A study of duodenal intraepithelial lymphocytic population and its relation to coeliac disease in a cohort of patients in the Nile delta of Egypt

Introduction 

Coeliac disease (CD) is a T-cell-mediated immune disease of the small intestine triggered by wheat gliadin. The availability of sensitive and specific serological tests has made it possible to assess the true prevalence of coeliac disease by detecting minimally symptomatic or even asymptomatic cases with atypical mucosal changes [1]. Recent epidemiological studies have provided evidence that this disorder is common in variable parts of the world, including the Asian continent [2]. The enteropathy of coeliac disease is predominantly proximal. The characteristic lesion is described as villous atrophy. Intraepithelial lymphocytes (IELs) are increased in number and may persist after resolution of other changes [3].

Increased intraepithelial infiltrate, crypt hyperplasia, and villous atrophy are the three basic features of coeliac enteropathy. Observing that the degree of any of these features can be variable, Marsh and Crowe demonstrated that mucosal lesions could range from architecturally normal mucosa, with just an increased number of IELs, to severe villous atrophy and crypt hyperplasia [4]. Even though duodenal (or jejunal) intraepithelial lymphocytosis, as an isolated finding, has a high sensitivity for CD, it lacks sufficient specificity for this entity, as elevated IEL levels can be observed in a variety of gastrointestinal (and extra-intestinal) diseases, as shown in Table 1 [5].

Table 1. Etiology of increased duodenal intraepithelial lymphocytes in the absence of gluten sensitivity [5].

	Etiology
	Food allergy
	Primary immunodeficiency diseases
	Viral enteritis
	Giardiasis
	Blind loop syndrome
	Tropical (postinfectious) sprue
	Crohn’s disease
	Autoimmune diseases
	NSAIDS
	Irritable bowel disease

The most preserved histological end of the pathological spectrum is represented by an uncommon form of coeliac disease, defined as potential coeliac disease (PCD). A diagnosis of PCD can be made in any gluten consuming patient in whom, in spite of the presence of ‘coeliac stigmata’ such as a high IELs count and/or positive endomysial antibodies (EMA), an architecturally normal small bowel mucosa can still be found. Interestingly, PCD can evolve into active CD. From the pathologist’s point of view, an increased number of IELs in an architecturally normal duodenal mucosa always suggests PCD [3].

Increased numbers of CD8+ intraepithelial lymphocytes, especially those expressing Gamma/Delta T-cell receptors (γ/δ-TCR), have been observed in the pathogenesis of active CD [6].

In H. pylori infection, an increased number of IELs can be observed with no alterations of villous architecture. Differentiating the various patterns of lymphocytic infiltrates in these cases from those observed in CD and other small bowel diseases is not possible on H&E examination [5].

Patients and methods 

Results 

Histopathological results: data summarised in Table 2 

Villous atrophy 

Group A cases included 6 cases with grade 3 villous atrophy (Fig. 1), 10 cases with grade 2 villous atrophy (Fig. 2) and 8 cases with grade 1 villous atrophy (Fig. 3). Group B included 8 cases with grade 2 villous atrophy (Fig. 4) and 12 cases with grade 1 villous atrophy. Group C cases were histologically normal (control group).

View Large Image Download to PowerPoint

Fig. 1. Group A case with grade 3 villous atrophy (H&E stain, 100×).

View Large Image Download to PowerPoint

Fig. 2. Group A case with grade 2 villous atrophy (H&E stain, 100×).

View Large Image Download to PowerPoint

Fig. 3. Group A case with grade 1 villous atrophy (H&E stain, 100×).

View Large Image Download to PowerPoint

Fig. 4. Group B case with grade 2 villous atrophy and increased IELs (H&E stain, 400×).

Lymphoid aggregates 

Group A cases with grades 3, 2, and 1 villous atrophy had +++, ++, and + lymphoid aggregates, respectively. Group B cases had variable positivity and the positive cases correlated with H. pylori infection. No case in group C had lymphoid aggregates.

Intraepithelial lymphocytes 

IELs were 52.3, 34.8, and 28.8 in group A cases with grades 3, 2, and 1 villous atrophy, respectively. IELs were 17.8 and 5 in groups B and C cases (Fig. 4).

Helicobacter pylori 

H. pylori was studied in the gastric biopsy of the examined cases. The results are presented in Table 3. H. pylori infection correlated with lymphoid aggregates in a statistically significant manner (p<0.05).

Table 2. The histopathological results of the studied cases.

	Groups	Endoscopic diagnosis	Villous atrophy grade	Lymphoid aggregates				IELs (mean)	H.p		Total
				−	+	++	+++		−	+
	Group A	Coeliac	3	0	0	2	4	52.3	4	2	6
		Coeliac	2	0	2	8	0	34.8	6	4	10
		Coeliac	1	0	4	4	0	28.8	8	0	8
	Group B	Suspected potential coeliac disease	2	0	0	2	6	21.5	0	8	8
			1	0	12	0	0	15.3	12	0	12
	Group C	Normal	Control	6	0	0	0	5	6	0	6

Table 3. Correlation between H. pylori infection and both lymphoid aggregates and villous atrophy.

		Lymphoid aggregates				Villous atrophy				Total
		0	1	2	3	0	1	2	3
	HP−ve	6	16	10	4	6	18	8	4	36
	HP+ve	0	2	6	6	0	2	10	2	14
	Total	6	18	16	10	6	20	18	6	50
	Sig.	0.03				0.1

Table 4. Mean intraepithelial and lamina propria lymphocytes number (immunohistochemical staining).

	Groups	IEL			LP												Total
		CD3 (mean)	CD8 (mean)	CD4 (mean)	CD3				CD8				CD4
					0	1	2	3	0	1	2	3	0	1	2	3
	Group A – grade 3 villous atrophy	37.3	25.3	9	0	0	0	6	0	0	6	0	0	0	2	4	6
	Group A – grade 2 villous atrophy	25.6	14	7	0	0	6	4	0	6	4	0	0	2	6	2	10
	Group A – grade 1 villous atrophy	22.5	12	4.3	0	2	2	4	0	2	6	0	0	4	4	0	8
	Group B	11.9	8.1	5.5	0	12	8	0	0	16	4	0	6	8	6	0	20
	Group C	1.7	0.7	2.7	6	0	0	0	6	0	0	0	6	0	0	0	6
	Sig	0.01	0.03	0.02	0.03				0.009				0.008				50

Immunohistochemical results (Table 4) 

Intraepithelial lymphocyte distribution and profile 

There was no predominant pattern of distribution of IELs along the villous length, with equal superficial/supranuclear and subnuclear distribution (Fig. 5, Fig. 6).

View Large Image Download to PowerPoint

Fig. 5. CD showing even distribution of CD3+ve IEL in the villous tip and base (streptavidin–biotin–DAB, 400×).

View Large Image Download to PowerPoint

Fig. 6. CD with villous tip distribution of CD3+ve IEL (streptavidin–biotin–DAB, 400×).

The mean number of CD3+ IELs was 37.3 and 22.5 in group A cases, with grade 3 villous atrophy and grade 1 villous atrophy, respectively (Fig. 7). In group B cases, the mean CD3+ IELs was 11.9, with the highest numbers (showing more than 20 CD3+ IELs) related to H. pylori infection.

View Large Image Download to PowerPoint

Fig. 7. Group A case with grade 3 villous atrophy showing large numbers of CD3+ve lymphocytes in both the epithelium and LP (streptavidin–biotin–DAB, 100×).

In relation to CD8 and CD4, the highest mean was in group A cases with grade 3 villous atrophy (Fig. 8). However, group B cases also showed a relatively high number (Fig. 9) compared to group C.

View Large Image Download to PowerPoint

Fig. 8. Group A with grade 3 villous atrophy case showing CD8+ve lymphocytes both in the epithelium and taking grade 2 in LP (streptavidin–biotin–DAB, 400×).

View Large Image Download to PowerPoint

Fig. 9. Group B case showing CD8+ve IEL and few in the LP taking grade 1 (streptavidin–biotin–DAB, 200×).

Lamina propria lymphocytes profile 

Grade 3 CD3+ lymphocytes in LP was present in all cases of group A with grade 3 villous atrophy (Fig. 7); grade 2 was observed in 40% of cases of group B, predominantly those with H. pylori infection.

Grade 2 CD8+ lymphocytes in LP was present in all cases of group A with grade 3 villous atrophy; grade 1 was observed in 80% of cases of group B (Fig. 8, Fig. 9).

Grade 3 CD4+ lymphocytes in LP was present in 66.6% of cases of group A with grade 3 villous atrophy; grade 2 was observed in 30% of cases of group B (Fig. 10). All of group B cases were H. pylori positive.

View Large Image Download to PowerPoint

Fig. 10. Group A case showing CD4+ve lymphocytes in the LP taking grade 3 (streptavidin–biotin–DAB, 100×).

Discussion 

Fry et al. [11] brought attention to the diagnostic significance of increased IELs in patients with coeliac disease in 1972. Intraepithelial lymphocytosis was found to be more reliable in detecting gluten sensitivity than the endoscopic appearance of small bowel mucosa.

Recent studies have, however, reported a low specificity of ‘isolated intraepithelial lymphocytosis’ for diagnosing coeliac disease [12].

A raised IELs count in an architecturally normal duodenal mucosa has always suggested PCD to the pathologist, but it now appears that there are many causes of increased IELs with normal or slightly affected mucosa rather than PCD [3].

In this study, the highest mean of IELs was in the CD of grade 3 villous atrophy; this gradually decreased with the decrease of atrophy grade. It was significantly lower in patients with PCD and lowest in the control group. It was also high in H. pylori patients, in accordance with the findings of Hasan et al. [13], who reported increased IELs in ulcer-associated and non-specific duodenitis.

Hayat et al. [14] also found raised IEL counts in the second part of the duodenum in patients with H. pylori-associated lymphocytic gastritis. They noticed a decrease in IEL counts after treatment of H. pylori infection, suggesting the possibility of underlying CD in these patients.

Memeo et al. [5] observed a high rate of duodenal intraepithelial lymphocytosis in patients with H. pylori gastritis who had otherwise normal villous architecture. We observed a mixed pattern of intraepithelial lymphocytic location, similar to that reported by Memeo et al. [5] and Goldstein and Underhill [15], but in contrast to Hasan et al. [13].

The lymphoid aggregates in the LP were also highest in CD of grade 3 villous atrophy. Their presence showed a significant correlation with H. pylori infection.

Our observations, in accordance with Mino and Lauwers [16], indicate a substantial overlap in IEL counts between CD and other diseases/conditions.

Kakar et al. [17] observed a significant association between isolated intraepithelial lymphocytosis and a variety of immunological diseases and NSAID use.

Ernst et al. [18] stated that the source of duodenal IELs in H. pylori infection is not known.

It has been shown that H. pylori stimulates B lymphocytes and causes an increase in their number, predominantly in the LP [19].

Hasan et al. [13] reported normalisation of IEL counts (<10/100 epithelial cells) after treatment with cimetidine. However, none of these patients had normal villous architecture.

In relation to IELs immunoprofile in this work, the CD cases with grade 3 villous atrophy showed that CD3+ cells were the most predominant, followed by CD8+ and CD4+ cells, respectively. The number of these cells gradually decreased in the subsequent groups, with a significant difference evident when compared with the normal group and with other groups.

Bedoya et al. [20] demonstrated the presence of CD3+ IELs in coeliac disease and H. pylori infection. They assumed that in undiagnosed patients, higher counts of CD8+ IELs may imply that the diagnosis is CD rather than H. pylori infection.

Our findings support the published data of Broide et al. [10], who stated that CD8+ IELs are involved significantly in the pathogenesis of CD.

In this work, CD3+, CD8+, and CD4+ cells were high in the lamina propria of cases of CD with grade 3 villous atrophy. CD4+ and CD8+ cells were significantly associated with H. pylori infection.

We can conclude that although the difference between PCD and non-coeliac controls was significant, the two groups overlapped considerably. As the purpose of this study was to alert pathologists to the frequent occurrence of duodenal intraepithelial lymphocyte expansions in individuals other than coeliac disease, the presence of patients with PCD should be considered. Also, infection with H. pylori and the overlap of the intraepithelial lymphocyte counts as well as the distribution patterns with those described for coeliac disease should be considered. However, it should be noted that these observations and the results of this study are preliminary in nature, due to the limited number of patients.

From the results of this work it can be recommended that for patients infected by H. pylori and suspicious for having atypical CD (negative serology with increased IELs with normal villous architecture), an increased number of CD8 IELs might guide the diagnosis of CD rather than H. pylori infection. On the other hand, an increased number of lymphoid follicles might guide to the diagnosis of H. pylori infection. Thus, it is important to include immunohistochemical analysis of CD8 lymphocytes in undefined cases of CD.