Review article: in vivo imaging by endocytoscopy

Authors


Professor H. Neumann, Department of Medicine I, University of Erlangen-Nuremberg, Ulmenweg 18, Erlangen 91054, Germany.
E-mail: helmut.neumann@uk-erlangen.de

Abstract

Aliment Pharmacol Ther 2011; 33: 1183–1193

Summary

Background  Endocytoscopy (EC) enables in vivo microscopic imaging at 1400-fold magnification, thereby allowing the analysis of mucosal structures at the cellular level. In contrast to fluorescence imaging with confocal laser endomicroscopy which allows analysis of mucosal structures up to 250 μm in depth, EC is based on the principle of contact light microscopy and only allows visualisation of the very superficial mucosal layer.

Aim  To systematically review the feasibility and diagnostic yield of EC for in vivo diagnosis of diseases.

Methods  A systematic search of the literature on diagnostic interventions in the gastrointestinal tract using EC was performed by searches in MEDLINE, Current Contents, PubMed, cross-references and references from relevant articles using the search terms ‘endocytoscopy’, ‘endocytoscope’, ‘magnification endoscopy’, ‘endocytoscopic imaging’, ‘virtual histology’ and ‘optical biopsy’. Only full manuscripts and case reports published in English were included.

Results  Overall twenty-nine relevant reports were identified. EC was feasible to detect oesophageal squamous cell cancer with sensitivity, specificity and accuracy of 95%, 84% and 82%, respectively. Moreover, EC reached excellent sensitivity and specificity for in vivo diagnosis of colon polyps (91% and 100%, respectively). Other diagnostic applications of EC included diagnosis of Barrett’s oesophagus, Helicobacter pylori, coeliac disease and small cell lung cancer. No serious complications of EC have yet been reported.

Conclusions  Endocytoscopy is a safe and effective new endoscopic imaging technique to obtain in vivo histology and guided biopsies with high diagnostic accuracy. Therefore, endocytoscopy has the potential to facilitate both diagnosis and patient management.

Introduction

Whatever you can do or dream you can, begin it. Boldness has genius, power, and magic in it’.   Johann Wolfgang von Goethe, German poet.

Recent advances in endoscopic imaging have substantially influenced the diagnostic approach of various gastrointestinal diseases and our understanding of disease pathogenesis. Traditionally, standard white light endoscopy only allows the investigation of mucosal surfaces and surrounding blood vessels at a magnification of about 70-fold. Subsequent histopathological analysis of biopsy specimens remained the gold standard for final diagnosis and corresponding patients’ management.

Emerging endoscopic imaging modalities including both, vital (e.g. methylene blue, cresyl violet) and virtual chromoendoscopy (Narrow band imaging, NBI; Fujinon intelligent colour enhancement, FICE; i-Scan) and magnification endoscopy enabled the endoscopist to visualise and interpret increasing mucosal details.1

By introducing confocal laser endomicroscopy in 2004 in vivo microscopic imaging at 1000-fold magnification became possible, now allowing the endoscopist to obtain real time in vivo optical biopsies during ongoing endoscopy.2 Currently, two endomicroscopy systems are available. One is integrated in the distal tip of a high-resolution standard video endoscope (iCLE; Pentax, Tokyo, Japan) and the other is probe-based and capable of passage through the accessory channel of a standard endoscope (pCLE; Cellvizio, Mauna Kea Technologies, Paris, France). Various studies have shown the potential of both types of endomicroscopy systems for in vivo diagnosis of different disease entities including oesophageal squamous cell carcinoma, Barrett’s oesophagus, Helicobacter pylori infection, coeliac disease, ulcerative colitis and colon polyps.3

Endocytoscopy (Olympus, Tokyo, Japan) was recently introduced and represents another emerging endoscopic imaging technique enabling real time in vivo diagnosis of cellular patterns at extremely high magnification. Recent published data indicate that endocytoscopy has the potential to change the current paradigm for the diagnosis of various diseases (Table 1). Nevertheless, to date no systematic review on this emerging endoscopic imaging technique has been performed. The aim of this systematic review was therefore to summarise the diagnostic yield and relevance of endocytoscopy in the clinical setting.

Table 1.   Current potential clinical applications of endocytoscopy
OesophagusOesophageal squamous cell carcinoma13–19
 Barrett′s oesophagus20
StomachHelicobacter pylori– infection21
 Signet ring cell carcinoma22
DuodenumCoeliac disease23, 24
ColonPolyps26–31
 Amoebic colitis32
LungSmall cell lung cancer33, 34
OthersBladder carcinoma35
 Intraoperative carcinoma diagnosis36, 38

Methods and techniques

A systematic search of the literature on diagnostic interventions in the gastrointestinal tract using endocytoscopy was performed by searches of MEDLINE, Current Contents, PubMed, cross-references and references from relevant articles using the search terms ‘endocytoscopy’, ‘endocytoscope’, ‘magnification endoscopy’, ‘endocytoscopic imaging’, ‘virtual histology’ and ‘optical biopsy’. Only papers and case reports published as full text in English were included. Abstracts were not considered for this review. The literature search was performed in December 2010. Overall, twenty-nine suitable manuscripts on endocytoscopy were finally identified.

Techniques of endocytoscopy

Endocytoscopy is based on the principle of contact light microscopy. The system consists of a fixed-focus, high-power objective lens that projects magnified images from the sampling site onto a charge-coupled device.4

Mucosal preparation

To enable ultra high-magnification imaging using the endocytoscope adequate bowel preparation and mucosal preparation is necessary. Based on its principle of contact light microscopy and in contrast to endomicroscopy, endocytoscopy only allows visualisation of the very superficial mucosal layer. The light focusing depth ranges from 0 to 50 μm. Therefore, mucolysis using N-acetylcysteine is recommended before the procedure to remove surface mucus.4 In addition, prestaining of the mucosa with an absorptive agent is mandatory. One recent study evaluated the optimal staining regimen for endocytoscopy (pEC; XEC-300) using different dye concentrations of crystal violet, methylene blue and toluidine blue.5 Optimal conditions for endocytoscopic mucosal imaging were found after staining with 1% methylene blue in the oesophagus and with 0.25% toluidine blue in the stomach and colon. The dyes are usually sprayed onto the mucosal surface using standard spraying catheters. After an appropriate time of exposure (about 60 s) to the dye, repeat washing of the mucosa is necessary to remove excess contrast dye before high-magnification imaging. While using absorptive contrast agents, for extended visualisation repeat staining of the mucosa may be needed.4

Noteworthily, previous tattooing of suspicious lesions (for example for surveillance examinations) may limit the use of endocytoscopy (pEC; XEC-300) due to decreased image quality because the black pigment is likely to interfere optically with staining of crypts.6

Procedure

Currently, two types of endocytoscopes are available (Table 2).4 One is integrated in the distal tip of a standard endoscope (iEC) and one is probe-based (pEC, Figure 1). Endocytoscopy images are displayed on a monitor at 30 frames per second, which corresponds to the frame rate during routine high-resolution video endoscopy.

Table 2.   Technical characteristics of different endocytoscopy devices
TypeProbe-based (pEC)Endoscope-based (iEC)
  1. EC, endocytoscope.

 XEC-300XEC-120XGIF-Q260EC1XCF-Q260EC1GIF-Y0002
Endoscope magnificationN/A×85×110×380
(×600 digital)
Endoscope working channel (mm)N/A2.83.22.8
Total length (cm)380103133103
Functional length (cm)240103133103
EC magnification×450 (14′ monitor)
×570 (19′ monitor)
×1100 (14′ monitor)
×1390 (19′ monitor)
×580×380 (×600 digital)
Field of view (μm)300 × 300120 × 120400 × 400700 × 600
(440 × 380 digital)
Horizontal resolution4.21.74.03.7
Outer diameter3.2 mm11.613.610.7
Imaging plane depth30 μm5 μm 50 μm
Figure 1.

 The probe-based endocytoscope is advanced through the working channel of a standard endoscope. Inserted image highlights the magnification probe. Images displayed with kind permission from Olympus, Hamburg, Germany.

Integrated endocytoscopy systems use two separated lenses and are integrated into an 80-fold magnification endoscope enabling 450-fold magnification (field of view 400 × 400 μm). Recently, another endocytoscopy system was introduced consisting of only one lens that can consecutively increase the magnification up to 380-fold magnification using a hand lever (GIF-Y0002; field of view 700 × 600 μm). For the first time, this new endoscope-generation enables continues magnification from standard overview to endocytoscopy therefore representing an ‘all-in-one’ scope. Upper endoscopes (XGIF-Q260EC1) are 103 cm and lower endoscopes (XCF-Q260EC1) 133 cm in length (Table 2).

Probe-based endocytoscopy consists of handheld miniprobes, which are capable of being inserted through the accessory channel of a standard endoscope. Two different pEC devices exist, providing either 450-fold (XEC 300F) or 1390-fold (XEC 120 U) magnification images on a 19-inch monitor. The probes are 380 cm in length and 3.2 mm in diameter thus requiring an accessory channel of 3.7 mm. By introducing the pEC-device through the working channel targeted optical biopsy under guidance of the mother endoscope becomes possible. In addition, two monitors can be used during the procedure; one displaying the endoscopic picture, the other, an in vivo histological picture. To avoid motion artefacts during the endocytoscopic observation of mucosal structures, a soft short (4 mm) plastic cap is recommended to be affixed at the tip of the endoscope (Figure 2). By using slight suctioning the probe is stabilised against the mucosal wall and high-magnification imaging in focus become possible. The horizontal observation field is given with 300 × 300 μm (0.09 mm2) for the 450-fold magnification probe and with 120 × 120 μm for the 1390-fold magnification probe.

Figure 2.

 After mucosal staining with methylene blue the probe-based endocytoscope is gently pushed against the mucosal surface to enable high-magnification imaging. Note the plastic cap at the distal end of the endoscope to reduce motion artefacts.

Interpretation

Interpretation of optical biopsies is based on several architectural details (e.g. epithelial structure), cellular features (e.g. size, arrangement of cells) and vascular patterns (e.g. size, leakage and tortuosity). In addition to these factors, endocytoscopy enables the assessment of cytological features, such as the density of cells, the size and shape of nuclei and the nucleus-to-cytoplasm ratio (Figure 3).4 These are inevitable features for the diagnosis of intraepithelial neoplasia (formerly termed dysplasia) and cancer. Additionally, by appraisal of the staining pattern differentiation of mucosal inflammatory cells become possible (personal unpublished data). Eberl et al. evaluated if endocytoscopy (pEC; XEC-120 and XEC-300) could predict histology in neoplastic lesions.7 Therefore, twenty-five patients with oesophageal lesions, twenty-eight patients with colonic lesions and twenty-three patients with gastric lesions were examined. The sensitivity and specificity for the evaluation of the blinded pathologist was 81% and 100%, respectively, in the oesophagus, 56% and 89% in the stomach, and 79% and 90% in the colon. Nevertheless, to date no prospective study has validated endocytoscopy-based image criteria. Importantly, if an endoscopist evaluated the endocytoscopic pictures in combination with the macroscopic endoscopic images sensitivity and specificity increased significantly (sensitivity 96%, specificity 95%).

Figure 3.

 Optical biopsy at 450-fold (Panel A) and 1390-fold (Panel B) magnification using the endocytoscope highlights density of cells, the size and shape of nuclei and the nucleus-to-cytoplasm ratio. Panel A shows normal gastric epithelium after toluidine blue staining. Microvessels are clearly visible (arrows). Panel B highlights normal colonic epithelium after toluidine blue staining.

Comparison of endocytoscopy to magnification endoscopy and endomicroscopy

Endocytoscopy, magnification endoscopy and endomicroscopy represent completely different endoscopic techniques and all require totally different endoscopes.

As mentioned above, endocytoscopy is based on the principle of contact light microscopy. In vivo imaging is achieved by a fixed-focus, high-power objective lens that projects images onto a charge-coupled device.4 Therefore, the depth of field is dependent on the absorption spectra and light penetration depth of the intestinal tissue and cannot be altered manually. Because absorption spectra and light penetration depth are themselves dependent on tissue structure (e.g. squamous epithelium, cylindric epithelium) and composition (e.g. inflammation, haemoglobin) the light focusing depth ranges theoretically from 0 μm up to 50 μm.8 Of note it is currently not possible to determine the exact level of endocytoscopic imaging. Furthermore to realise ultra high-magnification imaging up to 1400-fold, prestaining of the mucosa with either methylene blue (Akorn, Lake Forest, IL, USA) or toluidine blue (Dr. Franz Köhler Chemie, Bensheim, Germany) is strictly necessary. Consequently, endocytoscopy is able to identify cellular (e.g. visibility of nuclei) and architectural features of the surface epithelial layer. In contrast to endomicroscopy, endocytoscopy does not allow the visualisation of deeper layers of the epithelium and is therefore not suitable to evaluate the depth of invasion in early neoplasia. In addition, concerns about the use of methylene blue have risen, which may increase DNA damage in the tissue after illumination.9 As a consequence, methylene blue was recently withdrawn from the German market but is still available in other countries (i.e. in USA). Toluidine blue is approved by the Federal Institute for Drugs and Medical Devices for vital staining. To the best of our knowledge, no concerns regarding the safety of topically applied toluidine blue are known.

In contrast to endocytoscopy, magnification (also called ‘zoom’) endoscopy uses a movable lens controlled by the endoscopist to vary degree of magnification, ranging from 1.5-fold to 150-fold. Both, high-resolution and magnification features are integrated in one endoscope and enable the endoscopist to analyse the pit pattern (e.g. villiform or distorted in Barrett’s oesophagus) of lesions.10 Besides optical magnification endoscopy, modern video processors enable digital magnification features. These are based on special software algorithms which enlarge the original image. Because the resolution of the digital sensor is fixed, the software tries to interpolate additional pixels which markedly reduce the quality of the resulting image.

In contrast to endocytoscopy, confocal laser endomicroscopy is based on tissue illumination after application of fluorescence agents which can either be applied systemically (fluorescein) or topically (e.g. acriflavine, cresyl violet). Laser light is focused to a single point within a defined microscopic field of view. The laser light stimulates the fluorescence agent and light emanating from this point is focused through a pinhole to a detector while light emanating from outside the illuminated spot is excluded from detection. The grey scale image created is an optical section representing one focal plane within the specimen.11 The iCLE system offers the operator to manually adjust the imaging plane depth ranging from 0 to 250 μm using an optical slice thickness of 7 μm. Endomicroscopy offers in vivo imaging of cellular and subcellular structures at 1000-fold magnification compared to 1400-fold magnification of endocytoscopy.3 While concerns about the use of acriflavine have raised (potential mutagenic effects) one recent large multicentre study has addressed the safety of intravenous fluorescein for endomicroscopy in the gastrointestinal tract.12

Current applications of endocytoscopy

Oesophagus

Endoscopic imaging of oesophageal disorders is well established in clinical practice due to the increasing incidence of cancers at this location and their poor prognosis in advanced disease stages. Moreover, the opportunity of gentle curative endoscopic resection methods in early disease stages increases the need for imaging modalities which allow the detection of subtle mucosal changes with a high diagnostic yield (Table 3).

Table 3.   Summary of studies evaluating endocytoscopy for diagnosing oesophageal diseases
Author and referenceDisease entityStudy designPatientsResults
  1. EC, endocytoscopy; SCC, squamous cell carcinoma; PPV, positive predictive value; NPV, negative predictive value.

Kumagai13SCCDescriptive study12In cancerous mucosa, cell distribution and nucleus-to-cytoplasm ratio was irregular and density of cells increased.
Dumonceau14SCCCase report1 
Fujishiro15SCCCase report1 
Banerjee16SCCCase series3 
Inoue17SCCProspective study29PPV 94%, accuracy 82%
Fujishiro18SCCEx vivo pilot study27Endocytoscopic images closely correlated with conventional histology
Kumagai19SCCProspective study21Sensitivity 95%, specificity 84%
Pohl20Barrett′s oesophagusProspective study16Assessment of EC-images possible in 51% (× 450) and 22% (× 1125). PPV, NPV 0.29 and 0.87 (× 450) and 0.44 and 0.83 (× 1125). Interobserver kappa 0–0.45.

In 2004, Kumagai et al. observed the surface of the squamous epithelium in normal oesophageal mucosa and superficial oesophageal cancers using endocytoscopy (pEC; XEC-120 and XEC-300).13 Normal cells were homogeneously arranged, and the nucleus to cytoplasm ratio was uniform and low. In contrast, in oesophageal cancers, the cell distribution and nucleus-to-cytoplasm ratio was irregular and the density of cells was increased. However, no statistic analysis was done in this study.

In addition, three case reports confirmed the potential of endocytoscopy for in vivo diagnosis of early oesophageal squamous cell carcinoma.14–16 Using pEC with 450-fold magnification (pEC; XEC-300), it was possible to visualise typical neoplastic mucosal alterations like irregular and enlarged nuclei during ongoing endoscopy.

Inoue et al. conducted a pilot trial including twenty-nine consecutive patients with oesophageal lesions who were evaluated using endocytoscopy (iEC; XGIF-Q260EC1) after prestaining of the oesophageal mucosa with 0.5% methylene blue.17 The positive predictive value for malignancy was 94%, the sensitivity 83% and specificity 94% with an overall accuracy for differentiating between malignant and nonmalignant tissue of 82%.

To investigate whether endocytoscopic images of cancerous and noncancerous oesophageal mucosa correspond with conventional histopathology, a multicentre ex vivo pilot study was conducted.18 Overall, twenty-seven patients with oesophageal squamous cell cancers were included and evaluable pairs of an endocytoscopic image and a histological picture of the same site were obtained at twelve cancerous and fourteen normal areas. The pattern of the cellular arrangement and density and the size and shape of nuclei were morphologically identical between endocytoscopy (pEC; XEC-300) and conventional histopathology. However, the cytoplasm was not identifiable in endocytoscopic images. Mean (± s.d.) total numbers of nuclei per endocytoscopic image were 129 ± 14.8 at normal areas and 550 ± 66.5 at cancerous areas (P < 0.0001).

One recent published study aimed at evaluating whether endocytoscopic observation of oesophageal squamous cell carcinoma could replace histopathological examination of biopsy specimen.19 Therefore, fifty-seven iodine-unstained areas in resected specimens of the oesophagus from twenty-eight patients were examined. In addition, seventy-one lesions of oesophageal squamous cell carcinoma were analysed in vivo using endocytoscopy (pEC; XEC-120 and XEC-300). The sensitivity and specificity of endocytoscopy for malignant lesions was 95% and 84%, respectively. In 94% of cancerous lesions diagnosed using endocytoscopy further biopsy examinations were considered as unnecessary by the histopathologist.

Barrett’s oesophagus is strongly associated with gastroesophageal reflux disease and well recognised as a precursor lesion of oesophageal adenocarcinoma. Early detection of mucosal alterations is important to provide an efficient surveillance and endoscopic therapy in those patients.

One study conducted by Pohl and co-workers evaluated the feasibility of endocytoscopy (pEC; XEC-120 and XEC-300) in the surveillance of patients with Barrett’s oesophagus.20 Overall 166 biopsy sites from sixteen patients without visible lesions, who presented for Barrett surveillance were analysed using endocytoscopy and correlated to histopathology. Adequate assessment of endocytoscopy images was possible in 51% with magnification ×450 and in 22% with magnification ×1125. At most, 23% of images with lower magnification were interpretable to identify characteristics of neoplasia, and 41% with higher magnification. Interobserver agreement was fair with kappa values from <0 to 0.45. Positive and negative predictive values for high-grade intraepithelial neoplasia or cancer were 0.29 and 0.87, respectively, for magnification ×450 and 0.44 and 0.83, respectively, for magnification ×1125.

Key message

Endocytoscopy could detect oesophageal squamous cell carcinoma with good sensitivity and specificity.

In Barrett’s oesophagus endocytoscopy lacks sufficient image quality to be currently of assistance in identifying neoplastic areas.

Stomach

Helicobacter pylori (H. pylori) infection affects about fifty percentage of the general population (10–30% in industrialised countries and 80–90% in developing countries) and is strongly associated with gastroduodenal ulcer disease, gastric lymphoma and nonproximal gastric adenocarcinoma.

Kimura et al. cultured H. pylori ex vivo from gastric mucus of three patients suffering from gastric ulcer disease. Subsequently, H. pylori micro-organisms could be observed in the supernatant of the culture medium using endocytoscopy with 1100-fold magnification (pEC; XEC-300).21

In addition, Fasoli and co-workers published one case report highlighting the potential of endocytoscopy (XEC-300) for in vivo evaluation of gastric signet ring cell carcinoma. Endocytoscopy was reliable to identify disease-specific histological aspects and there was a strong concordance between in vivo and ex vivo histology.22

Key message

Currently, no studies are available concerning endocytoscopy-based in vivo diagnosis of H. pylori infection, hyperplastic and precancerous lesions of the stomach.

Duodenum

Coeliac disease represents an autoimmune enteropathy in genetically susceptible hosts induced by gluten. Histopathological diagnosis is based on villous atrophy, hyperplasia of intestinal crypts and an increased number of intraepithelial lymphocytes.

Up to now, two studies have evaluated the potential of endocytoscopy for in vivo diagnosis of coeliac disease.

In the first study, a total of 166 endocytoscopy (pEC; XEC-120 and XEC-300) recordings from forty patients with established (n = 32) or suspected coeliac disease were prospectively acquired and corresponded to conventional histopathology.23 When adjusted by multivariate analysis four criteria were identified to be significant predictors for Marsh III pathology: low number of villi per visual field [<3; odds ratio (OR) 9.1; 95% CI, 1.3–62.0], confluence of villi (OR 37.1; 95% CI, 1.3–1021.2), irregular epithelial lining (OR 10.9; 95% CI, 2.5–46.7) and inability to delineate loop capillaries (OR 14.9; 95% CI, 3.3–67.0). Nevertheless, none of these factors was a good predictor for Marsh I pathology.

Sixteen patients with documented coeliac disease and seven control patients were included in the second trial, evaluating the potential of endocytoscopy (pEC; XEC-300) in coeliac disease.24 Endocytoscopy visualised the presence of normal-appearing, long, thin villi, lined with clearly distinguishable surface epithelial cells, the presence of thick, shortened villi, reflecting partial villous atrophy and the total absence of villi and the presence of enlarged crypt orifices, reflecting total villous atrophy. Sensitivity and specificity of endocytoscopy for the diagnosis of villous atrophy was 88% and 100%, respectively.

Key message

In coeliac disease endocytoscopy enables accurate determination of severe abnormalities like villous atrophy. Nevertheless, mild changes cannot be assessed. Therefore, diagnosis and follow-up of patients should still be based on biopsies.

Colon

According to most national and international guidelines, colonoscopy is the recommended surveillance method of patients aged 50 years and older to exclude adenomatous polyps. However, most polyps are diminutive in size (1–5 mm) and harbour only a very low prevalence of cancer.25 Therefore, sending diminutive polyps to the histopathologist generally leads to a profusion of resources. By using real time in vivo histology, suspicious polyps could be identified for targeted biopsy or endoscopic polypectomy (Table 4).

Table 4.   Summary of studies evaluating endocytoscopy for diagnosing colon polyps
Author and referenceStudy designPatientsResults
  1. EC, endocytoscopy; PPV, positive predictive value; NPV, negative predictive value, LGIEN, low-grade intraepithelial neoplasia; HGIEN, high-grade intraepithelial neoplasia.

Sasajima26Prospective study113Accuracy 93%; kappa between EC and histopathology 0.91
Neumann27Case report1EC could detect focal HGIEN
Meroni28Case report1EC could detect tissue abnormalities in normal mucosa surrounding colorectal cancer
Cipolletta29Prospective study41Sensitivity 91%, Specificity 100%, Interobserver agreement 0.68
Rotondano30Prospective study49PPV for hyperplastic polyps, LGIEN, HGIEN and invasive cancer was 100%, 93%, 90% and 100%

Indeed, Sasajima et al. evaluated in a prospective study the usefulness of endocytoscopy (pEC; XEC-120 and XEC-300) for the differentiation of colorectal lesions in 113 consecutive patients.26 Endocytoscopy was able to differentiate normal colonic mucosa, hyperplastic polyps, low-grade and high-grade adenomas and invasive cancers. The overall accuracy of endocytoscopy was 93% with an excellent kappa value between endocytoscopy and the corresponding histopathological analysis of 0.910. The differential diagnosis between neoplastic and non-neoplastic lesions and between adenoma and invasive cancer was statistically significant (< 0.001).

In addition to this context, one recent published case report highlighted the potential of endocytoscopy (pEC; XEC-120) for in vivo diagnosis of very focal high-grade intraepithelial neoplasia in colonic polyps.27 This is of tremendous importance, because polyps often show only focal lesions which are surrounded by normal colonic mucosa.

Meroni et al. addressed this point in another case report.28 They performed endocytoscopy (pEC; XEC-300) in one surgical specimen obtained after resection of colon cancer and found focal abnormalities of colonic glands highly suggestive of aberrant crypt foci seven cm apart from the primary tumour within macroscopically normal mucosa.

Finally, Cipolletta et al. conducted a prospective study including forty-one consecutive patients with colorectal aberrant crypt foci.29 The mean duration of the endocytoscopy procedure was 44 ± 12 min (range 31–62 min) and it was possible to observe lesions at the cellular level and evaluate both cellular and structural atypia in vivo. In normal mucosa, crypts had preserved individuality and round-shaped contours. Nuclei were located at the basal third of the crypt in a single line, and the lumen was circular. In dysplastic aberrant crypt foci, crypt contours were polygonal, cell nuclei were elongated with pseudo stratification towards the luminal half of the crypt and irregularly arranged, and the lumen was linear. Sensitivity and specificity of endocytoscopy (pEC; XEC-300) for the detection of intra-epithelial neoplasia was 91% and 100%, respectively. Interobserver agreement between endoscopist and pathologist was good (weighted kappa 0.68; 95% CI 0.59–0.78).

The same group aimed to access the capability of endocytoscopy (pEC; XEC-300) in differentiating neoplastic from non-neoplastic lesions in the colorectum.30 Overall, fifty-two lesions were examined in 49 patients (17 polypoid and 35 nonpolypoid). Final pathological diagnosis was normal mucosa or hyperplastic polyp in ten cases, low-grade adenoma in twenty-nine, high-grade adenoma in eleven and submucosal invasive cancer in two cases. Positive predictive value of endocytoscopy was 100% for hyperplastic polyps, 93% for low-grade intraepithelial neoplasia, 90% for high-grade intraepithelial neoplasia and 100% for invasive cancer. In addition, endocytoscopy diagnosis correlated completely with histopathology in the differentiation between neoplastic and non-neoplastic lesions.

One recent case report identified endocytoscopy (pEC; XEC-300) as a potential method for imaging of a carcinoid tumour which was located in the terminal ileum. Nests of monomorphous cells separated by poor cellulated strands without glandular structures were evident on endocytoscopic imaging.31

Very recently, Hosoe and co-workers prospectively evaluated the feasibility of endocytoscopy with 450-fold magnification (iEC; CF-Y0001) for the in vivo diagnosis of amoebic parasites in five patients. The authors could successfully visualise amoebic trophozoites in all cases and therefore diagnose amoebic colitis in vivo during ongoing endoscopy.32

Key message

Endocytoscopy can predict the presence of preneoplastic and neoplastic lesions in the colon with high accuracy. One drawback is the prolongation up to 1 h of the duration of endoscopy.

Lung

Very recently, our group addressed the potential of endocytoscopy for in vivo diagnosis of small cell lung cancer during ongoing bronchoscopy in four patients.33 After topical application of methylene blue, endocytoscopy (pEC; XEC-120) was able to reliably identify numerous small blue cells with hyperchromatic nuclei (Figure 4). Corresponding histopathological analysis confirmed in vivo diagnosis of small cell lung cancer. This exciting initial experience is now evaluated in a larger ongoing prospective study.

Figure 4.

 Methylene-blue aided endocytoscopy in a patient with small cell lung cancer visualises numerous small blue cells with hyperchromatic nuclei.

Shibuya and co-workers performed endocytoscopy (pEC; XEC-300) in twenty-two patients (seven squamous cell carcinoma, 11 squamous dysplasia and four after PDT therapies). Both abnormal areas of interest and normal bronchial mucosa were stained with 0.5% methylene blue and afterwards examined with endocytoscopy at 570-fold magnification. In this study, endocytoscopy was useful to discriminate between normal bronchial epithelial cells, dysplastic cells and malignant cells during ongoing bronchoscopy.34

Others

Besides the above mentioned areas, endocytoscopy was also proven to be efficient in different disease entities.

In a pilot study including five patients, endocytoscopy was used to evaluate bladder carcinoma.35 Using the pEC system with 450-fold magnification (XEC-300), the cell structure and nuclear morphology of the tumours were identified and graded correctly in 80% of the cases.

In another case report endocytoscopy (not nearly specified system) was used to assist intraoperative diagnosis of carcinoma in a patient who was suffering from chronic pancreatitis. Diagnosis of cancer was based on the visualisation of cells with hyperchromatic nuclei and a high nucleus to cytoplasm ratio.36

Whereas it is often challenging to optimally sample the muscularis propria endoscopically for the diagnosis of muscle layer diseases, especially for motility disorders resulting from neuroenteric dysfunction, Sumiyama and co-workers designed an ex vivo and in vivo porcine animal study to access the muscular layer using endocytoscopy (pEC; XEC-120 and XEC-300).37 Cellular microstructures resembling spindle-shaped smooth muscle cells and neuron-like cells were visualised by endocytoscopy in vivo.

Endocytoscopy was used in another report of two patients for intraoperative diagnosis of disseminated malignancy. During diagnostic laparoscopy, endocytoscopy (pEC; XEC-300) of peritoneal nodules revealed malignant cells and in vivo diagnosis was confirmed by corresponding histopathology.38

In addition, after methylene blue staining onto normal rectal mucosa, endocytoscopy (pEC; XEC-300) revealed red blood cell flow within the microvasculature circulating through the arterioles.39

Key message

Potential applications of endocytoscopy were shown in different pilot studies including lung, bladder and intraoperative diagnosis of malignancy. Larger trials are now needed to confirm these interesting and promising preliminary results.

Conclusions

Modern endoscopy with advanced endoscopic imaging techniques has revolutionised the diagnosis and management of patients in the endoscopy suite. Besides other emerging endoscopic imaging techniques, including chromoendoscopy, magnification endoscopy and confocal laser endomicroscopy, endocytoscopy has become an inherent part of advanced endoscopic imaging. In different disease entities, endocytoscopy allowed the endoscopist to obtain real time in vivo histology during ongoing endoscopy with a high diagnostic yield and to guide subsequent biopsy acquisition for histopathological analysis. Up to now, no serious complications of endocytoscopy have been reported.

Currently, the spectrum of endocytoscopy ranges from oesophageal disorders, including oesophageal squamous cell carcinoma and Barrett’s oesophagus to coeliac disease and in vivo diagnosis of colon polyps and intestinal pathogens like Helicobacter pylori and amoebic trophozoites. Besides, the value of endocytoscopy was also proven for the in vivo diagnosis of small cell lung cancer and found to facilitate final histological diagnosis during diagnostic laparoscopy.

Nevertheless, most studies so far only include small sample sizes or are case reports. No multicentre trial evaluating the technique has yet been performed. In addition, as endocytoscopy is a developing technology the performance of different instruments varies widely. Up to date, no information regarding the learning curve or the inter-observer and intra-observer agreement of endocytoscopy are available. Therefore, endocytoscopy is a promising new development providing advanced endoscopic imaging. Nevertheless, multicentre studies evaluating the technique in larger settings are highly warranted.

In the near future, new ‘all-in-one’ scopes for stable ultra-high-magnification imaging and functional imaging could further improve the diagnostic applicability of endocytoscopy.

Acknowledgements

Declaration of personal interests: Helmut Neumann has served as a speaker for Pentax, Mauna Kea Technolgies and Essex. Helmut Neumann is an employee of University of Erlangen. Florian Fuchs has served as a speaker, a consultant and an advisory board member for Novartis, Roche, Boehringer Ingelheim, AstraZeneca, MSD, Lilly and Nycomed. Michael Vieth has served as a speaker, a consultant and an advisory board member for AstraZeneca, Falk, Pentax, Mauna Kea, Olympus and Malescci. Michael Vieth is an employee of Klinikum Bayreuth. Raja Atreya is an employee of University Hospital of Erlangen. Ralf Kiesslich has served as a speaker, a consultant and an advisory board member for Pentax Europe, Abbott, Shire, Falk, and has received research funding from Pentax Europe. Markus Neurath has served as a speaker, a consultant and an advisory board member for Essex Pharma, MSD, Abbott, Giuliani and Pentax, and has received research funding from the Deutsche Forschungsgemeinschaft. Declaration of funding interests: None.

Ancillary