- Email: [email protected]
Limited applicability of chromoendoscopy-guided confocal laser endomicroscopy as daily-practice surveillance strategy in Crohn’s disease
ORIGINAL ARTICLE: Clinical Endoscopy
Limited applicability of chromoendoscopy-guided confocal laser endomicroscopy as daily-practice surveillance strategy in Crohn’s disease Linda K. Wanders, MSc, MD,1 Teaco Kuiper, MD, PhD,1 Ralf Kiesslich, Prof Dr,2 John G. Karstensen, MSc, MD,3 Rupert W. Leong, MSc, MD,4 Evelien Dekker, Prof Dr,1 Raf Bisschops, Prof Dr5 Amsterdam, Netherlands; Leuven, Belgium
Background and Aims: Patients with longstanding ulcerative colitis have an increased risk for developing colorectal cancer (CRC). Although the risk for ulcerative colitis is well-established, for Crohn’s disease data are contradictory. This study aimed to determine the number of patients with Crohn’s disease with dysplasia who are undergoing surveillance and to assess the diagnostic accuracy of chromoendoscopy (CE) combined with integrated confocal laser endomicroscopy (iCLE) for differentiating dysplastic versus nondysplastic lesions. Methods: Patients with longstanding Crohn’s colitis undergoing surveillance colonoscopy were included in this multicenter, prospective, cohort study. Surveillance was performed with CE, and lesions were assessed with iCLE for differentiation. All lesions were removed and sent for pathology as the reference standard. Results: Between 2010 and 2014, a total of 61 patients with Crohn’s colitis were included in 5 centers. Seventytwo lesions, of which 7 were dysplastic, were detected in 6 patients (dysplasia detection rate 9.8%); none included high-grade dysplasia or cancer. Combined CE with iCLE for differentiating neoplastic from nonneoplastic lesions had accuracy of 86.7% (95% confidence interval [CI], 78.1-95.3), sensitivity of 42.9% (95% CI, 11.8-79.8), and specificity of 92.4% (95% CI, 80.9-97.6). For CE alone, this was 80.3% (95% CI, 70.7-89.9), 28.6% (95% CI, 5.1-69.7), and 86.4% (95% CI, 80.9-97.6). The study terminated early because of frequent failure of the endoscopic equipment. Conclusions: This study shows a low incidence of dysplastic lesions found during surveillance colonoscopy in patients with longstanding extensive Crohn’s colitis. The accuracy of both CE alone and CE in combination with iCLE was relatively good, although the sensitivity for both was poor. Because of frequent equipment failure, iCLE has limited applicability in daily practice as a surveillance strategy. (Gastrointest Endosc 2016;83:966-71.)
Longstanding ulcerative colitis (UC) increases a patient’s risk of developing colorectal dysplasia and colorectal cancer (CRC). Although this risk is well established for UC, conflicting data exist regarding this risk with extensive colitis in patients with Crohn’s disease (CD). The risk
estimates in patients with CD differ considerably from no increase in CRC risk to a relative risk of up to 3.3 for developing CRC.1-3 A potential explanation for this heterogeneity in CRC risk might be selection bias. Previous studies analyzing
Abbreviations: CD, Crohn’s disease; CE, chromoendoscopy; CLE, confocal laser endomicroscopy; CRC, colorectal cancer; IBD, inflammatory bowel disease; iCLE, integrated confocal laser endomicroscopy; UC, ulcerative colitis.
Received March 17, 2015. Accepted September 1, 2015.
DISCLOSURE: R. Kiesslich is a consultant for and is on paid advisory boards for Pentax. J.G. Karstensen received financial support from Pentax, Mauna Kea Technologies, and Boston Scientific. E. Dekker received financial support and has equipment on loan from Olympus and Pentax. R. Bisschops is supported by Research Foundation Flanders (FWO Vlaanderen) and received a speaker’s fee and research support from Pentax. All other authors disclosed no financial relationships relevant to this publication.
Current affiliations: Department of Gastroenterology and Hepatology, Academic Medical Centre, Amsterdam, The Netherlands (1), Department of Gastroenterology, Dr. Horst Schmidt Klinik, Wiesbaden, Germany (2), Gastro Unit, Division of Endoscopy, Copenhagen University Hospital Herlev, Herlev, Denmark (3), Gastroenterology and Liver Services, Bankstown-Lidcombe Hospital, Sydney, Australia (4), Department of Gastroenterology, University Hospitals Leuven, Leuven, Belgium (5). Reprint requests: R. Bisschops, MD, PhD, Department of Gastroenterology and Hepatology, University Hospital Leuven, Herestraat 49, 3000, Leuven, Belgium.
Copyright ª 2016 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2015.09.001
If you would like to chat with an author of this article, you may contact Dr Bisschops at [email protected].
966 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
Wanders et al
CRC risk in patients with CD included patients with different disease extents. Studies including patients with CD and colitis showed, however, a significant increase in CRC risk.2-3 Besides disease extent, several other risk factors, such as disease duration, family history of colon cancer, and concomitant primary sclerosing cholangitis, are known to increase the risk of developing dysplasia and cancer in patients with CD.4-6 Colonoscopic surveillance of patients with inflammatory bowel disease (IBD) is associated with early detection of premalignant lesions and subsequently to an improved prognosis.7 Guidelines therefore recommend surveillance in patients with IBD.8-10 It is known that dysplasia in patients with IBD can be subtle and flat and therefore more difficult to detect than dysplasia in an average-risk population.11 In addition, postinflammatory mucosal changes such as scarring and inflammatory pseudopolyps further complicate detection of dysplasia in these patients. Until recently, surveillance was performed by using white-light endoscopy with 4-quadrant biopsies every 10 cm. However, recent studies have proven that chromoendoscopy (CE) significantly increases lesion detection in IBD surveillance colonoscopies compared with white light endoscopy with quadrant-random biopsies every 10 centimeters.12 Current guidelines recommend the use of CE in surveillance of patients with longstanding colitis.8-10 During CE, a dye spray (eg, methylene blue or indigo carmine) is used to highlight the mucosal surface and to enhance the delineation of (early) dysplastic lesions. Although CE improves detection, it has not yet been proven that CE improves differentiation. Even endoscopists with expertise in IBD show great interobserver variation in differentiating lesions found during IBD surveillance colonoscopy, reflecting the difficulty in distinguishing dysplastic from nondysplastic lesions.13 Confocal laser endomicroscopy (CLE) is an advanced imaging technique providing highly magnified images of GI epithelium comparable to histopathology. Because of visualization of a limited surface area, CLE is specifically used for differentiation, rather than for detection. CLE can be integrated into a colonoscope (Pentax, Tokyo, Japan) (integrated confocal laser endomicroscopy [iCLE]) or can be applied as a probe system that passes through the working channel of any endoscope (Mauna Kea Technologies, Paris, France). A recent meta-analysis showed that CLE has the highest accuracy of all commercially available imaging techniques in differentiation of neoplastic from nonneoplastic lesions in an average-risk population, although mainly by experts.14 The aim of this study was to determine the number of patients with dysplasia undergoing surveillance colonoscopy for longstanding extensive Crohn’s colitis and to assess whether the combination of CE and iCLE improves differentiation of lesions found during surveillance colonoscopy. www.giejournal.org
Chromoendoscopy and iCLE in Crohn’s disease surveillance
METHODS Patients This study was performed in 5 academic medical centers. Patients with Crohn’s colitis undergoing surveillance colonoscopy were included in the study if they were aged 18 years, diagnosed with >50 cm of colon disease, and if they were in clinical remission, based on the Simple Endoscopic Score for Crohn’s Disease.15 Patients were excluded if they had undergone colon resections with <50% of their colons in situ, incomplete colonoscopy because of poor bowel preparation, disease activity or stenosis, contraindications for iCLE (allergy for intravenous fluorescein, pregnancy or breastfeeding, severe cardiopulmonary disease, or pre-existent renal disease), or noncorrectable coagulopathy that precludes taking biopsy specimens. This study was approved by the medical ethics committees of all participating centers (NTR2293).
Endoscopic equipment For CE and CLE, the endoscope-integrated system, iCLE (Pentax, Tokyo, Japan and Optiscan; Melbourne, Australia), was used. In this system, the confocal laser unit is integrated into the distal tip of a standard video endoscope, enabling iCLE imaging on a second monitor. The diameter of both the distal tip and the insertion tube is 12.8 mm. The iCLE imaging plane depth relative to the mucosal surface (ie, z-axis) can be controlled by using two buttons on the hand piece of the endoscope. During iCLE, a single-line laser beam with an excitation wavelength of 488 nm is delivered at the tissue. Confocal image data are collected at a scan rate of 0.8 frames per second. The CLE images have a lateral resolution of 0.7 mm. The field of view is 500 mm, and the range of the z-axis is 0 to 250 mm below the surface.
Endoscopic procedure All patients were prepared with either 4 L of polyethylene glycol solution or 2 L of polyethylene glycol solution plus ascorbic acid (Moviprep or Kleanprep, Norgine, Netherlands). The procedures were performed with the patients under conscious sedation by using intravenous midazolam and fentanyl or propofol at the discretion of the endoscopist. At the discretion of the endoscopists, butylscopolamine was given intravenously to reduce colon motility. During withdrawal, the entire colon was stained with a 0.1% methylene blue solution. Bowel preparation was determined as good (100% mucosa visible), moderate (>90% mucosa visible), or poor (<90% mucosa, even after extensive cleaning). Each colon segment was scrutinized for the presence of suspicious areas, mucosal irregularities, unusual ulcers, and strictures. All detected lesions were classified according to the macroscopic classification (Paris Classification16) of early GI dysplasia, scored for size, location in the colon, and distance from the anus. All lesions were assessed for Kudo pit pattern.17 The colon mucosa Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 967
Chromoendoscopy and iCLE in Crohn’s disease surveillance
Wanders et al
1 cm adjacent to each detected lesion also was scored for the Kudo pit pattern. The location of all detected lesions with respect to the extent of colitis (proximal to or within the inflammatory changed colon on endoscopy) was noted. Digital still images of all detected lesions were taken. Subsequently, after intravenous injection of fluorescein (10%), all lesions as well as their adjacent “normal” mucosa were inspected with iCLE and scored by the endoscopist as normal, regenerative, or dysplastic, according to the previously described Mainz criteria.18 Hereafter, targeted biopsies of detected lesion and 4 random biopsies of the adjacent colon mucosa were sent for histopathologic evaluation.
Endoscopists All procedures were performed by experienced endoscopists who have participated in previous studies with CE and iCLE in Amsterdam, Mainz, Leuven, Copenhagen, and Sydney.18-20 Therefore, all endoscopists were familiar with the techniques and should have performed beyond the learning curve.
Outcome measures The primary outcome measure of this study was the number of patients with dysplasia detected in a Crohn’s colitis surveillance group, expressed as dysplasia detection rate. The secondary outcome measure was diagnostic accuracy of CE and iCLE for the differentiation of dysplastic and nondysplastic mucosa, by using final histopathology as the reference standard diagnosis.
Figure 1. Flow diagram of patient inclusion.
Study design This was a multicenter, prospective, cohort study. The sample size calculation was based on the analysis by Friedman et al,1 in which the prevalence of dysplasia or cancer in patients with CD affecting at least one third of the colon was 7% on screening examination. A successful endoscopic surveillance protocol was defined as a similar dysplasia incidence of 7% and an upper limit of the 95% CI of 12%. The calculated sample size was 101 patients.
RESULTS Patient characteristics
Normal distributed data were described with the mean and the standard deviation, and data with a skewed distribution were described by using the median and the interquartile range. The diagnostic accuracy of CE and iCLE for differentiation was calculated by creating a 2 2 table comparing the index test result with the reference standard (ie, dysplastic or nondysplastic, according to histopathology). A positive index test result on CE and iCLE was defined as a dysplastic appearance during iCLE. The sensitivity, specificity, and overall accuracy were calculated and represented by a point estimator with the 95% confidence interval (CI). The difference in performance of CE and iCLE was compared with the Fisher exact test.
Between February 2010 and June 2014, a total of 82 patients with CD were recruited, of whom 61 were included (Fig. 1). The mean age was 48 years, and participants had a mean disease duration of 22 years by the time they were included in this study. In 2 patients, circumscribed dysplasia was detected in previous colonoscopies and endoscopically resected, and 2 patients had concomitant primary sclerosing cholangitis. Bowel preparation was considered as good in 60.7% of cases; all other cases scored moderate. Four patients had to be excluded because of poor bowel preparation. Other patients had to be excluded because of active inflammation (n Z 10), impassable stenosis (n Z 3), or technical failure of the endoscope (n Z 4) (Fig. 1). Because of frequent technical failure of the equipment, the study terminated early, and the calculated sample size was not reached. In 4 of 5 centers, the lens of the endomicroscope broke, and the equipment had to be sent to the manufacturer. Each repair took several weeks to months, which hindered patient recruitment and inclusion. The mean total procedure time was 40 minutes. The mean total inspection time including iCLE acquisition was 30 minutes, and without iCLE assessments, this was 10 minutes. Of all included patients, 34 had at least 1 lesion. In total, 72 lesions were detected. Of these lesions, 66 were examined with CE and 60 with iCLE. Seven dysplastic lesions were found in 6 patients, which accounts for a dysplasia detection rate of 9.8% (Table 1).
968 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
Histopathology Histologic samples are processed by using standard procedures and evaluated by one GI expert in each center. In case of discrepancy, discussion between two pathologists would lead to a consensus diagnosis. The pathologists were not aware of the detection technique and the findings on iCLE. Biopsy specimens were classified according to the Vienna criteria of GI epithelial neoplasia.21 The histologic diagnosis of all biopsy specimens served as the reference standard.
Statistics
Wanders et al
Chromoendoscopy and iCLE in Crohn’s disease surveillance
TABLE 1. Patient characteristics (N [ 61) Male, no (%)
TABLE 2. Lesion characteristics (N [72) 29 (47.5)
Size, mean ( SD), mm Paris classification,
16
6.8 ( 4.7)
%
Age at diagnosis, mean ( SD), y
26 ( 9.9)
Age at colonoscopy, mean ( SD), y
48 ( 11.5)
Is
22.2
Duration of colitis, mean ( SD), y
22 ( 7.8)
Isp
9.7
Primary sclerosing cholangitis, no. (%)
2 (3.3)
Ip
1.4
Previous dysplasia, no. (%)
2 (3.3)
IIa
38.9
IIb
9.7
37 (60.7)
IIc
0
24 (39.3)
IIaþc
40 ( 17)
Unknown
Colon preparation, no. (%) Good Moderate Total procedure time, mean ( SD), min Inspection time without iCLE, mean ( SD), min
10 ( 7)
Inspection time with iCLE, mean ( SD), min
30 ( 14)
0 18.1
Location, % Cecum
13.9
Lesions detected, no.
72
Ascending colon
11.1
Adenomas detected, no.
7
Transverse colon
19.4
Patients with lesions, no.
34
Descending colon
11.1
Patients with dysplasia, no.
6
Sigmoid colon
23.6
Dysplasia detection rate, %
9.8
Rectum
18.1
Missing
2.8
SD, Standard deviation; iCLE, integrated confocal laser endomicroscopy.
Histopathology, %
Lesion characteristics
Normal
21.8
The mean diameter of the detected lesions was 6.8 mm. Lesions were categorized as flat (Paris IIa/IIb16) in almost half of cases; none of the lesions were depressed. The lesions were equally distributed throughout the colon. None of the dysplastic lesions contained high-grade dysplasia or cancer (Table 2).
Hyperplastic
43.9
Inflammatory
24.6
Diagnostic accuracy In total, 66 of the 72 lesions were assessed by CE, and 60 of the 72 also were assessed by iCLE. The overall diagnostic accuracy of CE for predicting dysplasia was 80.3% (95% CI, 70.7%-89.9%), with a sensitivity of 28.6% (95% CI, 5.1%69.7%) and specificity of 86.4% (95% CI, 74.5%-93.6%). The positive predictive value was 20.0% (95% CI, 3.5%55.8%), and the negative predictive value was 91.1% (95% CI, 79.6%-96.7%) (Table 3). The overall diagnostic accuracy of iCLE in combination with CE for differentiating dysplastic from nondysplastic lesions was 86.7% (95% CI, 78.1%-95.3%), with a sensitivity of 42.9% (95% CI, 11.8%-79.8%) and specificity of 92.4% (95% CI, 80.9%-97.6%). The positive predictive value was 42.9% (95% CI, 11.8%-79.8%) and the negative predictive value was 92.5% (95% CI, 80.9%-96.6%) (Table 4). The difference in diagnostic accuracy between CE and iCLE in combination with CE was not statistically significant (Fisher exact test; P Z .47, 95% CI, 0.92-1.26).
Dysplasia
9.7
Low-grade dysplasia
100
High-grade dysplasia
0
Colorectal cancer
0
SD, Standard deviation.
Because of frequent mechanical failure of the iCLE equipment, this study had to be aborted early, therefore
the calculated sample size was not reached. This cohort showed that the dysplasia detection rate in patients with longstanding Crohn’s colitis is only 9.8%. In 61 patients, with a mean disease duration of 22 years and with previously proven colitis, only 7 dysplastic lesions were found in 6 patients. None of the lesions contained high-grade dysplasia or cancer. In all patients, endoscopic resection of the lesion was considered as sufficient. In addition, we have shown that the sensitivity of combined CE and iCLE is insufficient to differentiate neoplastic from nonneoplastic changes in this low-prevalence setting. Our findings confirm the results from previous studies showing a low prevalence of dysplasia in patients with Crohn’s colitis.2,3,22,23 Patients included in this study had extensive colon involvement of CD and should therefore be considered as the group with the highest risk of developing CRC. Although limited in size, our data concur with recent epidemiologic analysis of a large Danish database including more than 45,000 patients with IBD and reporting on the incidence of CRC. Unlike patients with UC, the incidence of CRC for patients with CD actually was lower than the risk for individuals without IBD. The
www.giejournal.org
Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 969
DISCUSSION
Chromoendoscopy and iCLE in Crohn’s disease surveillance
Wanders et al
authors suggested that this may be due to better surveillance and improved disease management.3 Another recent study showed that surveillance in patients with IBD diminishes the risk of developing CRC.7 It should be emphasized that the calculated sample size was not reached, which might have distorted the outcome of the study. Therefore, no firm conclusions on the dysplasia prevalence can be made because of the limited number of included patients. Larger cohort studies have shown that the cumulative risk of detecting high-grade dysplasia or cancer after a negative screening colonoscopy is 7%.1 Further prospectively collected data are necessary to make a better estimation of the prevalence of dysplastic lesions developing in patients with CD and to sort out the apparent contradiction in the currently available epidemiologic data. The secondary outcome measure of this study focused on the diagnostic accuracy for the differentiation of dysplastic and nondysplastic lesions. The overall accuracy was 80.3% for CE, and was not significantly different for iCLE (86.4%). Although the accuracy was relatively good, the sensitivity was poor for combined CE and iCLE. A sensitivity of 42.9% means that 57.1% of the dysplastic lesions were incorrectly diagnosed as nondysplastic by iCLE. Comparable outcomes were found in the pilot study by van den Broek et al,24 assessing the diagnostic accuracy of probe-based CLE in patients with longstanding UC who were undergoing surveillance. A possible explanation for this lower sensitivity in contradiction with the results from a recent meta-analysis could be that previously inflamed mucosa might be difficult to distinguish from dysplasia.11 The endoscopists assessing
the images are not as well trained as pathologists to distinguish inflammation from dysplasia. In particular, it has not been very well assessed how regenerative changes can be distinguished from dysplastic changes. In the pilot study by Kiesslich et al,25 the overall diagnostic accuracy for distinguishing dysplastic from nondysplastic changes is very good, but it is not clear in how many cases a regenerative pattern was found. Given the average small size of the lesions in our study, another explanation for the poor sensitivity could be inaccurate positioning of the tip of the endomicroscope, thereby generating a false-negative image. The latter also raises the issue of to what extent the final diagnosis by iCLE is influenced by the endoscopic appearance of the lesions. If the lesion looks diminutive, the endoscopist may be more prone to call it a nondysplastic lesion. With this respect, the lower resolution of the iCLE system may exert a negative effect on the diagnostic accuracy during chromoendoscopy. The main reason for early termination was frequent technical failure of the endoscopic equipment in 4 of the 5 participating centers. Initially, the study protocol was intended to include 101 patients within 1 year after the start of the study (February 2010). After more than 4 years, only 61 patients were included. During the study period the endoscopic equipment failed several times in 4 of 5 research centers, with lens distortion being the most common reason. When the equipment was away for repair, no patients could be included in the study, which was the main reason for delay. Although previous CLE results looked promising for differentiating lesions in an average-risk population, we believe that because of the technical limitations, iCLE in colitis surveillance has little additive clinical value in daily endoscopy practice.14 Besides the technical limitations, CLE in itself requires specialized skills for interpretation, positioning the endoscope onto the lesions, and even more robust bowel cleansing than other endoscopic surveillance techniques, because fecal remnants highly interfere with image acquisition and interpretation. In addition, it requires a longer inspection time than a surveillance colonoscopy done solely with CE. It is unclear what the exact future of this high-end technical achievement will be, but possible applications of the technique to predict disease relapse are a more promising clinically relevant application if it can help to tailor therapeutic strategies in patients.26,27 In conclusion, this study shows a low incidence of dysplastic lesions found during surveillance colonoscopy in longstanding extensive Crohn’s colitis. The accuracy of combined CE and iCLE was relatively good, although the sensitivity was poor. Because of frequent technical failure, prolonged procedure time, and insufficient sensitivity, iCLE has limited applicability in daily practice as surveillance strategy in patients with longstanding Crohn’s colitis.
970 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
TABLE 3. Diagnostic accuracy of CE alone Histopathology
Dysplastic CE
Nondysplastic
Dysplastic
Nondysplastic
2/7
8/59
PPV 20.0%
5/7
51/59
NPV 91.1%
Sensitivity 28.6%
Specificity 86.4%
Accuracy 80.3%
CE, Chromoendoscopy; PPV, positive predictive value; NPV, negative predictive value.
TABLE 4. Diagnostic accuracy of iCLE in combination with CE Histopathology
Dysplastic iCLE
Nondysplastic
Dysplastic
Nondysplastic
3/7
4/53
PPV 42.9%
4/7
49/53
NPV 92.5%
Sensitivity 42.9%
Specificity 92.5%
Accuracy 86.7%
iCLE, Integrated confocal laser endomicroscopy; PPV, positive predictive value; NPV, negative predictive value.
Wanders et al
Chromoendoscopy and iCLE in Crohn’s disease surveillance
REFERENCES 1. Friedman S, Rubin PH, Bodian C, et al. Screening and surveillance colonoscopy in chronic Crohn’s colitis: results of a surveillance program spanning 25 years. Clin Gastroenterol Hepatol 2008;6:993-8. 2. Jess T, Gamborg M, Matzen P, et al. Increased risk of intestinal cancer in Crohn’s disease: a meta-analysis of population-based cohort studies. Am J Gastroenterol 2005;100:2724-9. 3. Jess T, Simonsen J, Jørgensen KT, et al. Decreasing risk of colorectal cancer in patients with inflammatory bowel disease over 30 years. Gastroenterology 2012;143:375-81. 4. Rubin DT, Huo D, Kinnucan JA, et al. Inflammation is an independent risk factor for colonic neoplasia in patients with ulcerative colitis: a case-control study. Clin Gastroenterol Hepatol 2013;11:1601-8. 5. Askling J, Dickman PW, Karlén P, et al. Family history as a risk factor for colorectal cancer in inflammatory bowel disease. Gastroenterology 2001;120:1356-62. 6. Wang R, Leong RW. Primary sclerosing cholangitis as an independent risk factor for colorectal cancer in the context of inflammatory bowel disease: a review of the literature. World J Gastroenterol 2014;20: 8783-9. 7. Ananthakrishnan AN, Cagan A, Cai T, et al. Colonoscopy is associated with a reduced risk for colon cancer and mortality in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol. Epub 2014 Jul 17. 8. Annese V, Daperno M, Rutter MD, et al. European evidence based consensus for endoscopy in inflammatory bowel disease. European Crohn’s and Colitis Organisation. J Crohns Colitis 2013;7:982-1018. 9. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666-89. 10. Farraye FA, Odze RD, Eaden J, et al. AGA medical position statement on the diagnosis and management of colorectal neoplasia in inflammatory bowel disease. Gastroenterology 2010;138:738-45. 11. Rutter MD, Riddell RH. Colorectal dysplasia in inflammatory bowel disease: a clinicopathologic perspective. Clin Gastroenterol Hepatol 2014;12:359-67. 12. Subramanian V, Mannath J, Ragunath K, et al. Meta-analysis: the diagnostic yield of chromoendoscopy for detecting dysplasia in patients with colonic inflammatory bowel disease. Aliment Pharmacol Ther 2011;33:304-12. 13. Farraye FA, Waye JD, Moscandrew M, et al. Variability in the diagnosis and management of adenoma-like and non-adenoma-like dysplasia-
www.giejournal.org
14.
15.
16.
17. 18.
19.
20. 21. 22.
23. 24.
25.
26.
27.
associated lesions or masses in inflammatory bowel disease: an Internet-based study. Gastrointest Endosc 2007;66:519-29. Wanders LK, East JE, Uitentuis SE, et al. Diagnostic performance of narrowed spectrum endoscopy, autofluorescence imaging, and confocal laser endomicroscopy for optical diagnosis of colonic polyps: a meta-analysis. Lancet Oncol 2013;14:1337-47. Daperno M, D'Haens G, Van Assche G, et al. Development and validation of a new, simplified endoscopic activity score for Crohn’s disease: the SES-CD. Gastrointest Endosc 2004;60:505-12. Participants in the Paris Workshop. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon. Gastrointest Endosc 2002;58(suppl 6):S3-43. Kudo S, Rubio CA, Teixeira, et al. Pit pattern in colorectal neoplasia: endoscopic magnifying view. Endoscopy 2001;33:367-73. Kiesslich R, Burg J, Vieth M, et al. Confocal laser endoscopy for diagnosing intraepithelial neoplasias and colorectal cancer in vivo. Gastroenterology 2004;127:706-13. Leong RW, Nguyen NQ, Meredith CG, et al. In vivo confocal endomicroscopy in the diagnosis and evaluation of celiac disease. Gastroenterology 2008;135:1870-6. Bisschops R, Bergman J. Probe-based confocal laser endomicroscopy: scientific toy or clinical tool? Endoscopy 2010;42:487-9. Schlemper RJ, Riddell RH, Kato Y, et al. The Vienna classification of gastrointestinal epithelial neoplasia. Gut 2000;47:251-5. Katsanos KH, Stamou P, Tatsioni A, et al. Northwest Greece IBD Study Group. Prevalence of inflammatory bowel disease related dysplasia and cancer in 1500 colonoscopies from a referral center in northwestern Greece. J Crohns Colitis 2011;5:19-23. Basseri RJ, Basseri B, Vassilaki ME, et al. Colorectal cancer screening and surveillance in Crohn’s colitis. J Crohns Colitis 2012;6:824-9. van den Broek FJ, van Es JA, van Eeden S, et al. Pilot study of probebased confocal laser endomicroscopy during colonoscopic surveillance of patients with longstanding ulcerative colitis. Endoscopy 2011;43:116-22. Kiesslich R, Goetz M, Lammersdorf K, et al. Chromoscopy-guided endomicroscopy increases the diagnostic yield of intraepithelial neoplasia in ulcerative colitis. Gastroenterology 2007;132:874-82. Lim LG, Neumann J, Hansen T, et al. Confocal endomicroscopy identifies loss of local barrier function in the duodenum of patients with Crohn’s disease and ulcerative colitis. Inflamm Bowel Dis 2014;20: 892-900. Li CQ, Liu J, Ji R, et al. Use of confocal laser endomicroscopy to predict relapse of ulcerative colitis. BMC Gastroenterol 2014;14:45.
Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 971
Limited applicability of chromoendoscopy-guided confocal laser endomicroscopy as daily-practice surveillance strategy in Crohn’s disease Linda K. Wanders, MSc, MD,1 Teaco Kuiper, MD, PhD,1 Ralf Kiesslich, Prof Dr,2 John G. Karstensen, MSc, MD,3 Rupert W. Leong, MSc, MD,4 Evelien Dekker, Prof Dr,1 Raf Bisschops, Prof Dr5 Amsterdam, Netherlands; Leuven, Belgium
Background and Aims: Patients with longstanding ulcerative colitis have an increased risk for developing colorectal cancer (CRC). Although the risk for ulcerative colitis is well-established, for Crohn’s disease data are contradictory. This study aimed to determine the number of patients with Crohn’s disease with dysplasia who are undergoing surveillance and to assess the diagnostic accuracy of chromoendoscopy (CE) combined with integrated confocal laser endomicroscopy (iCLE) for differentiating dysplastic versus nondysplastic lesions. Methods: Patients with longstanding Crohn’s colitis undergoing surveillance colonoscopy were included in this multicenter, prospective, cohort study. Surveillance was performed with CE, and lesions were assessed with iCLE for differentiation. All lesions were removed and sent for pathology as the reference standard. Results: Between 2010 and 2014, a total of 61 patients with Crohn’s colitis were included in 5 centers. Seventytwo lesions, of which 7 were dysplastic, were detected in 6 patients (dysplasia detection rate 9.8%); none included high-grade dysplasia or cancer. Combined CE with iCLE for differentiating neoplastic from nonneoplastic lesions had accuracy of 86.7% (95% confidence interval [CI], 78.1-95.3), sensitivity of 42.9% (95% CI, 11.8-79.8), and specificity of 92.4% (95% CI, 80.9-97.6). For CE alone, this was 80.3% (95% CI, 70.7-89.9), 28.6% (95% CI, 5.1-69.7), and 86.4% (95% CI, 80.9-97.6). The study terminated early because of frequent failure of the endoscopic equipment. Conclusions: This study shows a low incidence of dysplastic lesions found during surveillance colonoscopy in patients with longstanding extensive Crohn’s colitis. The accuracy of both CE alone and CE in combination with iCLE was relatively good, although the sensitivity for both was poor. Because of frequent equipment failure, iCLE has limited applicability in daily practice as a surveillance strategy. (Gastrointest Endosc 2016;83:966-71.)
Longstanding ulcerative colitis (UC) increases a patient’s risk of developing colorectal dysplasia and colorectal cancer (CRC). Although this risk is well established for UC, conflicting data exist regarding this risk with extensive colitis in patients with Crohn’s disease (CD). The risk
estimates in patients with CD differ considerably from no increase in CRC risk to a relative risk of up to 3.3 for developing CRC.1-3 A potential explanation for this heterogeneity in CRC risk might be selection bias. Previous studies analyzing
Abbreviations: CD, Crohn’s disease; CE, chromoendoscopy; CLE, confocal laser endomicroscopy; CRC, colorectal cancer; IBD, inflammatory bowel disease; iCLE, integrated confocal laser endomicroscopy; UC, ulcerative colitis.
Received March 17, 2015. Accepted September 1, 2015.
DISCLOSURE: R. Kiesslich is a consultant for and is on paid advisory boards for Pentax. J.G. Karstensen received financial support from Pentax, Mauna Kea Technologies, and Boston Scientific. E. Dekker received financial support and has equipment on loan from Olympus and Pentax. R. Bisschops is supported by Research Foundation Flanders (FWO Vlaanderen) and received a speaker’s fee and research support from Pentax. All other authors disclosed no financial relationships relevant to this publication.
Current affiliations: Department of Gastroenterology and Hepatology, Academic Medical Centre, Amsterdam, The Netherlands (1), Department of Gastroenterology, Dr. Horst Schmidt Klinik, Wiesbaden, Germany (2), Gastro Unit, Division of Endoscopy, Copenhagen University Hospital Herlev, Herlev, Denmark (3), Gastroenterology and Liver Services, Bankstown-Lidcombe Hospital, Sydney, Australia (4), Department of Gastroenterology, University Hospitals Leuven, Leuven, Belgium (5). Reprint requests: R. Bisschops, MD, PhD, Department of Gastroenterology and Hepatology, University Hospital Leuven, Herestraat 49, 3000, Leuven, Belgium.
Copyright ª 2016 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2015.09.001
If you would like to chat with an author of this article, you may contact Dr Bisschops at [email protected].
966 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
Wanders et al
CRC risk in patients with CD included patients with different disease extents. Studies including patients with CD and colitis showed, however, a significant increase in CRC risk.2-3 Besides disease extent, several other risk factors, such as disease duration, family history of colon cancer, and concomitant primary sclerosing cholangitis, are known to increase the risk of developing dysplasia and cancer in patients with CD.4-6 Colonoscopic surveillance of patients with inflammatory bowel disease (IBD) is associated with early detection of premalignant lesions and subsequently to an improved prognosis.7 Guidelines therefore recommend surveillance in patients with IBD.8-10 It is known that dysplasia in patients with IBD can be subtle and flat and therefore more difficult to detect than dysplasia in an average-risk population.11 In addition, postinflammatory mucosal changes such as scarring and inflammatory pseudopolyps further complicate detection of dysplasia in these patients. Until recently, surveillance was performed by using white-light endoscopy with 4-quadrant biopsies every 10 cm. However, recent studies have proven that chromoendoscopy (CE) significantly increases lesion detection in IBD surveillance colonoscopies compared with white light endoscopy with quadrant-random biopsies every 10 centimeters.12 Current guidelines recommend the use of CE in surveillance of patients with longstanding colitis.8-10 During CE, a dye spray (eg, methylene blue or indigo carmine) is used to highlight the mucosal surface and to enhance the delineation of (early) dysplastic lesions. Although CE improves detection, it has not yet been proven that CE improves differentiation. Even endoscopists with expertise in IBD show great interobserver variation in differentiating lesions found during IBD surveillance colonoscopy, reflecting the difficulty in distinguishing dysplastic from nondysplastic lesions.13 Confocal laser endomicroscopy (CLE) is an advanced imaging technique providing highly magnified images of GI epithelium comparable to histopathology. Because of visualization of a limited surface area, CLE is specifically used for differentiation, rather than for detection. CLE can be integrated into a colonoscope (Pentax, Tokyo, Japan) (integrated confocal laser endomicroscopy [iCLE]) or can be applied as a probe system that passes through the working channel of any endoscope (Mauna Kea Technologies, Paris, France). A recent meta-analysis showed that CLE has the highest accuracy of all commercially available imaging techniques in differentiation of neoplastic from nonneoplastic lesions in an average-risk population, although mainly by experts.14 The aim of this study was to determine the number of patients with dysplasia undergoing surveillance colonoscopy for longstanding extensive Crohn’s colitis and to assess whether the combination of CE and iCLE improves differentiation of lesions found during surveillance colonoscopy. www.giejournal.org
Chromoendoscopy and iCLE in Crohn’s disease surveillance
METHODS Patients This study was performed in 5 academic medical centers. Patients with Crohn’s colitis undergoing surveillance colonoscopy were included in the study if they were aged 18 years, diagnosed with >50 cm of colon disease, and if they were in clinical remission, based on the Simple Endoscopic Score for Crohn’s Disease.15 Patients were excluded if they had undergone colon resections with <50% of their colons in situ, incomplete colonoscopy because of poor bowel preparation, disease activity or stenosis, contraindications for iCLE (allergy for intravenous fluorescein, pregnancy or breastfeeding, severe cardiopulmonary disease, or pre-existent renal disease), or noncorrectable coagulopathy that precludes taking biopsy specimens. This study was approved by the medical ethics committees of all participating centers (NTR2293).
Endoscopic equipment For CE and CLE, the endoscope-integrated system, iCLE (Pentax, Tokyo, Japan and Optiscan; Melbourne, Australia), was used. In this system, the confocal laser unit is integrated into the distal tip of a standard video endoscope, enabling iCLE imaging on a second monitor. The diameter of both the distal tip and the insertion tube is 12.8 mm. The iCLE imaging plane depth relative to the mucosal surface (ie, z-axis) can be controlled by using two buttons on the hand piece of the endoscope. During iCLE, a single-line laser beam with an excitation wavelength of 488 nm is delivered at the tissue. Confocal image data are collected at a scan rate of 0.8 frames per second. The CLE images have a lateral resolution of 0.7 mm. The field of view is 500 mm, and the range of the z-axis is 0 to 250 mm below the surface.
Endoscopic procedure All patients were prepared with either 4 L of polyethylene glycol solution or 2 L of polyethylene glycol solution plus ascorbic acid (Moviprep or Kleanprep, Norgine, Netherlands). The procedures were performed with the patients under conscious sedation by using intravenous midazolam and fentanyl or propofol at the discretion of the endoscopist. At the discretion of the endoscopists, butylscopolamine was given intravenously to reduce colon motility. During withdrawal, the entire colon was stained with a 0.1% methylene blue solution. Bowel preparation was determined as good (100% mucosa visible), moderate (>90% mucosa visible), or poor (<90% mucosa, even after extensive cleaning). Each colon segment was scrutinized for the presence of suspicious areas, mucosal irregularities, unusual ulcers, and strictures. All detected lesions were classified according to the macroscopic classification (Paris Classification16) of early GI dysplasia, scored for size, location in the colon, and distance from the anus. All lesions were assessed for Kudo pit pattern.17 The colon mucosa Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 967
Chromoendoscopy and iCLE in Crohn’s disease surveillance
Wanders et al
1 cm adjacent to each detected lesion also was scored for the Kudo pit pattern. The location of all detected lesions with respect to the extent of colitis (proximal to or within the inflammatory changed colon on endoscopy) was noted. Digital still images of all detected lesions were taken. Subsequently, after intravenous injection of fluorescein (10%), all lesions as well as their adjacent “normal” mucosa were inspected with iCLE and scored by the endoscopist as normal, regenerative, or dysplastic, according to the previously described Mainz criteria.18 Hereafter, targeted biopsies of detected lesion and 4 random biopsies of the adjacent colon mucosa were sent for histopathologic evaluation.
Endoscopists All procedures were performed by experienced endoscopists who have participated in previous studies with CE and iCLE in Amsterdam, Mainz, Leuven, Copenhagen, and Sydney.18-20 Therefore, all endoscopists were familiar with the techniques and should have performed beyond the learning curve.
Outcome measures The primary outcome measure of this study was the number of patients with dysplasia detected in a Crohn’s colitis surveillance group, expressed as dysplasia detection rate. The secondary outcome measure was diagnostic accuracy of CE and iCLE for the differentiation of dysplastic and nondysplastic mucosa, by using final histopathology as the reference standard diagnosis.
Figure 1. Flow diagram of patient inclusion.
Study design This was a multicenter, prospective, cohort study. The sample size calculation was based on the analysis by Friedman et al,1 in which the prevalence of dysplasia or cancer in patients with CD affecting at least one third of the colon was 7% on screening examination. A successful endoscopic surveillance protocol was defined as a similar dysplasia incidence of 7% and an upper limit of the 95% CI of 12%. The calculated sample size was 101 patients.
RESULTS Patient characteristics
Normal distributed data were described with the mean and the standard deviation, and data with a skewed distribution were described by using the median and the interquartile range. The diagnostic accuracy of CE and iCLE for differentiation was calculated by creating a 2 2 table comparing the index test result with the reference standard (ie, dysplastic or nondysplastic, according to histopathology). A positive index test result on CE and iCLE was defined as a dysplastic appearance during iCLE. The sensitivity, specificity, and overall accuracy were calculated and represented by a point estimator with the 95% confidence interval (CI). The difference in performance of CE and iCLE was compared with the Fisher exact test.
Between February 2010 and June 2014, a total of 82 patients with CD were recruited, of whom 61 were included (Fig. 1). The mean age was 48 years, and participants had a mean disease duration of 22 years by the time they were included in this study. In 2 patients, circumscribed dysplasia was detected in previous colonoscopies and endoscopically resected, and 2 patients had concomitant primary sclerosing cholangitis. Bowel preparation was considered as good in 60.7% of cases; all other cases scored moderate. Four patients had to be excluded because of poor bowel preparation. Other patients had to be excluded because of active inflammation (n Z 10), impassable stenosis (n Z 3), or technical failure of the endoscope (n Z 4) (Fig. 1). Because of frequent technical failure of the equipment, the study terminated early, and the calculated sample size was not reached. In 4 of 5 centers, the lens of the endomicroscope broke, and the equipment had to be sent to the manufacturer. Each repair took several weeks to months, which hindered patient recruitment and inclusion. The mean total procedure time was 40 minutes. The mean total inspection time including iCLE acquisition was 30 minutes, and without iCLE assessments, this was 10 minutes. Of all included patients, 34 had at least 1 lesion. In total, 72 lesions were detected. Of these lesions, 66 were examined with CE and 60 with iCLE. Seven dysplastic lesions were found in 6 patients, which accounts for a dysplasia detection rate of 9.8% (Table 1).
968 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
Histopathology Histologic samples are processed by using standard procedures and evaluated by one GI expert in each center. In case of discrepancy, discussion between two pathologists would lead to a consensus diagnosis. The pathologists were not aware of the detection technique and the findings on iCLE. Biopsy specimens were classified according to the Vienna criteria of GI epithelial neoplasia.21 The histologic diagnosis of all biopsy specimens served as the reference standard.
Statistics
Wanders et al
Chromoendoscopy and iCLE in Crohn’s disease surveillance
TABLE 1. Patient characteristics (N [ 61) Male, no (%)
TABLE 2. Lesion characteristics (N [72) 29 (47.5)
Size, mean ( SD), mm Paris classification,
16
6.8 ( 4.7)
%
Age at diagnosis, mean ( SD), y
26 ( 9.9)
Age at colonoscopy, mean ( SD), y
48 ( 11.5)
Is
22.2
Duration of colitis, mean ( SD), y
22 ( 7.8)
Isp
9.7
Primary sclerosing cholangitis, no. (%)
2 (3.3)
Ip
1.4
Previous dysplasia, no. (%)
2 (3.3)
IIa
38.9
IIb
9.7
37 (60.7)
IIc
0
24 (39.3)
IIaþc
40 ( 17)
Unknown
Colon preparation, no. (%) Good Moderate Total procedure time, mean ( SD), min Inspection time without iCLE, mean ( SD), min
10 ( 7)
Inspection time with iCLE, mean ( SD), min
30 ( 14)
0 18.1
Location, % Cecum
13.9
Lesions detected, no.
72
Ascending colon
11.1
Adenomas detected, no.
7
Transverse colon
19.4
Patients with lesions, no.
34
Descending colon
11.1
Patients with dysplasia, no.
6
Sigmoid colon
23.6
Dysplasia detection rate, %
9.8
Rectum
18.1
Missing
2.8
SD, Standard deviation; iCLE, integrated confocal laser endomicroscopy.
Histopathology, %
Lesion characteristics
Normal
21.8
The mean diameter of the detected lesions was 6.8 mm. Lesions were categorized as flat (Paris IIa/IIb16) in almost half of cases; none of the lesions were depressed. The lesions were equally distributed throughout the colon. None of the dysplastic lesions contained high-grade dysplasia or cancer (Table 2).
Hyperplastic
43.9
Inflammatory
24.6
Diagnostic accuracy In total, 66 of the 72 lesions were assessed by CE, and 60 of the 72 also were assessed by iCLE. The overall diagnostic accuracy of CE for predicting dysplasia was 80.3% (95% CI, 70.7%-89.9%), with a sensitivity of 28.6% (95% CI, 5.1%69.7%) and specificity of 86.4% (95% CI, 74.5%-93.6%). The positive predictive value was 20.0% (95% CI, 3.5%55.8%), and the negative predictive value was 91.1% (95% CI, 79.6%-96.7%) (Table 3). The overall diagnostic accuracy of iCLE in combination with CE for differentiating dysplastic from nondysplastic lesions was 86.7% (95% CI, 78.1%-95.3%), with a sensitivity of 42.9% (95% CI, 11.8%-79.8%) and specificity of 92.4% (95% CI, 80.9%-97.6%). The positive predictive value was 42.9% (95% CI, 11.8%-79.8%) and the negative predictive value was 92.5% (95% CI, 80.9%-96.6%) (Table 4). The difference in diagnostic accuracy between CE and iCLE in combination with CE was not statistically significant (Fisher exact test; P Z .47, 95% CI, 0.92-1.26).
Dysplasia
9.7
Low-grade dysplasia
100
High-grade dysplasia
0
Colorectal cancer
0
SD, Standard deviation.
Because of frequent mechanical failure of the iCLE equipment, this study had to be aborted early, therefore
the calculated sample size was not reached. This cohort showed that the dysplasia detection rate in patients with longstanding Crohn’s colitis is only 9.8%. In 61 patients, with a mean disease duration of 22 years and with previously proven colitis, only 7 dysplastic lesions were found in 6 patients. None of the lesions contained high-grade dysplasia or cancer. In all patients, endoscopic resection of the lesion was considered as sufficient. In addition, we have shown that the sensitivity of combined CE and iCLE is insufficient to differentiate neoplastic from nonneoplastic changes in this low-prevalence setting. Our findings confirm the results from previous studies showing a low prevalence of dysplasia in patients with Crohn’s colitis.2,3,22,23 Patients included in this study had extensive colon involvement of CD and should therefore be considered as the group with the highest risk of developing CRC. Although limited in size, our data concur with recent epidemiologic analysis of a large Danish database including more than 45,000 patients with IBD and reporting on the incidence of CRC. Unlike patients with UC, the incidence of CRC for patients with CD actually was lower than the risk for individuals without IBD. The
www.giejournal.org
Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 969
DISCUSSION
Chromoendoscopy and iCLE in Crohn’s disease surveillance
Wanders et al
authors suggested that this may be due to better surveillance and improved disease management.3 Another recent study showed that surveillance in patients with IBD diminishes the risk of developing CRC.7 It should be emphasized that the calculated sample size was not reached, which might have distorted the outcome of the study. Therefore, no firm conclusions on the dysplasia prevalence can be made because of the limited number of included patients. Larger cohort studies have shown that the cumulative risk of detecting high-grade dysplasia or cancer after a negative screening colonoscopy is 7%.1 Further prospectively collected data are necessary to make a better estimation of the prevalence of dysplastic lesions developing in patients with CD and to sort out the apparent contradiction in the currently available epidemiologic data. The secondary outcome measure of this study focused on the diagnostic accuracy for the differentiation of dysplastic and nondysplastic lesions. The overall accuracy was 80.3% for CE, and was not significantly different for iCLE (86.4%). Although the accuracy was relatively good, the sensitivity was poor for combined CE and iCLE. A sensitivity of 42.9% means that 57.1% of the dysplastic lesions were incorrectly diagnosed as nondysplastic by iCLE. Comparable outcomes were found in the pilot study by van den Broek et al,24 assessing the diagnostic accuracy of probe-based CLE in patients with longstanding UC who were undergoing surveillance. A possible explanation for this lower sensitivity in contradiction with the results from a recent meta-analysis could be that previously inflamed mucosa might be difficult to distinguish from dysplasia.11 The endoscopists assessing
the images are not as well trained as pathologists to distinguish inflammation from dysplasia. In particular, it has not been very well assessed how regenerative changes can be distinguished from dysplastic changes. In the pilot study by Kiesslich et al,25 the overall diagnostic accuracy for distinguishing dysplastic from nondysplastic changes is very good, but it is not clear in how many cases a regenerative pattern was found. Given the average small size of the lesions in our study, another explanation for the poor sensitivity could be inaccurate positioning of the tip of the endomicroscope, thereby generating a false-negative image. The latter also raises the issue of to what extent the final diagnosis by iCLE is influenced by the endoscopic appearance of the lesions. If the lesion looks diminutive, the endoscopist may be more prone to call it a nondysplastic lesion. With this respect, the lower resolution of the iCLE system may exert a negative effect on the diagnostic accuracy during chromoendoscopy. The main reason for early termination was frequent technical failure of the endoscopic equipment in 4 of the 5 participating centers. Initially, the study protocol was intended to include 101 patients within 1 year after the start of the study (February 2010). After more than 4 years, only 61 patients were included. During the study period the endoscopic equipment failed several times in 4 of 5 research centers, with lens distortion being the most common reason. When the equipment was away for repair, no patients could be included in the study, which was the main reason for delay. Although previous CLE results looked promising for differentiating lesions in an average-risk population, we believe that because of the technical limitations, iCLE in colitis surveillance has little additive clinical value in daily endoscopy practice.14 Besides the technical limitations, CLE in itself requires specialized skills for interpretation, positioning the endoscope onto the lesions, and even more robust bowel cleansing than other endoscopic surveillance techniques, because fecal remnants highly interfere with image acquisition and interpretation. In addition, it requires a longer inspection time than a surveillance colonoscopy done solely with CE. It is unclear what the exact future of this high-end technical achievement will be, but possible applications of the technique to predict disease relapse are a more promising clinically relevant application if it can help to tailor therapeutic strategies in patients.26,27 In conclusion, this study shows a low incidence of dysplastic lesions found during surveillance colonoscopy in longstanding extensive Crohn’s colitis. The accuracy of combined CE and iCLE was relatively good, although the sensitivity was poor. Because of frequent technical failure, prolonged procedure time, and insufficient sensitivity, iCLE has limited applicability in daily practice as surveillance strategy in patients with longstanding Crohn’s colitis.
970 GASTROINTESTINAL ENDOSCOPY Volume 83, No. 5 : 2016
www.giejournal.org
TABLE 3. Diagnostic accuracy of CE alone Histopathology
Dysplastic CE
Nondysplastic
Dysplastic
Nondysplastic
2/7
8/59
PPV 20.0%
5/7
51/59
NPV 91.1%
Sensitivity 28.6%
Specificity 86.4%
Accuracy 80.3%
CE, Chromoendoscopy; PPV, positive predictive value; NPV, negative predictive value.
TABLE 4. Diagnostic accuracy of iCLE in combination with CE Histopathology
Dysplastic iCLE
Nondysplastic
Dysplastic
Nondysplastic
3/7
4/53
PPV 42.9%
4/7
49/53
NPV 92.5%
Sensitivity 42.9%
Specificity 92.5%
Accuracy 86.7%
iCLE, Integrated confocal laser endomicroscopy; PPV, positive predictive value; NPV, negative predictive value.
Wanders et al
Chromoendoscopy and iCLE in Crohn’s disease surveillance
REFERENCES 1. Friedman S, Rubin PH, Bodian C, et al. Screening and surveillance colonoscopy in chronic Crohn’s colitis: results of a surveillance program spanning 25 years. Clin Gastroenterol Hepatol 2008;6:993-8. 2. Jess T, Gamborg M, Matzen P, et al. Increased risk of intestinal cancer in Crohn’s disease: a meta-analysis of population-based cohort studies. Am J Gastroenterol 2005;100:2724-9. 3. Jess T, Simonsen J, Jørgensen KT, et al. Decreasing risk of colorectal cancer in patients with inflammatory bowel disease over 30 years. Gastroenterology 2012;143:375-81. 4. Rubin DT, Huo D, Kinnucan JA, et al. Inflammation is an independent risk factor for colonic neoplasia in patients with ulcerative colitis: a case-control study. Clin Gastroenterol Hepatol 2013;11:1601-8. 5. Askling J, Dickman PW, Karlén P, et al. Family history as a risk factor for colorectal cancer in inflammatory bowel disease. Gastroenterology 2001;120:1356-62. 6. Wang R, Leong RW. Primary sclerosing cholangitis as an independent risk factor for colorectal cancer in the context of inflammatory bowel disease: a review of the literature. World J Gastroenterol 2014;20: 8783-9. 7. Ananthakrishnan AN, Cagan A, Cai T, et al. Colonoscopy is associated with a reduced risk for colon cancer and mortality in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol. Epub 2014 Jul 17. 8. Annese V, Daperno M, Rutter MD, et al. European evidence based consensus for endoscopy in inflammatory bowel disease. European Crohn’s and Colitis Organisation. J Crohns Colitis 2013;7:982-1018. 9. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666-89. 10. Farraye FA, Odze RD, Eaden J, et al. AGA medical position statement on the diagnosis and management of colorectal neoplasia in inflammatory bowel disease. Gastroenterology 2010;138:738-45. 11. Rutter MD, Riddell RH. Colorectal dysplasia in inflammatory bowel disease: a clinicopathologic perspective. Clin Gastroenterol Hepatol 2014;12:359-67. 12. Subramanian V, Mannath J, Ragunath K, et al. Meta-analysis: the diagnostic yield of chromoendoscopy for detecting dysplasia in patients with colonic inflammatory bowel disease. Aliment Pharmacol Ther 2011;33:304-12. 13. Farraye FA, Waye JD, Moscandrew M, et al. Variability in the diagnosis and management of adenoma-like and non-adenoma-like dysplasia-
www.giejournal.org
14.
15.
16.
17. 18.
19.
20. 21. 22.
23. 24.
25.
26.
27.
associated lesions or masses in inflammatory bowel disease: an Internet-based study. Gastrointest Endosc 2007;66:519-29. Wanders LK, East JE, Uitentuis SE, et al. Diagnostic performance of narrowed spectrum endoscopy, autofluorescence imaging, and confocal laser endomicroscopy for optical diagnosis of colonic polyps: a meta-analysis. Lancet Oncol 2013;14:1337-47. Daperno M, D'Haens G, Van Assche G, et al. Development and validation of a new, simplified endoscopic activity score for Crohn’s disease: the SES-CD. Gastrointest Endosc 2004;60:505-12. Participants in the Paris Workshop. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon. Gastrointest Endosc 2002;58(suppl 6):S3-43. Kudo S, Rubio CA, Teixeira, et al. Pit pattern in colorectal neoplasia: endoscopic magnifying view. Endoscopy 2001;33:367-73. Kiesslich R, Burg J, Vieth M, et al. Confocal laser endoscopy for diagnosing intraepithelial neoplasias and colorectal cancer in vivo. Gastroenterology 2004;127:706-13. Leong RW, Nguyen NQ, Meredith CG, et al. In vivo confocal endomicroscopy in the diagnosis and evaluation of celiac disease. Gastroenterology 2008;135:1870-6. Bisschops R, Bergman J. Probe-based confocal laser endomicroscopy: scientific toy or clinical tool? Endoscopy 2010;42:487-9. Schlemper RJ, Riddell RH, Kato Y, et al. The Vienna classification of gastrointestinal epithelial neoplasia. Gut 2000;47:251-5. Katsanos KH, Stamou P, Tatsioni A, et al. Northwest Greece IBD Study Group. Prevalence of inflammatory bowel disease related dysplasia and cancer in 1500 colonoscopies from a referral center in northwestern Greece. J Crohns Colitis 2011;5:19-23. Basseri RJ, Basseri B, Vassilaki ME, et al. Colorectal cancer screening and surveillance in Crohn’s colitis. J Crohns Colitis 2012;6:824-9. van den Broek FJ, van Es JA, van Eeden S, et al. Pilot study of probebased confocal laser endomicroscopy during colonoscopic surveillance of patients with longstanding ulcerative colitis. Endoscopy 2011;43:116-22. Kiesslich R, Goetz M, Lammersdorf K, et al. Chromoscopy-guided endomicroscopy increases the diagnostic yield of intraepithelial neoplasia in ulcerative colitis. Gastroenterology 2007;132:874-82. Lim LG, Neumann J, Hansen T, et al. Confocal endomicroscopy identifies loss of local barrier function in the duodenum of patients with Crohn’s disease and ulcerative colitis. Inflamm Bowel Dis 2014;20: 892-900. Li CQ, Liu J, Ji R, et al. Use of confocal laser endomicroscopy to predict relapse of ulcerative colitis. BMC Gastroenterol 2014;14:45.
Volume 83, No. 5 : 2016 GASTROINTESTINAL ENDOSCOPY 971