probe-tissue angle were insignificant. The spectra from normal and diseased tissues revealed only subtle differences. For example, initial ex vivo studies on Barrett’s esophagus samples (207 nondysplastic; 53 dysplastic) demonstrated slight differences in spectral line-shapes in the 1100–1800/cm range, but no specific prominent changes in peak in-

New technologies for endoscopic diagnosis

S93 tensity or position. Sophisticated computational techniques such as principle component analysis and artificial neural networks are being investigated; these have the advantage that they analyze the full Raman spectrum and, hence, the full diagnostic content in order to distinguish subtle spectral differences among these spectra. One type of neural network analysis differentiated esophageal dysplasia from metaplasia with a sensitivity of 77% and a specificity of 93%. These diagnostic algorithms are being refined to improve upon these results and, furthermore, to give the best discriminating power in classifying the various dysplastic grades (indefinite versus LGD versus HGD). Once optimized, these algorithms can be used in a prospective fashion to assess the potential of Raman spectroscopy for Barrett’s tissue differentiation. Moreover, the recent demonstration of the feasibility of obtaining in vivo Raman spectra of the GI tract (Fig. 6) is a critical step in initiating systematic clinical trials to determine the diagnostic accuracy of Raman spectroscopy in BE. Although Raman microscopic imaging can be done ex vivo, the weakness of the signals may never allow realtime endoscopic in vivo Raman imaging.

LIGHT ELASTIC SCATTERING SPECTROSCOPY Principles Light scattering spectroscopy (LSS) is, in principle, a relatively simple optical technique that provides structural information about tissue. Elastic scattering spectroscopy is based on white light reflectance, in which photons incident on tissue are back scattered without a change in wavelength. The measurement of the relative back-scattering intensity over the wavelength spectrum of visible light has been shown to be sensitive to both tissue scatterers, such as cell nuclei and mitochondria and tissue absorbers such as hemoglobin. By separating the contribution of multiply scattered light from deep in the tissue (using spectral subtraction or crosspolarization) information can be obtained primarily from the mucosa. Changes in the density and/or size of scatterers associated with tissue transformation can be measured from the fine structure in the spectra and correlated with corresponding histopathological diagnosis. As a tissue becomes dysplastic the nuclei enlarge and become crowded and LSS is able to measure the size distributions of epithelial cell nuclei, providing direct quantitative measurement of nuclear enlargement, crowding, and hyperchromaticity that can aid in clinical diagnoses. The main advantages of this



Figure 4 In vivo example of (a) WLE and (b) autofluorescence endoscopic imaging and (c) corresponding histopathology of Barrett’s esophagus. Images were taken using Xillix LIFE-GI imaging system (437 nm excitation, green (490–540 mn) and red (> 590 mn) emission).

S94

RS DaCosta et al. LSS is a limited sampling volume by the optical probe. Nevertheless, rapid (< 1 s) spectroscopic readings can be obtained with LSS,40–42 and development of LSS imaging is in progress.

Clinical experience Wallace et al. assessed the potential of in vivo LSS to detect epithelial nuclear crowding and enlargement in BE.41 A component of the spectra (representative of elastically scattered light from epithelial nuclei) was extracted mathematically and used to estimate nuclear size and number. Dysplasia was diagnosed if 30% or more of the nuclei exceeded ~10 mm in diameter. Using this threshold, the sensitivity and specificity for detecting dysplasia (LGD and HGD) were both 90%. Further studies testing the capabilities to differentiate between various grades of dysplasia are awaited.21,41

OPTICAL COHERENCE TOMOGRAPHY Principles

Figure 5 In vivo example of (a) WLE image and (b) corresponding fluorescence image of Barrett’s esophagus using low dose (2 mg/kg, 6 h) oral ALA-induced PpIX. Normal squamous mucosa appears cyan, while selective uptake of ALA in Barrett’s mucosa causes bright red fluorescence outlining the Barrett’s field.

non-invasive method arise from a significantly stronger signal collected in real time (< 1 s for a spectroscopic reading) and the use of white light, instead of a laser. This results in a simplified and less expensive spectroscopic detection system. Unfortunately, similar to point fluorescence spectroscopy, the primary drawback of

Optical coherence tomography (OCT) is a novel biomedical imaging technique, which enables crosssectional imaging with high spatial resolution (~10– 20 mm) and in near real time. Optical coherence tomography is analogous to B-scan ultrasonography, but the images are formed by detecting light (as opposed to sound) that is back-reflected from subsurface tissue microstructures. The resolution of current OCT systems is nearly 10-fold greater than that of highfrequency endoscopic ultrasound (EUS), high enough to identify microscopic features such as villi, glands, crypts, lymphatic aggregates and blood vessels. The price to be paid for this remarkable resolution is a limited imaging depth of about 2 mm. Optical coherence tomography is based on the principle of low-coherence interferometry. Briefly, lowcoherence infrared light is split by a 50/50 beamsplitter, with half of the light directed towards the tissue and half towards a moveable mirror whose exact location is known. Light reflected from the mirror and tissue is recombined in the beamsplitter and directed to a detector. Using a light source with limited coherence length, an interference signal is produced only when the light that returns from the tissue travels the same distance as the light returning from the mirror. The interference signal depends on the magnitude of the back-reflected light intensity from tissue microstructures at that particular depth. By varying the distance of the mirror, the tissue can be scanned at varying depths to produce an axial scan (A-scan). A 2-D image (B-scan) is then obtained by performing repeated A-scan measurements at different lateral positions as the optical probe is scanned across tissue. The axial (depth) resolution is determined by the coherence length of the light source, whereas the lateral resolution is determined by the focusing beam optics. In current endoscopic

New technologies for endoscopic diagnosis

S95

6 5

Intensity (au)

4 3 2 1 0 −1 −2 −3 −4 800

1000

1200 1400 1600 wavelength shift (cm−1)

1800

2000

Figure 6 In vivo examples of mean Raman spectra collected from normal squamous mucosa, Barrett’s esophagus and highgrade dysplasia/adenocarcinoma in Barrett’s. (Red line), Normal mucosa; (blue line), intestinal metaplasia; (black line), high grade/adenocarcinoma.

OCT prototypes, these are about 10 and 25 mm, respectively. The current in vivo endoscopic OCT probes (2–2.4 mm diameter) are similar to EUS catheter probes, but with varying designs allowing end-on scanning, linear scanning (along the longitudinal axis of the esophagus) or 360° radial scanning. For images to be in focus, the working distance should be within 1 mm from the surface. With probe-tissue contact, compression artefacts may be seen. Image acquisition rates of 4 frames per second, avoiding motion blurring, have been achieved to date, although in principle real-time imaging is possible and development is in progress.

Clinical experience Using a linear scan OCT, Bouma et al. clearly delineated, in situ, the normal esophageal layers up to the muscularis propria, which were correlated with ex vivo specimens.43 In BE, the most obvious features were an absence of the layered structure of normal esophagus and abnormal glandular morphology. The lamina propria, muscularis mucosa, and deeper structures were typically not seen due to loss of signal from light scattering within the metaplastic epithelium. Barrett’s adenocarcinoma was characterized by the presence of pronounced architectural disorganization. Pitris et al. examined the feasibility of OCT for highresolution imaging of gastrointestinal malignancies with ex vivo imaging of normal squamous esophagus, BE, squamous carcinoma, ulcerative colitis, normal colon and colonic adenocarcinoma.44 Stratified squamous epithelium of the normal esophagus was clearly visible and contrasted with the disorganized and nonuniform nature of the mucosal layers of Barrett’s eso-

phagus and squamous carcinoma. The columnar epithelial morphology as well as other mucosal structures in normal colon were also distinct. In contrast, disorganization of the normal mucosal layers and ulcerative lesions were identified in UC and colon adenocarcinomas. The ability of OCT to image tissue microstructure at high resolutions showed that it may be useful for minimally invasive assessment of the GI tract and evaluation of early neoplastic changes. Using a radial scan OCT, Sivak et al. demonstrated different layers (not yet precisely identified) of normal esophagus. The evaluation of BE using this scanning technique was not reported. Although these OCT techniques produced high-resolution images of the esophagus, and one of them clearly demarcated the structural alterations that may be seen with adenocarcinoma, they have not yet provided diagnostic information relevant to the detection of dysplasia in BE.45 In its current form and resolution, OCT will likely localize areas displaying architectural distortion for biopsy, but will be limited in staging dysplasia and in differentiating inflammatory/reactive changes from dysplastic lesions. However, one can speculate that OCT may determine tumor invasion through muscularis mucosa into the submucosa better than EUS, and hence improve upon the T-staging for selecting patients for endoscopic cure. Improvements in both axial and lateral resolutions are expected, with the development of better light sources and focusing beam optics, respectively. Newer sources can improve the axial resolution to subcellular levels (< 5 mm), which may allow dysplastic cells to be resolved. Currently, these sources are cumbersome and expensive, but are a starting point for the development of high power, compact and affordable lasers. Incidentally, one drawback of the linear scanning OCT technique is its small sampling volume, similar to point spectroscopy.The maintenance of the OCT probe at a specific site, amid peristaltic motion, also appears challenging, so that contact with tissue (with potential compression artefacts) may be necessary. Optical coherence tomography is not likely to replace biopsy and histopathology, but it may bring endoscopy and histopathology together in a single technology. The future OCT endoscopist will need to be familiar with histopathological changes accompanying tissue transformation, which may initiate a new breed of ‘endoscopic pathologists’.

CHROMOENDOSCOPY Principles Chromoendoscopy should be a very simplistic and widely used technique for enhancing mucosal detail by spraying a variety of colored solutions, and has received a lot of attention in the last 5 years in the Western world. The stains are mainly absorptive dyes (methylene blue and Lugol’s) or contrast agents (indigo carmine). These agents enhance detection of subtle mucosal irregularities that may not be colored or may be slightly depressed. The stains are usually sprayed with specially

S96 designed catheters, can be ingested or given as enemas. The effects of these stains can be visualized either by conventional white light endoscopes or by video endoscopes with image enhancement, or high power magnification. Absorptive stains are taken up selectively by epithelial cells. The common agents are Lugol’s solution and methylene blue. Toluidine blue and cresyl violet (also called gentian violet) are less commonly used.

Clinical experience Lugol’s solution has been shown to be effective in identifying early squamous cell cancer in the esophagus.46 Lugol’s is a mixture of potassium iodide and iodine. Lugol’s reacts with glycogen in non-keratinized squamous epithelium and results in a dark greenish-brown color.The normal esophageal mucosa is richly endowed with glycogen. Dysplastic areas are deficient in glycogen and are therefore non-stained and readily seen as pale yellow islands. Inflammation also causes nonstaining areas. Japanese endoscopists have used Lugol’s spraying to increase the yield of early esophageal cancer in patients who already have a diagnosis of head and neck cancer. Ina et al. screened a large number of patients with head and neck cancers comparing conventional white light endoscopy followed by Lugol’s spraying.47 In this study, eight superficial cancers were detected. Five of the eight were not recognized by conventional WLE and only identified with Lugol’s. Although Lugol’s gave high sensitivity, it was also associated with a high rate of false-positive unstained lesions that were not dysplastic. Most of these lesions were caused by esophagitis. This mapping of dysplastic lesions with Lugol’s can also enhance the accuracy in carrying out submucosal resections. Vital staining with Lugol’s is also useful in surveying the esophageal mucosa after ablative therapies as for Barrett’s high-grade dysplasia with PDT. These residual non-staining islands can then be thermally ablated. Methylene blue is taken up by actively absorbing epithelial cells of the small intestine and colon. It stains the intestinal metaplasia that is pathognomonic for Barrett’s. It will not stain nonabsorptive epithelium of the stomach as is found in the cardia and fundus, or normal squamous mucosa of the esophagus. Spraying with methylene blue is a little more complicated and time intensive. The Barrett’s segment must first be cleared of mucus. This is accomplished by spraying the surface with 10% acetylcysteine. This is left in contact for 2 min, then sprayed with 1 or 2% methylene blue, which is left undisturbed for up to 3 min. Then vigorous washing with saline is carried out to clear excess dye. Methylene blue staining has almost a 100% correlation with Barrett’s epithelium. Therefore, biopsies can be more efficiently and economically directed than with the traditional fourquadrant biopsy protocol. The big question is whether methylene blue staining leads to detection of more areas of dysplasia? These areas are thought to be identified by

RS DaCosta et al. zones of decreased staining within a field of homogeneous blueness. These color changes are unfortunately subjective and not standardized. Further studies will be needed to clarify this problem. All studies have been carried out with conventional video endoscopes. The use of image enhancement or magnification endoscopes to improve dysplasia detection continues to be problematic. Methylene blue is also used for its surface demarcation to access the mosaic pattern in patients with celiac disease. Methylene blue spraying is also used in the colon to demarcate aberrant crypt foci but must be used with magnification scopes.29 Adenomatous aberrant crypts are thought to be precursor lesions for adenomas. Another vital stain is indigo carmine.This is inert and not absorbed. Most of its reported use has been to detect diminutive, flat and depressed lesions in the colon using magnification endoscopes. Kudo et al. have reported extensively and developed a classification system of pit patterns, where pit pattern 1 is normal and pit pattern 5 is interpreted as dysplasia and submucosal invasion.48 Indigo carmine is sprayed at concentrations from 0.1 to 1%. It gives a bluish contrast to the intestinal mucosa and clearly demonstrates the folds and crevices, and facilitates identification of flat and depressed adenomas.49 A recent landmark publication by Saitoh et al. reports on a selected group of 211 American patients undergoing screening colonoscopy where spraying with 0.08% indigo carmine increased by 65% the diagnosis of flat and depressed lesions using a conventional nonmagnifying colonoscope. Although most of these lesions were hyperplastic polyps and diminutive, a small proportion (5%) had dysplasia.50 Indigo carmine has been promoted initially by Japanese endoscopists using magnification colonoscopes. They report the ability to accurately predict when a diminutive lesion is hyperplastic or adenomatous in about 90% of the time. Their assumption was that hyperplastic lesions had no malignant potential, could be ignored and, therefore, not be biopsied or removed. We believe that all lesions regardless of their interpretation with magnification, as long as they are 1 cm or larger, should be removed, because the magnification capabilities are not 100% and some of these lesions may be serrated adenomas. Magnification endoscopy is not suitable for screening but only to better define small lesions previously suspected with dye spraying. Indigo carmine spraying has improved detection of dysplasia in Barrett’s using a magnification gastroscope.51 In conclusion, chromoendoscopy with vital staining should be a simple and widespread technique to augment the endoscopist’s diagnostic ability. These agents are safe, inexpensive and can be used with a conventional endoscope. However, they are also associated with subjectivity and poor standardization. Magnification may not be required. Whether high-resolution videoscopes will improve the diagnosis of dysplasia using dye spraying remains to be determined. Lately, computer software manipulation of contrastdyeenhanced endoscopic images has been presented; however, it is not yet clear whether these will be useful

New technologies for endoscopic diagnosis

S97

tools or simply endoscopic ‘toys’. Where chromoendoscopy fits with presumed expensive and sophisticated optically based systems, such as fluorescence, remains to be seen.

CONFOCAL FLUORESCENCE MICROENDOSCOPY Principles Distinct tissue structures can be visualized using fluorescence microscopy. Conventional epifluorescence microscopy allows the examination of thin unstained frozen sections (~5 mm thick), or tissue sections stained with fluorescently labeled agents. However, a major limitation of epifluorescence microscopy is the difficulty in examining thicker specimens, such as whole frozen or fresh biopsies, as structures above or below the focal plane (out-of-focus light) cause blurring. Confocal fluorescence microscopy (CFM), on the other hand, offers a clear advantage over conventional epifluorescence microscopy. Pinholes are placed confocally at the source and detector to reduce the ‘out-offocus’ light and the image is formed point-by-point by a scanning laser beam. This results in high-resolution fluorescent optical ‘slices’ through the sample. By varying the axial position of the focal plane through the specimen, a stack of optical sections can be collected without physical sectioning of the tissue sample enabling the complex 3-D fluorescence structure of the sample to be determined. Confocal fluorescence microscopy may be used to image the fluorescence light from endogenous or exogenous fluorophores as well as the reflected light from tissues. The maximum interrogation depth depends on the excitation wavelength, the optical properties of the tissue, and the collection efficiency of the detection optics. Generally, CFM of tissue biopsies (~1–3 mm3) can image as deep as 100–300 mm, using 488 nm excitation light. We have demonstrated the use of 3-D CFM imaging of ex vivo normal esophagus, Barrett’s esophagus, gastric and colon biopsies, optically sectioning these from the luminal surface, yielding 3-D rendered volume of the mucosa. Figure 7 shows examples for (a) normal squamous esophagus (b) Barrett’s esophagus, and (c) normal gastric mucosa. The surface topologies are very different for each tissue type. The total field of view in each image is ~0.65 mm2. The differences in the autofluorescence emission patterns of each tissue type are 

Figure 7 3D confocal fluorescence micrograph of fresh ex vivo esophageal, Barrett’s and gastric whole biopsies.Views are from the luminal surface of (a) normal squamous esophagus (b) Barrett’s esophagus and (c) gastric antrum. 458 nm excitation, green (500–530 mn) and red (> 585 nm) emission. Note the squamous epithelia and dermal papillae (DP) in the normal esophagus, the disorganized glandular structure and increased red fluorescence in the Barrett’s mucosa, and the individual pits in the gastric mucosa. (Scale bar 200 mm).

S98

RS DaCosta et al.

also clearly seen, with the fluorescence from Barrett’s mucosa exhibiting more red fluorescence than either the squamous or gastric mucosa. The surface microarchitectural details of each tissue type are clearly seen as well, allowing differentiation between each type. In other studies, we used blue light excitation (~458 nm) CFM to compare the tissue fluorescence sources and microdistributions within each tissue layer of a given tissue type, and to compare the differences between tissue types (Fig. 8). In squamous mucosa, the stratified arrangement of the squamous cells can be seen, despite their weak green fluorescence intensity. The surface layer of cells was slightly more fluorescent than lower mucosal cells. Strong green fluorescence from collagen fibers was observed in the submucosa. In Barrett’s mucosa, numerous small red fluorescent ‘granules’ were observed in the lamina propria, and shown by immunohistochemistry to be macrophages and endocrine cells. Glands in the Barrett’s mucosa are observed to contain weakly green fluorescent cells. The submucosal layer was highly green fluorescent due to structural collagen. In the gastric tissue sample, gastric pits were clearly observed, and a band immediately below the luminal surface was weakly fluorescent. Very bright orange-red fluorescent cells are observed in the lower mucosa, and were shown by immunohistochemistry to be enzyme-secreting cells and endocrine cells. Submucosal collagen is the most intensely fluorescent tissue component. Confocal fluorescence microscopy has shown that squamous mucosa, Barrett’s mucosa and gastric mucosa differ significantly in autofluorescence features whereas the submucosal layers of each tissue type are similar. This suggests that a fluorescence imaging device that was sensitive to ‘mucosal autofluorescence’ alone may differentiate between these tissue types and be able to identify early intramucosal lesions. Preliminary studies on ex vivo tissue biopsies demonstrated that is was possible to achieve superior image quality and depth resolution in the mucosal layer of the upper and lower GI.52 Inoue et al. reported the use of confocal fluorescence microscopy to obtain immediate microscopic images from fresh specimens of gastrointestinal mucosa.53 Briefly, fresh untreated mucosal specimens from the esophagus, stomach and colon, obtained by endoscopic pinch biopsy, polypectomy or endoscopic mucosal resection (EMR) were fixed in normal saline and examined by CFM with 488 nm excitation, in reflectance 

Figure 8 False-color confocal autofluorescence in longitudinal cross sections of (a) normal esophagus (b) Barrett’s esophagus, and (c) gastric antrum frozen tissue sections (437 nm excitation; green = 505–580 nm, and red = 580–700 nm). Squamous mucosa appears weakly green fluorescent. Barrett’s mucosa appears more red fluorescent with highly fluorescent cells in lamina propria. The glandular structure of Barrett’s can be seen. Intensely orange-red autofluorescent cells (likely parietal cells) can be seen in the normal gastric mucosa.While the connective tissue matrix (collagen) of the submucosa fluoresces bright green in all samples. (M: mucosa; SM: submucosa)

New technologies for endoscopic diagnosis

Figure 9 Example of in vivo fluorescence image using anticc49 monoclonal antibody labeled with NIR fluorophore for selective targeting of a xenografted human colonic tumor in a nude mouse implanted with LS174T human colon adenocarcinoma cell line. Note the bright false-color fluorescence at the tumor sight, and at the site of injection in the tail-vein.

mode. Images were compared with conventional H&E staining, analyzing the nucleus-to-cytoplasm ratios.The overall diagnostic accuracy of CFM for cancer was 89.7%. This study showed it was possible to obtain immediate microscopy of fresh tissue biopsies, which were uncut. The obvious advantages of ‘blur-free’ fluorescence imaging and 3-D optical sectioning capabilities of ex vivo biological tissues has made CFM an attractive concept for in vivo fluorescence endoscopic imaging. To date there has been little published information on confocal endoscopic devices. Fiberoptic confocal imaging (FOCI) was developed by Optiscan Inc. (Victoria, Australia) for subsurface microscopy of the colon in vivo.54 This system has been used to take optical sections of the mucosal surface of the rat colon in vivo using a prototype rigid probe following topical application of a fluorescent dye, with the colon surgically exposed. The laser beam scanning mechanism is a dual high-speed galvanometer scanner fitted with super reflective dielectric-coated front surface silvered mirrors, giving a lateral resolution of 220 nm, and an axial resolution of 450 nm using a 60 ¥, 1.4 numerical aperture (NA) oil objective lens at 488 nm. The images have a 100-mm field of view (12.5 mm at 8 ¥ zoom) with scanning speeds of up to 16 frames/s. This prototype is

S99 believed to be the predecessor to a true CFM endoscopic imaging device, but this has not been reported to date. Advancements in silicon-based microelectronic micromachined systems (MEMS) may allow further miniaturization of the confocal scanning mechanisms for endoscopy. A miniaturized confocal laser scanning microscope has been reported by Laser- und MedizinTechnologie GmbH, Berlin, Germany. This uses a 2MEM scanning unit to produce a 2-D scan. Typical technical characteristics of the device are display field (0.7 ¥ 0.7 mm2), optical resolution (~2 mm), scan area 2048 ¥ 2048 pixels, and the physical dimensions are 11 ¥ 14.5 ¥ 35 mm3. Other developments in MEMSbased confocal endoscopy are in progress.55–57 The research group led by Dr Arthur Gmitro, at the University of Arizona, has developed a real-time confocal fluorescence endoscopic imaging device using 488 nm argon laser light. This uses a 30 000 fiberoptical imaging bundle with an active area of 720 mm. A miniature objective and focusing mechanism are used at the distal end of the catheter. The field of view is ~430 mm2, and the fiber-limited lateral resolution is 2 mm. Focusing is accomplished via a hydraulic mechanism that moves the distal end of the fiber relative to the lens. A widefield imaging channel is included for an overall view of the tissue, comprising four illumination fibers and a smaller imaging fiber bundle and a color charged coupled device (CCD) camera. Experiments were performed in cell cultures, ex vivo tissue samples, and in mouse and rat animal models (i.e. rat liver stained with the fluorescent dye acridine range). Unpublished results (http://www.optics.arizona.edu/gmitro/) using high quantum yield fluorescence dyes are impressive. Despite the promise of confocal fluorescence microendoscopic technologies, and the continual technical advances that permit further miniaturization, this technology is yet to be demonstrated in vivo in human endoscopic trials.

IMMUNO-RELATED FLUORESCENCE CONTRAST AGENTS Principles Fluorescent dyes are routinely coupled to tumor-related or tumor-specific antigens for immunohistochemical staining of biopsied tissues. This concept may be extended to enhance the contrast in tumor-to-normal tissue endoscopic imaging by targeting tumors with monoclonal antibodies. For example, radiopharmacology has, for the past two decades, used the high specificity of antigen–antibody reactivity and the differential expression of tumor-associated antigen relative to normal tissues to target tumor cells selectively with radiotherapeutic agents.58 Development of ‘immunophotodetection’ agents involves the conjugation of a fluorophore dye to a monoclonal antibody or other tumor-targeting moiety. Typically, the dyes are near-infrared emitting, have high fluorescence emission efficiencies, produce fluorescence

S100 that is detectable through millimeter thicknesses of tissues, and have adequate stability for labeling in vivo.59 Recent improvements in monoclonal antibodies and their derivatives, the development and commercial availability of near-infrared emitting fluorophores and the availability of high sensitivity digital cameras in this spectral region have made tumor localization using fluorescence practical. Optimal fluorescent dyes can be selected based on their photophysical and spectral properties independent of their tumor-localizing properties.60 Recent studies in animals have demonstrated that fluorophore labeling of monoclonal antibodies produces adequate sensitivity and improved resolution. For example, in separate studies, far red-emitting cyaninebased dyes were conjugated to the tumor-targeting monoclonal antibodies, anticarcinoembryonic antigen (CEA) and monoclonal antibody (MAb) E48 directed against squamous cell carcinoma. These were injected into nude mice bearing xenografted human tumors which overexpressed the respective antigen.61,62 Specific tumor localization ranged from 15 to 30% of the injected dose per gram of tumor at 24 h post injection, and clearly detectable red fluorescence contrasted the tumors against the minimal-to-non-fluorescent normal tissues. No detectable red fluorescence was observed in control mice injected with conjugated irrelevant antibodies or free dye. However, many difficulties remain with this approach. Until recently, most monoclonal antibodies were raised in non-human hosts (mice) resulting in an immune response against them when used in patients. This not only causes the antibodies to be quickly eliminated, but also forms immune complexes that cause damage to the kidneys.63 However, ‘humanized’ monoclonal antibodies are now becoming available. Studies have shown that antibody bound in human tumors does not exceed 10–5 of the administered dose per gram of tumor, therefore requiring large amounts of injected conjugated monoclonal antibody, long exposure times and high sensitivity to achieve adequate tumor brightness and contrast.59 The major limitation is due to the pharmacokinetic properties of conjugated whole antibodies. Ramjiawan et al. demonstrated specific binding of a fragment of antihuman antibody with broad cancer specificity, conjugated to an NIR-emitting dye (Cy5.5) in a mouse xenograft model.64 Here, peak fluorescence intensity was detected with a high sensitivity CCD camera 2 h post injection. The presence and distribution of the conjugated fragment revealed that about 16% and 73% of the agent was located in the tumor and the kidneys, respectively. Use of smaller antibody fragments produces rapid tumor uptake, better penetration (at the expense of reduced circulation time), more homogeneous tumor penetration and reduced immunogenicity.65,66 Fluorescent dyes can also be targeted to tumor tissues by means other than monoclonal antibodies. For example, Weissleder et al. coupled a NIR fluorophore to a biocompatible polymer which was taken up by tumor cells via pinocytosis.67 The intracellular release of the fluorophore in tumor-bearing mice by the protease cathepsin D in the tumor resulted in a fluorescence

RS DaCosta et al. signal detected in vivo in subnanomolar quantities and at depths sufficient for clinical imaging. Thus, they demonstrated that specific enzyme activity in a tumor could be imaged by fluorescence contrast agents in vivo. Currently, work in our laboratory is assessing the utility of colonic mucins as a possible target for colonic adenomas and adenocarcinomas. Briefly, nude mice were xenografted with human colonic adenocarcinoma LS174T cells. The monoclonal antibody cc49, which recognizes a tumor-associated glycoprotein antigen, was conjugated to a commercial fluorescent dye (AlexaFluor647; Molecular Probes, Eugene, OR, USA) at a molar ratio of 6 mol dye : 1 mol antibody, as was an irrelevant monoclonal antibody against (M195, antiCD33). The conjugated relevant and irrelevant antibodies and free dye of equivalent concentration were administered i.v., each at a dose of ~30 mg per mouse. Using 630 nm excitation, and a high sensitivity CCD camera with fluorescence filters, preliminary results demonstrated distinct contrast enhancement of the tumor compared with surrounding normal tissues using the labeled cc49, compared with control autofluorescence images.Tumor visualization was apparent as early as 2 h in the cc49, but the maximum contrast was at 48 h postinjection (Fig. 9). Conjugated irrelevant antibody and free dye did not produce detectable florescence contrast at the tumor site. Hence, this demonstrated the selective targeting of fluorescence dye to tumorassociated mucins, resulting in the enhanced fluorescence detection of small (~4–5 mm diameter) xenografted human colonic tumors.

Clinical experience Preliminary work has been done on ex vivo human tissues using indocyanine green (ICG)-labeled antibody against epithelial membrane antigen (EMA) and an infrared flexible endoscope, with 768 nm excitation and 807 nm detection.68 This demonstrated fluorescent antibody staining of ex vivo normal human esophageal tissues, but did not evaluate diseased tissues. Preliminary evaluation of fluorescence contrast agents in patients has been reported in a limited number of studies. Early vascular changes were assessed in Crohn’s disease in a prospective endoscopic study of 10 asymptomatic patients with unconjugated 10% sodium fluorescein.69 Fluorescence endoscopy was used to evaluate the mucosal microcirculation of the neoterminal ileum in relation to endoscopic recurrence in patients who had undergone ileo-colonic resection for Crohn’s disease. The fluorescence observed may reflect vasodilation associated with inflammation, or genuine microvascular lesions. Correlation with histology suggested that these early vascular lesions were secondary to the inflammatory process.69 Tatsuta et al. labeled anti-CEA monoclonal antibodies with fluorescein isothiocyanate (FITC) to study ex vivo human gastric lesions. Fluorescein isothiocyanate has a high fluorescence efficiency and favorable excitation and detection wavelengths (~488 nm excitation, ~520 emission).70 In vitro fluorescent images of 42 gastric lesions were obtained including 20 early carci-

New technologies for endoscopic diagnosis nomas and 10 advanced cancers. The conjugated antibody was applied topically. Twenty-seven (90%) of 30 tumors showed positive fluorescence after 60 min, with no false-positives, whereas only two of five cancers (40%) could be detected earlier than 60 min, which resulted in 50% false-negatives for 30 min. To remove gastric mucus and improve the binding of the tumor with the labeled antibody, pretreatment with a mixture of proteinases, sodium bicarbonate and dimethylpolysiloxane was used. In vivo, this would add another 90 min to the endoscopic examination. Computerassisted processing was used to enhance the fluorescence images. No significant relationship between positive fluorescence and tumor type or stage was found. However, positive fluorescence could not be demonstrated in benign gastric lesions. In 1998, Keller et al. coined the term ‘immunoscopy’ in a report on the detection of colorectal carcinomas and villous adenomas in surgically resected tissue samples. Fluorescence from FITC-labeled anti-CEA antibody was detected using a sensitive filtered photographic camera in 27 of 28 cancers and in one of two adenomas, as well as in six of 18 normal controls, giving a sensitivity of 93% and specificity of 67%.71 To date, the only published report of fluorescenceconjugated monoclonal antibodies to be used in humans in vivo used a monoclonal fluoresceinated antiCEA conjugate to detect human colon carcinoma.72 Following i.v. administration of the mouse–human chimeric anti-CEA monoclonal antibody, ex vivo photodetection of the tumors was accomplished in six patients, and in vivo fluorescence rectosigmoidoscopic imaging in one patient. Upon laser irradiation, clearly detectable heterogeneous green fluorescence from the dye–antibody conjugate was visually observed on all six tumors; almost no such fluorescence was detectable on normal mucosa. Tissue autofluorescence, emitted from both tumor and normal mucosa, was subtracted by realtime image processing. The overall results demonstrate the feasibility of tumor immunophotodiagnosis at the clinical level. However, despite these encouraging initial results, several important issues must be resolved. Selection of the best tumor-associated targets (i.e. monoclonal antibodies, peptides, enzymes) is not clear, and the possibilities are seemingly endless. For example, antigens expressed on the cell surface such as growth factor receptors, mucins and cell adhesion molecules can be targeted by their respective fluorescence-conjugated antibodies, as can intracellular markers such as enzymes.73,74 It is possible that each segment of the GI tract will have its own specific diagnostically relevant markers. Additionally, simultaneous localization of multiple reagents is made possible by labeling multiple NIR fluorochromes; thus background subtraction and differential labeling of multiple tumor-associated components can be performed. Difficulties in using the fluorochrome labels are mainly related to light scattering and absorption in tissues, but detection of small tumors at depths of several millimeters should be feasible. Given the limitations in current fluorescence endoscopic imaging in detecting very early GI lesions or preventing false-positives due to confounding con-

S101 current conditions (i.e. inflammation), these developments significantly complement existing fluorescence endoscopy.

THE OPTIMAL TECHNIQUE Several new optically based techniques are being evaluated to enhance the diagnostic capability of clinical GI endoscopy. The ideal system should function in real time and combine excellent focal and diagnostic accuracy with wide mucosal area surveillance. A major issue is how the detection of dysplasia and intramucosal cancer will ultimately fit into the treatment algorithm: for example, who and/or what should be treated with endoscopic ablation, chemoprevention or resective surgery? Treatment will be markedly affected by accurate staging of lesions, via super high-resolution ultrasound or OCT. Short of replacing conventional biopsy, such technologies should provide guidance in locating optimal sites for targeted biopsy and be able to monitor ablative therapies such as photodynamic therapy. In this regard, fluorescence endoscopic imaging with its wide field of view, has already detected early lesions, scars, and demonstrated reliability in differentiating hyperplastic versus adenomatous polyps in vivo, and so appears most appealing and practical for screening. Additionally, fluorescence endoscopy does not require dye spraying and is relatively fast. However, many issues, such as optimal excitation and emission wavelength(s), confounding background metaplastic fluorescence (false-positives), and artefacts due to motility, remain unresolved. Additionally, it is not clear if exogenous fluorophores, such as prodrugs like ALA, will be necessary to achieve clinically useful sensitivity and specificity. Despite its very high molecular specificity, Raman spectroscopy suffers the same weakness as all point spectroscopies in that its clinical use is limited by practicality. This is also the case for LSS, which has shown promise in differentiating dysplasia (LGD and HGD) from Barrett’s esophagus based on nuclear size and density. However, used adjunctively with imaging techniques which survey large tissue surfaces for targeting suspicious lesions, the molecular specificity of Raman spectroscopy or the sensitivity to subcellular scattering features of LSS may be useful for in situ diagnosis.These combinations are yet to be attempted. Optical coherence tomography is attractive, but current OCT prototypes have several limitations that prevent their use as a stand-alone technique for surveillance. The main clinical advantage of OCT is the ability to stage mucosal disease, as a means of identifying those patients where dysplasia and intramucosal cancer does not penetrate into the submucosa and, therefore, would be ideal for curative endoscopic therapy. Although it has the potential of yielding histological details this resolution has not yet been achieved in a real-time endoscopic system. Additionally, OCT will only be applicable for viewing small areas of the GI tract. However, with anticipated improvements in resolution (subcellular level) and speed, OCT may become the technique of choice for surveillance and staging in the future.

S102 At the moment, CFM has only been demonstrated on ex vivo human GI tissues including normal, metaplastic and preneoplastic lesions in the esophagus, stomach and colon. Distinct fluorescence differences have been found between normal and abnormal mucosal tissues in each organ, yet this is likely not to be diagnostically useful in endoscopic fluorescence imaging, as the already weak mucosal fluorescence is overwhelmed by very strong fluorescence from lower GI tissue layers. Currently it is the technology that is preventing the clinical utility of ‘confocal microendoscopy’. All point spectroscopic techniques, as well as magnification endoscopy, are inherently limited by the small tissue area they sample. However, they contain more detailed information about tissue than any imaging system, which may translate into more accurate tissue differentiation. Rather than competing with an imaging system, the ‘best’ instrument for surveillance may combine spectroscopy and imaging. For instance, a lesion could be detected by fluorescence imaging or OCT, and its dysplastic nature characterized by Raman spectroscopy. However, in this era of cost containment, such an approach may be prohibitive. Moreover, all of these expensive optical modalities will need to be compared against cheaper and equally promising alternatives such as chromoendoscopy, for which the dye is cheap and colonoscopes are readily available. However, dye spraying is labor intensive. By far, the least reported method to date is the use of immune-related fluorescence contrast agents. A limited number of ex vivo studies have demonstrated relative GI tumor selectivity with highly fluorescent conjugated antibodies to well-known tumor-associated biomarkers. Such contrast agents have also been evaluated in a very limited number of patients with encouraging enhancement of tumor contrast. There are important technical issues to be resolved, for example, finding the optimum ‘site- and pathology-specific’ biomarkers, conjugate design, optimizing the relative tumor uptake, cost and safety issues. However, advances in tumor-associated GI biomarkers, conjugation biochemistry, safety assessments, and fluorescence imaging hardware and software continue. This technology also offers the means of aiding our fundamental understanding of disease processes in the GI tract on a molecular level. It is conceivable that in the future, molecular-targeted fluorescence endoscopic imaging will allow earlier detection and characterization of GI disease, and may offer assessment of treatment effects. ‘Optical biopsy’ refers to tissue diagnosis based on in situ optical measurements, which would eliminate the need for tissue removal. The above-mentioned optical techniques are striving towards this goal, but none are likely to replace conventional biopsy and histopathological interpretation in the near future. Although they demonstrate potential for better diagnosis, these modalities are still in their infancy, with future technological refinement and large-scale clinical trials needed to assess their utility and limitations. To date, no commercial systems have withstood the test of comparative clinical trials. Ultimately, whether these optical tech-

RS DaCosta et al. niques will become part of standard clinical endoscopic practice or remain on the sidelines can be summed up in two questions: how much better will they perform and at what cost?

REFERENCES 1 DaCosta RS, Lilge L, Kost J et al. Confocal fluorescence microscopy/macroscopy and microspectrofluorimetry analysis of human colorectal tissues. J. Anal. Morph. 1997; 4: 24–9. 2 Zonios GI, Cothren RM, Arendt JT et al. Morphological model of human colon tissue fluorescence. IEEE Trans. Biomed. Eng. 1996; 43: 113–22. 3 Marchesini R, Pignoli E, Tomatis S et al. Ex vivo optical properties of human colon tissue. Lasers Surg. Med. 1994; 15: 351–7. 4 Cothren RM, Richards-Kortum R, Sivak MV et al. Gastrointestinal tissue diagnosis by laser-induced fluorescence spectroscopy at endoscopy. Gastrointest. Endosc. 1990; 36: 105–11. 5 Panjepour M, Overholt BF, VO-Dinh T, Haggitt RC, Edwards DA, Buckley FR III. Endoscopic fluorescence detection of high-grade dysplasia in Barrett’s esophagus. Gastroenterology 1996; 111: 93–101. 6 Peng Q, Berg K, Moan J, Kongshaug M, Nesland JM. 5-Aminolevulinic acid-based photodynamic therapyprinciples and experimental research. Photochem. Photobiol. 1997; 65: 235–51. 7 Martin A, Tope WT, Grevelink JC et al. Lack of selectivity of protoporphyrin-IX fluorescence for basal-cell carcinoma after topical application of 5-aminolevulinic acid: implications for photodynamic treatment. Arch. Dermat. Res. 1995; 287: 665–74. 8 Loh CS, MacRobert AJ, Bedwell J, Regula J, Krasner N, Bown SG. Oral versus intravenous administration of 5¢aminolaevulinic acid for photodynamic therapy. Br. J. Cancer 1993; 68: 41–51. 9 Battle AMC. Porphyrins, porphyrias, cancer and photodynamic therapy: a model of carcinogenesis. J. Photochem. Photobiol. B. 1993; 20: 5–22. 10 Fijan S, Honigsmann H, Ortel B. Photodynamic therapy of epithelial skin tumors using delta aminolevulinic acid and desferrioxamine. Br. J. Dermatol. 1995; 133: 282–8. 11 Rick K, Sroka R, Stepp H et al. Pharmacokinetics of 5¢Aminolevulinic acid-induced protoporphyrin IX in skin and blood. J. Photochem. Photobiol. B. 1997; 40: 313–19. 12 Webber J, Kessel D, Fromm D. Side effects and photosensitization of human tissue after aminolevulinic acid. J. Surg. Res. 1997; 68: 31–7. 13 Lange N, Jichlinski P, Zellweger M et al. Photodetection of early human bladder cancer based on the fluorescence of 5-aminolaevulinic acid hexylester-induced protoporphyrin IX. a pilot study. Br. J. Cancer 1999; 80: 185–93. 14 Messmann H, Knuchel R, Baumer W et al. Endoscopic fluorescence detection of dysplasia in patients with Barrett’s esophagus, ulcerative colitis, or adenomatous polyps after 5-aminolevulinic acid-induced protoporphyrin IX sensitization. Gastrointest. Endosc. 1999; 49: 97–100.

New technologies for endoscopic diagnosis 15 Kapadia CR, Cutruzzola FW, O’Brien KM et al. Laserinduced fluorescence spectroscopy of human colonic mucosa: detection of adenomatous transformation. Gastroenterology 1990; 99: 150–7. 16 Richards-Kortum R, Rava RP, Fitzmaurice M et al. Spectroscopic diagnosis of colonic dysplasia. Photochem. Photobiol. 1991; 53: 777–86. 17 Schomacker KT, Frisoli JK, Compton CC et al. Ultraviolet laser-induced fluorescence of colonic polyps. Gastroenterology 1992; 102: 1155–60. 18 Cothren RM, Sivak MV, Van Dame J et al. Detection of dysplasia at colonoscopy using laser-induced fluorescence: A blinded study. Gastrointest. Endosc. 1996; 44: 168– 76. 19 Panjehpour M, Overholt BF, Schmidhammer JL et al. Spectroscopic diagnosis of esophageal cancer: new classification model, improved measurement system. Gastrointest. Endosc. 1995; 41: 577–81. 20 Mycek MA, Schomacker KT, Nishioka NS. Colonic polyp differentiation using time-resolved autofluorescence spectroscopy. Gastrointest. Endosc. 1998; 48: 390–4. 21 Georgakoudi I, Jacobson BC, Van Dam J et al. Fluorescence, reflectance, and light-scattering spectroscopy for evaluating dysplasia in patients with Barrett’s esophagus. Gastroenterology 2001; 120: 1620–9. 22 Wagnieres GA, Star WM, Wilson BC. In vivo fluorescence spectroscopy and imaging for oncological applications. Photochem. Photobiol. 1998; 68: 603–32. 23 Haringsma J, Tytgat GN, Yano H et al. Autofluorescence endoscopy: feasibility of detection of GI neoplasms unapparent to white light endoscopy with an evolving technology. Gastrointest. Endosc. 2001; 53: 642–50. 24 DaCosta RS, Wilson BC, Marcon NE. Light-induced fluorescence endoscopy of the gastrointestinal tract. Gastrointest. Endosc. Clin. North Am. 2000; 10: 37–69. 25 Yano H, Iishi H, Tatsuta M. Diagnosis of early gastric cancers by endoscopic autofluorescence imaging system [Abstract]. Endoscopy 1996; 28: S29. 26 Messmann H, Kullmann F, Wild T et al. Detection of dysplastic lesions by fluorescence in a model of colitis in rats after previous photosensitization with 5-aminolaevulinic acid. Endoscopy 1998; 30: 333–8. 27 Bird RP. Role of aberrant crypt foci in understanding the pathogenesis of colon cancer. Cancer Lett. 1995; 29: 55–71. 28 Takayama T, Katsuki S, Takahashi Y et al. Aberrant crypt foci of the colon as precursors of adenoma and cancer. N. Engl. J. Med. 1998; 339: 1277–84. 29 Riddell RH, Goldman H, Ransohoff DF et al. Dysplasia in inflammatory bowel disease: standardization classification with provisional clinical applications. Hum. Pathol. 1983; 14: 931–68. 30 Messmann H, Mikvy P, Montan S et al. Endoscopic and microscopic fluorescence studies in patients with ulcerative colitis after 5-aminolevulinic acid sensitization [Abstract]. Gastroenterology 1997; 108: A506. 31 van den Boogert J, Houtsmuller AB, de Rooij FWM et al. Kinetics, localization, and mechanism of 5-aminolevulinic acid-induced porphyrin accumulation in normal and Barrett’s like rat esophagus. Laser Surg. Med. 1999; 24: 3–13. 32 Kriegmair M, Baumgartner R, Knuechel P et al. Fluorescence photodetection of neoplastic urothelial lesions fol-

S103

33

34

35

36

37 38

39

40 41

42 43

44

45

46

47

48

49 50

lowing intravesical instillation of 5-aminolevulinic acid. Urology 1994; 44: 836–41. Leunig A, Rick K, Stepp H et al. Fluorescence imaging and spectroscopy of 5-aminolevulinic acid-induced protoporphyrin IX for the detection of neoplastic lesions of the oral cavity. Am. J. Surg. 1996; 172: 674–7. Stummer WS, Stocker S, Wagner S et al. Intraoperative detection of malignant glioma by 5-ALA induced porphyrin fluorescence. Neurosurgery 1998; 42: 518–26. Marcus SL, Sobel RS, Golub AL et al. Photodynamic therapy (PDT) and photodiagnosis (PD) using endogenous photosensitization induced by 5-aminolevulinic acid (ALA): current clinical and developmental status. J. Clin. Laser Med. Surg. 1996; 14: 59–66. Shim MG,Wilson BC.The effects of ex vivo handling procedures on the near-infrared Raman spectra of normal mammalian tissues. Photochem. Photobiol. 1996; 63: 662– 71. Hanlon EB, Manoharan R, Koo TW et al. Prospects for in vivo Raman spectroscopy. Phys. Med. Biol. 2000; 45: 1–59. Shim MG, Wilson BC. Development of an in vivo Raman spectroscopic system for diagnostic applications. J. Raman Spectrosc. 1997; 28: 131–42. Shim MG, Wilson BC, Marple E, Wach M. A study of fiber-optic probes for in vivo medical Raman spectroscopy. Appl. Spectrosc. 1999; 53: 619–27. Bohorfoush AG. Tissue spectroscopy for gastrointestinal diseases. Endoscopy 1996; 28: 372–80. Wallace MB, Perelman LT, Backman V et al. Endoscopic detection of dysplasia in patients with Barrett’s esophagus using light-scattering spectroscopy. Gastroenterology 2000; 119: 677–82. Backman V, Wallace MB, Perelman LT et al. Detection of preinvasive cancer cells. Nature 2000; 406: 35–6. Bouma BE, Tearney GJ, Compton CC, Nishioka NS. High-resolution imaging of the human esophagus and stomach in vivo using optical coherence tomography. Gastrointest. Endosc. 2000; 51: 467–74. Pitris C, Jesser C, Boppart SA, Stamper D, Brezinski ME, Fujimoto JG. Feasibility of optical coherence tomography for high-resolution imaging of human gastrointestinal tract malignancies. J. Gastroenterol. 2000; 35: 87–92. Sivak MV, Kobayashi Y, Izatt JA et al. High-resolution endoscopic imaging of the GI tract using optical coherence tomography. Gastrointest. Endosc. 2000; 51: 474–9. Endo M, Sakakibara N, Suzuki H et al. Observation of esophageal lesions with the use of endoscopic dyes. Prog. Dig. Endosc. 1972; 1: 34. Ina H, Shibuya H, Ohashi I et al. The frequency of a concomitant early esophageal cancer in male patients with oral and oropharyngeal cancer: Screening results using Lugol dye endoscopy. Cancer 1994; 73: 2038–41. Kudo S, Tamura S, Nakajima T, Yamano H, Kusaka H, Watanabe H. Diagnosis of colorectal tumorous lesions by magnifying endoscopy. Gastrointest. Endosc. 1996; 44: 8–14. Kudo S, Tamure S, Nakajima T et al. Depressed type of colorectal cancer. Endoscopy 1995; 27: 54–7. Saitoh Y, Waxman I, West AB et al. Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population. Gastroenterology 2001; 120: 1657–65.

S104 51 Sharma P, Weston AP, Sampliner RE. Magnification chromoendoscopy for the detection of intestinal metaplasia and dysphasia in Barrett’s oesophagus [Abstract]. Gastrointest Endosc 2001; 53: AB62. 52 DaCosta R. Mechanisms of fluorescence endoscopy of the human colon. MSc. Thesis, University of Toronto, Toronto, Ontario, Canada. 2000. 53 Inoue H, Igari T, Nishikage T, Ami K, Yoshida T, Iwai T. A novel method of virtual histopathology using laserscanning confocal microscopy in-vitro with untreated fresh specimens from the gastrointestinal mucosa. Endoscopy 2000; 32: 439–43. 54 Delaney PM, King RG, Lambert JR, Harris MR. Fibre optic confocal imaging (FOCI) for subsurface microscopy of the colon in vivo. J. Anat. 1994; 184: 157–60. 55 Dickensheets DL, Kino GS. Micro-machined scanning confocal optical microscope. Opt. Lett. 1996; 21: 764–6. 56 Lane P, Dlugan ALP, Richards-Kortum R, MacAuley C. Fiber-optic confocal microscopy using a spatial light modulator. Opt. Lett. 2000; 25 (24): 1780–2. 57 Lane P, Dlugan ALP, MacAuley C. DMD-enabled confocal microscopy. In: Tuchin VV, Izatt JA, Fujimotto JG, eds. Coherence Domain Optical Methods in Biomedical Science and Clinical Applications V. Proceedings of BIOS 2001, 23–24 January 2001, San Jose, CA, USA. Bellingham WA: SPIE Publishers, 2001. 58 Reilly RM. Radioimmunotherapy of malignancies. Clin. Pharm. 1991; 10: 359–75. 59 Ballou B, Fisher GW, Hakala TR, Farkas DL. Tumor detection and visualization using cyanine fluorochromelabeled antibodies. Biotechnol. Prog. 1997; 13: 649–58. 60 Haringsma J, Tytgat GNJ. Fluorescence and autofluorescence. Bailliere’s Clin. Gastroenterol. 1999; 13: 1–10. 61 Pelegrin A, Folli S, Buchegger F, Mach JP, Wagnieres G, van den Bergh H. Antibody-fluorescein conjugates for photoimmunodiagnosis of human colon carcinoma in nude mice. Cancer 1991; 67: 2529–37. 62 Folli S, Westermann P, Braichotte D et al. Antibodyindocyanin conjugates for immunophotodetection of human squamous cell carcinoma in nude mice. Cancer Res. 1994; 54: 2643–9. 63 Nussbaum S, Roth HJ. Human anti-mouse antibodies: pitfalls in tumor marker measurement and strategies for

RS DaCosta et al.

64

65

66

67

68

69

70

71

72

73

74

enhanced assay robustness; including results with Elecsys CEA. Anticancer Res. 2000; 20: 5249–52. Ramjiawan B, Maiti P, Aftanas A et al. Noninvasive localization of tumors by immunofluorescence imaging using a single chain Fv fragment of a human monoclonal antibody with broad cancer specificity. Cancer 2000; 89: 1134–44. Yokota T, Milenic DE,Whitlow M, Schlom J. Rapid tumor penetration of a single-chain Fv and comparison with other immunoglobulin forms. Cancer Res. 1992; 52: 3402–8. Wu A,Williams LE, Zieran L et al. Anti-carcinoembryonic antigen (CEA) diabody for rapid tumor targeting and imaging. Tumor Targeting 1999; 4: 47–58. Weissleder R, Tung CH, Mahmood U, Bogdanov A Jr. In vivo imaging of tumours with protease activated nearinfrared fluorescent probes. Nat. Biotechnol. 1999; 17: 375–8. Muguruma N, Ito S, Hayashi S et al. Antibodies labeled with fluorescence-agent excitable by infrared rays. J. Gastroenterol. 1998; 33: 467–71. Maunoury V, Mordon S, Geboes K et al. Early vascular changes in Crohn’s disease: an endoscopic fluorescence study. Endoscopy 2000; 32: 700–5. Tatsuta M, Iishi H, Ichii M et al. Diagnosis of gastric cancers with fluorescein-labeled monoclonal antibodies to carcinoembryonic antigen. Lasers Surg. Med. 1989; 9: 422–6. Keller R, Winde G, Eisenhawer C et al. Immunoscopy—a technique combining endoscopy and immunofluorescence for diagnosis of colorectal cancer. Gastrointest. Endosc. 1998; 47: 154–61. Folli S, Wagnieres G, Pelegrin A et al. Immunophotodiagnosis of colon carcinomas in patients injected with fluoresceinated chimerical antibodies against carcinoembryonic antigen. Proc. Natl Acad. Sci. USA 1992; 89: 7973–7. Krishnadath KK, Reid BJ, Wang KK. Biomarkers in Barrett esophagus. Mayo Clin. Proc. 2001; 76: 438– 46. Srivastava S, Verma M, Henson DE. Biomarkers for early detection of colon cancer. Clin. Cancer Res. 2001; 7: 1118–26.