The work described in this manuscript was performed at the Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern, Bern, Switzerland. Parts of this paper were presented as an abstract at the ECVIM Congress in Glasgow, 2005 (title “No changes in histological scoring, total number of infiltrating cells, and number of T cells after treatment in dogs with chronic enteropathies”).
Clinical Signs, Histology, and CD3-Positive Cells before and after Treatment of Dogs with Chronic Enteropathies
Article first published online: 30 JUL 2008
Copyright © 2008 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 22, Issue 5, pages 1079–1083, September–October 2008
How to Cite
Schreiner, N.M.S., Gaschen, F., Gröne, A., Sauter, S.N. and Allenspach, K. (2008), Clinical Signs, Histology, and CD3-Positive Cells before and after Treatment of Dogs with Chronic Enteropathies. Journal of Veterinary Internal Medicine, 22: 1079–1083. doi: 10.1111/j.1939-1676.2008.0153.x
- Issue published online: 25 AUG 2008
- Article first published online: 30 JUL 2008
- Submitted July 22, 2007; Revised November 2, 2007; Accepted April 17, 2008.
- Clinical score;
- Histological score;
Background: Histopathology is widely used for the diagnosis of inflammatory bowel disease in dogs. Variations in lesions and unavailability of uniform grading systems limit the usefulness of histologic examination.
Hypothesis: CD3 cell numbers in chronic enteropathies of dogs correlate with clinical activity of the disease and with severity of histopathologic changes.
Animals: Nineteen client-owned dogs examined because of chronic diarrhea, vomiting, or both.
Methods: Samples of duodenal and colonic mucosa were collected endoscopically before and after treatment. Dogs that responded to a hypoallergenic diet were grouped as food-responsive diarrhea dogs (FRD, n = 10). Dogs with no clinical improvement after 10 days of treatment then received prednisolone (immunosuppressive doses) and were grouped as steroid-responsive diarrhea dogs (SRD, n = 9). Histopathologic assessment with a standardized grading system was performed retrospectively on the intestinal samples. Histologic score, total number of infiltrating cells, and CD3-positive cells were counted and compared with the clinical scoring.
Results: No statistically significant difference was detected among histologic grading, total number of cells in the lamina propria, and T-cell numbers in biopsies before and after treatment in either group (FRD and SRD).
Conclusions and Clinical Importance: Currently used histopathologic grading scores, total numbers of cells, and numbers of CD3-positive cells did not allow differentiation between FRD and SRD and did not correlate with clinical response to therapy. Based on these results, new grading scores assessing other criteria than total cell numbers and CD3-positive cells should be evaluated in the future.
Inflammatory bowel disease (IBD) is a collective term for disorders associated with persistent or recurrent signs of gastrointestinal disease in dogs. It is characterized by histologic evidence of intestinal inflammation, mainly by infiltration of lymphocytes and plasma cells, of undetermined cause. Despite being the most common cause of chronic intestinal disease, its pathogenesis is still poorly understood.1 Wide variations of the histologic changes within the gastrointestinal tract suggest that IBD is not a single disease, and therefore the term idiopathic chronic enteropathy has recently been used more widely. Only dogs in which histologic evidence of inflammation is found without an apparent underlying cause can be diagnosed as chronic idiopathic enteropathy.2 In addition, there is wide agreement that most instances of chronic enteropathy can be subdivided by their response to a successive treatment regimen encompassing elimination diet (diet-responsive disease), antibiotic treatment (antibiotic-responsive disease), and treatment with immunosuppressive doses of corticosteroids (true idiopathic IBD).1–5
Histologic interpretation of biopsies from dogs with IBD varies greatly between pathologists.6 Several grading systems for histologic biopsies have been described and are still used, but only 1 grading system has been published.7 Some studies show that immunohistochemistry may be a more objective method to quantify intestinal inflammation in the canine gut. Lamina propria T-cell numbers, particularly T cells expressing the αβ T-cell receptor and CD4, were found to be increased in duodenal biopsies of dogs with IBD, as well as IgG+ plasma cells, macrophages, and granulocytes. CD3+ T lymphocytes are increased in the epithelial compartment in dogs with IBD versus healthy controls.8 However, there are no differences in duodenal CD3 cell numbers when dogs with food-responsive disease or antibiotic-responsive disease are compared with dogs needing glucocorticoid treatment to control their disease.8 CD3 cell infiltrates in the duodenum of dogs with IBD decrease after treatment with cyclosporine, implicating a role for CD3 cells in the pathogenesis of at least some cases of IBD in dogs.9 It is unclear whether the infiltration of lymphocytes and especially of CD3-positive cells is a crucial element for the development of disease and, furthermore, if we are currently interpreting intestinal biopsies using reliable, clinically relevant assessment criteria.
The purpose of this study was therefore to test the hypothesis that CD3 infiltrates in chronic enteropathies of dogs are correlated with clinical activity of disease as assessed by objective clinical grading (Canine Chronic Enteropathy Clinical Activity Index, CCECAI)5 as well as with the severity of histopathologic changes (histologic score), not only before but also after treatment.
Materials and Methods
Some information on clinical findings and outcome of the dogs in this study has previously been reported.5 Sample number statistics relying on previously published data8 predicted that 9–10 paired samples in each group would suffice to detect a statistically and clinically significant difference in the total number of cells and CD3 cells in the lamina propria. We included 10 dogs with food-responsive disease and 9 dogs with steroid-responsive disease in the study. Dogs of various breeds were referred to the Department of Clinical Veterinary Medicine, Vetsuisse Faculty, University of Bern/CH, for the investigation of daily or intermittent diarrhea, vomiting, or both or weight loss of longer than 6 weeks duration. All dogs had fenbendazole (50 mg/kg once a day for 3 consecutive days) administered, and all medications were withdrawn for at least 2 weeks before the sample collection.
Complete blood cell count, chemistry panel, serum trypsin-like immunoreactivity, vitamin B12 and folic acid concentrations, urinalysis, fecal analysis (parasitic and bacterial), and abdominal ultrasound were initially used to exclude systemic diseases as well as parasitic or bacterial diseases.
Preparation for endoscopic examination was performed by withholding food for 72 hours and administrating a colonic electrolyte lavage solution by gastric intubation. Gastroduodenoscopy and colonoscopy were performed with a flexible video endoscopea (KA and FG). Ten to 15 samples of duodenal and colonic mucosa were collected endoscopically with disposable biopsy forceps. Tissue samples were immediately placed in formalin, processed by routine methods, and embedded in paraffin wax. A 2nd endoscopic examination was performed with the same protocol 4 weeks after the 1st endoscopic examination in food-responsive diarrhea dogs (FRD, n = 10) and 10 weeks after the 1st endoscopic examination in steroid-responsive diarrhea dogs (SRD, n = 9). At the time the 2nd endoscopic examination was performed, the SRD group had not received prednisolone for 2 weeks.
The immunohistochemical procedures used have been described previously.10 Polyclonal rabbit anti-human CD3 antibodyb was used as a primary antibody to stain CD3-positive cells on the formalin-fixed material in ready-to-use dilutions and incubated for 60 minutes in a humid chamber. For negative controls, rabbit IgGb was used instead of the primary antibody. Further materials used for the antigen-antibody complex were the secondary antibody (biotinylated mouse anti-rabbit)c and Streptavidin-biotin-horseradish peroxidase complex.c
All collected formalin-fixed tissues were stained with hematoxylin and eosin, coded, and analyzed by a single pathologist (AG). A histopathologic grading system (0 = normal, 1 = mild, 2 = moderate, and 3 = severe) was used, which is based on the only published scoring system available.11 Grading was assigned as follows: mild IBD lesions were those with cellular infiltrates but without architectural distortion or mucosal epithelial immaturity, moderate lesions had cellular infiltrates accompanied by mucosal epithelial immaturity or solitary epithelial necrosis or both, and severe lesions consisted of cellular infiltrates accompanied by multifocal epithelial necrosis or extensive architectural distortion with epithelial immaturity.
Examination of the Sections
The protocol for examination of the sections followed the ones previously published.10,12 Sections were coded and analyzed blindly by two of the investigators (NMS and KA), taking the mean number of counted cells from both observers to insure consistency. Samples were examined using a Leica microscope linked to a digital camera.d Photos were taken with 40 × objectives and transferred to a computer. A standard area of lamina propria was delineated on the computer screen, excluding epithelium, lymphatics, and larger blood vessels. The total number of cells in the lamina propria as well as CD3-positive T cells were determined in at least 10 biopsies each in the tip of the duodenum (area 1), the base of the duodenal villi (area 2), and at the base of the colonic crypts. Five randomly chosen regions were counted for each area and an average cell count was calculated. Results were expressed as cells/10,000 μm2. Intraepithelial cell counts were attempted but abandoned because only a few biopsies in each dog had enough intact epithelium available to make reliable counts.
Treatment was started with commercially available salmon, trout, and rice-based elimination diete after the 1st endoscopic examination. Dogs that responded within 10 days to the diet were kept on it for at least 14 weeks and were grouped as FRD dogs (n = 10). Dogs that had no clinical improvement after a week on the diet alone additionally received prednisolone (1 mg/kg PO twice daily for 10 days, 0.5 mg/kg PO q12h for 10 days, 0.5 mg/kg PO q24h for 10 days, 0.5 mg/kg PO q48h for 10 days) and were grouped as SRD dogs (n = 9). Thus, dogs were separated into 2 groups according to their initial response to treatment with elimination diet. Although it is possible that some dogs in the SRD group may have responded to a longer dietary trial, the clinical decision was made to treat these dogs relatively early with steroids because they had severe disease.
Severity of clinical signs was assessed before and after treatment with the CCECAI, a numeric scoring based on 9 clinical signs: attitude/activity, appetite, vomiting, feces consistency, feces frequency, weight loss, serum albumin concentration, peripheral edema or ascites, and pruritus. A total score from 0 to 3 indicates clinically insignificant disease, 4–5 mild IBD, 6–8 moderate IBD, 9–11 severe IBD, and >12 very severe IBD.5 CCECAI scores were assigned by two of the investigators (KA and FG), taking care that all dogs were scored by the same clinician before and after treatment to be as consistent as possible in our clinical assessment.
All statistical calculations were made by a commercially available software.f Sample number predictions were performed before starting the study. Previously published numbers of total cells and T cells and the highest standard deviation in the duodenal lamina propria of dogs with chronic enteropathies were taken as a guideline to predict clinically significant differences between the groups and before and after treatment (60 ± 30 CD3-positive cells/10,000 μm2 versus 40 ± 30 CD3-positive cells/10,000 μm2 in healthy endoscopic controls8). With a 95% confidence interval and a power of 80%, 9–10 paired samples are sufficient to detect a statistically significant and clinically relevant difference. Normally distributed data are reported as mean ± standard deviation; nonnormally distributed data are reported as median and ranges. Data for total cell count, CD3-positive cells in the lamina propria before and after treatment, and the age of the dogs were normally distributed with equal variances and differences between means and were compared using an unpaired t-test. Differences between means of total numbers of cells in the lamina propria as well as T cells within the same group were compared using a paired t-test. Correlations between the CCECAI and histologic grading, total number of infiltrating cells, and number of T cells were performed by Fisher's exact test for categorical data, and with the Spearman rank test for continuous data. Statistical significance was set at P < .05.
Ten dogs with FRD and 9 dogs with SRD were included in the study. The FRD group consisted of 4 females and 6 males. Two of the dogs were German Shepherd Dogs; other breeds were 1 Bernese Mountain Dog, 1 Border Collie, 1 Great Dane, 1 Whippet, 1 Leonberger, 1 West Highland White Terrier, and 2 mixed breed dogs. The SRD group consisted of 3 females and 6 males. Breeds included were 2 Dachshunds, 2 mixed breed dogs, 1 German Shepherd Dog, 1 Rottweiler, 1 Shar Pei, 1 Mastiff, and 1 Yorkshire Terrier.
The age of the 19 dogs ranged from 0.5 to 11.6 years (mean 4.7 ± 3.2). The mean age of the FRD dogs was 2.2 years ± 1.6, and the mean age of the SRD dogs 6.9 years ± 2.9, with the SRD group being significantly older than the FRD group (P= .002).
The median CCECAI score of the FRD group was 6 before the 1st endoscopy with a range from 4 to 8, and decreased to 1 after treatment (P= .002; range 0–3). In the SRD group the median CCECAI score was 10 before the 1st endoscopy (range 6–15), and decreased to 3 (range 0–6) after treatment with prednisolone (P= .008). The CCECAI was significantly higher before treatment in the SRD group compared with the FRD group (P= .005).
Immunohistochemistry and Histopathology
Statistical analysis of these data revealed no difference in number of total cells or T cells between the SRD group and the FRD group before treatment was started (Tables 1 and 2). In addition, there was no statistically significant difference between the total number of cells in the lamina propria and the total number of T cells before and after treatment in either group.
|Total Cell Numbers in Lamina Propria|
|Duodenal Villus Tip Area||P-Value Paired t-Test||Duodenal Villus Base Area||P-Value Paired t-Test||Colonic Crypt Area||P-Value|
|FRD before treatment||96.6 ± 7.3||88.5 ± 9.5||42.5 ± 7.8|
|FRD after treatment||96.5 ± 11.5||.8||80.9 ± 9.6||.6||45.2 ± 9.6||.4|
|SRD before treatment||92.2 ± 13.1||90.2 ± 10.3||49.8 ± 7.3|
|SRD after treatment||95.9 ± 11.8||.3||91.5 ± 13.5||.6||63.3 ± 23.3||.2|
|T-Cell Numbers in Lamina Propria|
|Duodenal Villus Tip Area||P-Value Paired t-Test||Duodenal Villus Base Area||P-Value Paired t-Test||Colonic Crypt Area||P-Value|
|FRD before treatment||12.8 ± 9.2||6.6 ± 6.6||1.8 ± 1.3|
|FRD after treatment||11.7 ± 9.5||.9||5.8 ± 5.6||.6||2.2 ± 1.8||.4|
|SRD before treatment||6.9 ± 7.0||5.4 ± 4.5||2.4 ± 1.6|
|SRD after treatment||7.8 ± 10.9||.6||6.1 ± 7.2||.8||3.8 ± 3.3||.2|
The median histologic severity score of endoscopically collected biopsies did not change after treatment in either group. The median score was 2 before and after treatment in the duodenal samples of both the FRD and the SRD group, with a median of 1.6 (range 0–2) before and 2 (range 1–3) after treatment in the FRD group (P= .6) and 2.3 (range 2–3) and 2.1 (range 2–3) in the SRD group (P= .4). In the colonic samples, the median score was 1.5 (range 0–3) before and 1 (range 0–2) after treatment in the FRD group (P= .09), and 1.1 (range 0–3) before and 1.4 after treatment (range 0–3) in the SRD group (P= .8). Moreover, the histologic scoring, total cell counts, and T-cell numbers showed no correlation with the CCECAI (histological grading—CCECAI: Fisher's exact test, P= .12; total cell counts in lamina propria—CCECAI: Spearman's rank correlation, r2= 0.1, P= .12; T-cell numbers—CCECAI: Spearman's rank correlation, r2= 0.3, P= .09).
The dogs in the FRD group responded very well and quickly to treatment: all dogs had a CCECAI of 0–3 within 10 days of starting elimination diet, which is considered a complete response in all cases (score of 0–3 indicates clinically insignificant disease13). However, in the SRD group, the response may be considered more variable. There were 6 dogs that went from a severe CCECAI to either insignificant or mild disease after treatment, 1 dog with a moderate CCECAI, which reduced to insignificant disease after treatment, and 2 dogs that went from very severe disease activity to moderate disease. The latter 2 dogs could be considered to have had a partial response; however, 7/9 dogs in that group have responded well to steroid treatment. Although similar variable success rates using steroid treatment in dogs with chronic enteropathies have been published before,4,13 the fact that 2 dogs responded partially to steroid treatment makes interpretation of the results of this study more difficult. According to power calculations, this study included barely sufficient dogs to test our hypothesis. This power should be considered when negative results are interpreted.
Although care was taken to ensure that all dogs were assessed by the same clinicians, we had 2 different clinicians assessing dogs during the study period, which might have resulted in variability in the CCECAI scoring between observers.
The SRD dogs in this study were significantly older than the FRD dogs. This is consistent with other studies where IBD was found to be most common in middle-aged dogs1 and even though no age predisposition is reported for FRD, up to one third of the cases can occur in dogs <1 year old.1
Our results revealed no change in the total number of cells as well as numbers of CD3 T cells in the lamina propria and subjective severity of inflammatory infiltrates after treatment in either group (FRD and SRD). Furthermore, the median histologic severity score of endoscopically collected biopsies did not change after treatment in either group. Therefore, this study suggests that histologic scores, total number of infiltrative cells, and CD3 immunohistochemistry do not allow differentiation between FRD and SRD.
The finding that total numbers of lamina propria cells were not different among dogs with IBD, ARD (antibiotic-responsive disease), or FRD has been previously reported by German et al.8 In this previous study, the total number of cells and CD3 cell numbers in the duodenal lamina propria were higher than the ones reported in our study. No clinical information was given in the study by German et al8 and it is possible that the dogs reported had more severe diseases and therefore higher total cell numbers and CD3 cells than our dogs. Numbers of duodenal CD3 cells in the lamina propria were also reported to be significantly different between IBD dogs and controls in the previous study. However, there were no follow-up data available on the same dogs after treatment. Because the main purpose of our study was to correlate clinical activity with CD3 cell numbers in the duodenum and colon before and after standardized treatment, we did not include a control group of healthy dogs to reproduce the findings. German et al8 have also found differences in the intraepithelial compartment between IBD and ARD versus healthy controls. The intraepithelial cells in the duodenum and colon of the dogs were not counted in this study, because many of the biopsies did not have an intact epithelial cell layer that could be reliably assessed. Because this is a frequent finding in endoscopic biopsies, measurement of intraepithelial cell counts may be difficult to assess in a routine manner on histopathology.
The results of our study raise a number of questions: Are our duodenum and colon specimens representative for the disease and are the methods we currently use accurate to interpret intestinal biopsies? The authors feel that in the current study every effort was made to obtain a sufficient number of biopsies of the duodenal as well as the colonic mucosa from each dog. In addition, all biopsies were taken using disposable biopsy forceps (used separately for duodenum and colon in each dog) in an effort to reduce artifacts. The noninvasive procedure has some other drawbacks compared with full thickness biopsies: infectious foci in parts of the small intestine where the endoscope does not reach could be missed and the histopathologic differentiation between IBD and lymphoma has been shown to be difficult in certain cases.14 However, the follow-up of the patients (telephone conversations with owners, referring veterinarians, or both) even after the 2nd endoscopy did not give any evidence of the possibility that a focal area of intestinal lymphoma had been missed.
The number of lamina propria cells and infiltrating CD3-positive cells does not seem to be crucial, neither for severity of the disease nor for therapeutic response. This finding is unexpected. Based on these results, histologic criteria for diagnosis, grading of disease severity, and assessment of therapeutic response in dogs with chronic enteropathies may need to be reevaluated.
In our study, the 2nd endoscopy was performed 4 (FRD) and 10 weeks (SRD) after starting treatment. Further follow-up by endoscopies might have shown whether the intestinal mucosa stays inflamed long term in chronic enteropathies in dogs. Recently published data of dogs with lymphocytic-plasmacytic enteritis confirmed our findings that clinical improvement is not always followed by a significant improvement of the histopathologic lesions.15 The authors also mention the lack of a standardized protocol for histologic evaluation of the tissue samples and suggest the development of a histologic grading system based mainly on mucosal architectural and epithelial alterations as opposed to primarily emphasizing on the cellular infiltrate.15
In conclusion, only 1 study was published so far to reveal histopathologic changes in dogs with chronic enteropathies before and after therapy.15 CD3-positive cells were thought to be of clinical significance in canine chronic enteropathies. This small study shows that the total number of infiltrating cells and the number of T cells in the biopsies do not change before and after treatment. More research is needed to unveil the mechanisms responsible for the disease and to find a new histopathologic grading system that might help correlate clinical response to treatment strategies. A gastrointestinal standardization group has, with the support of the World Small Animal Veterinary Association (WSAVA), proposed to develop a standardized histologic evaluation system that enables uniform diagnosis, staging of disease, and the subsequent development of controlled clinical trials for the treatment of canine and feline gastrointestinal disorders.16 This will be especially helpful to compare dogs from different studies and on different treatment protocols.
aOlympus flexible video-endoscope GIF-1T30, Volketswil, Switzerland
bBiomeda, Foster City, CA
cDAKO, Carpinteria, CA
dMicroscope DM LB2, Camera DFC 320, Leica Microsystems, Wetzlar, Germany
ePurina Veterinary Diets LA, Salmon and Rice, Société des Produits Nestlé SA, Vevey, Switzerland
fNCSS Statistical Software version 2005, Kaysville, UT
- 1Small intestinal disease in dogs. In: EttingerSJ, FeldmanEC, ed. Textbook of Veterinary Internal Medicine. St Louis, MO: Elsevier Saunders; 2005:1332–1378.,
- 16WSAVA Gastrointestinal Standardisation Group. http://www.wsava.org/StandardizationGroup.htm.