Coeliac Disease

Histopathology

Dr Vincenzo Villanacci

Department of Pathology

Spedali Civili

Brescia

Italy

METHODOLOGY APPROACH TO DUODENAL BIOPSY

The biopsies that the pathologist receives nowadays are all performed by endoscopic examination,

which, in addition to the duodenum, makes it possible to explore other districts of the gastro-

intestinal tract. Biopsies performed through the use of the Crosby-Watson capsule by peroral

route are now considered outdated and are no longer performed.

Here are some points which require a close working relationship between the endoscopist, the

endoscopy-room nurse, the pathology laboratory technician and the pathologist.

Site of the biopsy

Biopsy by endoscopy is always performed in the second and third duodenal portion, as the bulb

and the proximal duodenum can be a source of incorrect assessments; we recommend at least 4

biopsies, 2 for each of the areas mentioned above. Performing a biopsy only in the duodenal bulb

may be a source of error, or may at least greatly reduce the sensitivity of the examination, and

hence is strongly discouraged (1).

Orientation of the biopsy sample

This is essential for proper histological assessment.

Positioning of biopsies on cellulose acetate filters is advisable, with benefits for:

A) the laboratory staff, since with a simple 90°rotation it is possible to embed the combined

biopsy-filter

B) for subsequent histological evaluation by the pathologist.

The method based on experience acquired at St Mark's Hospital in London ensures histological

samples in which it is possible to analyze the mucosa and, if necessary, the submucosa of the

removed tissue, thus respecting the normal anatomical relationship between the different layers

of the intestinal wall. In particular, after initial experience with filters that needed to be cut with

obvious waste of time and human resources, a comprehensive kit has been developed, on which

three easily-detachable cut filters with a bevelled end shaped like a clarinet mouthpiece are

already fixed (Fig.1),

After the fixing stage, the filter-biopsy combination is processed and then embedded. During this

last phase the technician rotates the filter-biopsy combination 90 degrees in order to place the

samples in their natural position.

After cutting, the biopsies are placed on a slide and, if necessary, the position of the beveled end is

marked on the label, to indicate the first biopsy.

When properly carried out, this method is of great benefit to the pathologist, but also to the

technician who during the embedding phase does not have to search for the individual biopsies,

which are sometimes fragmented and have no guiding landmark.

The use of cellulose acetate filters allows perfect adhesion of the biopsies, avoiding their

dispersion in the fixation medium. These filters also do not react chemically with the fixatives and

reagents used during the processing of the sample; during the cutting phase they do not offer

resistance to the blade and, unlike tissue paper, they do not fray.

This method, which can be applied on all segments of the gastro-intestinal tract, has led to

considerable diagnostic and economic benefits by reducing the time, the number of embeddings,

and consequently the number of sections to be cut and colored.

For its obvious advantages, the use of the kit is strongly recommended

(Fig.1)

Stains

It is sufficient to stain with hematoxylin-eosin, possibly associated with an Alcian Blue-PAS, to

assess all the necessary morphological elements (one or two sections can be used for

immunohistochemical assessment if necessary)

HISTOPATHOLOGICAL ASPECTS OF NORMAL AND PATHOLOGICAL

DUODENAL MUCOSA

Microscopy

From the viewpoint of optical microscopy we will first consider the appearance of normal

intestinal mucosa, then the progressive histological patterns that can be found in celiac disease.

Normal intestinal mucosa

Villi

digitiform appearance with the ratio between the height of the villi and of the crypts always in

favor of the villus (3 / 1 or more).

Enterocytes

normal height with 29-34 micron clear brush-border

Intra-epithelial lymphocytic infiltrate

The number of intra-epithelial lymphocytes (T Lymphocytes) is subject to individual variability. The

majority of normal subjects have less than 20 lymphocytes per 100 epithelial cells; based on the

experiences of Hayat (2) and Veress (3), a count of more than 25 lymphocytes per 100 epithelial

cells is currently considered as pathological, while between 25 and 30 is considered borderline and

worthy of study as a comparison with clinical and laboratory data. A value of 40 lymphocytes per

100 epithelial cells can be considered as frankly pathological but is rarely found in routine practice.

The intra-epithelial lymphocyte count is very important and should always be done, especially in

the initial lesions, following the indications given below:

A) always count the T lymphocytes with the help of immunohistochemical investigations using

anti-CD3 antibodies;

B) evaluate the biopsies perfectly oriented with a precise alignment of the surface-coating

epithelial cells;

C) do the count both in the apical portions and along the edges of the villi; it is important to

have accurate and reproducible fields. Counts done only on the apical portions have proved

unreliable. (FIG. 2A-2B-2C-2D)

Glandular crypts

The crypts basically have the task of performing a regenerative function, which means it is possible

to find evidence of mitosis; the normal range is usually one mitosis per crypt. Alongside the

epithelial cells are endocrine cells, goblet cells and Paneth cells, but these have no value as regards

the diagnosis of celiac disease.

Lamina propria

Plasma cells, eosinophils, histiocytes, mast cells and lymphocytes are normally found in the lamina

propria. Neutrophils are generally absent, except in cases of active duodenitis with possible gastric

metaplasia closely related to HP infection.

The cellular component mainly consists of plasma cells and lymphocytes, the latter sometimes in

the form of lymphoid aggregates and eosinophilic granulocytes whose value must never be

greater than 60 for 10 fields of vision examined at 40 x.

PATHOLOGICAL INTESTINAL MUCOSA

(Marsh Classification)

(Type I or infiltrative lesion)

1) Villi architecturally within normal morphological limits (normal villa / crypt ratio 3 / 1)

2) Increased number of intraepithelial lymphocytes (greater than 25-30 per 100 epithelial cells)

(Fig.3-3B-3C-3D)

(Type II or hyperplastic lesion)

1) Villi architecturally within normal morphological limits (like type 1)

2) Increased number of intraepithelial lymphocytes (greater than 25-30 per 100 epithelial cells)

3) Hyperplasia of the glandular elements (regenerative aspect of the glandular elements

highlighted by the reduced muciferous activity and increased number of mitoses)

(Type III or destructive lesion)

1) Varying degrees of villous atrophy associated with hyperplasia of glandular crypts;

2) Reduced surface enterocyte height, with irregular brush-border and sometimes cytoplasmic

vacuoles;

3) Increased number of intraepithelial lymphocytes (like type I and II lesions).

The combination of the three factors described above is consistent with a diagnosis of celiac

disease or gluten-sensitive enteropathy, which should be considered as synonyms. These 3

patterns, albeit schematic, represent the histological lesions seen in celiac disease and it is

important to consider them as dynamic and progressive both in one direction and the other and

not static, since they depend on how much gluten the patient has been exposed to and for how

long.

The following diagram summarizes the above description. (From Marsh GUT 1990, amended) (4)

Infiltrative

(Type 1)

Hyperplastic

(Type 2)

Destructive

(Type 3)

Close relatives of celiac patients Celiac patients exposed to

moderate amounts of gluten

Untreated celiac disease

Treated celiac patients exposed to

minimal amounts of gluten

Dermatitis herpetiformis without

clinical enteropathy

Treated celiac patients exposed to

large amounts of gluten

Dermatitis herpetiformis without

clinical enteropathy

Dermatitis herpetiformis with

clinical enteropathy

Tropical enteropathy

Tropicale sprue

This classification is universally recognized for the diagnosis of celiac disease, and extensively

validated; the only point worthy of observation and critical analysis is that the cases with mild,

moderate or severe atrophy (total villous flattening) are all grouped together in a single category:

the type 3 lesion.

An amendment to this classification has been proposed by Oberhuber et al. (5), who divided the

Marsh type lesion 3 into three subgroups.

3a mild villous atrophy and pathological increase of intraepithelial lymphocytes.

3b moderate villous atrophy and pathological increase of intraepithelial lymphocytes. (Fig.4-4B)

3c total villous atrophy and pathological increase of intraepithelial lymphocytes (Fig 5A-5B)

Without prejudice to all the other morphological criteria described above, this classification

provides a better description of the spectrum of lesions that may occur both in celiac disease

patients on a normal diet and in those on a gluten-free diet.

Along the same lines, and in an attempt to simplify and standardize the work of pathologists and

facilitate the relationship between pathologists and clinicians, a new version of the histological

classification has recently been proposed (6-7); in particular, the lesions that characterize celiac

disease have been divided into two categories

Non-atrophic (grade A) and atrophic (grade B)

Grade A lesions are characterized by a pathological increase in intraepithelial lymphocytes, best

recognized by the use of immunohistochemical techniques.

Grade B lesions are further subdivided into

Grade B1

in which the villus / crypt ratio is less than 3 / 1, with villi still identifiable, and

Grade B2

in which the villi are no longer identifiable.

Table 1 compares the two classification systems

Marsh-Oberhuber classification Corazza-Villanacci classification

Type 1

Type 3a

Type 3c

Grade A

Grade B1

Grade B2

Type 2

Type 3b

DIAGNOSIS

The diagnosis represents the culmination of what is described above and must comprise all the

morphological requirements so as to allow a direct, clear and simple understanding of the

morphological situation of the duodenal mucosa by the clinician.

A two-step proposal is presented below:

A: assessment of the morphological pattern divided according to description and diagnosis.

The description should report, in sequence, the same morphological elements listed above,

namely: villous trophism, number of intraepithelial lymphocytes, features and glandular structures

of the lamina propria, concluding with compatibility or non-compatibility with the pattern of a

celiac patient based on complete clinical and laboratory data.

The final diagnosis should not use the term “celiac disease” but only the histologic abnormalities

found; it is the pathologist’s task to provide an interpretation of a certain moment in the

development of a disease, while the final diagnosis is entirely up to the adult or pediatric clinician.

As a brief addition to the above we propose that the term sub-atrophy, in itself unclear and

misleading, should no longer be used. Instead it is better to specify whether the villI are normal or

atrophic, and in the latter case, the degree of atrophy, from mild to moderate to severe. In the

event of severe atrophy it is possible to use the term total or severe atrophy. Scores should not be

attributed to the individual morphological elements as they are too subjective and of little or no

use for the final diagnosis.

B: Checklist

ASSESSMENT FORM

Name _______ Surname__________ Sex M__ F__ Date of Birth_____ / _____ / _______

1st Biopsy___ No. ___ Control___________ No.__________

No. of biopsies _____ ______ Positioning correct not correct _______

Villi: normal Atrophy _____ ____ mild_____ moderate_____ severe_____

Villus / crypt ratio Normal [3 / 1] altered ______ _______

Intraepithelial lymphocytes: normal_____ increased______

(less than 25-30 lymphocytes/100 epithelia) (more than 25-30 lymphocytes/100 epithelia)

Evaluation with CD3________________________________________________________________

Glands: normal hyperplastic _______ _________

Lamina propria____________________________________________________________________

Diagnosis: (according to Oberhuber Marsh) (New proposed classification)

Type 1 Grade ______ A______

Type 2 ______

Type 3a Grade B1_____

Type 3b ______

Type 3c ______ Grade B2_____

Comments :______________________________________________________________________

________________________________________________________________________________

IMMUNOHISTOCHEMISTRY

One of the key points in the diagnosis of celiac disease is the number of intraepithelial

lymphocytes, which are CD3 and CD8 positive T lymphocytes; in pathological conditions, their

number should be more than 25 lymphocytes per 100 epithelial cells (border-line value 25-30).

The definition of a precise cut-off between normal, abnormal and border-line pathological lesions

is of particular importance given the increase in celiac disease diagnosed in the early/subclinical

stage.

The counts can be performed reasonably well on the normal and irreplaceable hematoxylin - eosin

but we suggest, especially in the initial forms, that an immunohistochemical assessment should

always be carried out with monoclonal CD3 which often allows for a more accurate display of

lymphocytes, following a series of procedures (see the section on intraepithelial lymphocyte

infiltration). Evaluation with CD8 may also help, and is particularly useful in cases of elderly

subjects where it is possible to find refractory forms which do not respond to diet, regarded by

many as pre-lymphomatous and in which the expression of CD8 may be negative with respect to

the "norm". (8).

As frozen material is available, immunohistochemical typing for the gamma-delta receptor of T

lymphocytes can be carried out; in normal conditions this receptor is not expressed by more than

2-3% of T lymphocytes while in celiac disease it may reach 20-30% - a particularly useful marker in

initial lesions. This assessment is, however, based on the use of frozen material and is not

therefore recommended in routine practice.

DIFFERENTIAL DIAGNOSIS

The above summarizes the morphological lesions which may occur with celiac disease and where

the pathologist clearly has a key role, if only to exclude the possibility of clinically suspected

malabsorption which may have as also be:

• Parasitic (Giardia lamblia, Cryptosporidium, Microsporidium)

• Infectious (Whipple's disease)

• viral (cytomegalovirus, herpes virus)

• idiopathic (Crohn's disease)

• neoplastic.

The most important problem today in the diagnosis of celiac disease is represented by early

lesions, i.e. normal villi with a pathologic increase in intraepithelial T lymphocytes. This issue is

appropriately dealt with in the excellent review by Brown et al. (9), in which the summary table

below

shows that in addition to celiac disease, there are a number of pathological conditions that have

the same morphological aspect as celiac disease in its early stages, i.e .normal villous architecture

but with a pathological increase of IELs (> 25-30/100 epithelial cells) (lesion type 1 according to

Marsh, Grade A according to the new proposed classification). These conditions include

hypersensitivity to other foods (milk, cereals, soybeans, fish, etc.), infections (Hp, Giardia ...), the

use of drugs, immunodeficiencies and immunodysregulation (Hashimoto thyroiditis, systemic

lupus erythematosus, rheumatoid arthritis, etc.) and, not least, chronic idiopathic inflammatory

bowel colitis or colitis with a different etiology, such as lymphocytic and collagenous colitis.

The question that we must therefore ask is: How can we discriminate between different

pathological conditions where the morphology is essentially superimposable? A proper clinical

evaluation based on histological and laboratory data is crucial. We must not forget that a diagnosis

of celiac disease is a "marker" which remains throughout life with obvious therapeutic and

behavioral relapses. The table above helps understand how important the need for collaboration

between the pathologist and endoscopist is in the detection of other conditions, such as infection

with Giardia lamblia or other parasites, the possibility of presentations of immunodeficiencies

morphologically superimposable on celiac disease and not least the localization of Crohn's disease

or particular forms of enteritis within the sphere of untreatable diarrhea, such as autoimmune

enteritis, tufting enteropathy, a disease caused by atrophy of the microvilli, and cases of graft-

versus-host disease, all conditions in which the morphological element is fundamental.

Three conditions, however, deserve special mention:

• Forms of so-called "autoimmune enteritis" possible in children with immunological deficiency

(Common variable immunodeficiency, X linked Agammaglobulinemia) in which the intestinal

biopsy may be fully comparable to the pattern of celiac disease. (10))

• Damage by drugs: there is increasing evidence in the literature showing that the use of drugs,

especially non-steroidal anti-inflammatory drugs (NSAID), are capable of causing morphological

alterations identical to those of celiac disease, and it is therefore important to keep this

possibility in mind in cases of elderly patients, especially when the serological markers are all

negative. (9)

• The possibility that concurrent infection with Helicobacter pylori in the stomach can produce a

morphological pattern very similar to that of initial lesions of celiac disease as recently

reported. (11)

COMPLICATIONS THAT CAN BE CONFIRMED HISTOLOGICALLY

Unlike what occurs in children, there is considerable evidence that celiac disease in adults,

especially if diagnosed late and even more so if not dealt with by a timely and rigorous gluten-free

diet, is burdened by a higher mortality rate than in the general population.

The removal of gluten from the diet therefore determines not only an improvement of the

histological and clinical aspects, but also prevents the complications which must always be

suspected if an adult patient continues to be unwell, despite the diet.

These complications are due to:

• Collagenous sprue: the patient does not respond to diet and histology shows fibrous tissue in

the intestinal wall at the level of the superficial subepithelial layer. This morphological pattern

is very similar to the condition of collagenous colitis described in the colon, where the thickness

of the connective band best highlighted with Masson’s trichrome is more than 15 millimicrons,

although this is a very rare event is described in the literature.

• Refractory Sprue: this condition reproduces the same clinical picture as collagenous sprue but

can be identified by immunohistochemical staining, demonstrating that T lymphocytes, which

in normal conditions express CD3 and CD8, in this case present only the expression of CD3 and

not of CD8 (8).

• Ulcerative jejunoileitis: presence of extensive ulceration of the intestinal mucosa, often related

to refractory sprue

• Lymphoma: this is the most serious complication and should always be suspected when

histology shows a prevalence of atypical monomorphous lymphocytic elements. In these cases

it is useful to carry out immunophenotyping of the lymphoid population, which is almost always

type T (12 -13 - 14).

SUMMARY

ANATOMOPATHOLOGICAL ASPECTS

What are the "certainties" in the diagnosis of celiac disease

An obvious prerequisite for certainty in the diagnosis of celiac disease (CD) from the anatomo-

pathological point of view is the observation and respect of a number of key points:

1. Close collaboration between clinicians, laboratory technicians and endoscopists

2. An adequate number of biopsies (at least 4, 2 in the distal and 2 in the proximal

duodenum)

3. Correct orientation of the biopsy (the use of pre-cut cellulose acetate filters)

4. Sufficient clinical information.

5. Excellent quality of the biopsy samples

• With these details it is clear that "certainty" in the diagnosis of CD is only possible if the

villous atrophy is associated with a pathological increase in the number of intraepithelial

lymphocytes (value exceeding 25-30/100 epithelial cells). In this situation, by applying the

three classifications now known and validated (Marsh, Marsh-Oberhuber and Corazza-

Villanacci) there is no problem in the diagnosis and comparison with clinical and

laboratory data.

• The degree of atrophy should be certain and not merely pseudo-atrophy due to incorrect

orientation and cutting of the villi. Assessment of the number of intraepithelial

lymphocytes is useful in these cases and "must" always be pathological (> 25-30/100

epithelial cells), best evaluated both with H & E staining and with immunohistochemistry

staining for CD3.

• Attention should be paid to biopsies taken from the duodenal bulb, where the presence

of Brunner glands can lead to a false diagnosis of atrophy; biopsies of the bulb should

always be compared with those taken from the distal portions, especially in the early

stages of the disease, which has a progression of the pathological process in a cranio-

caudal direction.

• If there are varying degrees of atrophy, these should all be described, not just the most

severe degree. An assessment of compatibility should only be included in the description

of the case, while the term “celiac disease” should be avoided in the final diagnosis, which

should be limited to a description, giving the clinician a precise "snapshot" of the state of

the duodenal mucosa. The final diagnosis of CD should be made solely and exclusively by

the pediatric or adult clinical gastroenterologist.

What are the "doubts" in the diagnosis of celiac disease

The points that cause doubt and require caution on the part of the pathologist in the diagnosis of

CD are clearly represented by the cases in which there are initial lesions

(Marsh 1-2, and Grade A in accordance with the new proposed classification); in these cases it is

necessary to:

1. Carefully assess the orientation of the biopsies

2. Consider whether the villus/crypt ratio of at least 3 / 1 is respected

3. Carefully count the number of lymphocytes in the surface coating epithelium

4. Always carry out additional immunohistochemical evaluation with CD3

5. Compare the clinical and laboratory data.

• The two key elements that must be assessed are the absence of atrophy and the

increase in the number of intra-epithelial lymphocytes; it is therefore crucial to always

associate immunohistochemical evaluation with CD3. The presence or absence of

hyperplasia of the glandular elements is totally irrelevant for practical and therapeutic

purposes.

• Do not forget that the "slide" is proof of the assessment by the pathologist and as such

can be compared and re-assessed by other colleagues and specialists; it must also be

strongly emphasized that the histological assessment must be conducted solely by the

pathologist and not by other "specialists".

• As with "certain" cases, it is even more important in doubtful cases to merely express

an opinion of possible compatibility with CD in the description, describing only the

histological aspects in the final diagnosis.

• Exclude, if possible, a concurrent infection due to Helicobacter Pylori (it is advisable to

always take biopsies of the antral and oxyntic gastric mucosa ), immunodeficiencies,

parasitic infections, allergies to other dietary factors and use of drugs.

• In doubtful situations in which the final pattern is unclear, it is useful to bear in mind

the excellent review by Brown et al.: the pathologist must be sure that he is faced with

a pathological condition unequivocally demonstrated by the increase in the number of

intraepithelial T lymphocytes confirmed by the evaluation with CD3. The final diagnosis

will be based on comparison of the clinical and laboratory data.

• In pediatric patients in the first year of life, the possibility of intolerance to cow's milk

proteins should not be forgotten; in such cases the eosinophilic granulocyte count may

be useful (pathologic value above 60 for 10 fields of vision at 40 X)

REFERENCES

1) Pais WP, Duerksen DR, Pettigrew NM, Bernstein CN.

How many duodenal biopsy specimens are required to make a diagnosis of celiac disease?

Gastrointest Endosc. 2008 Jun;67(7):1082-7.

2) Hayat M, Cairns A, Dixon MF, O'Mahony S.

Quantitation of intraepithelial lymphocytes in human duodenum: what is normal?

J Clin Pathol. 2002 May;55(5):393-4.

3) Veress B, Franzen L, Bodin L, Borch K.

Duodenal Intraepithelial Lymphocyte Count Revisited

Scand J Gastroenterol 2004 Feb;39(2):138-44.

4) Marsh MN.

Grains of truth: evolutionary changes in small intestinal mucosa in response to environmental

antigen challenge.

Gut. 1990 Jan;31(1):111-4. Review.

5) Oberhuber G, Granditsch G, Vogelsang H.

The histopathology of coeliac disease: time for a standardized report scheme for pathologists.

Eur J Gastroenterol Hepatol. 1999 Oct;11(10):1185-94. Review.

6) Corazza GR, Villanacci V

Coeliac Disease. Some considerations on the histological classification.

Journal of Clinical Pathol. 2005; 58, 573-574

7) Corazza Gr, Villanacci V, Zambelli C, Milione M, Luinetti O, Vindigni C, Chioda C, Albarello L, Bartolini

D, Donato F.

Comparison of the interobserver reproducibility with different histologic criteria used in celiac

disease.

Clin Gastroenterol Hepatol 2007 Jul;5(7):838-43.

8) De Mascarel A, Belleannée G, Stanislas S, Merlio C, Parrens M, Laharie D, Dubus P, Merlio JP.

Mucosal intraepithelial T-lymphocytes in refractory celiac disease: a neoplastic population with a

variable CD8 phenotype.

Am J Surg Pathol. 2008 May;32(5):744-51

9) Brown I, Mino-Kenudson M, Deshpande V, Lawers GY.

Intraepithelial lymphocytosis in architecturally preserved proximal small intestinal mucosa: an

increasing diagnostic problem with a wide differential diagnosis.

Arch Pathol Lab Med 2006 (130): 1020-5.

10) Washington K, Stenzel TT, Buckley RH, Gottfried MR.

Gastrointestinal pathology in patients with common variable immunodeficiency and X- inked

agammaglobulinemia.

Am J Surg Pathol. 1996 Oct;20(10):1240-52.

11) Memeo L, Jhang J, Hibshoosh H, Green PH, Rotterdam H, Bhagat G.

Duodenal intraepithelial lymphocitosis with normal villous architecture: common occurrence in H.

pylori gastritis.

Mod Pathol 2005 Aug; 18 (8):1134-44

12) Du MQ, Isaacson PG.

First steps in unraveling the genotype of enteropathy-type T-cell lymphoma.

Am J Pathol. 2002 Nov;161(5):1527-9

13) Dickson BC, Streutker CJ, Chetty R.

Coeliac disease: an update for pathologists

J Clin Pathol 2006 (10): 1008-16

14) Serra S, Jani PA.

An approach to duodenal biopsies

J Clin Pathol 2006 (59): 1133-1150

PICTURES

(Figs. 2A-2B-2C-2D) Normal Duodenal Mucosa : villus/crypt ratio 3/1. Intrapeithelial lymphocytes within

the normal range

(Figs. 3A-3B) Type 1/2 infiltrative lesion according to Marsh Oberhuber.Grade A new classification

2A

A

2B

3A

A

3B

(Figs. 4A-4B) Type 3A-3B lesion according to Marsh Oberhuber. Grade B1 new classification

4A

A

4B

A

(Fig.5A-5B) Type 3C lesion according to Marsh Oberhuber. Grade B2 new classification

5A

A

5B

A

Digestive and Liver Disease 43S (2011) S385–S395

Coeliac disease: The histology report

Vincenzo Villanacci a,*, Paola Ceppab, Enrico Tavani c, Carla Vindigni d, Umberto Volta e

On behalf of the “Gruppo Italiano Patologi Apparato Digerente (GIPAD)” and of the “Società Italianadi Anatomia Patologica e Citopatologia Diagnostica”/International Academy of Pathology,

Italian division (SIAPEC/IAP)aDepartment of Pathology, Spedali Civili, Brescia, Italy

bSurgical Department, Integrated Morphological and Methods Section of Pathological Anatomy, University of Genova, Genova, ItalycDepartment of Pathology, G. Salvini Hospital Rho, Rho, Italy

dDepartment of Pathology and Human Oncology, University of Siena, Siena, ItalyeDepartment of Diseases of the Digestive System and Internal Medicine, Policlinico S. Orsola – Malpighi, Bologna, Bologna, Italy

Abstract

To this day intestinal biopsy is justly considered the “gold standard” for the diagnosis of coeliac disease (CD). The aim of the authorsin setting up these guidelines was to assist pathologists in formulating a more precise morphological evaluation of a duodenal biopsy in thelight of clinical and laboratory data, to prepare histological samples with correctly oriented biopsies and in the differential diagnosis with otherpathological entities and complications of the disease. A further intention was to promote the conviction for the need of a close collaborativerelationship between different specialists namely the concept of a “multidisciplinary team”.© 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

Keywords: Coeliac disease; GIPAD report; T lymphocytes; Malabsorption

1. Introduction

These guidelines are intended as an aid for pathologists,in order to allow a more precise morphological evaluationof duodenal biopsies in the light of clinical and laboratorydata. The aim is to arrive at the conviction of the need fora close collaborative relationship between different specialistssuch as adult or pediatric gastroenterologists, endoscopists,laboratory staff, endoscopy room nurses, pathology laboratorytechnicians and pathologists. This implies the creation ofa “multidisciplinary team” led by a gastroenterologist who,

List of abbreviations: CD, coeliac disease; tTGA, antitransglutaminase anti-bodies; EMA, antiendomysial antibodies; AGA, antigliadin antibodies; IEL,intra epithelial lymphocytes.* Correspondence to: Vincenzo Villanacci, MD, Department of Pathology,Spedali Civili, Piazzale Spedali Civili 1, 25100 Brescia Italy.

E-mail address: villanac@alice.it (V. Villanacci).

1590-8658/$ – see front matter © 2011 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

based on the most recent acquisitions regarding the diagnosisand pathogenesis of coeliac disease, is the only specialist thatcan make the final diagnosis of coeliac disease. A numberof specific points in the diagnostic process will thus be dealtwith, considering issues and concerns of differential diagnosiswith other similar diseases before the final diagnosis can bereached.

This document is an update of the “guidelines” publishedby the Italian Group of Digestive Disorders (GIPAD) in 1998[1].

The document, after a brief historical and epidemiologicaladdress, considers the following points:• Clinical and laboratory aspects• The methodological approach to duodenal biopsies• Aspects of normal and pathological duodenal mucosa• Diagnosis• Differential diagnosis• Complications that can be confirmed histologically.

S386 V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395

2. Brief history

The first descriptions of coeliac disease can be found in thefirst century A.D. when the physician Celsus introduced theLatin term “coeliac” to indicate a diarrhea-like disease. Later,in 250 A.D., Areteo Cappadocia described the clinical signsof a prolonged intestinal disease that was very difficult totreat, using the Greek word koiliakos to identify “those whosuffer in their intestines”. In 1856, Francis Adams translatedthis Greek word into English, coining the term “coeliac”. Afew years later, in 1888, Samuel Gee described the detailedsymptoms of this condition both in adults and in children, pre-dicting that the only treatment consisted of an appropriate diet,with few items derived from flour. Only halfway through thetwentieth century, however, did it become clear that coeliacdisease occurs in some individuals following the ingestionof wheat proteins, which damage the intestinal mucosa. Thesystematic description of the histopathological alterations ofcoeliac disease (CD) is mainly due to the work of Marsh[2,3]. Today we know that CD is a chronic, immune-mediateddisease occurring in genetically predisposed individuals dueto an intolerance to gluten-containing foods and, in particular,to some of its proteins, called gliadins. This intolerance leadsto abnormal immune response, which is followed by a chronicinflammation of the small intestinal mucosa with progressivedisappearance of intestinal villi.

3. Epidemiology

The disease has a variable incidence, which in Europeis estimated between 0.3 and 1.2%, similar percentages arereported in North America and Australia. Recently, a highprevalence has been reported in people of Northern Africa(5–6% in populations of Western Sahara). In Italy the mostrecent statistics estimate a prevalence of 1/100, and each yearabout 5000 new cases are diagnosed. CD, once considered adisease of childhood, can affect individuals of all ages, with apreference for females (male/female ratio 1:2).

4. Clinical and laboratory aspects

4.1. Clinical aspects of CD

The variety of clinical manifestations which coeliac diseasemay present complicates its recognition. A correct diagnosiscan not rely on a single test, but requires a precise recon-struction of a puzzle, whose pieces are represented by theclinical, serological, genetic and histological aspects. Theevaluation of all these factors, apart from genetics, must takeplace while the patient is still on a diet containing gluten,since a gluten-free diet changes the clinical, serological andhistological pattern, making it impossible to recognize thecharacteristic aspects of disease.

The significant improvement in our knowledge of intoler-ance to gluten has made it possible to identify the so-called

risk groups in which, on the basis of intestinal and extrain-testinal symptoms, and of the presence of any associateddiseases and familiarity, the possibility of coeliac diseasemust be investigated [4,5].

These risk groups are made up of:1. Subjects in whom coeliac disease is strongly suspected

(cases with severe malabsorption and with highly predic-tive associated diseases):• malabsorption syndrome with repeated diarrhea-like

bowel movements, abdominal pain and marked weightloss;

• dermatitis herpetiformis, also called coeliac disease ofthe skin, since in practically all cases there is more orless severe gluten-dependent intestinal damage.

2. Subjects in whom coeliac disease is moderately suspected(cases with atypical or extraintestinal symptoms and asso-ciated diseases):• atypical gastro-intestinal symptoms (dyspepsia, consti-

pation, vomiting and intestinal subocclusion);• extraintestinal symptoms (anemia – most often due to a

lack of iron but also to a lack of folic acid and vitaminB12), hyposomia, oral ulcers, hypertransaminasemia,osteopenia or osteoporosis, tooth enamel abnormalities,hemorrhagic syndrome due to vitamin K malabsorption,changes in the female reproductive system (late menar-che, early menopause, recurrent miscarriage, prematurelabour);

• associated diseases (diabetes mellitus type 1, Hashimotothyroiditis, Graves’ disease, selective IgA deficiency,alopecia areata, piebald skin, psoriasis, Addison’s dis-ease, Systemic Lupus Erythematosus, polymyositis,rheumatoid arthritis, cerebellar ataxia, epilepsy withor without cerebral calcifications, peripheral neuropa-thy, autoimmune hepatitis, primary biliary cirrhosis,idiopathic dilated cardiomyopathy, Berger’s disease,Down’s syndrome, Turner’s syndrome, Williams’ syn-drome).

3. 1st degree relatives of coeliac patients (high familiarityof coeliac disease that is present in 4–17% of 1st degreerelatives of coeliac patients, but may also be found in highproportions in 2nd degree relatives).A major role in the diagnostic process of coeliac disease

is played by serology, which allows identification within theat-risk groups of the subjects who should undergo intestinalbiopsy. While making it clear that no positive antibody testallows a diagnosis of coeliac disease without the necessaryconfirmation provided by an intestinal biopsy, some of theantibody markers show such a high diagnostic accuracy (withlevels of sensitivity and specificity >95%) that they are highlypredictive of coeliac disease.

4.2. Antibody markers

• IgA class antitransglutaminase antibodies (tTGA) are thetests with the highest sensitivity for coeliac disease (98%)with specificity estimated at around 90%. High titres ofIgA class tTGA (>5 times the cut-off) are almost always

V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395 S387

the expression of coeliac disease, while false positives(about 10%) almost always present medium-low titres (<2times the cut-off)

• IgA class antiendomysial antibodies (EMA), while havinga lower sensitivity compared to IgA class tTGA (90%vs. 98%), show an almost absolute specificity for coeliacdisease. Antibody titres from >1:40 correlate with agreater severity of intestinal lesions, and low titre positives(1:5) are often an expression of infiltrative lesions of theintestinal mucosa (type 1), suggestive of coeliac disease.

• IgA class antigliadin antibodies (AGA) are now an obsoletetest with levels of sensitivity and specificity significantlylower than tTGA and EMA, and the search for theirpresence is useful only in early childhood (children aged<2 years); because they are the first antibodies to appear,they show a higher sensitivity than other tests in this agegroup. Positivity for IgA AGA associated with negativityfor EMA and tTGA is almost never an expression ofcoeliac disease in adults and in children aged >2 years.

• With regard to the IgG class of antibodies, their use shouldbe restricted to patients with selective IgA deficiency,because only in this subgroup of patients is the responseindicative of coeliac disease. The recommended test is thedetection of tTGA (+ AGA in children aged <2years). Thenegativity of IgA tTGA with very low values (<0.1 AU)always suggests the presence of a selective IgA deficiencyand indicates a search for IgG class tTGA.

4.3. Indications for biopsy in suspected coeliac disease

Intestinal biopsies taken from the first and second duodenalportion remain an essential means of confirming the diagnosisof coeliac disease. To retain its diagnostic validity, it isfundamental for the patient to be on a normal diet containinggluten at the time of the biopsy (often due to incorrectinformation, patients may have already been on a gluten-freediet for some time when they undergo the biopsy).

4.4. Who should undergo an intestinal biopsy?

• Subjects with positive serology characterized by the pres-ence of IgA class antitransglutaminase and antiendomysialantibodies, and children younger than 2 years with isolatedIgA AGA positivity. In some cases, the detection of verylow antibody titres, particularly for IgA tTGA (test with10% false positives), in the absence of EMA, suggestsmonitoring the patient for some time and re-testing beforeproceeding with an endoscopy investigation.

• Subjects with deficiency of IgA positive for IgG tTGA (andeven children aged <2 years with positivity for AGA IgGwith or without IgG tTGA) should also undergo intestinalbiopsy.

• Subjects in whom coeliac disease is strongly suspected,in whom a severe malabsorption syndrome is present,irrespective of antibody test results (in practice it should beperformed even if all the antibodies are negative), preciselybecause in highly symptomatic subjects it is possible to

find that they are serologically negative for coeliac disease.• According to NIH guidelines [7], a duodenal biopsy is not

essential in patients with dermatitis herpetiformis if thediagnosis is supported by immunofluorescence detectionof granular deposits of IgA in the dermis. In these cases,in fact, the gluten-dependent intestinal damage is alwayspresent in a more or less severe form and gluten-free dietwill lead to resolution of the skin lesions.When the results of the intestinal biopsy and serological

tests are consistent, the clinician is able to make the diagnosisof coeliac disease. The diagnosis is confirmed with theresolution of the clinical symptoms and negative serologytests after a reasonable period of strict gluten-free diet(usually 12 months). Therefore, provided that the clinicalsituation has improved and serological tests have becomenegative as a result of following this diet, an intestinal biopsyafter gluten withdrawal is no longer considered essential forthe definitive diagnosis of coeliac disease, not only in children[6], but also in adults [7].

4.5. Genetic testing

Coeliac disease is closely associated with histocompatibil-ity antigens (HLA) DQ2 and DQ8. Practically all patientswith coeliac disease are positive for one or both of theseHLAs or for a fraction of the heterodimer, but genetic testingis never diagnostically significant since at least 30% of thegeneral population present the same HLAs as coeliac patients.1. When the genetic test should be performed:

• In cases where there is a discrepancy between serologyand histology.

• In 1st degree relatives to assess the genetic predisposi-tion to coeliac disease.

2. Significance of genetic testing:• The main clinical significance of genetic testing is to

exclude a diagnosis of coeliac disease in the absenceof HLA-DQ2 (and its fractions) and -DQ8 in cases ofdiagnostic doubt.

• Exclusion of predisposition to coeliac disease in familymembers of coeliac patients in the absence of HLA-DQ2 (and fractions) and -DQ8.

4.6. Clinical notes

A close collaboration between pathologist and clinicianis essential in order to address the concerns relating tothe diagnosis of coeliac disease especially in cases that aredifficult to evaluate.

The information that the clinician should provide thepathologist may be summarized as follows:• Details of the patient’s diet (normal or gluten-free, in the

second case specifying how long the patient has been on agluten-free diet).

• Level of clinical suspicion: high or moderate based on thesymptoms.

• Whether the patient has a family history of coeliac disease(defining the degree of relationship).

S388 V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395

• Serology with absolute (EMA), high (tTGA) or low(AGA) predictability. Always specify the antibody class,the presence of selective IgA deficiency, and the antibodytitre if available (with its cut-off).

• Genetic test results (if performed in accordance with theindications), especially with regard to DQA1 05, DQB1 02and DQB1 0302.

5. Approach to duodenal biopsy: the method

The biopsies that the pathologist receives nowadays areall performed by endoscopic examination, which, in additionto the duodenum, makes it possible to explore other districtsof the gastro-intestinal tract. Biopsies performed through theuse of the Crosby-Watson capsule by perioral route are nowconsidered outdated and are no longer performed.

Here are some points which require a close workingrelationship between the endoscopist, the endoscopy-roomnurse, the pathology laboratory technician and the pathologist.

5.1. Site of the biopsy

Biopsy by endoscopy is always performed in the secondand third duodenal portion, as the bulb and the proximalduodenum can be a source of incorrect assessments; we rec-ommend at least 4 biopsies, 2 for each of the areas mentionedabove. Performing a biopsy only in the duodenal bulb may bea source of error, or may at least greatly reduce the sensitivityof the examination, and hence is strongly discouraged [8].

5.2. Orientation of the biopsy sample

This is essential for proper histological assessment.

Positioning of biopsies on cellulose acetate filters isadvisable, with benefits for:1. the laboratory staff, since with a simple 90° rotation it is

possible to embed the combined biopsy-filter;2. for subsequent histological evaluation by the pathologist.

The method based on experience acquired at St Mark’sHospital in London ensures histological samples in whichit is possible to analyze the mucosa and, if necessary, thesubmucosa of the removed tissue, thus respecting the normalanatomical relationship between the different layers of theintestinal wall. In particular, after initial experience withfilters that needed to be cut with obvious waste of time andhuman resources, a comprehensive kit has been developed, onwhich three easily-detachable cut filters with a bevelled endshaped like a clarinet mouthpiece are already fixed (Fig. 1).

After the fixing stage, the filter-biopsy combination isprocessed and then embedded. During this last phase thetechnician rotates the filter-biopsy combination 90 degrees inorder to place the samples in their natural position.

After cutting, the biopsies are placed on a slide and, ifnecessary, the position of the bevelled end is marked on thelabel, to indicate the first biopsy.

Fig. 1. Cellulose acetate filter with three easily-detachable cut filters.

When properly carried out, this method is of great benefitto the pathologist, but also to the technician who during theembedding phase does not have to search for the individualbiopsies, which are sometimes fragmented and have noguiding landmark.

The use of cellulose acetate filters allows perfect adhesionof the biopsies, avoiding their dispersion in the fixationmedium. These filters also do not react chemically withthe fixatives and reagents used during the processing of thesample; during the cutting phase they do not offer resistanceto the blade and, unlike tissue paper, they do not fray.

This method, which can be applied on all segments ofthe gastro-intestinal tract, has led to considerable diagnosticand economic benefits by reducing the time, the number ofembeddings, and consequently the number of sections to becut and colored.

For its obvious advantages, the use of the kit is stronglyrecommended.

5.3. Stains

It is sufficient to stain with Haematoxilin & Eosin,possibly associated with an Alcian Blue-PAS, to assess all thenecessary morphological elements (one or two sections can beused for immunohistochemical assessment if necessary).

6. Histopathological aspects of normal and pathologicalduodenal mucosa

6.1. Normal intestinal mucosa

Villi: Digitiform appearance with the ratio between theheight of the villi and of the crypts always in favor of thevillus (3 : 1 or more).

Enterocytes: Normal height with 29–34 μm clear brush-border.

Intra-epithelial lymphocytic infiltrate: The number of intra-epithelial lymphocytes (T lymphocytes) is subject to individ-ual variability. The majority of normal subjects have less than20 lymphocytes per 100 epithelial cells; based on the experi-ences of Hayat [9] and Veress [10], a count of IEL between25 and 29/100 epithelial cells is considered borderline andpathological over 30/100 epithelial cells.

V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395 S389

Fig. 2. Normal duodenal mucosa: villus/crypt ratio 3 :1. Intraepithelial lymphocytes within the normal range.

The intra-epithelial lymphocyte count is very importantand should always be done, especially in the initial lesions,following the indications given below:• always count the T lymphocytes with the help of immuno-

histochemical investigations using anti-CD3 antibodies;• evaluate the biopsies perfectly oriented with a precise

alignment of the surface-coating epithelial cells;• do the count both in the apical portions and along

the edges of the villi; it is important to have accurateand reproducible fields. Counts done only on the apicalportions have proved unreliable (Fig. 2A–D).Glandular crypts: The crypts basically have the task of

performing a regenerative function, which means it is possibleto find evidence of mitosis; the normal range is usually one

mitosis per crypt. Alongside the epithelial cells are endocrinecells, goblet cells and Paneth cells, but these have no value asregards the diagnosis of coeliac disease.

Lamina propria: Plasma cells, eosinophils, histiocytes,mast cells and lymphocytes are normally found in the laminapropria. Neutrophils are generally absent, except in casesof active duodenitis with possible gastric metaplasia closelyrelated to Helicobacter pylori infection.

The cellular component mainly consists of plasma cells andlymphocytes, the latter sometimes in the form of lymphoidaggregates and eosinophilic granulocytes whose value mustnever be greater than 60 for 10 fields of vision examined at40×.

S390 V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395

7. Pathological intestinal mucosa

7.1. Basic lesions

The histological diagnosis of CD consists of an integratedassessment of the following elementary lesions:• Increased intraepithelial T lymphocytes: a value between

25 and 29 IEL/100 enterocytes is considered border-linevalue; >30 IEL/100 enterocytes represents a pathological“lymphocytosis”.

• Decreased enterocyte height, flattening of enterocytes, in-tracytoplasmic vacuolation as well as reduction or absenceof brush-border are possible but not specific.

• Crypt hyperplasia: extension of the regenerative epithelialcrypts associated with changes in the presence of morethan 1 mitosis per crypt.

• Villous atrophy: decrease in villous height, alteration ofnormal crypt/villous ratio (3:1) until total disappearanceof villi. This assessment requires proper orientation of thebiopsies.None of these elementary lesions of CD is exclusive; the

diagnosis of CD is based on the identification of histologicallesions accompanied by clinical and serological consistentdata. On the basis of the presence of one or more of theseelementary lesions the histopathology of CD is subdividedinto different diagnostic categories according to the Marshclassification [2].

Fig. 3. Type 1/2 infiltrative lesion according to Marsh-Oberhuber. Grade A new classification.

Marsh classification

Type 1 or infiltrative lesion1. Villi architecturally within normal morphological limits

(normal villa/crypt ratio 3:1);2. Increased number of intraepithelial lymphocytes (greater

than 25–30 per 100 epithelial cells) (Fig. 3A–D).

Type 2 or hyperplastic lesion1. Villi architecturally within normal morphological limits

(like type 1);2. Increased number of intraepithelial lymphocytes (greater

than 25–30 per 100 epithelial cells) (like type 1);3. Hyperplasia of the glandular elements (regenerative aspect

of the glandular elements highlighted by the reducedmuciferous activity and increased number of mitoses).

Type 3 or destructive lesion1. Varying degrees of villous atrophy associated with hyper-

plasia of glandular crypts;2. Reduced surface enterocyte height, with irregular brush-

border and sometimes cytoplasmic vacuoles;3. Increased number of intraepithelial lymphocytes (like type

1 and 2 lesions).

The combination of the three factors described above isconsistent with a diagnosis of coeliac disease or gluten-sensitive enteropathy, which should be considered as syn-onyms. These three patterns, albeit schematic, represent thehistological lesions seen in coeliac disease and it is important

V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395 S391

Fig. 4. From Marsh [2], amended.

to consider them as dynamic and progressive both in one di-rection and the other and not static, since they depend on howmuch gluten the patient has been exposed to and for how long.

Figure 4 summarizes the above description.This classification is universally recognized for the diag-

nosis of coeliac disease, and extensively validated; the onlypoint worthy of observation and critical analysis is that thecases with mild, moderate or severe atrophy (total villousflattening) are all grouped together in a single category: thetype 3 lesion.

Fig. 5. Type 3a–3b lesion according to Marsh-Oberhuber. Grade B1 new classification.

Fig. 6. Type 3c lesion according to Marsh-Oberhuber. Grade B2 new classification.

An amendment to this classification has been proposed byOberhuber et al. [11], who divided the Marsh type lesion 3into three subgroups.3a mild villous atrophy and pathological increase of intraep-

ithelial lymphocytes.3b moderate villous atrophy and pathological increase of

intraepithelial lymphocytes (Fig. 5A, B).3c total villous atrophy and pathological increase of intraep-

ithelial lymphocytes (Fig. 6A, B).

S392 V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395

Without prejudice to all the other morphological criteriadescribed above, this classification provides a better descrip-tion of the spectrum of lesions that may occur both in coeliacdisease patients on a normal diet and in those on a gluten-freediet.

Along the same lines, and in an attempt to simplify andstandardize the work of pathologists and facilitate the rela-tionship between pathologists and clinicians, a new versionof the histological classification has recently been proposedby Corazza and Villanacci [12,13]; in particular, the lesionsthat characterize coeliac disease have been divided into twocategories: Non-atrophic (grade A) and atrophic (grade B).

Grade A lesions are characterized by a pathologicalincrease in intraepithelial lymphocytes, best recognized by theuse of immunohistochemical techniques.

Grade B lesions are further subdivided into:Grade B1 in which the villus/crypt ratio is less than 3:1,

with villi still identifiable, andGrade B2 in which the villi are no longer identifiable:

8. Diagnostic criteria

The diagnosis represents the culmination of what isdescribed above and must comprise all the morphologicalrequirements so as to allow a direct, clear and simpleunderstanding of the morphological situation of the duodenalmucosa by the clinician.

A two-step proposal is presented below:

8.1. Assessment of the morphological pattern dividedaccording to description and diagnosis

The description should report, in sequence, the same mor-phological elements listed above, namely: villous trophism,number of intraepithelial lymphocytes, features and glandularstructures of the lamina propria, concluding with compatibil-ity or non-compatibility with the pattern of a coeliac patientbased on complete clinical and laboratory data.

The final diagnosis in the case of an atrophic lesion,culminating in a “consistent” with a CD with atrophiclesions (type 3a, 3b or 3c); in the case of non-atrophiclesions culminating in finding attributable to intraepitheliallymphocytosis, stressing that these injuries are “suggestivefor” but not exclusive of CD and should therefore necessarilybe placed in the right clinical setting and supported by aserological confirmation.

As a brief addition to the above we propose that theterm sub-atrophy, in itself unclear and misleading, should nolonger be used. Instead, it is better to specify whether thevilli are normal or atrophic, and in the latter case, the degreeof atrophy, from mild to moderate to severe. In the event ofsevere atrophy it is possible to use the term total or severeatrophy. Scores should not be attributed to the individualmorphological elements as they are too subjective and of littleor no use for the final diagnosis.

8.2. The histology report: checklist

Name Sex M/F Date of Birth / /

1st Biopsy No. Control No.

No. of biopsies Oriented Non-Oriented

Villi: normal atrophy mild moderate severe

Villus/crypt ratio: normal [3 :1] altered

Intraepithelial lymphocytes: normal increased

(normal: less than 25–30 lymphocytes/100 epithelia;increased: more than 25–30 lymphocytes/100 epithelia)

Evaluation with CD3

Glands: normal hyperplastic

Lamina propria

DIAGNOSISOberhuber-Marsh: New grading system:

Type 1 Grade A

Type 2

Type 3a Grade B1

Type 3b

Type 3c Grade B2

Comments:

9. Immunohistochemistry

One of the key points in the diagnosis of coeliac diseaseis the number of intraepithelial lymphocytes, which are CD3and CD8 positive T lymphocytes; in pathological conditions,their number should be more than 25 lymphocytes per 100epithelial cells (border-line value 25–30).

The definition of a precise cut-off between normal, ab-normal and border-line pathological lesions is of particularimportance given the increase in coeliac disease diagnosed inthe early/subclinical stage.

The counts can be performed reasonably well on thenormal and irreplaceable hematoxylin-eosin but we suggest,especially in the initial forms, that an immunohistochemicalassessment should always be carried out with monoclonal

V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395 S393

CD3 which often allows for a more accurate display oflymphocytes, following a series of procedures (see the sectionon intraepithelial lymphocyte infiltration). Evaluation withCD8 may also help, and is particularly useful in cases ofelderly subjects where it is possible to find refractory formswhich do not respond to diet, regarded by many as pre-lymphomatous and in which the expression of CD8 may benegative with respect to the “norm” [14].

As frozen material is available, immunohistochemical typ-ing for the gamma-delta receptor of T lymphocytes can be car-ried out; in normal conditions this receptor is not expressed bymore than 2–3% of T lymphocytes while in coeliac disease itmay reach 20–30% – a particularly useful marker in initial le-sions. This assessment is, however, based on the use of frozenmaterial and is not therefore recommended in routine practice.

10. Differential diagnosis

The above summarizes the morphological lesions whichmay occur with coeliac disease and where the pathologistclearly has a key role, if only to exclude the possibility ofclinically suspected malabsorption which may also be:• Parasitic (Giardia lamblia, Cryptosporidium, Microsporid-

ium)• Infectious (Whipple’s disease)• Viral (cytomegalovirus, herpes virus)• Idiopathic (Crohn’s disease)• Neoplastic.

The most important problem today in the diagnosis ofcoeliac disease is represented by early lesions, i.e. normal villiwith a pathologic increase in intraepithelial T lymphocytes.This issue is appropriately dealt with in the excellent reviewby Brown et al. [15], summarized in Table 1.

Table 1 shows that in addition to coeliac disease, thereare a number of pathological conditions that have the samemorphological aspect as coeliac disease in its early stages, i.e.normal villous architecture but with a pathological increase of

Table 1Causes of proximal small intestinal intraepithelial lymphocytosis with nor-mal villus architecture*

• Gluten sensitivity†

• Non-gluten food hypersensitivity (e.g., cereals, cow’s milk, soy products,fish, rice, and chicken)

• Infections (e.g., viral enteritis, Giardia, Cryptosporidia, Helicobacterpylori)†

• Bacterial overgrowth• Drugs (e.g., NSAIDs)†

• Immune dysregulation (e.g., Hashimoto thyroiditis, rheumatoid arthritis,SLE, autoimmune enteropathy)†

• Immune deficiency (e.g., IgA deficiency, CVID)• Infammatory bowel disease• Lymphocytic and collagenous colitis

*NSAIDs, non-steroidal anti-inflammatory drugs; SLE, systemic lupus ery-thematosus; IgA, immunoglobulin A; CVID, common variable immunodefi-ciency.†Most common associations.

IELs (>25–30/100 epithelial cells) (lesion type 1 according toMarsh, Grade A according to the new proposed classification).These conditions include hypersensitivity to other foods(milk, cereals, soybeans, fish, etc.), infections (Helicobacterpylori, Giardia, etc.), the use of drugs, immunodeficienciesand immunodysregulation (Hashimoto thyroiditis, systemiclupus erythematosus, rheumatoid arthritis, etc.) and, not least,chronic idiopathic inflammatory bowel colitis or colitis witha different etiology, such as lymphocytic and collagenouscolitis.

The question that we must therefore ask is: How can wediscriminate between different pathological conditions wherethe morphology is essentially superimposable? A proper clin-ical evaluation based on histological and laboratory data iscrucial. We must not forget that a diagnosis of coeliac diseaseis a “marker” which remains throughout life with obvioustherapeutic and behavioral relapses. The table below helpsunderstand how important the need for collaboration betweenthe pathologist and endoscopist is in the detection of otherconditions, such as infection with Giardia lamblia or otherparasites, the possibility of presentations of immunodeficien-cies morphologically superimposable on coeliac disease andnot least the localization of Crohn’s disease or particularforms of enteritis within the sphere of untreatable diarrhea,such as autoimmune enteritis, tufting enteropathy, a diseasecaused by atrophy of the microvilli, and cases of graft-versus-host disease, all conditions in which the morphologicalelement is fundamental.

Three conditions, however, deserve special mention:• Forms of so-called “autoimmune enteritis” possible in

children with immunological deficiency (common vari-able immunodeficiency, X linked agammaglobulinemia) inwhich the intestinal biopsy may be fully comparable to thepattern of coeliac disease [16].

• Damage by drugs: there is increasing evidence in theliterature showing that the use of drugs, especially non-steroidal anti-inflammatory drugs (NSAID), are capableof causing morphological alterations identical to those ofcoeliac disease, and it is therefore important to keep thispossibility in mind in cases of elderly patients, especiallywhen the serological markers are all negative [15].

• The possibility that concurrent infection with Helicobacterpylori in the stomach can produce a morphological patternvery similar to that of initial lesions of coeliac disease asrecently reported [17].

11. Complications that can be confirmed histologically

Unlike what occurs in children, there is considerableevidence that coeliac disease in adults, especially if diagnosedlate and even more so if not dealt with by a timely andrigorous gluten-free diet, is burdened by a higher mortalityrate than in the general population.

The removal of gluten from the diet therefore determinesnot only an improvement of the histological and clinicalaspects, but also prevents the complications which must

S394 V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395

always be suspected if an adult patient continues to be unwell,despite the diet.

These complications are due to:• Collagenous sprue: The patient does not respond to diet

and histology shows fibrous tissue in the intestinal wallat the level of the superficial subepithelial layer. Thismorphological pattern is very similar to the conditionof collagenous colitis described in the colon, where thethickness of the connective band best highlighted withMasson’s trichrome is more than 15 millimicrons, althoughthis is a very rare event is described in the literature.

• Refractory sprue: This condition reproduces the sameclinical picture as collagenous sprue but can be identifiedby immunohistochemical staining, demonstrating that Tlymphocytes, which in normal conditions express CD3 andCD8, in this case present only the expression of CD3 andnot of CD8 [14].

• Ulcerative jejunoileitis: Presence of extensive ulceration ofthe intestinal mucosa, often related to refractory sprue.

• Lymphoma: This is the most serious complication andshould always be suspected when histology shows a preva-lence of atypical monomorphous lymphocytic elements. Inthese cases it is useful to carry out immunophenotyping ofthe lymphoid population, which is almost always type T[18–20].

12. Summary

12.1. What are the “certainties” in the diagnosis of coeliacdisease?

An obvious prerequisite for certainty in the diagnosis ofcoeliac disease (CD) from the anatomo-pathological point ofview is the observation and respect of a number of key points:

1. Close collaboration between clinicians, laboratory techni-cians and endoscopists.

2. An adequate number of biopsies (at least 4, 2 in the distaland 2 in the proximal duodenum).

3. Correct orientation of the biopsy (the use of pre-cutcellulose acetate filters).

4. Sufficient clinical information.5. Excellent quality of the biopsy samples.

– With these details it is clear that “certainty” in thediagnosis of CD is only possible if the villous atrophyis associated with a pathological increase in the numberof intraepithelial lymphocytes (value exceeding 25–30/100epithelial cells). In this situation, by applying the threeclassifications now known and validated (Marsh, Marsh-Oberhuber and Corazza-Villanacci) there is no problem inthe diagnosis and comparison with clinical and laboratorydata.

– The degree of atrophy should be certain and not merelypseudo-atrophy due to incorrect orientation and cuttingof the villi. Assessment of the number of intraepithelial

lymphocytes is useful in these cases and “must” always bepathological (>25–30/100 epithelial cells), best evaluatedboth with H&E staining and with immunohistochemistrystaining for CD3.

– Attention should be paid to biopsies taken from theduodenal bulb, where the presence of Brunner glands canlead to a false diagnosis of atrophy; biopsies of the bulbshould always be compared with those taken from thedistal portions, especially in the early stages of the disease,which has a progression of the pathological process in acranio-caudal direction.

– If there are varying degrees of atrophy, these should all bedescribed, not just the most severe degree. An assessmentof compatibility should only be included in the descriptionof the case, while the term “coeliac disease” should beavoided in the final diagnosis, which should be limited to adescription, giving the clinician a precise “snapshot” of thestate of the duodenal mucosa. The final diagnosis of CDshould be made solely and exclusively by the pediatric oradult clinical gastroenterologist.

12.2. What are the “doubts” in the diagnosis of coeliacdisease?

The points that cause doubt and require caution on thepart of the pathologist in the diagnosis of CD are clearlyrepresented by the cases in which there are initial lesions(Marsh 1–2, and Grade A in accordance with the newproposed classification); in these cases it is necessary to:

1. Carefully assess the orientation of the biopsies.2. Consider whether the villus/crypt ratio of at least 3:1 is

respected.3. Carefully count the number of lymphocytes in the surface

coating epithelium.4. Always carry out additional immunohistochemical evalua-

tion with CD3.5. Compare the clinical and laboratory data.

– The two key elements that must be assessed are the absenceof atrophy and the increase in the number of intra-epitheliallymphocytes; it is therefore crucial to always associateimmunohistochemical evaluation with CD3. The presenceor absence of hyperplasia of the glandular elements istotally irrelevant for practical and therapeutic purposes.

– Do not forget that the “slide” is proof of the assessmentby the pathologist and as such can be compared andre-assessed by other colleagues and specialists; it must alsobe strongly emphasized that the histological assessmentmust be conducted solely by the pathologist and not byother “specialists”.

– As with “certain” cases, it is even more important indoubtful cases to merely express an opinion of possiblecompatibility with CD in the description, describing onlythe histological aspects in the final diagnosis.

– Exclude, if possible, a concurrent infection due to Heli-cobacter pylori (it is advisable to always take biopsies of

V. Villanacci et al. / Digestive and Liver Disease 43S (2011) S385–S395 S395

the antral and oxyntic gastric mucosa), immunodeficien-cies, parasitic infections, allergies to other dietary factorsand use of drugs.

– In doubtful situations in which the final pattern is unclear,it is useful to bear in mind the excellent review by Brownet al.: the pathologist must be sure that he is faced with apathological condition unequivocally demonstrated by theincrease in the number of intraepithelial T lymphocytesconfirmed by the evaluation with CD3. The final diagnosiswill be based on comparison of the clinical and laboratorydata.

– In pediatric patients in the first year of life, the possibilityof intolerance to cow’s milk proteins should not beforgotten; in such cases the eosinophilic granulocyte countmay be useful (pathologic value above 60 for 10 fields ofvision at 40×).

Acknowledgements

We are indebted to the President of the Italian Associationof Celiac Disease Elisabetta Tosi, the President of theFoundation of the Italian Association of Celiac DiseaseAdriano Pucci, and the General Director of the ItalianAssociation of Celiac Disease for the encouragement andsupport to our work.

Conflict of interest

The authors have no conflict of interest to report.

References

[1] Chiarelli S, Villanacci V. Celiachia. Requisiti diagnostici minimi per ladiagnosi istopatologica (GIPAD). Pathologica 1998; 90:809–13.

[2] Marsh MN. Grains of truth: evolutionary changes in small intestinalmucosa in response to environmental antigen challenge. Gut 1990;31:111–4.

[3] Marsh MN. Gluten, major histocompatibility complex, and the smallintestine. Gastroenterology 1992;102:330–54.

[4] Calabrò A, Catassi C, De Vitis I, et al. Documento di inquadramentoper la Diagnosi ed il Monitoraggio della Celiachia e relative Patolo-

gie associate. A cura del Comitato Scientifico Nazionale (CSN) edell’Associazione Italiana Celiachia (AIC) – Ministero della Salute –Gazzetta Ufficiale 7 febbraio 2008 n° 32, S.O.

[5] Tonutti E, Visentini D, Bizzaro N, et al. Linee guida per la diagnosidi laboratorio e istologica della malattia celiaca. Riv Ital Med Lab2005;1:1–35.

[6] Walker-Smith J, Guandalini S, Schmitz J, et al. Revised criteria fordiagnosis of coeliac disease. Report of Working Group of EuropeanSociety of Paediatric Gastroenterology and Nutrition. Arch Dis Child1990;65:909–11.

[7] NIH Consensus Development Conference on Celiac Disease. 28–30June 2004. http://consensus.nih.gov/2004/2004CeliacDisease118html.htm.

[8] Pais WP, Duerksen DR, Pettigrew NM, et al. How many duodenalbiopsy specimens are required to make a diagnosis of celiac disease?Gastrointest Endosc 2008;67:1082–7.

[9] Hayat M, Cairns A, Dixon MF, et al. Quantitation of intraepitheliallymphocytes in human duodenum: what is normal? J Clin Pathol2002;55:393–4.

[10] Veress B, Franzen L, Bodin L, et al. Duodenal Intraepithelial Lympho-cyte Count Revisited Scand J Gastroenterol 2000;39:138–44.

[11] Oberhuber G, Granditsch G, Vogelsang H. The histopathology ofcoeliac disease: time for a standardized report scheme for pathologists.Eur J Gastroenterol Hepatol 1999;1:1185–94.

[12] Corazza GR, Villanacci V. Coeliac Disease. Some considerations onthe histological classification. J Clin Pathol 2005;58:573–4.

[13] Corazza GR, Villanacci V, Zambelli C, et al. Comparison of theinterobserver reproducibility with different histologic criteria used inceliac disease. Clin Gastroenterol Hepatol 2007;5:838–43.

[14] De Mascarel A, Belleannée G, Stanislas S, et al. Mucosal intraepithelialT-lymphocytes in refractory celiac disease: a neoplastic population witha variable CD8 phenotype. Am J Surg Pathol 2008; 32:744–51.

[15] Brown I, Mino-Kenudson M, Deshpande V, et al. Intraepithelial lym-phocytosis in architecturally preserved proximal small intestinal mu-cosa: an increasing diagnostic problem with a wide differential diagno-sis. Arch Pathol Lab Med 2006;130:1020–5.

[16] Washington K, Stenzel TT, Buckley RH, et al. Gastrointestinal pathol-ogy in patients with common variable immunodeficiency and X-linkedagammaglobulinemia. Am J Surg Pathol 1996;20:1240–52.

[17] Memeo L, Jhang J, Hibshoosh H, et al. Duodenal intraepithelial lym-phocitosis with normal villous architecture: common occurrence in H.pylori gastritis. Mod Pathol 2005;18:1134–44.

[18] Du MQ, Isaacson PG. First steps in unraveling the genotype ofenteropathy-type T-cell lymphoma. Am J Pathol 2002;161:1527–9.

[19] Dickson BC, Streutker CJ, Chetty R. Coeliac disease: an update forpathologists J Clin Pathol 2006;10:1008–16.

[20] Serra S, Jani PA. An approach to duodenal biopsies. J Clin Pathol2006;59:1133-11.