SEARCH

SEARCH BY CITATION

Keywords:

  • inflammatory bowel disease;
  • Crohn's disease;
  • confocal laser endomicroscopy;
  • fluorescein;
  • advanced imaging

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Background:

Confocal laser endomicroscopy (CLE) allows microscopic imaging within the mucosal layer of the gut during ongoing endoscopy. Different studies have addressed the potential of CLE for in vivo diagnosis of ulcerative colitis and microscopic colitis. However, there are no data on the utility of CLE for in vivo diagnosis of Crohn's disease (CD). The aim was to assess the clinical utility of CLE in patients with CD and to determine whether disease activity can be graded using CLE.

Methods:

Consecutive patients with and without CD were enrolled. The colonic mucosa was examined by standard white-light endoscopy followed by CLE. The features seen on CLE were compared between CD patients and controls.

Results:

In all, 76 patients with CD were screened, of whom 54 patients were included in the present study. Eighteen patients without inflammatory bowel disease (IBD) served as controls. A significantly higher proportion of patients with active CD had increased colonic crypt tortuosity, enlarged crypt lumen, microerosions, augmented vascularization, and increased cellular infiltrates within the lamina propria. In quiescent CD, a significant increase in crypt and goblet cell number was detected compared with controls. Based on our findings, we propose a Crohn's Disease Endomicroscopic Activity Score (CDEAS) for assessing CD activity in vivo.

Conclusions:

CLE has the potential to significantly improve diagnosis of CD compared with standard endoscopy. These findings should be evaluated in future prospective trials to assess the value of this newly developed CLE score for prediction of disease course and therapeutic responses. (Inflamm Bowel Dis 2012;)

Inflammatory bowel diseases (IBDs) comprise two major forms of chronic intestinal disorders: Crohn's disease (CD) and ulcerative colitis (UC). Recent evidence suggests that IBD results from an inappropriate inflammatory response to the commensal microflora in a genetically susceptible host environment.1, 2

It is estimated that CD affects about 500,000 people in North America, while the prevalence of CD in Northern Europe ranges between 27–48 patients per 100,000 people.3, 4 It is well established that patients with IBD suffer from an increased risk for intestinal and extraintestinal cancer. In this context, the duration and anatomic extent are well established risk factors for colon cancer development.5–7 Therefore, surveillance colonoscopy is considered the gold standard in diagnosing intraepithelial neoplasia (formerly termed dysplasia) and cancer in IBD patients. Additionally, national and international guidelines recommend ileocolonoscopy as an outcome parameter for biological therapies or to access therapeutic efficacy.8–10

After a certain time of disease duration, most consensus groups recommend yearly surveillance colonoscopies and the obtainment of a specific number of random biopsies in addition to targeted biopsies from suspicious lesions for cancer surveillance.9, 10 Previously, it was reported that to exclude neoplasia in colonic mucosal biopsies at least 56 nontargeted jumbo-forceps biopsies have to be taken.11

In addition, regenerative hyperplasia in chronic IBD forms extensive polyposis. While these pseudopolyps appeared to be more frequent in UC than in CD, little is known about the mechanisms that trigger their growth and it is still under debate if these polyps harbor a neoplastic risk.12

However, while there is no doubt on the imperative of endoscopy in IBD patients, final diagnosis of intraepithelial neoplasia and disease activity still remained the domain of histopathology.

Recently, confocal laser endomicroscopy (CLE) has emerged as a new endoscopic imaging modality enabling real-time in vivo histology of the cellular mucosal layer at a magnification of 1000-fold.13, 14 Endomicroscopy is based on tissue illumination after application of an exogenous fluorescence agent. These agents can either be applied systemically (fluorescein) or topically (acriflavine hydrochloride, cresyl violet) to the intestinal mucosa. The use of fluorescence agents enables imaging of mucosal structures and emphasizes the imaging of features such as the capillary, architectural, and cellular patterns. Recent reports have shown that endomicroscopy has the capability to encourage histopathological diagnosis in different gastrointestinal diseases including Barrett's esophagus, celiac disease, and colon polyps.15–17

Although previous studies have also shown the utility of endomicroscopy in diagnosing ulcerative and microscopic colitis, the role of endomicroscopy in patients with CD has not been explored yet.18, 19

The hypothesis of this study was that endomicroscopy could reveal mucosal findings in patients with CD not visualized by standard endoscopy and comparable to conventional histopathological analysis. Thus, the aim of this trial was to prospectively access the clinical utility of endomicroscopy in patients with CD with or without macroscopic visible lesions.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Screening, Inclusion, and Exclusion Criteria

Consecutive patients with and without CD who underwent colonoscopy (EC-3870 CIFK, Pentax, Tokyo, Japan) for the evaluation of their symptoms were prospectively included between October 2009 and August 2011 at the endoscopy unit at the University of Erlangen-Nuremberg. All patients signed informed consent to participate in this study after the endoscopist and attending physician had explained the procedure in detail to them. The study was approved by the local ethical committee and government authorities and was conducted according to the Declaration of Helsinki. The UMIN Clinical Trials Registry identification number for this study is NCT01102855.

Subjects were enrolled if they met the following inclusion criteria: more than 18 years of age, ability to provide written informed consent, and known CD. Patients with one or more of the following criteria were excluded from the study: inability to provide written informed consent, severe uncontrolled coagulopathy, impaired renal function, pregnancy or breast feeding, active gastrointestinal bleeding, known allergy to fluorescein, and residence in institutions.

CD activity was assessed using the Crohn's Disease Endoscopic Index of Severity (CDEIS). The score calculates disease activity according to five endoscopically visualized segments (rectum, sigmoid/left colon, transverse colon, right colon, ileum) and endoscopic findings, including ulcer size, extent of disease (surface involved by disease and surface involved by ulcerations), and stenosis.20, 21 C-reactive protein (CRP) was determined in every patient at the time of endoscopy.

Control patients underwent colonoscopy due to colorectal cancer screening according to current national and international guidelines. Patients suffering from infectious diarrhea were not included.

Structure and Working of the Endomicroscopy System

Currently, two CE (European Conformity) certified and FDA (Food and Drug Administration)-approved endomicroscopy systems are available including an integrated endoscopy system (iCLE; Pentax) and a probe-based system (pCLE, Cellvizio, Mauna Kea Technologies, Paris, France).13, 14

The integrated system consists of a conventional high-resolution white-light video endoscope in which a confocal fluorescence microscope is integrated into the distal tip. The confocal lens is slightly prominent at the distal tip at the 7 o'clock position and can be clearly visualized in the videoendoscopic image. Therefore, it is possible to macroscopically target suspicious areas under videoendoscopic control. The system enables the endoscopist to manually adjust the scan rate between 1.6 frames per second at a resolution of 1024 × 512 pixels and 0.8 frames per second at a resolution of 1024 × 1024 pixels (1 megapixel) with dynamically adjustable depth of scanning ranging from 0–250 μm with control to ≈4-μm increments. The optical slice thickness is 7 μm, with lateral and axial resolution of 0.7 μm and a confocal image field of view (FOV) of 475 × 475 μm. Additionally, it is possible to manually adjust the laser power between 0–1000 μW.13, 14

The pCLE-system represents a stand-alone confocal probe that is capable of passage through the accessory channel of a standard video endoscope and utilizes a fixed laser power and fixed imaging plane depth to explore the mucosal microarchitecture. Confocal images are streamed at a rate of 12 frames per second, making it possible to obtain real-time videos of the intestinal mucosa. A special computer algorithm (“mosaicing”) allows reconstruction of single video frames either in real-time or postprocessed with an increased FOV (4 × 2 mm). In addition, postprocessing using the Cellvizio Viewer (Mauna Kea Technologies) enables virtual staining of mucosal structures to further enhance tissue contrast. In this study the ColoFlexUHD probe was used. The probe requires an accessory channel of 2.8 mm and has a lateral resolution of 1 μm with a FOV of 240 μm and fixed image plane depth varying between 55–65 μm.13, 14

Endomicroscopy Procedure

All colonoscopies were performed by two experts in confocal imaging (H.N. and J.M.). Endomicroscopy procedures were performed under conscious sedation (e.g., midazolam hydrochloride and pedithin hydrochloride) with constant monitoring of vital signs using either iCLE or pCLE utilizing the ultrahigh definition probe.

iCLE images were collected at a frame rate of 0.8/s at 1024 × 1024 pixels or 1.6/s at 1024 × 512 pixels. Initially, the whole colon and terminal ileum were analyzed with standard high-resolution white-light endoscopy. In order to evaluate the presence of mucosal lesions we assessed the CDEIS in each patient as described before.20, 21 For definition of quiescent and active CD we defined a CDEIS cutoff value of 6. Consequently, a CDEIS >6 defined active CD. Subsequent to macroscopic examination, 5 mL of fluorescein sodium 10% (Alcon Laboratories, Fort Worth, TX) was administered intravenously in every patient to optimize tissue contrast. First, confocal images were analyzed in real time. After the endoscopic procedure, images were reviewed offline to zoom in on details (iCLE), which allowed a higher magnification of the mucosa (≈10,000 fold) or by using the Cellvizio Viewer for virtual staining of mucosal structures to enhance tissue contrast.

Normal mucosa and pathological lesions were evaluated according to the Mainz confocal pattern classification for iCLE and according to the Miami classification for pCLE.22, 23

After endomicroscopy, biopsies were taken from both macroscopically normal and abnormal mucosa and specimens were fixed in 4% buffered formalin for subsequent histopathological analysis. Findings seen on review of the confocal images were compared with histology reviewed by a specialized gastrointestinal-pathologist.

Pilot Study

Initially, 10 patients (six in the active phase, four in the quiescent phase) with CD and two controls were enrolled in a pilot study to evaluate, grade, and classify the endomicroscopic images. Based on these initial evaluations, alterations in the following features seen on endomicroscopy (but not on standard white-light endoscopy) were identified as characteristic for active CD compared with inactive disease and controls.

  • 1
    Crypts (number, tortuosity, crypt lumen). Based on a previous publication, a normal crypt number was defined as being eight crypts per FOV.18 Tortuosity was defined as the crypts having cobblestone behavior.
  • 2
    Presence of microerosions. Microerosions were defined as erosions that were not visible at standard endoscopy but on endomicroscopy.
  • 3
    Vascularity within the lamina propria.
  • 4
    Goblet cells (number). These cells were defined according to previously published data.15, 18
  • 5
    Cellular infiltrate within the lamina propria.

Afterwards, these factors were recorded and evaluated in the prospective study.

Prospective Study

Confocal images reflecting these five features were obtained from both macroscopically normal and abnormal mucosa. The images were then evaluated by the investigators blinded to the patient groups. The maximum, minimum, and average numbers of crypts/field were recorded for each patient. Additionally, tortuosity of crypts and extent of crypt lumens were recorded. Furthermore, other findings such as the presence of microerosions, vascularity within the lamina propria, number of goblet cells, and cellular infiltrate within the lamina propria were recorded.

Interobserver and Intraobserver Agreement

We performed a post-hoc evaluation of inter- and intraobserver agreement for the endomicroscopic findings in this study population. Therefore, a dataset of 300 iCLE and 300 pCLE images that depicted the endomicroscopic findings in these patients was evaluated by two blinded investigators. Inter- and intraobserver reproducibility was measured based on comparison of still images (either for pCLE) between two investigators. To evaluate the intraobserver agreement, the investigator who had assessed the images at the first occasion reassessed all images in a blinded manner. To evaluate the interobserver agreement, a second investigator assessed all images in a blinded fashion, and findings were compared with those of the first investigator.

Statistical Analysis

All statistical analysis were done using PASW Statistics 18 (SPSS, Chicago, IL). The t-test was used for all continues variables, such as age to determine whether differences between any two groups existed. A two-sided P-value < 0.05 was considered significant. For comparisons of proportions, such as crypt architecture, presence of microerosions, and vascular pattern between the different groups we used the chi-square test. If the validity of the chi-square test was in question (because of small number of expected events in a cell), Fisher's exact test was used instead. To evaluate the impact of image interpretation for diagnosis of CD, we calculated the positive and negative predictive values. The median in this study is presented for nonnormally distributed variables and the mean for normally distributed variables. The range indicated the range between the minimum and maximum values. Correlation between endomicroscopy diagnosis and histopathology was determined using kappa statistics. Therefore, the strength of rater agreement was categorized according to the definition proposed by Landis and Koch24: 0–2, slight; 0.21–0.4, fair; 0.41–0.6, moderate; 0.61–0.8, substantial; 0.81–1, almost perfect.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Between October 2009 and August 2011 a total of 76 patients with CD were screened to participate in the endomicroscopy study. Twenty-two patients did not meet the inclusion criteria. Four of these patients had poor bowel preparation and one patient suffered from strong abdominal pain during colonoscopy, resulting in abortion of the endoscopic procedure. Furthermore, nine patients had endoscopically not passable stenoses, three patients had active upper gastrointestinal bleeding, and five patients declined to participate in the study. Additionally, two patients were excluded from the final analysis because of poor image quality precluding an accurate evaluation of endomicroscopy features. In addition, 18 control patients without any history of IBD or diarrhea presenting for surveillance colonoscopy were included.

Overall, 27 patients with quiescent and 27 patients with active CD were eligible for the final analysis of endomicroscopic findings. The mean age of the study group was 39 years (standard deviation [SD] 13.74; range 18–72 years) and 52% of the participants were female. The control group consisted of 18 patients with a mean age of 68 years (SD 17.16; range 28–85 years), of whom nine patients (50%) were female. Patients with CD were significantly younger compared with controls (P < 0.001). However, no significant difference was observed with regard to gender or ethnicity.

CLE Patterns in Controls

In all control patients endomicroscopy visualized regular colonic architecture with crypts appearing normal in shape and size (Fig. 1). Evaluation of microvascular patterns revealed normal vessel structures without fluorescein leakage. There were no signs of increased cellular infiltrates within the lamina propria and no microerosions present. Goblet cells were regularly shaped and showed normal structure and frequency.

thumbnail image

Figure 1. Endomicroscopy findings in controls compared with patients with quiescent and active CD. Different panels highlight endomicroscopic findings as described in the different segments of the figure. In control patients normal colonic epithelium with regular crypt architecture, cellular infiltrates, and vascularity was found. Microerosions were not present and goblet cells were regular in size and shape. In patients with quiescent CD an increased number of colonic crypts and increased crypt tortuosity was found. Crypt lumens were dilated and appeared white according to fluoresceine sodium leakage. Microvessels were dilated and augmented but showed no signs of leakage. Moreover, total number of goblet cells, which appeared dark, was increased. Cellular infiltrates within the lamina propria were only slightly increased in about one-third of patients. In patients with active CD an overall irregular colonic architecture with a decreased number of colonic crypts that have a tortuous appearance and dilated crypt lumens became visible. All patients with CD demonstrated microerosions and an increased cellular infiltrate within the lamina propria. With regard to the inflammatory activity, endomicroscopy identified an increased vascularity with fluorescein leakage and a decreased number of goblet cells.

Download figure to PowerPoint

Comparison Between Patients with Quiescent CD and Controls

In a subgroup analysis, patients with quiescent CD, defined as a CDEIS ≤6, were compared with control patients. Patients with CD showed a significantly increased number of colonic crypts (67% vs. 11%, respectively, P = 0.001), increased crypt tortuosity (52% vs. 11%, respectively, P = 0.02), and dilated crypt lumens (52% vs. 6%, respectively, P = 0.008) compared with the control group. They were also more likely to show an increased vascularity (44% vs. 6%, respectively, P = 0.027) and an increased number of goblet cells (63% vs. 6%, respectively, P < 0.001). In contrast, there was no difference between the two groups with regard to the presence of microerosions and cellular infiltrates within the lamina propria (Fig. 1). These findings and their sensitivity and specificity are shown in Table 1. Multivariate analysis revealed that an increased number of goblet cells was the best predictor for diagnosis of quiescent CD (odds ratio [OR], 28.9; 95% confidence interval [CI]: 4.5–186.1).

Table 1. Endomicroscopy Findings in Patients with Quiescent Crohn's Disease and Controls
Endomicroscopy FindingsCrohn's Disease n = 27Controls n = 18P-valuesSensitivity (%)Specificity (%)Odds Ratio (95% CI)
  1. CI, confidence interval; ns, not significant.

Crypt number increased, n (%)18 (67)2 (11)0.001678916 (3.2-80.3)
Crypt number decreased, n (%)4 (15)1 (6)ns15943 (0.04-238)
Crypt tortuous, n (%)14 (52)2 (11)0.0252898.6 (1.6-47.3)
Crypt lumen dilated, (%)14 (52)1 (6)0.008529418 (2.6-130.2)
Microerosions, n (%)4 (15)0 (0)ns15100n/a
Cellular infiltrate increased, n (%)8 (30)0 (0)ns30100n/a
Vascularity increased, n (%)12 (44)1 (6)0.027449413.6 (1.7-106.6)
Goblet cells increased, n (%)17 (63)1 (6)<0.001639428.9 (4.5-186.1)
Goblet cells decreased, n (%)2 (7)0 (0)ns7100n/a

Comparison Between Patients with Active CD and Controls

In this subgroup analysis, patients with active CD, defined as a CDEIS >6, were more likely to show a decreased number of colonic crypts (93% vs. 6%, respectively, P < 0.001), increased crypt tortuosity (100% vs. 0%, respectively, P < 0.001), and dilated crypt lumen (100% vs. 17%, respectively, P < 0.001). Furthermore, CD patients were more likely to demonstrate microerosions (100% vs. 0%, respectively, P < 0.001), an increased cellular infiltrate within the lamina propria (100% vs. 0%, respectively, P < 0.001), an increased vascularity (100% vs. 11%, respectively, P < 0.001), and a decreased number of goblet cells (85% vs. 6%, respectively, P < 0.001; Fig. 1). Table 2 highlights these findings with their sensitivity and specificity. Table 3 depicts the comparisons between active and quiescent CD.

Table 2. Endomicroscopy Findings in Patients with Active Crohn's Disease and Controls
Endomicroscopy FindingsCrohn's Disease n = 27Controls n = 18P-valuesSensitivity (%)Specificity (%)
  1. CI, confidence interval; ns, not significant.

Crypt number increased, n (%)1 (4)2 (11)ns089
Crypt number decreased, n (%)25 (93)1 (6)<0.0019394
Crypt tortuous, n (%)27 (100)0 (0)<0.001100100
Crypt lumen dilated, (%)27 (100)3 (17)<0.00110083
Microerosions, n (%)27 (100)0 (0)<0.001100100
Cellular infiltrate increased, n (%)27 (100)0 (0)<0.001100100
Vascularity increased, n (%)27 (100)2 (11)<0.00110089
Goblet cells increased, n (%)2 (7)0 (0)ns7100
Goblet cells decreased, n (%)23 (85)1 (6)<0.0018594
Table 3. Comparison of Endomicroscopy Findings Between Patients with Quiescent and Active (CDEIS > 6) Crohn's Disease
Endomicroscopy FindingsQuiescent Crohn's Disease n=27Active Crohn's Disease n=27P-values
  1. CI, confidence interval.

Crypt number increased, n (%)18 (67)1 (4)<0.001
Crypt number decreased, n (%)4 (15)25 (93)<0.001
Crypt tortuous, n (%)14 (52)27 (100)0.002
Crypt lumen dilated, (%)14 (52)27 (100)0.002
Microerosions, n (%)4 (15)27 (100)<0.001
Cellular infiltrate increased, n (%)8 (30)27 (100)<0.001
Vascularity increased, n (%)12 (44)27 (100)<0.001
Goblet cells increased, n (%)17 (63)2 (7)<0.001
Goblet cells decreased, n (%)2 (7)23 (85)<0.001

Comparison Between CRP and CDEAS

The Crohn's Disease Endomicroscopic Activity Score (CDEAS) strongly correlated with CRP (Table 4). Median CRP was 2.8 mg/L (normal value <5) in patients with quiescent CD and 24.5 in patients with active CD (P = 0.005). Comparison of CDEAS between patients with quiescent and active CD revealed a median score of 2 and 5, respectively (Table 4). This difference was statistically highly significantly different (P < 0.001).

Table 4. Correlation of C-reactive Protein and Crohn's Disease Endomicroscopic Activity Score (CDEAS) in Different Groups
 Quiescent Crohn's Disease n=27Active Crohn's Disease n=27P-values
CRP, median2.824.50.005
CDEAS, median25<0.001

Interobserver and Intraobserver Agreements

Interobserver and intraobserver agreements were calculated between two investigators. Therefore, number of colonic crypts, crypt tortuosity, crypt lumen, presence of microerosions, vascularity, cellular infiltrates within the lamina propria, and number of goblet cells were included in this calculation. The interobserver agreement for endomicroscopy findings in patients with CD ranged from moderate (crypt lumen, microerosions, goblet cell number) to substantial (tortuosity of crypts, vascularity) to almost perfect (number of colonic crypts, cellular infiltrate). Table 5 summarizes interobserver agreement for endomicroscopy findings in patients with CD.

Table 5. Interobserver Agreement of Endomicroscopic Findings in Patients with Crohn's Disease
Endomicroscopy FindingsK Value (95% CI)Rating
  1. CI, confidence interval. Rating according to Landis and Koch.24

Number of crypts0.943Almost perfect
Tortuosity of crypts0.764Substantial
Crypt lumen0.415Moderate
Microerosions0.511Moderate
Vascularity0.753Substantial
Goblet cells0.508Moderate
Cellular infiltrate0.830Almost perfect

The intraobserver agreement ranged from moderate (crypt lumen) to substantial (crypt tortuosity, presence of microerosions, vascularity, cellular infiltrate, and goblet cell number) to almost perfect for the number of colonic crypts. Table 6 highlights intraobserver agreement for endomicroscopy findings in patients with CD. Highest κ-values for interobserver and intraobserver agreement were found for the number of colonic crypts (0.943 and 0.950, respectively).

Table 6. Intraobserver Agreement for Endomicroscopy Findings in Patients with Crohn's Disease
Endomicroscopy FindingsK Value (95% CI)Rating
  1. CI, confidence interval. Rating according to Landis and Koch.24

Number of crypts0.950Almost perfect
Tortuosity of crypts0.728Substantial
Crypt lumen0.559Moderate
Microerosions0.674Substantial
Vascularity0.753Substantial
Goblet cells0.614Substantial
Cellular infiltrate0.607Substantial

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Endoscopy has emerged as the gold standard for assessing disease activity and extent of inflammation in patients with CD.25–31 Nevertheless, standard white-light endoscopy appears to be an insensitive test for the diagnosis of CD in the quiescent or even mild phase of the disease.32, 33 Therefore, only histopathology is currently able to reliably predict disease severity in these settings.28, 29 In the present study we aimed to overcome this limitation of current endoscopic techniques using endomicroscopy.

CLE is a new, emerging endoscopic imaging modality enabling the detection of subsurface structures, in part comparable to some aspects of histomorphology.13, 14, 34, 35 Although CLE has been previously used for assessment of inflammation and neoplastic lesions in patients with UC, few data on the use of this technique in CD are available.18, 36 However, Kiesslich et al37 showed that cell shedding within the colonic mucosal layer leads to epithelial discontinuities, denoted gaps, in the terminal ileum of patients with CD. Furthermore, colonic bacteria were visualized using fluorescein and confocal endomicroscopy in CD and a retrospective evaluation showed that patients with CD have more frequently intramucosal bacteria than control patients.38 However, there was no correlation between the number of mucosal bacteria and disease activity.

In the present study we used CLE for the first time to systematically determine disease activity in CD. It was found that CLE could readily identify architectural changes in patients with active CD similar to established histopathological criteria.39, 40 Specifically, in active CD the following parameters were significantly different from the control group: crypt morphology (number of colonic crypts, crypt tortuosity, crypt lumen), presence of microerosions, vascularity, cellular infiltrate within the lamina propria, and number of goblet cells. Furthermore, differences between active and quiescent CD were noted. In particular, we could show that crypt atrophy could be quantified by counting crypts per FOV with a high specificity in patients with active CD but not in patients with quiescent disease. This is in accordance with histopathologic findings showing increased cellular infiltrates combined with an increased distance between the crypts in active CD.39, 40 Importantly, the number of crypts per FOV was significantly increased in quiescent CD as compared with controls, reflecting epithelial regeneration in inactive disease. Furthermore, a decrease of goblet cells with reduction of mucin was found to have a 100% specificity for active CD compared with control patients. Concerning quiescent CD, multivariate analysis even showed that an increased number of goblet cells served as the best predictor for a diagnosis of CD. A major criterion for chronic disease was crypt distortion that was a typical feature in CD, with a high interobserver agreement and 100% sensitivity. Microerosions and cellular infiltrates could be identified to be an endomicroscopic sign of active CD. Importantly, increased vascularity was typical for active CD but not quiescent CD, defining an endomicroscopic feature beyond established histomorphologic findings. While we did not quantify fluorescein leakage in our study because we aimed at establishing morphological aspects, we observed that this phenomenon was present in active CD and was not found in control patients. Interestingly, in the present study we found that the crypt lumen was enlarged in some patients with CD. However, the interobserver agreement was only moderate. This finding is probably explained by the fact that we defined no cutoff value for this parameter. Nevertheless, this parameter should be taken into account in future studies comparing different forms of IBD, because one recent study showed that enlarged crypt lumen may be an important marker for active UC.18

Based on our findings we proposed a CDEAS consisting of six parameters: crypt number (increased or decreased), crypt distorsion, microerosions, cellular infiltrate, vascularity, and number of goblet cells (increased or decreased). By assignment of one point for each given parameter, the score ranges from 0 to 8. In order to validate the CDEAS as an activity parameter of CD we compared CDEAS with CRP as an already established biological inflammatory marker of CD.41 We found strong correlation of CDEAS and CRP, thereby highlighting the potential of CDEAS for assessment of CD activity.

We have also shown that the differentiation between quiescent and active CD can be defined by several parameters; notably endomicroscopy was able to differentiate between quiescent CD with macroscopic normal mucosa and controls when taking into account an increased number of goblet cells and crypts. Additionally, we found a remarkable number of colonic segments without macroscopic inflammation that had a histologic and endomicroscopic evidence of inflammation.

Taken together, we have shown that endomicroscopy is able to predict disease severity even in quiescent CD with high kappa values and accuracy. We also found that endomicroscopy reveals inflammation in colonic segments where standard white-light endoscopy exhibited uninflamed mucosa. Thus, CLE is more sensitive compared with standard white-light endoscopy and further studies will reveal if activity assessment of CD by endomicroscopy will have a better predictive value in the light of disease course and response to medical therapy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We thank Professor Ralf Kiesslich and Dr. Martin Goetz from the University of Mainz for helpful discussions. Author contributions: Helmut Neumann: study idea, study design, patient selection, patient inclusion, endomicroscopy, interpretation of data, statistics, article writing, final revision of the article; Michael Vieth: interpretation of data, article writing, final revision of the article; Raja Atreya: endomicroscopy, patient inclusion, final revision of the article; Jürgen Siebler: patient inclusion, final revision of the article; Martin Grauer: patient inclusion, final revision of the article; Thomas Bernatik: patient inclusion, final revision of the article; Markus F. Neurath: study idea, study design, interpretation of data, article writing, final revision of the article; Jonas Mudter: article writing, patient selection, patient inclusion, endomicroscopy, final revision of the article, interpretation of data.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES