Clinicopathological and Microbiological Profile of Patients with Chronic Diarrhoea and Malabsorption with Special Emphasis on Duodenal Biopsies: A Study from Tertiarycare Hospital in South India

Background: Chronic diarrhoea and malabsorption continue to be a major health problem worldwide. Due to conflicting reports regarding the clinicohistological spectrum of malabsorption, this study was conducted to study the spectrum of malabsorption. Objective: To study the morphologic spectrum of duodenal biopsies and stool microbiological profile in patients with chronic diarrhoea and malabsorption. Materials and Methods: The study included 106 patients aged above 18 years, presenting with chronic diarrhoea or malabsorption. Duodenal mucosal biopsy obtained was evaluated for villous atrophy, crypt architecture and Intraepithelial lymphocyte count (IEL) by H&E and immunohistochemistry for CD3. Stool samples collected were subjected to wet mount and culture examination. Serological testing for antibodies to coeliac sprue were also performed. Results: This descriptive study included 57 males and 49 females. Duodenal biopsy spectrum consisted of chronic duodenitis (48), duodenitis associated with autoimmune conditions (20), histomorphology of coeliac sprue (5), duodenitis with villous atrophy (14), only increased IEL (11) and miscellaneous (8). CD3 stained sections (n=90) offer a slightly better advantage over H & E-stained sections in detecting IEL. Stool microscopy and culture were positive in 6 cases. Coeliac sprue antibodies were positive in 1 case with Marsh type 1 lesion. Conclusion: Duodenal biopsy is a useful parameter in the study of malabsorption which must be adjuncted by microbiological and serological tests. Coeliac sprue is a rare entity in our population as per this study.


Introduction:
Chronic diarrhoea and malabsorption continue to be a major health problem worldwide. (1) Malabsorption is an important cause of mortality and morbidity in the tropics. (2) The current trends suggest that coeliac disease and inflammatory bowel disease are also emerging as potential causes of malabsorption. (3) Previously it was thought to be more common in the western population with an incidence of about 1%. (4) One of the investigations in the evaluation of malabsorption and chronic diarrhoea is the study of small intestinal biopsy (5). Duodenal biopsy may show villous and crypt abnormalities.
Increased intraepithelial lymphocytes (IEL) may be increased in varying cases such as peptic duodenitis, inflammatory bowel disease etc. and may be a nonspecific finding.(6) Serological testing for coeliac disease antibodies such as anti-gliadin, anti endomysial and anti-tissue transglutaminase help in distinguishing between the close mimics of coeliac sprue despite overlapping serology between coeliac sprue and autoimmune enteropathy. (7) We intended to study the morphologic spectrum of duodenal biopsies adjuncted by Immunohistochemistry (IHC) for IEL with CD3 antibody and to assess the microbiological profile of stool samples in patients with chronic diarrhoea and malabsorption. We also wished to study the prevalence of coeliac sprue by serological testing for antibodies seen in coeliac disease.

Materials and Methods
A descriptive study was conducted in the Department of Pathology, Microbiology and Medical Gastroenterology, JIPMER from June 2016 to May 2018. After approval by the institutional ethics committee, patients aged more than 18 years presenting with chronic diarrhoea and/ or malabsorption were recruited for the study. After obtaining informed consent from the participants, demographic details and clinical characteristics like features of nutritional deficiency were noted. Hemogram, biochemical and serological parameters were also collected. Criteria applied for malabsorption included chronic diarrhoea lasting for ≥ 4 weeks, anaemia defined by haemoglobin < 10g/dl, steatorrhea, oedema/ hypoalbuminemia/ hypoproteinaemia, clinical features of nutritional deficiencies and autoimmune diseases. (8) Duodenal biopsy was obtained endoscopically in the Department of Medical Gastroenterology. It was received in 10% neutral buffered formalin in a glass vial. The tissues were subjected to standard processing protocols and embedded in paraffin wax, sectioned at 4µm thickness and stained with Haematoxylin and Eosin (H&E) stain. Histopathological parameters like villous architecture and atrophy, crypt architecture, lamina propria infiltrate and IEL count per 100 enterocytes were observed. Villous atrophy was classified as mild, moderate/ marked and total /severe corresponding to Villous to Crypt (V:C) ratio of 2:1, 1:1 and 0:1 respectively (9) and Modified Marsh Oberhuber classification was applied (10). Marsh 1 lesions show increased IEL, Marsh 2 lesions show crypt hyperplasia and Marsh 3 lesions show villous atrophy in addition to Marsh 1 and 2 lesions. Crypt architecture was analysed as either normal or hyperplastic wherein hyperplastic crypts were elongated with increased number of mitoses. (11) Lamina propria infiltrate was graded subjectively into three grades as mild, moderate and severe (12).
IHC for CD3 (Dako, polyclonal) was done to analyse the IEL count and complemented with the counts on H&E-stained sections. They were interpreted as normal (0-25), borderline (26-29) or increased (≥ 30) for the entire villous per 100 enterocytes (6,13,14) and normal (0-5) or increased (≥ 6) at the villous tip per 20 enterocytes (15). Stool samples were collected and were subjected to wet mount microscopy and culture examination. At the same time blood samples were collected in a serum tube. They were centrifuged and serum was separated and preserved in 3 aliquots of 0.5ml each at -80˚C. Serum ELISA was done to detect the presence of IgA anti gliadin, IgA anti endomysial (EMA) and IgA anti tissue transglutaminase (tTG) antibodies (BIOCODON).
A total of 106 patients were included in the study after excluding patients with biopsy showing neoplastic lesions. Statistical analysis was done using IBM -SPSS software version 19.0. Level of agreement of intraepithelial lymphocyte count by haematoxylin and eosinstained sections were done by kappa statistics. All the statistical analysis was carried out at 5% level of significance with p value < 0.05 as statistically significant.

Results
This descriptive study included 106 consecutive patients presenting with diarrhoea or malabsorption during the study period. The morphologic spectrum of duodenal biopsies in patients with chronic diarrhoea and malabsorption was grouped as follows: (Table  1).  Figure  1A), moderate/marked atrophy in 7 (6.6%) patients ( Figure 1B) and severe/total atrophy in 5 (4.7%) patients ( Figure 1C). Crypt hyperplasia was seen in 14 (13.2%) of patients ( Figure 1D) and cryptitis was observed in three (2.8%) patients. Mild lamina propria infiltrate was seen in 30 (28.3%), moderate infiltrate in 48 (45.3%) and dense infiltrate in 28 (26.4%) patients. Brunner gland hyperplasia was observed in one patient who presented with chronic diarrhoea ( Figure  2A) and endoscopy showed mucosal nodularity. Incidentally, stool wet mount examination of this patient showed cysts of Entamoeba. Also, one of the duodenal biopsies revealed cross sections of adult forms of nematode in the crypts and lamina propria ( Figure 2B). Rheumatoid arthritis and seronegative arthritis were the most common autoimmune disorders encountered in this study. 2 cases of rheumatoid arthritis showed villous atrophy with a coeliac disease like pattern. Remaining cases showed chronic inflammation with a duodenitis like pattern. There were 5 cases in the study showing features of villous atrophy with increased IEL (Histomorphology suggestive of coeliac disease). Of these, 3 cases showed moderate villous atrophy, and a single case each of mild and total villous atrophy. The mean IEL for 100 enterocytes was found to be 14.53 ± 8.49. Increased IEL was observed in 12 (11.3%) patients with borderline increase in 4 patients (3.8%) and 8 (7.5%) patients had markedly increased IEL (>30/100 enterocytes) ( Figure  2C). IHC for CD3 was performed in 90 patients ( Figure 2D). IEL counts were increased in 15 (16.7%) and increased in borderline range in 2 (2.2%) and normal in 73 (81.1%) cases. The level of agreement between IEL on H & E-stained sections and CD3 stained section was done using Kappa statistics. Kappa value was 0.644 in our study indicating a moderate agreement. This indicates that CD3 stained sections offer a slightly better advantage over H & E-stained sections in detecting IEL.

Discussion
Chronic diarrhoea is defined as a change in stool consistency or frequency for more than or equal to four weeks (16,17). Malabsorption occurs due to failure of nutrient absorption because of mucosal disease or structural disorder (18). Tropical sprue, coeliac sprue, infections and immunodeficiencies are the common causes of malabsorption. The criteria for malabsorption used in this study was similar to a previous study conducted by Balasubramanian P et al (8).
The age range in the present study was 19-90 years with a mean of 40.49 ± 15.30 years however majority of patients were in the age range of 18-30 years. Other similar studies from India showed a mean range of 37.8 ± 13.2 years and 36.5 ±12 .6 years (1,3). In a study from CMC Vellore, the age range was 31.9 ± 16 years (2). The gender distribution in the present study was similar to a Lucknow based study which showed a male dominance with 61% (3). Another study from Mumbai showed a male dominance with 49% (1).
Chronic diarrhoea and anaemia were the most common manifestations accounting for 81.1% and 65.1% respectively. In two other studies on duodenal biopsy in malabsorption, 84% and 71 % patients presented with chronic diarrhoea (8,19). Nutritional deficiency has been noted in 36.8% of patients in contrast to 12.8% of patients in a previous study (8). The mean haemoglobin observed was 10.3 ± 2.87 g/dl. Although, no specific histological features were identified in anaemic patients, this study is justified in subjecting patients with iron deficiency anaemia to routine endoscopic biopsy as this screening helps in identifying an additional 6.63% of patients with malabsorption (20). The mean differences in haemoglobin was statistically significant between atrophy and normal villous architecture group by chi square test with a p value of 0.015. In this study hypoproteinaemia was observed in 42.5% of patients and low serum albumin levels in 33% of cases. In two other studies hypoalbuminemia was seen in 40.3% and 12% of cases (2,21). The most common endoscopic findings in the present study were either normal mucosa or scalloped duodenal mucosal folds. A study on endoscopic features in coeliac disease documented that mosaic pattern and scalloped folds have maximum sensitivity and specificity for suspecting coeliac disease. (22) However, in this study, only 7 patients with scalloped folds on endoscopy showed villous atrophy. There was no significant association between the endoscopic findings and villous atrophy with a p value of .146.
In the present study, there was mild villous atrophy in 11.3% of patients, moderate villous atrophy in 6.6% and severe atrophy in only 4.7%. A Mumbai based study, demonstrated more moderate villous atrophy (24%) than mild villous atrophy (18%) and similar percentage of severe villous atrophy. ( The incidence of coeliac disease in Indian population is just about 1%.(26) In our study, we encountered 5 cases with histology of coeliac disease (Marsh type 3 lesion). However, serology for three antibodies in coeliac disease were negative. The possible reasons for negative serology may be that they were partially treated or have refractory coeliac disease or seronegative coeliac sprue. (27) One case was positive for IgA anti gliadin, and IgA anti EMA antibodies and high normal IgA tTG. Duodenal biopsy revealed borderline increase in IEL but there was no evidence of crypt hyperplasia or villous atrophy. This patient could indicate an early form of potential coeliac disease with Marsh type 1 lesion. This patient symptomatically improved after strict gluten free diet. Thus, the incidence of coeliac disease in our population is relatively low even when adjuncted by serological tests.
Stool microscopy and culture yielded positive results in 6 cases each and they were independent of endoscopic and histological findings. One of the reasons attributed for the low yield of positive stool microbiological profile could be due to assessment made on a single stool sample. It has been suggested that a minimum of three to six stools samples are necessary for parasite identification in stool.(28) Also more advanced tests of stool examination like stool antigen detection by ELISA and other molecular tests could be employed in future studies to achieve better comparative results.

Limitations
The criteria employed for suspecting cases of malabsorption should have been more stringent with the use of faecal fat estimation and urine D -xylose estimation. This study focused only on the morphological and serological analysis. The study employed IgA antibodies for serological testing and there is a small possibility of IgA deficient individuals.

Conclusion
Duodenal biopsy in the study of malabsorption is a useful parameter, as 54.7% of duodenal biopsies showed abnormal findings in the form of villous atrophy, increased IEL and duodenitis. The use of serological tests to identify the prevalence of coeliac disease indicated that coeliac disease is still rare in our population despite few recent studies documenting its increasing prevalence. We concluded that a constellation of clinical/ endoscopic/ serologic/ microbiologic /histopathologic findings is mandatory before we arrive at a final diagnosis.