section 3 chapter 18 disease and ulcerative colitis and ulcerative colitis k. geboes goal to review...

IntroductionUlcerative colitis (UC) and Crohn’s disease (CD) arechronic inflammatory bowel diseases of unknown cause.Clinically, both conditions usually begin gradually, butthey can start abruptly and sometimes even present asfulminant disease.The clinical course is characterized byexacerbations and spontaneous or drug-induced remis-sions. UC primarily affects the mucosa of the largebowel, while CD is a transmural disease that can affectthe whole gastrointestinal tract.The various clinical pat-terns are reflected in the microscopic features observedin biopsies obtained from surgical specimens or duringendoscopy. Endoscopic mucosal biopsies will not showall the characteristic features of CD. Review of biopsies,in combination with clinical, laboratory, radiographicand endoscopic observations, is used for the diagnosis ofUC and CD and the differentiation from other condi-tions. Such differentiation is important because a precisediagnosis is essential for appropriate treatment.1–3 Acorrect diagnosis is possible in the large majority ofpatients; still, in some patients with acute onset or fulmi-nant disease, a precise diagnosis may be difficult to reach,resulting in a temporary diagnosis of indeterminatecolitis (IC). Biopsies also allow assessment of diseaseactivity and identification of pre-cancerous lesions andcancer. In clinical practice, often only rectal biopsies areobtained for diagnosis. However, in contrast with UC,the rectum is not always involved in CD and lesions in CD frequently occur in a background of normalmucosa.Therefore it is more appropriate to take multipleendoscopic biopsies in different segments of the colon(and ileum) during the initial work-up of a patient pre-senting with inflammatory diarrhea or during follow-upand screening for cancer. Multiple biopsies permit an in-depth analysis of the distribution of inflammation andare essential for the recognition of dysplasia. It also

255

SECTION 3

CHAPTER

1 8Histopathology of Crohn’sDisease and Ulcerative ColitisK. GEBOES

G o a lTo review the important histologic features required for the

diagnosis, assessment of disease activity and early

detection of malignancy. The variability of features with

time and treatment and difficult differential diagnostic

problems will be discussed.

K e y p o i n t s• Histopathology can help to solve many diagnostic

problems, especially when multiple biopsies of the colon

and ileum are available.

• Exacerbations and remissions are reflected by mucosal

inflammation and activity of variable intensity, including

healing.

• Multiple biopsies are also indicated for the early

detection of pre-cancerous lesions and cancer in

ulcerative colitis and Crohn’s disease.

K e y r e f e r e n c e s• Schumacher G, Kollberg B, Sandstedt B. A prospective

study of first attacks of inflammatory bowel disease and

infectious colitis. Histologic course during the 1st year

after presentation. Scand J Gastroenterol 1994;

29:318–332.

• Jenkins D, Balsitis M, Gallivan S et al. Guidelines for the

initial biopsy diagnosis of suspected chronic idiopathic

inflammatory bowel disease. The British Society of

Gastroenterology Initiative. J Clin Pathol 1997; 50:

93–105.

• Shepherd NA. Pathological mimics of chronic

inflammatory bowel disease. J Clin Pathol 1991;

44:726–733

S E C T I O N 3

IBD4E-18(255-276) 03/04/2003 10:32 AM Page 255

increases diagnostic accuracy by 5–41%. Carcinoma can occur during the course of the disease as a late complication.

The Normal IntestineThe gastrointestinal (GI) tract is a hollow tube consistingof three layers: mucosa, submucosa, muscularis propriaand loose areolar tissue covered by mesothelium wherethe tract borders on the body cavity (serosa). In thecolon, the surface of the mucosa is flat and its architec-ture is characteristic with crypts as straight tubes, in par-allel alignment. The crypt base rests upon a layer ofsmooth muscle cells, the muscularis mucosae, whichseparates the mucosa from the submucosal connectivetissue.The distance between the crypts and the internaldiameter of the crypts is constant.The crypt architectureis maintained throughout the colon, except in the pres-ence of lymphoid collections, in zones of transition tosmall intestinal mucosa (ileocecal valve) or to squamousepithelium (the anorectum) and in normally occurringgrooves in the surface. In the small intestine, the surfaceis irregular due to the presence of finger-like villi ofuniform size and shape. In the ileum, the villi are tallerand the crypts less deep compared with the jejunum.Surface, villi and crypts are covered with a single-celllayer of columnar epithelium, separated from the con-nective tissue lamina propria by the basement membranecomplex. The epithelial cells form a heterogeneousgroup with mature cells (absorptive enterocytes andgoblet cells) lining the surface, villi and upper part of thecrypts, immature crypt cells including stem cells in thebase of the crypts and endocrine cells. Paneth cells arepresent in the base of the crypts in the small intestineand caecum. Epithelial cells have a barrier and secretoryfunction and are involved in the immune response(secretion of SIgA). In the colon they do not constitu-tively express the major histocompatibility class II anti-gens (MHC). Epithelial cell turnover ranges between 2and 8 days.4–6

The lamina propria extends from the subepithelialbasement membrane complex to the muscularismucosae. It is composed of extracellular matrix, fibro-blasts and various types of leukocytes. The presence ofleukocytes is necessary because the mucosa is continu-ously challenged by luminal antigens, which results in a‘controlled’ or ‘physiologic’ inflammation. Lymphocytesare the largest subtype. According to the location, theyare subdivided into intraepithelial lymphocytes (IEL),lamina propria lymphocytes (LPL) and lymphocytesorganized in follicles in association with epithelial cells(lymphoepithelial complexes). The latter may extendacross the muscularis mucosae (Fig. 18.1).They consti-

tute the mucosa associated lymphoid tissue (MALT)where the immune response is induced. GALT (gutassociated lymphoid tissue) is especially prominent inthe appendix and terminal ileum where it forms thePeyer’s patches along the anti-mesenteric border. Fourcompartments are distinguished in Peyer’s patches: thefollicle, dome, follicle-associated epithelium and inter-follicular regions. The specialized follicle-associatedepithelium, overlying lymphoid aggregates is distinctfrom the surrounding villous epithelial surfaces. Itcharacteristically has fewer goblet cells and containsmembranous cells or M cells.The M cell is a specializedepithelial cell that transports luminal antigens, thusallowing access to immunocompetent cells. It plays akey role in mucosal-based immunity and antigen toler-ance. IELs are present in-between the epithelial cellslining the surface. They are mainly CD8+ T lympho-cytes. LPLs are B cells (15–40%) and T cells (40–90%)and a limited number of natural killer cells. B cells aremainly present as plasma cells with a predominance ofIgA over IgM and IgG containing cells.The majority ofthe T cells are CD4+ cells (65%). A second importantsubtype of leukocytes in the lamina propria are the cellsof the monocyte/macrophage lineage. In the colon theyare diffusely present in the subepithelial part of thelamina propria. They are a heterogeneous group com-posed of cells having more phagocytic properties andcells equipped for antigen presentation. They appearoften as foamy histiocytes.7,8 Other myeloid cells thatnormally reside in the lamina propria are eosinophilsand mast cells. Neutrophils should not be present.Fibroblasts are located randomly, distributed throughoutthe lamina propria and in the most superficial portionand the pericryptal fibroblast sheet, tightly apposed tothe subepithelial basement membrane complex.4,5

C H A P T E R 1 8 . H I S T O P A T H O L O G Y O F C R O H N ’ S D I S E A S E A N D U L C E R A T I V E C O L I T I SS

EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

256

Fig. 18.1 Normal colon. Mucosal lymphoid follicle extendingtowards the submucosa. The germinal centre is stimulated (H&E × 25).

IBD4E-18(255-276) 03/04/2003 10:32 AM 56

An adequate immune response implies migration ofimmune cells, cell recognition and interaction of cellsrequiring adhesion and deadhesion. Proteins involved inleukocyte recruitment and migration belonging to thelarge family of ‘adhesion molecules’ are therefore vari-ably expressed on endothelial cells and leukocytes.Molecules involved in cell interactions can also beidentified. Examples of these are proteins such as CD40,CD40 ligand (CD40L), CD28 and CD80 and CD86.These molecules act as ligand-receptor pairs. They canbe expressed on T cells (CD40L) and monocytes actingas antigen presenting cells (CD40) and by doing so,induce cell activation.4,5

In the submucosa, the ganglionated plexuses ofMeissner and Henle of the ENS are found. Ganglia arecomposed of enteroglial cells and neurones.The muscu-laris propria is composed of two layers of smoothmuscle separated by a thin layer of connective tissue inwhich the ganglionated myenteric plexus (Auerbach’s)can be observed.

Ulcerative Colitis

Gross features

UC starts from the rectum, spreads proximally and incontinuity, involving a variable length of the colon.Pancolitis normally stops abruptly at the ileocecal valve,but in some cases a limited distal ileitis, called backwashileitis, is observed.9 These ileal lesions are in continuitywith colonic lesions and characterized microscopicallyby diffuse inflammatory lesions with regular shorteningof the villi. In patients with limited UC, the transitionfrom diseased to normal mucosa is usually gradual andonly occasionally abrupt. Atypical presentations can beobserved. Some patients have left-sided involvement ofthe colon and cecal (cecal patch) or appendicealinvolvement.10 The affected sites are separated bynormal mucosa. Diffuse duodenitis and extensiveinvolvement of the upper small bowel can occur inseverely ill UC patients.11

The gross appearance varies with the activity of thedisease. Lesions are usually limited to the mucosa.Stenoses, fistulas and significant thickening of the wallare rare. In the acute form, the mucosal surface is wetand glaring from blood and mucus with numerouspetechial hemorrhages. Ulcers of various sizes canappear. They may be small, rounded and superficial ormore irregular and somewhat geographic in configura-tion. Fissuring ulcers are not seen, except in some casesof toxic megacolon. Ulcers can become more extensiveand undermine the mucosa so that mucosal bridgeswith an underlying inflammatory infiltrate develop.

Multiple confluent ulcerations provoke denudation ofthe mucosa. The serosal side may appear congested,often to a larger extent than the mucosal lesions.

Following healing of mucosal ulcers, elevated sessilereddish nodules – pseudopolyps – appear in an other-wise flat surface. They are typically small and multiple,but may have a filiform configuration (Fig. 18.2).Thereis some variability in the prevalence of the lesionsdepending upon the colonic segment involved. Pseudo-polyps are common in the sigmoid and descendingcolon and rare in the rectum where mucosal friabilitypredominates. In the more advanced stages, the entirebowel becomes fibrotic, narrowed and shortened.During remission, the mucosa may become normalagain. Healing often occurs in an irregular way leadingto a discontinuous, heterogeneous aspect of the mucosa,which can be confused with CD.Treatment may increasethe heterogeneous nature of the lesions. In children,lesions may initially be heterogeneous.Topical treatmentof the rectum can induce complete rectal healing. UCwith rectal sparing, although rare, must therefore not beconfused with CD.12 In severe, active forms, the entirecolon, or a segment, may become dilated (toxic mega-colon). Inflammation may extend towards the submu-cosa. In such cases, the wall is thin and perforation canoccur.

Microscopic features of ulcerative colitis

Diagnostic featuresThe microscopic pattern of UC is characterized by aninflammatory reaction with special distribution andstructural abnormalities of the mucosa (Table 18.1).Inflammation is characterized by increased intensity of

U L C E R A T I V E C O L I T I S

257

SECTION 3

Fig. 18.2 Ulcerative colitis. Inflammatory pseudopolyps inulcerative colitis are typically multiple. They can have a filiformconfiguration (H&E × 1).

IBD4E-18(255-276) 03/04/2003 10:32 AM Page 257

the lamina propria cellular infiltrate with alterations ofthe composition and changes in distribution. Thenormal lamina propria infiltrate is located in the upperpart of the mucosa and this pattern persists in infectiouscolitis. In UC, the infiltrate is more extensive andextends diffusely towards the deeper part (transmucosal)(Fig. 18.3). Accumulation of plasma cells near themucosal base, in-between the crypt base and the muscu-laris mucosae (basal plasmacytosis), is common. Focal ordiffuse basal plasmacytosis (combined with crypt distor-tion) is a strong predictor for the diagnosis of chronicidiopathic inflammatory bowel disease (IBD) and occursin over 70% of the patients.2

The presence of neutrophils, indicating a change inthe composition of the inflammatory infiltrate, is anotherimportant feature. When combined with unequivocalepithelial cell damage, it indicates disease activity.13 Thisfeature can be assessed with good reproducibility.Neutrophils within epithelial structures, such as the crypt wall (cryptitis), or the crypt lumen and wall (cryptabscesses) or in association with crypt damage (cryptdestruction) are helpful for the diagnosis, but the predic-tive value of these features is limited. Cryptitis and cryptabscesses can indeed also be seen in infectious colitis,Crohn’s colitis and diversion colitis. In UC however, theyare generally more common, being present in 41% of thecases while in CD they are found in 19% of the cases.2

Eosinophils may be so numerous as to suggest eosi-nophilic colitis, particularly in chronic disease or whenthe disease is quiescent. This is partly explained by the reduction of other inflammatory cells, induced bymedical treatment.

Structural changes include an irregular surface or avilliform surface and a disturbed crypt architecture.Thiscan be assessed with good reproducibility. Overall, anirregular surface is present in approximately 60% ofcases with UC.14 Crypt alterations are more commonand more widespread; they are observed in 57–100% of

cases. Low-power examination is important for the dif-ferential diagnosis with CD where similar architecturalalterations are less common (27–71%) and less diffuse.2

They can however, also be seen after surgery, in drug-induced lesions and in rare forms of chronic infectiouscolitis such as chronic Shigella dysentery. Crypt distor-tion includes shortened crypts that become widelyseparated from the underlying muscularis mucosae,crypt drop-out and especially prominent crypt budding(branching crypts, bifid crypts) (Fig. 18.4). Mucosalatrophy is a combination of crypt drop-out and short-ening of crypts. These alterations can be evaluated onperpendicular sections (preferentially), but also on trans-verse or tangential sections. For the latter, one can relyon differences in intercryptal distance and variability ofthe internal crypt diameter (Fig. 18.5).15


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

258

Table 18.1 Microscopic features suggestive for adiagnosis of ulcerative colitis

Architecture

Severe crypt architectural distortionSevere widespread decreased crypt densityFrankly villous surface

InflammatoryHeavy diffuse transmucosal lamina propria cell increaseDiffuse basal plasmacytosis

MiscellaneousIncreased intensity of the alterations towards the distal colonSevere mucin depletionPaneth-cell metaplasia distal to the hepatic flexure

Fig. 18.3 Ulcerative colitis. The microscopic diagnosis ofulcerative colitis is based upon widespread mucosal distortion(shortened and branching crypts) and transmucosalinflammation with basal plasmacytosis. Activity can berecognized by the presence of neutrophils infiltrating the wall ofsome crypts (H&E × 10).

IBD4E-18(255-276) 03/04/2003 10:33 AM Page 258

Several other features may help to establish a diagno-sis of UC or to evaluate the severity of the condition.These include mucosal ulcerations and erosions, mucindepletion, Paneth-cell metaplasia and diffuse thickeningof the muscularis mucosae (Fig. 18.6). Erosions andwidespread surface epithelial damage are more commonin UC than in CD. Severe, almost total mucin depletionis another feature that distinguishes UC from CD,where preserved mucin secretion is more common.Paneth cells in crypts distal to the ascending colonsuggest IBD, usually UC.

Granulomas are not present in UC but isolated giantcells or a histiocytic reaction around a ruptured crypt,mimicking granuloma formation, can occasionally be

seen. ‘Cryptolytic lesions’ or pericryptal granulomaspresent a special diagnostic problem.According to severalauthors these granulomas are not reliable for a diagnosisof CD and occur in UC.2,16Yet, in a series of 22 patientswith suspected IBD, this lesion was present in colorectalbiopsies in 14 cases. Ten patients were subsequentlyfound to have CD.17

The diagnostic sensitivity and specificity of theabove-mentioned features has been assessed mainly onrectal biopsies. UC is also characterized by a special dis-tribution of inflammation and architectural distortionwith increasing intensity from the proximal towards thedistal colon. For proper assessment of this feature, exam-ination of multiple biopsies obtained in different seg-ments is needed.

Variation in time: evolutionThe pattern of the various microscopic features varies in time and depends upon the severity of the disease.In the early, acute phase, crypts are often still regular inshape and size.The most characteristic feature is mucindepletion, associated with neutrophils infiltrating cryptand surface epithelium and inducing crypt abscesses andsecondary crypt destruction.The cellular infiltrate in thelamina propria is homogeneously increased in intensityand mixed in composition. It may show a transmucosaldistribution. Crypt architectural abnormalities appearonly during the evolution of the disease. The naturalhistory of crypt distortion has been examined in a studycomparing the evolution of the histology in patientsafter first attacks of inflammatory bowel disease andinfectious colitis. Biopsies were taken before any treat-ment was given (0–15 days), between 16–30 days,between 1–4 months and between 4–10 months. Cryptdistortion started in the second period being present in

U L C E R A T I V E C O L I T I S

259

SECTION 3

Fig. 18.4 Ulcerative colitis. Distortion of the mucosalarchitecture can persist in quiescent ulcerative colitis. The cryptsare widely separated from the underlying muscularis mucosae.Some of the crypts are branching. (H&E × 10).

Fig. 18.6 Ulcerative colitis. Quiescent phase characterized by anabnormal architecture and diffuse thickening of the muscularismucosae. (H&E × 10)

Fig. 18.5 Ulcerative colitis. Transverse section of the mucosashowing bifid crypts characterized by the presence of doublelumina. The diameter of the lumina and the intercryptal space arevariable. (H&E × 25)

IBD4E-18(255-276) 03/04/2003 10:33 AM Page 259

almost 25% of the patients that eventually had IBD. Inthe period from 4–10 months, almost 75% of theseshowed crypt distortion.1 Treatment may induce somereversibility.

Basal plasmacytosis also becomes more prominent inthe course of the disease, mainly after two weeks ofsymptom duration. It is seen in 54% of biopsiesobtained two weeks after the start of the symptoms andin more than 80% in samples obtained after 4 weeks.During the first attack of IBD, it is commonly the onlyfinding. It can start as focal plasmacytosis. From 16 daysonwards, focal plasmacytosis develops into diffuseplasmacytosis. This feature is usually also present inbiopsies obtained during a flare-up and may predictrelapse.18 During the natural course, the cellular laminapropria infiltrate can remain homogeneously increasedand mixed in composition with a predominance oflymphocytes and plasma cells. The distribution patternmay, however, become patchy.19 These data show that itmay be important to obtain further biopsies one monthfollowing the start of symptoms and treatment, if thediagnosis is not clear during the initial examination.

The inactive, quiescent phase of UC is characterizedby the presence or persistence of more severe crypt archi-tectural abnormalities, while mucin secretion returns tonormal.The lamina propria infiltrate is mononuclear andmay be moderately increased in intensity. In a morechronic phase, lymphoid aggregates and follicles, mainlysituated in the deeper part of the mucosa, will develop orincrease in number.

Disease activityHistology as a tool for the measurement of diseaseactivity of UC was introduced in the 1950s.The altera-tions in intensity and composition of the lamina propriainfiltrate, which can be observed in routinely processedsections stained with hematoxylin and eosin, allow a dis-tinction between active, inactive and quiescent disease.Active disease is defined by the presence of neutrophilsin association with epithelial cell damage.13 Inactivechronic disease is defined as the presence of architec-tural changes and an increase of lamina propria mono-nuclear cells. Quiescent disease means the presence ofstructural changes without alterations in intensity andcomposition of the lamina propria infiltrate. Over theyears, several microscopic scores for the assessment ofdisease activity in UC have been developed, generallyfor study purposes (Table 18.2).20,21 The reproducibilityof activity scores has not been studied extensively, butthe limited data available show good agreement betweendifferent observers.22 The microscopic pattern of activitygenerally correlates well with endoscopic features ofseverity, although in endoscopically inactive disease,microscopic features of activity may persist.The activity

scores may be used to either document disease evolu-tion, or to assess clinical efficacy in therapeutic trials.Histological assessment of disease activity could beimportant for the prediction of relapse. Clinical trialshave indeed shown that persistent active inflammationand basal plasmacytosis are associated with an increasedfrequency of relapse (Fig. 18.7).23

In fulminant disease, mucosal inflammation mayextend towards the submucosa and focally even towardsthe muscularis propria (proportionate inflammation).


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

260

Table 18.2 Example of a scoring system for theassessment of severity in ulcerative colitis.Subgrades are defined for each grade.Examples are given for grade 0 and 1

Grade 0 Structural (architectural change)Subgrades 0.0 No abnormality

0.1 Mild abnormality0.2 Mild or moderate diffuse or

multifocal abnormalities0.3 Severe diffuse or multifocal

abnormalities

Grade 1 Chronic inflammatory infiltrateSubgrades 1.0 No increase

1.1 Mild but unequivocal increase1.2 Moderate increase1.3 Marked increase

Grade 2 Lamina propria neutrophils and eosinophils2A Eosinophils2B Neutrophils

Grade 3 Neutrophils in epithelium

Grade 4 Crypt destruction

Grade 5 Erosion or ulceration

Fig. 18.7 Ulcerative colitis. Mucosal atrophy with loss of crypts.Neutrophils are still present in the lumen and wall of one of thecrypts indicating persistent activity. (H&E × 10). (With permissionfrom Geboes et al. Gut 2000; 47: 404–409.)

IBD4E-18(255-276) 03/04/2003 10:33 AM Page 260

Transmural lymphoid hyperplasia is however highlyunusual.

A significant association between disease activity andnumbers of immunoglobulin-containing cells in thelamina propria has been demonstrated.24 Upregulationof adhesion molecules and other markers is also relatedwith disease activity. Several studies have shownincreased or aberrant colonic epithelial cell expressionof HLA-DR, MHC class II molecules and reduction ordisappearance with different types of treatment. Theexpression of activation markers and the assessment ofimmunoglobulin-containing cells are however notuseful in routine practice.

Medical treatmentTopical and systemic medical treatments have a majorimpact upon histology in UC. Mucosal injury and neu-trophils can diminish substantially within 4 weeks, oreven disappear. Normalization with disappearance ofarchitectural alterations of the crypts and surfacetogether with reduction of the cellular lamina propriainfiltrate takes more time, but can occur (Fig. 18.8).25

In many cases, the inflammatory features become dis-continuous and heterogeneous. This may make a differ-ential diagnosis between UC and CD difficult. It istherefore important to know treatment and its effectsupon histology when analysing a biopsy.26

PouchitisTreatment of ulcerative colitis with total colectomy andcreation of a continent ileoanal anastomosis is notuncommonly accompanied by inflammation of the ilealpouch. Histologically there is normally slight mucosalinflammation, but occasionally crypt abscesses, ulcera-tion and blunting of the villi are seen. Sequential biop-

sies of pouches suggest that villous atrophy and chronicinflammation of the small intestinal mucosa may be fol-lowed by colonic metaplasia, so that pouchitis ultimatelyrepresents a form of ulcerative colitis in metaplastic ilealmucosa. Granulomas and transmural inflammation,which are considered to be diagnostic for CD, can beobserved in pouchitis, following surgery for ulcerativecolitis. Review of the original surgical specimen and ofany other material available is essential before consider-ing changing a diagnosis of ulcerative colitis intoCrohn’s disease.27

Crohn’s Disease

Gross features

CD can affect different segments of the GI tract. Theappearances are similar at all levels.The terminal ileumand proximal colon are the most common sites, fol-lowed by the anorectum and colon. Perianal diseasevaries between 14–76%. Involvement of the uppergastrointestinal tract is uncommon. The length of thesegments involved is variable and the lesions are sepa-rated by uninvolved ‘skip areas’. Macroscopic lesions are apparent on the mucosal and serosal side of thebowel wall. In the Vienna classification proposed by anInternational Working Party for the World Congress ofGastroenterology in 1998, a distinction is made betweeninflammatory disease, stricturing disease – defined asconstant luminal narrowing – and penetrating disease,defined by the occurrence of intra-abdominal or peri-anal fistulas, inflammatory masses and/or abscesses.Fistulas are commonly associated with strictures andtherefore the clinical distinction is often limited to two types: the perforating type and non-perforatingtype with predominantly mucosal lesions. The mucosalappearance is usually heterogeneous. Lesions of differentsize are simultaneously present.The mucosa may appearnormal or may show multiple small (1–2 mm in size)punctiform, rounded nodules or superficial erosionsknown as ‘aphthoid lesions’. Over a period of time, theerosions become confluent and give rise to larger longi-tudinal ulcers, known as serpiginous ulcers.28 The com-bination of longitudinal and transverse ulceration in anedematous mucosa induces a characteristic ‘cobblestone’aspect. Ulcerations are more common on the mesen-teric border of the small intestine. They can becomedeeply situated fissuring ulcers reaching the muscularispropria or pass through the muscularis and give rise toabcesses or fistulas between involved segments andadjacent organs or nearby uninvolved loops. Fistulas aredefined as abnormal communications between thelumen of the gut and the mesentery and/or another

C R O H N ’ S D I S E A S E

261

SECTION 3

Fig. 18.8 Ulcerative colitis. Mucosal biopsy obtained 6 weeksafter the start of medical treatment with 5-ASA showing limitedmucosal distortion and disappearance of activity. (H&E × 10)

IBD4E-18(255-276) 03/04/2003 10:33 AM Page 261

hollow organ or the abdominal wall and skin. Histo-logically, they are composed of granulation tissue, sur-rounding a lumen, which is mostly filled up by nucleardebris and inflammatory cells, in particular neu-trophils.29 Granulomas are present in approximately 25%of the perianal fistulas or abscesses. Fistulas are oftenassociated with strictures. Strictures are characterized byluminal narrowing and bowel wall thickening with orwithout prestenotic dilatation. In high-grade stenosis,the luminal diameter is less than 0.5 cm.They are oftenassociated with severe ulcerations with complete cir-cumferential loss of the mucosa (Fig. 18.9). In rare cases,CD can present a macroscopic picture of the left colonwith continuous involvement, reminiscent of UC, inassociation with ileal disease.30

Being a transmural disease, the bowel wall is thick-ened with involvement of the submucosa, the muscu-laris propria, the subserosa and mesenteric fat. Theserosal surface reveals prominent distended blood vesselsand may show fibrinous exudate with or without adhe-sions to adjacent loops. Mesenteric fat partially sur-rounds the intestine, extending from the mesentericattachment anteriorly and posteriorly corresponding tothe involved segment.This phenomenon, known as ‘fatwrapping’, is specific for CD. It is observed in 75% ofthe surgical specimens (Fig. 18.10). Fat wrapping isdefined on a transverse section of the intestine anddefined as being present when more than 50% of theintestinal circumference is affected. The correspondingmesentery is usually thickened and retracted. Fibrousstrands are present in the mesenteric fat, irradiating fromthe intestine and surrounding thickened, hypertrophiedfat lobules.31 The mesenteric lymph nodes are com-monly swollen, but the number and size of the swollenlymph nodes in CD is not different from that observedin UC.32

Inflammatory pseudopolyps of the colon and smallintestine (in approximately 20% of the cases), identicalto those described in UC can be observed in CD.Theseare usually tall mucosal outgrowths measuring a fewmillimeters in length, but giant forms have beendescribed. Such giant forms are more common in CDthan in UC. In extinguished cases, the characteristiclesions may no longer be present.The specimen showsmerely neuromuscular and vascular lesions and a properdiagnosis may be difficult.

Microscopic features

GranulomaGranulomas in histological sections are a key feature ofCD.33 It should however be remembered that granulo-mas can occur in other conditions, especially infectiousdiseases, such as tuberculosis and Yersinia pseudo-tuberculosis and occasionally in drug-induced colitis. Agranuloma is defined as a collection of monocyte/macrophage cells and other inflammatory cells with orwithout giant cells (Fig. 18.11).The macrophages appearas large cells with abundant pale eosinophilic cytoplasmand large oval nucleus. They are arranged in clusters.Because of this epithelial cell-like morphology, they are called epithelioid cells. The cells can be closelypacked together, and have a sarcoid-like aspect but a‘loose’ expanded form of granuloma is more commonin CD (Fig. 18. 12). Central necrosis and caseation arerare findings and should raise suspicion of tuberculosis.Giant cells may contain calcified conchoid bodies.27,34

Associated inflammatory cells are lymphocytes, usuallyCD4+ T cells, showing expression of CD28, the ligand


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

262

Fig. 18.9 Crohn’s disease. Small intestine: high-grade stenosis.The mucosa is ulcerated. Lymphoid hyperplasia is present in thedeeper parts of the bowel wall. (H&E × 1)

Fig. 18.10 Crohn’s disease. Small intestine: fat wrapping is acharacteristic feature of Crohn’s disease. It is characterized by theovergrowth of mesenteric fat.

IBD4E-18(255-276) 03/04/2003 10:34 AM Page 262

for the B7-related cell surface proteins CD80 (B7-1)and CD86 (B7-2).

The granuloma has to be distinguished from themicrogranuloma, which is smaller and composed of his-tiocytes.They are smaller than the epithelioid cells in thegenuine granuloma. The number of histiocytes in themicrogranuloma is also smaller than in the granuloma(7–18 versus 25–90) while the number of lymphocytes iscomparable (4–11 versus 2–15). Microgranulomas areusually situated in the upper part of the mucosa.The fre-quency in CD is not well established, but seems to varybetween 12% and 24% of endoscopic biopsies obtainedin patients with CD. The exact meaning of the micro-granuloma is still unclear.They seem more common ininactive disease.35 Granulomas must also be distinguishedfrom granulomatous crypt abscesses, a crypt abscess witha giant cell and with or without an excess of histiocytes,and from cryptolytic granulomas.The latter are definedby the rupture of the epithelial lining and the presenceof histiocytes with or without giant cells. While thislesion seems more common in CD it has also beenreported in genuine UC.

Granulomas can be detected in otherwise healthymucosa or in inflamed tissue.They develop in all layersof the intestines from the mucosa to the serosa but aremost frequent in the submucosa.They are also commonin draining lymph nodes being present in approximately20–50% of the cases. Rarely does the granulomatousinflammation affect extraintestinal sites, such as the skin,liver, lungs, eyes and ovaries.

The frequency of finding granulomas in CD variesbetween 15% and 85%, but is rarely higher than50–60%. The results depend highly on tissue sampling(number of biopsies, number of sections examined,endoscopic or surgical samples). For surgical samples thefrequency varies between 15–82% and for endoscopicsamples between 3–56%. The highest numbers areobserved in children, both in surgical series (82%) andin endoscopic series. In pediatric CD, the incidence ofgranulomas is two-fold compared with adults but it isreduced after the second year of the disease and afterthe age of 16 years. The lowest number comes from asurgical series composed of older patients. In general,granulomas are more common in the distal colon andrectum. Granulomas are as common in CD of the uppergastrointestinal tract as in the ileum and colon.The fre-quency of detection for the stomach and duodenumvaries between 3% and 58% and is higher when biopsysamples are taken in macroscopic lesions.

The number of examinations can influence the fre-quency of detection of granulomas. A granuloma wasfound in 23% and 47% respectively, of the patients inwhom one (2.5 ± 1.4 biopsy samples) and four colono-scopies were performed (8.0 ± 1.8 samples). Increasing


263

SECTION 3

Fig. 18.11 Crohn’s disease. Rectal biopsy: a granuloma, socalled because of the round appearance, is a collection ofepithelioid cells with giant cells as in the picture or without. (H&E × 10)

Fig. 18.12 Crohn’s disease. In Crohn’s disease the granulomashave often a loose aspect as illustrated by this lesion laying closeto a ganglion of the myenteric plexus. (H&E × 25)

IBD4E-18(255-276) 03/04/2003 10:34 AM Page 263

the number of biopsies increased the diagnostic yield.For 1–6 biopsy samples, the diagnostic frequency variesbetween 11–47%. Six biopsy samples seems an optimalnumber.When multiple serial sections are examined inthe pathology laboratory the frequency of detectionfurther increases with 50%.

While the diagnostic value of the granuloma in CDis generally accepted, its clinical and prognosticsignificance remain unclear. Several studies have exam-ined the relation between the presence of granulomasand prognosis looking at the recurrence rate, either clin-ically or the risk of a new surgical intervention. Theresults are conflicting. In 8 of 14 studies the presence ofgranulomas had no influence upon the outcome. Inthree series recurrence was diminished and in threeother series recurrence rate was increased.34

Mucosal lesionsEARLY MUCOSAL LESIONS

Various types of early microscopic lesions have beendescribed in CD.They occur as focal lesions, in a back-ground of normal mucosa in contrast with UC wherediffuse epithelial necrosis is seen. Early lesions in CDinclude epithelial patchy necrosis, the aphthoid ulcer ormucosal microulcerations (loss of 1–6 cells). Other fea-tures include the occurrence of a naked surface of thedome area overlying a mucosal lymph follicle and loss ofM cells. Ulcers at the base of crypts with neutrophilsstreaming into the bowel lumen, which leads in a laterphase to mountain peak ulcers, villous abnormalitiesand damage of small capillaries (including capillarythrombi) with subsequent loss of surface epithelial cells(summit lesion) have been reported (Figs 18.13 and18.14).35–39 In general, limited necrosis of surfaceepithelial cells is common while crypt epithelial cells arerarely involved (an exception is the ulcer at the cryptbase). Overall, biopsies of early lesions do not yieldessential diagnostic information. An exception to thismay be the aphthoid ulcer.This ulcer has a predilectionfor the epithelium overlying the lymphoid follicles,although it can occur in other sites too. According tosome studies, biopsies of aphthoid ulcers show morecommonly the characteristic granulomas, which arediagnostic for CD.34

DIAGNOSTIC FEATURES

Although the degree of mimicry with UC can be high,the presence of aphthoid ulcers, fissure ulcers, transmuralinflammation, fistulas, lymphangiectasia, fibrous strictur-ing and neural changes is predominantly a feature of CD.Six of these features proposed cannot, however, be reli-ably detected on endoscopic samples. A diagnosis of CDon endoscopic samples of the colon therefore reliesheavily on the identification of the microscopic features

of IBD. It has been suggested that the diagnosis of CDshould be based upon the presence of an epithelioidgranuloma with one other feature suggestive or diagnos-tic for IBD, or the presence of three other features in theabsence of granulomas, provided that specific infectionhas been excluded.40 The mucosal lesions of IBD consistof epithelial alterations and a cellular inflammatoryresponse.The former include cytological and architecturalchanges, indicative of damage and repair. The cellularinflammatory response consists of changes in intensity ofthe infiltrate, alterations in composition and changes inthe distribution pattern.These features have been exam-ined in a number of studies of rectal and colorectal biop-sies and surgical specimens. Only a limited number haveacceptable sensitivity, specificity and predictive value and are sufficiently reproducible (Table 18.3). Featuresthat favor CD are epithelioid granulomas, relativelyunchanged crypts or segmental distribution of cryptatrophy and crypt distortion together with discontinuous


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

264

Fig. 18.13 Crohn’s disease. Aphthoid ulcer in the ileum: earlymucosal ulcer, centrally located and appearing as a mountain topulcer. (H&E × 4)

IBD4E-18(255-276) 03/04/2003 10:34 AM Page 264

focal or patchy inflammation and mucin preservation inthe epithelium at an ulcer edge, and the presence of amixture of normal samples (skip lesions) and inflamedsamples in a set of biopsies obtained in the samearea.1,2,14,41–43 Focal inflammation is defined as a smallcollection of inflammatory cells in otherwise normalmucosa. It can present as a focal aggregation of neutrophilpolymorphs in a pericryptal position, a lesion which issuggestive for CD but not pathognomonic. Patchyinflammation is diagnosed when the mucosal backgroundshows inflammation of varying intensity. The diagnosticvalue of patchy inflammation is limited because a similarpattern can be seen in UC in long-standing disease orfollowing treatment.Another reliable feature which helpsto distinguish CD from UC in favor of CD is the pres-ence of neuronal changes (Fig. 18.15).42 These arehowever only rarely observed in endoscopic biopsysamples because of the superficial nature of the latter.

Most of the studies examining the features of IBD havebeen performed in adults. Separating CD and UC mightbe different in children and adolescents. Discontinuousinflammation and density of infiltration prevail in chil-dren while the diffuse type of infiltration is morecommon in CD in adults. However, granulomas are morecommon in children, being present in 26% of the biop-sies and 42% of the patients.

Ileal biopsies have rarely been included in studiesexamining the microscopic features of CD.The presence


265

SECTION 3

Fig. 18.14 Crohn’s disease. Early mucosal lesions in CD can beassociated with damage of small capillaries as illustrated here bythe presence of a fibrin plug in the lumen of a small vessel(immunohistochemistry × 40).

Table 18.3 Microscopic features with sufficientreproducibility, discriminative and predictivevalue, useful for the differential diagnosisbetween normal and IBD, acute infectiouscolitis (AIC) and IBD and CD, and UC

Mucosal architectureMucosal surface, normal, irregular, villousCrypt atrophy (shortened, widely spaced crypts)Distorted, dilated, branching crypts

Inflammatory changesBasal plasmacytosis, increase in cells in basal third of laminapropriaIncreased lamina propria cellularity (round cells andneutrophils)Basal lymphoid aggregates

Specific featuresEpithelioid granulomaBasal giant cellsExcess histiocytes in lamina propria

Features related with activity (separating IBD from normal)Neutrophils in surface epitheliumNeutrophils in crypt epitheliumUlceration

Fig. 18.15 Crohn’s disease. Crohn’s disease is characterized bythe presence of granulomas and by hyperplasia of thesubmucosal nerves, sometimes called ‘neuromatous lesion’. (H&E × 10)

IBD4E-18(255-276) 03/04/2003 10:35 AM Page 265

of ileal lesions is however another key lesion, whichallows to discriminate between UC and CD. Importantdiagnostic microscopic features of CD are architecturalabnormalities of the villi (irregularity and blunting orbroadening), preserved mucin secretion or increasedmucin production (hypercrinia) by epithelial cells,mucoid or pseudopyloric metaplasia, active chronicinflammation and the presence of granulomas.44 Pseudo-pyloric gland metaplasia is however a non-specific featurerelated with ulceration, healing and the ulcer associatedcell lineage (Fig. 18.16).45

Immunohistochemical analysis of mucosal biopsiesreveals expression of various cytokines and other media-tors, such as interleukin-10, interleukin-12, interleukin-15 and tumor necrosis factor alfa (TNFα ). Some markersmay be more specific for CD but increased expressionindicates mainly ongoing immune response and has nodiagnostic value at present.

The value of multiple, multi-step biopsiesThe importance of stepwise biopsies for the diagnosis ofeither CD or UC has received little attention in theliterature. Most studies have focused on single rectalbiopsies. Discontinuous inflammation in stepwise biop-sies does not distinguish CD from UC in adults. It canbe found in approximately 30% of patients with long-standing UC but it discriminates in children and inpatients with a short history of the disease.The presenceof an inflammatory infiltration with decreasing intensityfrom caecum to rectum favors a diagnosis of CD.12,46 Incombination with ileal biopsies, multiple colonic biop-sies are useful for the diagnosis.

Disease activity: effect of treatmentMicroscopic assessment of disease activity in CD isdifficult because of the segmental and transmural char-acter of the disease. Yet, several scoring systems havebeen designed and even used in clinical trials. Thesescores are usually based on the microscopic analysis ofmultiple biopsies from different segments. The correla-tion between endoscopic findings, clinical indices ofdisease activity and microscopic scores is variable.47 Thisreflects the fact that mucosal biopsies are not representa-tive of the transmural inflammation and that clinical fea-tures are not always reliable indicators of disease activityat the tissue level. Mucosal biopsies however do reflectmucosal inflammation. Monotherapy with corticos-teroids or 5-ASA has only a limited influence uponmucosal lesions in CD, although long-term combina-tion treatment may induce normalization. Reduction ofepithelioid granulomas with persistent chronic inflamma-tion has been reported following long-term treatmentwith prednisolone, sulfasalazine and 6-mercaptopurine.In contrast, polymeric diet and immunomodulatorytherapy with azathioprine and infliximab can pro-foundly influence the histologic lesions and inducehealing.48–50

The submucosa, muscularis propria and serosaLESIONS OF THE ENTERIC NERVOUS SYSTEM

Abnormalities of the enteric nervous system (ENS) arecommon in CD.The major structural abnormalities areirregular hypertrophy and hyperplasia of nerve fibersand alterations of neuronal cell bodies and enteric glialcells in ganglia of the submucosa and plexus myenteri-cus. The abnormalities of the nerve fibers are mostprominent in the submucosa, mainly in involved areas.They have been called ‘neuromatous lesions’. Mucosalnerve fiber hypertrophy is only seen in areas overlyingsubmucosal fiber hypertrophy and cannot reliably beassessed on routinely hematoxylin and eosin-stainedslides. Nerve fiber hypertrophy is commonly associatedwith inflammation and granulomas in the plexus


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

266

Fig. 18.16 Crohn’s disease. Ileal biopsies can help to establishthe diagnosis of Crohn’s disease. The villi are irregular. Mucoid or pseudopyloric metaplasia indicates previous ulceration. (H&E × 10)

IBD4E-18(255-276) 03/04/2003 10:35 AM Page 266

myentericus are not uncommon. A three-fold increasein the number of ganglion cells of the ileal myentericplexus was recorded in a series of 24 cases with CD.Allthese features were shown to be highly suggestive for adiagnosis of CD with a significant difference betweenUC and CD and they can be detected with greatreliability.51

Ultrastructural studies have shown that nerve fibersor axons appear as swollen, empty structures.52 Thespecificity of this lesion for CD has however not beenconfirmed. A relative increase in myenteric neuronescontaining nitric oxide synthase (NOS), vasoactiveintestinal peptide (VIP) and pituitary adenylate cyclaseactivating peptide (PACAP) and an abnormal pattern of VIP containing fibers have been recognized in CD.53

Immunohistochemistry revealed also increased MHCclass II (HLA-DR) membranous expression on entericglial cells in macroscopically involved and uninvolvedareas. The enhanced expression is positively correlatedwith the local intensity of the cellular inflammatoryinfiltrate. Induction of HLA-DP and DQ antigens onglial cells is restricted to areas of moderate and highinflammatory activity.51

VASCULAR LESIONS

Inflammatory cell infiltration of blood vessels and oblit-erative lesions have been observed with microscopy insurgical samples from patients with CD in a number ofstudies. The frequency varies between 3% and 85%.The frequency of vascular granulomatous inflamma-tion varies between 3–100% of the cases examined(Fig. 18.17).54 Granulomas are not only found in associ-ation with blood vessels. They may even be morecommon in lymphatic channels. While the nature andthe exact frequency of granulomatous vasculitis in CDremain to be determined, its occurrence and diagnostic

significance are unquestionable. Lymphangiectasia in thesubmucosa is another common finding in CD in addi-tion to granulomatous vasculitis.55

LYMPHOID HYPERPLASIA

Lymphoid aggregates and nodules with or without ger-minal centres are common in the mucosa and evenmore numerous in the submucosa in CD. They occuralso in the muscularis propria, in the vicinity of themyenteric plexus and in the subserosa. This transmurallymphoid hyperplasia is believed to be characteristic forCD and is not observed in UC or IC.56

EDEMA, FIBROSIS, FIBROMUSCULAR

OBLITERATIONS, STRICTURES

Widening of the submucosa by edema is common inCD and probably a sign of active disease. Strictures area common complication of CD and an indication forsurgery in approximately 50% of all CD patients.Histological studies of strictures show marked expan-sion of the muscularis mucosae by an irregular increasein the number of smooth muscle cells in this layer andthe presence of large amounts of collagen, laminin andtenascin. Types V and III collagen can be abundantlypresent (Fig. 18.18). Mast cells can be numerous. Inaddition to the expansion of the muscularis mucosaeper se, islands of smooth muscle cells and myofibroblastsare noted in disorganized fashion in the adjacent sub-mucosa. Smooth muscle cell fibers may interconnectthe muscularis mucosae and propria. The submucosacontains dense masses of large amounts of collagenousmaterial and the luminal margin of the muscularispropria is disorganized with numerous collagenous


267

SECTION 3

Fig. 18.17 Crohn’s disease. Granulomatous vasculitis is anotherlesion, which is not uncommon in Crohn’s disease. (H&E × 10)

Fig. 18.18 Crohn’s disease. Small intestine. Fibrosis is a commoncomplication of Crohn’s disease. It is characterized by abnormaldeposition of collagens, illustrated here by the presence ofcollagen V in the edges of a mucosal ulceration(immunohistochemistry × 10).

IBD4E-18(255-276) 03/04/2003 10:35 AM Page 267

septae invaginating its ragged edges and extending intothe muscularis. In general the muscularis propria is alsogreatly increased in thickness, although it may be inter-rupted in certain areas.57,58

Upper gastrointestinal diseaseUpper GI disease is present in approximately 8–40% ofthe cases.The variability is explained by the methods ofexamination such as symptomatology, endoscopy withbiopsy or radiology, used to evaluate upper GI involve-ment. Of all the sites within the GI tract, which may beaffected by CD, the esophagus has been suggested as theone least likely involved. Early changes include micro-scopic disease or aphthous ulcerations.The latter shouldnot be confused with other conditions, such as herpeticesophagitis. Biopsies are the method of choice for thedifferential diagnosis and the finding of granulomasimplies CD. More extensive disease includes ulcerationand stricture formation. Fissuring ulceration or fistulaformation is uncommon. Most cases of esophageal CDare associated with CD elsewhere in the gut, but thereare a few reports of cases in which CD initially pre-sented in the oesophagus.59

In the stomach, three major patterns may be distin-guished. First, reactive gastritis or Helicobacter pylori-associated gastritis not related with CD, minor but characteristic lesions known as focally enhanced gastri-tis, and CD of the stomach with histological evidence ofgranulomatous infiltration.3 Focally enhanced gastritis ischaracterized by a focal infiltration of CD3+ lympho-cytes and CD68+ histiocytes with granulocytes. It wasfound in 48% of a series of 75 CD patients and in only0.8% of the controls.60 Granulomatous infiltration isobserved in 2–10% of CD patients (Fig. 18.19). Besidestypical granulomas, the gastric wall may also show trans-mural, often focal lymphoid infiltrates, lymphangiec-tasias, submucosal and subserosal fibrosis, fissures andulceration. Generally, CD of the stomach is associatedwith disease in more typical sites. More than 80% ofpatients with gastric CD have duodenal bulb involve-ment. The finding of a granuloma in a gastric biopsyshould not automatically be considered as CD. Granulo-matous gastritis can have many other causes, as shown ina study of 71 patients in which CD was finally diag-nosed in only 52%.61

Granulomatous appendicitisCrohn’s disease affecting the appendix is well docu-mented and is present in up to 25% of cases.The diseaseis also reported as being isolated to the appendix. Thehistological features of CD in the appendix includethickening of the wall, transmural inflammation, patchylymphoid aggregates throughout the wall, fissured ulcer-ation and epithelioid granulomas. Isolated ‘granuloma-

tous’ inflammation in the appendix is however not nec-essarily CD. It may be the result of a number of diverseand unrelated causes such as parasites, foreign bodyreaction and bacterial infections.

Indeterminate ColitisThe term ‘colitis indeterminate’ was initially proposedfor a small group of cases in which there was difficultyin distinguishing CD from UC (and even infectiousdiseases) in the excised specimen. The difficulty wasdue to the fact that features of both conditions werepresent in the same specimen. Most of the cases hadfulminant disease and the classification was essentiallytemporary before a final diagnosis was established.56

Later, ‘indeterminate colitis’ (IC) was applied as a tem-porary classification to cases where a definite diagnosiswas not possible with endoscopic samples because ofabsence of diagnostic features discriminating betweenCD and UC.The diagnostic accuracy of a single rectal


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

268

Fig. 18.19 Crohn’s disease. Stomach. Gastric mucosal biopsycontaining two characteristic granulomas. (H&E × 10)

IBD4E-18(255-276) 03/04/2003 10:35 AM Page 268

biopsy is estimated to be 37–77% for CD patients and67–70% for UC patients. Multiple endoscopic biopsiesof the colon can already solve many problems for CD.In difficult cases with pancolitis, the differential diagno-sis can sometimes be solved by analysis of ileal biopsiesor gastric biopsies. A special type of focal gastritis hasbeen observed in CD. Indeterminate colitis is bydefinition a disease limited to the colon.The morpho-logic features of ‘indeterminate colitis’ diagnosed onendoscopic samples include lesions suggestive or diag-nostic for IBD such as (minimal) architectural distor-tion, inflammatory features (usually transmucosalinflammation) which do not allow a firm distinctionbetween UC and CD because of their patchy nature,and absence of small bowel involvement.While only aminority of patients have a diagnosis of IC, it is impor-tant to realize that such a diagnosis has certain conse-quences. It means that the patient has IBD and thatother conditions, especially infections, are less likely orindeed excluded. Relapse of symptoms due to compli-cations, such as cytomegalovirus (CMV) infection isequally excluded (Fig. 18.20). A diagnosis of IC alsomeans that the patient needs careful follow-up in orderto reach a definite diagnosis.

Differential Diagnosis

Ulcerative colitis or Crohn’s disease

Several studies have examined a large variety of mucosalbiopsy criteria for the distinction between UC and CDand shown wide overlap.2 The situation is differentwhen surgical specimens are involved, because CD ischaracterized by transmural inflammation, whereas UCis a mucosal disease. In most studies examining endo-scopic mucosal biopsies, the criteria, which discriminatebetween UC and CD, have a sensitivity and specificityexceeding only 75%.The histological diagnosis is largelybased on the finding of granulomas in addition to thepresence of criteria for the diagnosis of IBD, especiallymucosal distortion. Segmental distribution of crypts orcrypt atrophy, segmental distribution of mucin deple-tion, mucin preservation at the edge of an ulcer or incrypts with surrounding neutrophils, the occurrence offocal inflammation simultaneously with severe diffuse orpatchy inflammation in a set of biopsies, have a signifi-cant discriminative value in favor of CD. Furthermore,in contrast to UC, typical CD shows a segmental orfocal distribution of the disease in the bowel and focalor patchy inflammation within a biopsy specimen.Thesefeatures can reliably be assessed with good agreementbetween observers. In contrast, Paneth-cell metaplasiahas only limited discriminative value while cryptitis and

crypt abscesses have no significant discriminative value(Table 18.4).

Infectious and drug-related colitis

A variety of infectious or drug-related colitides may clin-ically mimic CD or UC, especially those cases withabrupt clinical onset. Acute inflammatory diarrheaaccompanied by fever and evidence in stool of aninflammatory process such as pus, mucus and blood, anddysentery, a more severe manifestation of inflammatorydiarrhea with, in addition, abdominal pain cramps andrectal tenesmus, are indeed often due to infectious agents.These can be divided into two subgroups – organismsthat elicit an inflammatory process by penetrating themucosa, and those that elaborate cytotoxins withoutinvading the cells. Worldwide, invasive bacterial agentsconstitute the most important subgroup of etiologicagents. A precise diagnosis is essential for a proper treat-

D I F F E R E N T I A L D I A G N O S I S

269

SECTION 3

Fig. 18.20 Ulcerative colitis. Relapse of disease symptoms inulcerative colitis can be due to infection with cytomegalovirus.The virus can be recognized by typical nuclear inclusions. (H&E × 25)

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 269

ment. Histology, in collaboration with other methodsplays an important role in establishing a correct diagno-sis.41 The histology of infectious-type colitis is variable,due to the natural course of the disease and to differencesin the virulence of the bacterial strain and the reaction ofthe host, but in general it shows a picture of activeinflammation with normal mucosal architecture. Anormal biopsy or mild edema can be seen when bacteriaare non-invasive. Active inflammation may be mild orsevere with extensive necrosis. In mild or moderatedisease, neutrophils are found in the upper part of the lamina propria. Cryptitis, with neutrophils usually inthe upper, more luminal part, superficial crypt abscessesand erosions are common. The lesions show a focal orpatchy distribution. Neutrophils are more common in theearly phase of the disease. Residual lesions appear later.An increase of plasma cells is observed 7–10 days afterthe initial onset of the disease. Increased numbers of IgA-and IgM-containing plasma cells are found in the mucosain patients with Campylobacter colitis, in contrast withpatients with active UC, who show increases of IgA andIgG.The crypts remain parallel, but they are often smallerat the upper part. The distinction between IBD andinfective type colitis relies mainly on the absence of fea-tures (architectural and basal plasmacytosis) which directtowards a diagnosis of chronic idiopathic inflammatorybowel disease. In rare occasions, such as Entamoeba his-tolytica or CMV infection, the pathogen can be identified.CMV infection constitutes a special situation. It may bepresent in the early stage of the disease of UC but it canalso be responsible for relapse of symptoms or cause pou-chitis. Histology is the best method for identification ofthis infection. Relapse of symptoms can also be the result

of infection with Clostridium difficile. The differentialdiagnosis of CD includes mainly bacterial diseases due toYersinia enterocolitica and Campylobacter jejuni, which canboth affect the ileum and proximal colon (Fig. 18.21).Special attention is required for tuberculosis and amoebi-asis because of the indolent nature of these infections andthe negative effect of a treatment with corticosteroids orimmunosuppressants.Tuberculosis is commonly found inthe terminal ileum and ileocecal region but, in contrast toCD, it is rare in the large intestine. Unlike the longitudi-nal serpiginous ulcers of CD, tuberculous ulcers tend tobe circumferential or transverse. Examination of ileal orcolonic biopsies can reveal specific features. Granulomas


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

270Fig. 18.21 Yersinia pseudotuberculosis infection. Ileal biopsyshowing a characteristic granuloma with central necrosis. Thelesion must be differentiated from tuberculosis and Crohn’sdisease. (H&E × 10)

Table 18.4 Frequency of detection (%) of the reproducible features for different diagnoses

Normal Infective IBD UC CD

ArchitectureIrregular surface 0–5 0–7 39 63 24Branched crypts 0–5 75 63–83 39–67Crypt shortening 0–5 0 29–78 12–37

Chronic inflammationIncreased basal cellularity 0 0 63 62Increased lamina propria cellularity 0–19 30 89–93 76–92 72–81Granulomas 0 0–2 25–27 0–5 21–100Discontinuous inflammation 7 10 26

Epithelial alterationsMucin depletion 17 35–69 5–57Mucin preservation at ulcer edge – +++

Neuronal changes 2 75

IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn’s disease.

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 270

are usually multiple and large and can be confluent.Positive Ziehl staining suggests intestinal tuberculosis.Central necrosis of a lymph follicle is suggestive ofYersinia infection. Salmonella, Shigella and Campylobacterinfections can mimic UC.

A medication history is indispensable in the investiga-tion of patients with diarrhea. A variety of drugs giveneither topically or systematically can cause colitis, ileo-colitis or proctitis. Some of these may mimic UC. Non-steroidal anti-inflammatory drug (NSAID)-inducedlesions present a significant clinical problem. Usually thepatients are elderly. NSAIDs can induce small bowel andcolonic lesions. Colitis induced by NSAIDs can presentas non-specific colitis, de novo colitis, reactivation ofquiescent inflammatory bowel disease, allergic colitis(with eosinophils), constipation with perforation (ster-coral ulcer), non-specific ulceration and fistulas relatedto diverticulosis. The microscopic distinction betweenactive UC and drug-induced colitis is usually not sodifficult because NSAID-induced lesions lack majorinflammation although architectural alterations of themucosa are common.The distinction with inactive UCis therefore more difficult (Fig. 18.22). In NSAID-induced pathology, apoptosis of epithelial cells may becommon in the crypts (as in graft-versus-host disease).62

Other drugs that can cause colitis or proctitis are penicil-lamine, gold salts, isotretinoin and methyldopa. Oral con-traceptives and cocaine tend to mimic ischemic colitis.

Diverticular disease-associated colitis

A clinical syndrome of chronic colitis, localized to thesigmoid colon and occurring in association with diver-ticular disease, has been recognized repeatedly. Mucosalbiopsies of this condition show features of IBD including

a distorted crypt architecture and basal plasmacytosis.Colonic mucosa both proximal and distal is usuallynormal. The inflammatory features can mimic UC orCD. The relation of the inflammatory reaction withgenuine IBD is unclear. Overall, it seems that diverticulardisease-associated chronic colitis will precede the onsetof conventional ulcerative proctitis and colitis or genuineCD in only a minority of cases.63

Diversion colitis

Diversion colitis is an inflammatory process that occursin the bypassed colonic segment after surgical diver-sion of the fecal stream. In a defunctioned, previouslynormal rectum, mild inflammation is already apparentat 3 months. When continuity is restored, the changesdisappear. Histologic abnormalities show a spectrum ofchanges ranging from mild colitis to those seen in severeactive ulcerative colitis. Lesions include aphthous ulcers,crypt distortion, atrophy and abscesses, a villous colonicsurface, mucin granulomas and a mixed inflammatoryinfiltrate with patchy lymphoid hyperplasia.The latter isa particular feature of this condition. All these findingsmay mimic IBD. When the original process leading todiversion of the fecal stream is not an inflammatorybowel disease the histological diagnosis of diversioncolitis is easy. However, when there is underlying UC,the differential diagnosis is difficult. Features favoringdiversion colitis are variation in severity of the lesionsseen on multiple biopsies taken from one single area andthe absence of morphologic changes in the proximal(not bypassed) segment.

It has been shown that the defunctioned rectumfrom patients suffering from unequivocal ulcerativecolitis who have undergone proximal colonic resectionmay show transmural inflammation, fissures and epithe-lioid granulomas.3 While these features would suggest adiagnosis of CD, the temptation to change the underly-ing diagnosis must be resisted, because the changes mayrepresent further manifestations of diversion colitis.Only follow-up of the patient and the eventual appear-ance of lesions in the small intestine may force a changeof diagnosis. In general, it seems that the conditionworsens in UC, while diversion is favorable in CD.

Lymphoid hyperplasia

Lymphoid hyperplasia is observed in chronic ulcerativecolitis per se, in a separate entity, described as lymphoidfollicular proctitis and in diversion colitis.

Microscopic colitis: collagenous and lymphocytic colitis

Collagenous colitis is a disease of the large intestine,characterized by the presence of a thickened collagen

D I F F E R E N T I A L D I A G N O S I S

271

SECTION 3

Fig. 18.22 Colitis caused by non-steroidal anti-inflammatorydrugs. There is mucosal ulceration and distortion of the cryptarchitecture. The lesions are however discontinuous andinflammation is moderate. (H&E × 4)

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 271

layer underneath the intercryptal surface epitheliumand, circumferentially, around the upper part of thecrypts and by increased mucosal inflammation. For aproper diagnosis, it is necessary to obtain several (at leastthree) colonic biopsies, because the thickness of thesubepithelial collagen table varies throughout the colon.To reduce false positivity a minimal value of 10 µm has been proposed. Plasma cells and lymphocytes areincreased in the lamina propria and eosinophils may befocally numerous. Neutrophils are occasionally seen and crypt abscesses are rare. Surface epithelial cells areflattened or cuboidal and sometimes desquamated andgoblet cells are reduced in number and size.Yet overall,the colonic mucosa has a preserved or only minimallydistorted crypt architecture. It should not be confusedwith rare forms of amyloid colitis, where the sub-epithelial thickening is due to deposition of amyloid(Fig. 18.23).

The microscopic picture of lymphocytic colitis ischaracterized by diffuse lesions with an increase ofinflammatory cells in the epithelium and lamina propria.The infiltrate is mixed in composition with eosinophils,neutrophils (cryptitis and crypt abscesses), lymphocytes,plasma cells and mast cells.The number of intraepitheliallymphocytes is significantly increased (24.6 ± 3 for 100epithelial cells; normal value 4.6 ± 1.5), whereas a non-significant increase of interepithelial lymphocytes isnoted in UC and CD. Epithelial mucin depletion iscommon in lymphocytic colitis.

Microscopic colitis may be associated with primaryileal villous atrophy (PIVA), without alterations in theupper small intestine. Microscopic colitis is clinically

and morphologically different from IBD. Yet, in somecases, collagenous or lymphocytic colitis can precede (orfollow) genuine UC or CD.

Endometriosis

Ileal endometriosis may present with acute, chronic orrecurrent distal small bowel obstruction, in the sameway as Crohn’s ileitis.The differential diagnosis with CDmay be very difficult and is mostly not possible pre-operatively. Intestinal endometriosis may mimic CD, butcan also be associated with CD. Endometriotic depositsin the bowel can induce microscopic mucosal lesionsresembling inflammatory bowel disease.

Miscellaneous

Microscopic lesions similar to CD have been reportedin ileal and colonic biopsies from patients with reactivearthritis and ankylosing spondylitis. Only a minority of these patients developed genuine CD. Less commondiseases, such as the colitis of Behçet’s disease and other vasculitides can mimic CD. Endoscopic biopsiesare often inconclusive, except when small vessels are involved. In chronic granulomatous disease andHermansky–Pudlak syndrome, a colitis complete withgranulomas, which is indistinguishable from CD, canoccur.The diagnostic feature is the presence of charac-teristic histiocytes in both mucosa and submucosa.Occasionally, malignancies may simulate CD on radio-logical examinations. The most common examples ofthese are in the colon: the poorly differentiated signet-ring cell carcinoma and metastases in the wall, and in the small intestine: a neuroendocrine cancer of the terminal ileum. Mucosal biopsies may show mildinflammatory features and architectural alterations. Thefinding of such features must be interpreted cautiouslyand, as always, histology has to be combined with theclinical history and other data in order to reach adefinite diagnosis and to solve the differential diagnosticproblem.

Dysplasia and CancerEndoscopy with multiple biopsies is used for second-ary prevention and in surveillance programmes for theearly detection of colorectal cancer in patients withlong-standing UC. This is based on the finding that precancerous lesions can be identified in biopsies.These lesions are called ‘dysplasia’. More recently, theterm ‘intraepithelial neoplasia’ has been proposed.Dysplasia was defined by an international InflammatoryBowel Disease–Dysplasia Morphology Study Group as‘unequivocal, non-invasive (confined within the base-


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

272

Fig. 18.23 Amyloid colitis. Collagenous colitis, one of thesubtypes of microscopic colitis is characterized by thickening ofthe subepithelial collagen table. The lesion can be mimicked inamyloid colitis as illustrated in this figure of a colonic biopsystained with Congo red and showing the red color of thesubepithelial amyloid band. (Congo red × 10)

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 272

ment membrane), neoplastic transformation of the epi-thelium excluding all reactive changes’.64 This definitionstresses the nature and origin of the lesion, but itsidentification relies upon the recognition of morpho-logical features resulting from cytological and architec-tural changes in routinely processed and hematoxylinand eosin-stained sections. Accumulated experiencefrom several prospective studies shows that 6–10 differ-ent biopsies from different sites might be sufficient todetect significant dysplasia. Generally, the histologicalfeatures of dysplasia in IBD are comparable with thoseseen in colonic adenomas in non-colitic patients andinclude cytological criteria, such as variations in nuclearposition, size and chromatin pattern combined witharchitectural distortion (Fig. 18.24). Although the mor-phological spectrum forms a continuum, dysplasia isdivided into different categories according to the sever-ity of the alterations, because of different implicationsfor patient management (Table 18.5). In the most com-

monly used classification, proposed by the InternationalStudy Group, two grades are distinguished.The category‘low-grade dysplasia’ includes cases of mild andmoderate dysplasia. The category ‘high-grade dysplasia’includes some cases of moderate dysplasia (particularlythose with prominent architectural alteration) and allcases of severe dysplasia and carcinoma in situ. Thegrade of dysplasia is always determined by the featuresof the most dysplastic portion.The distinction betweenhigh- and low-grade dysplasia is dependent primarilyon the degree of cytologic alterations. This classifica-tion appears to be reproducible, although in general, itappears that interobserver agreement is better for high-grade dysplasia.

Surveillance studies have shown that a distinction hasto be made between IBD-related dysplasia, especially thepolypoid or raised type also known as dysplasia associ-ated lesion or mass (DALM) and sporadic adenomaoccurring in IBD. While the latter is unequivocally aneoplastic and hence a dysplastic lesion, its developmentis unrelated to the underlying IBD. Furthermore, recentevidence suggests that the molecular pathogenesis ofIBD-associated dysplasia is different in terms of orderand timing of genetic events. DALM in IBD should infact be defined as a poorly circumscribed protrudinglesion surrounded by dysplastic flat mucosa. It shouldalso be distinguished from ALM (adenoma-like mass),when there is no surrounding dysplasia.65 The differentialdiagnosis between polypoid IBD-associated dysplasia andinflammatory pseudopolyps or dysplasia in inflammatorypseudopolyps is another problem. Dysplasia in inflamma-tory polyps is rare. The major problem is that thesepolyps frequently contain areas of residual regenera-tion and it may be difficult to make a clear distinctionbetween unequivocally neoplastic changes and regenera-tion.The differential diagnosis between reparative lesionsand ‘genuine’ dysplasia in general may be difficult.Whereas high-grade dysplasia can usually reliably be dis-tinguished, the distinction between repair and low-gradedysplasia may present a serious problem. Reparative cells of the surface may be distinguished from ‘genuine’dysplastic cells because of their shape (cuboidal or lowcolumnar versus crowded and high columnar) and

D Y S P L A S I A A N D C A N C E R

273

SECTION 3

Table 18.5 Biopsy classification of dysplasia(intraepithelial neoplasia) in inflammatorybowel disease

NegativeIndefinitePositive

Low-grade dysplasiaHigh-grade dysplasia

Fig. 18.24 Ulcerative colitis. Low-grade dysplasia, on the rightpanel of the figure is characterized by the presence of tallcolumnar cells with increased staining of the cytoplasm andelongated nucleus. (H&E × 10)

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 273

because mucus secretion in dysplastic cells is oftenabnormal. In addition, the presence of inflammationorients also to reactive changes. Some of the morerecently developed treatment modalities may indeeddecrease the inflammatory reaction before completion ofreparative phenomena. This discrepancy can result infollow-up biopsies showing little or no inflammatoryactivity and marked regeneration. The latter representspseudodysplasia and should not be mistaken for dyspla-sia. Because of the diagnostic problems related to dyspla-sia other markers have been examined and tested toimprove the reliability of the diagnosis. Changes innuclear DNA content, reflecting gross chromosomalalterations can be identified with flow cytometry.Inactivation of tumour suppressor genes, such as p53 canbe identified using immunohistochemistry for thedemonstration of mutant p53 protein expression. Boththe detection of aneuploidy and overexpression of p53can be useful for a correct diagnosis (Fig. 18.25).

Colorectal cancer in UC is predominantly located inthe sigmoid and rectum.The prognosis is similar to theone found in the general population.The tumour stageis essential for the prognosis.

Carcinoma is also a well-recognized complication ofCD, where it can develop in the small intestine and inthe colon in areas where there is or has been inflamma-tory disease.The carcinomas are often multiple.They areusually adenocarcinomas and are often associated withthe presence of dysplasia in the colon and in the smallintestine, suggesting that the development of malig-nancy in CD is similar to that of UC.66 The histologicalfeatures of dysplasia are identical to those seen in UCand the same classification can be used. Patients withCD may also be at an increased risk for the develop-

ment of perianal squamous carcinoma and both UC andCD appear more than usually susceptible to malignantlymphoma.

REFERENCES

1 Schumacher G, Kollberg B, Sandstedt B. A prospective study of firstattacks of inflammatory bowel disease and infectious colitis.Histologic course during the 1st year after presentation. Scand JGastroenterol 1994; 29:318–332.

2 Jenkins D, Balsitis M, Gallivan S, et al. Guidelines for the initial biopsydiagnosis of suspected chronic idiopathic inflammatory boweldisease. The British Society of Gastroenterology Initiative. J Clin Pathol1997; 50:93–105.

3 Shepherd NA. Pathological mimics of chronic inflammatory boweldisease. J Clin Pathol 1991; 44:726–733.

4 Levine DS, Haggitt RC. Normal histology of the colon. Am J SurgPathol 1989; 13:966–984.

5 Segal GH, Petras RE. Small intestine. In: Sternberg SS ed. Histology forpathologists. New York: Raven: 1992:547–572.

6 Chang EB. Intestinal epithelial function and response to mucosalinjury. In: Kirsner JB ed. Inflammatory bowel diseases. Philadelphia: WB Saunders; 2000:3–19.

7 Bejarano PA, Aranda-Michel J, Fenoglio-Preiser C. Histochemical andimmunohistochemical characterization of foamy histiocytes(muciphages and xanthelasma) of the rectum. Am J Surg Pathol 2000;24:1009–1015.

8 Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis.Gastroenterology 1998; 115:182–250.

9 Both H, Torp-Pedersen S, Kreiner S, et al. Clinical appearance atdiagnosis of ulcerative colitis and Crohn’s disease in a regional patientgroup. Scand J Gastroenterol 1983; 18:987–991.

10 D’Haens G, Geboes K, Peeters M, et al. Patchy cecal inflammationassociated with distal ulcerative colitis: a prospective endoscopicstudy. Am J Gastroenterol 1997; 92:1275–1279.

11 Valdez R, Appelman H, Bronner MP, et al. Diffuse duodenitis associatedwith ulcerative colitis. Am J Surg Pathol 2000; 24:1407–1413.

12 Bernstein CN, Shanahan F, Anton PA, et al. Patchiness of mucosalinflammation in treated ulcerative colitis: a prospective study.Gastrointest Endosc 1995; 42:232–237.

13 Goldman H. Colonic biopsy in inflammatory bowel disease. SurgPathol 1991; 4:3–24.

14 Seldenrijk CA, Morson BC, Meuwissen SGM, et al. Histopathologicalevaluation of colonic mucosal biopsy specimens in chronicinflammatory bowel disease: diagnostic implications. Gut 1991;32:1514–1520.

15 Rubio CA, Johansson C, Kock Y. A quantitative method of estimatinginflammation in the rectal mucosa.III. Chronic ulcerative colitis. Scan JGastroenterol 1982; 17:1083–1087.

16 Dundas SAC, Dutton J, Skipworth P. Reliability of rectal biopsy indistinguishing between chronic inflammatory bowel disease andacute self–limiting colitis. Histopathol 1997; 31:60–66.

17 Lee FD, Maguire C, Obeidat W, et al. Importance of cryptolytic lesionsand pericryptal granulomas in inflammatory bowel disease. J ClinPathol 1997; 50:148–152.

18 Nostrant TT, Kumar NB, Appelman HD. Histopathology differentiatesacute self-limited colitis from ulcerative colitis. Gastroenterology 1987;92:318–328.

19 Kleer CG, Appelman HD. Ulcerative colitis: patterns of involvement incolorectal biopsies and changes with time. Am J Surg Pathol 1998;22:983–989.


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

274

Fig. 18.25 Ulcerative colitis. Dysplasia in ulcerative colitis.Positive nuclear staining for p53 is clearly present in somesamples and absent in others (immunohistochemistry × 4).

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 274

20 Powell-Tuck J, Day DW, Buckell NA, et al. Correlation between definedsigmoidoscopic appearances and other measurements of diseaseactivity in ulcerative colitis. Dig Dis Sci 1982; 27:533–537.

21 Florèn CH, Benoni C, Willén R. Histologic and colonoscopicassessment of disease extension in ulcerative colitis. Scand JGastroenterol 1987; 22:459–462.

22 Geboes K, Riddell R, Öst Ä, et al. A reproducible grading scale forhistological assessment of inflammation in ulcerative colitis. Gut 2000;47:404–409.

23 Riley SA, Mani V, Goodman MJ, et al. Microscopic activity in ulcerativecolitis: what does it mean? Gut 1991; 32:174–178.

24 Keren DF, Appelman HD, Dobbins WO III, et al. Correlation ofhistopathologic evidence of disease activity with the presence ofimmunoglobulin-containing cells in the colons of patients withinflammatory bowel disease. Hum Pathol 1984; 15:757–763.

25 Odze R, Antonioli D, Peppercorn M, et al. Effect of topical 5-amino-salicylic acid (5-ASA) therapy on rectal mucosal biopsymorphology in chronic ulcerative colitis. Am J Surg Pathol 1993;17:869–875.

26 Geboes K. Pathology of inflammatory bowel diseases (IBD): variabilitywith time and treatment. Colorectal Dis 2001; 3:2–12.

27 Sanders DSA. The differential diagnosis of Crohn’s disease andulcerative colitis. Baillière’s Clin Gastroenterol 1998; 12:19–33.

28 Rutgeerts P, Geboes K, Vantrappen G, et al. Natural history of recurrentCrohn’s disease at the ileocolonic anastomosis after curative surgery.Gut 1984; 25:665–672.

29 Oberhuber G, Stangl P C, Vogelsang H, et al. Significant association ofstrictures and internal fistula formation in Crohn’s disease. VirchowsArch 2000; 437:293–297.

30 Vasiliauskas EA, Plevy SE, Landers CJ, et al. Perinuclear anti-neutrophilcytoplasmic antibodies in patients with Crohn’s disease define aclinical subgroup. Gastroenterology 1996; 110:1810–1819.

31 Sheehan AL, Warren BF, Gear MWL, et al. Fat-wrapping in Crohn’sdisease: pathological basis and relevance to surgical practice Br J Surg1992; 79:955–959.

32 Cook MG. The size and histological appearances of mesenteric lymphnodes in Crohn’s disease. Gut 1972; 13:970–972.

33 Chambers TJ, Morson BC. The granuloma in Crohn’s disease. Gut 1979;20:269–274.

34 Heresbach D, Heresbach-Le Berre N, Ramée MP, et al. Fréquence etvaleur pronostique du granulome épithélioïde au cours des maladiesinflammatoires chroniques de l’intestin. Gastroenterol Clin Biol 1999;23:1376–1387.

35 Rotterdam H, Korelitz B I, Sommers SC. Microgranulomas in grosslynormal rectal mucosa in Crohn’s disease. Am J Clin Pathol 1977;67:550–554.

36 Dvorak A, Connell AB, Dickersin GR. Crohn’s disease: a scanningelectron microscopic study. Hum Pathol 1979; 10:165–177.

37 Fujimura Y, Hosobe M, Kihara T. Ultrastructural study of M cells fromcolonic lymphoid nodules obtained by colonoscopic biopsy. Dig DisSci 1992; 37:1089–1098.

38 Tanaka M, Riddell RH. The pathological diagnosis and differentialdiagnosis of Crohn’s disease. Hepatogastroenterol 1990; 37:18–31.

39 Sankey EA, Dhillon AP, Anthony A, et al. Early mucosal changes inCrohn’s disease. Gut 1993; 34:375–381.

40 Lennard-Jones JE. Crohn’s disease: Definition, pathogenesis andaetiology. In: Wright R, ed. Recent advances in gastrointestinalpathology. London: WB Saunders; 1980: 173–189.

41 Surawicz CM, Belic L. Rectal biopsy helps to distinguish acuteself–limited colitis from idiopathic inflammatory bowel disease.Gastroenterology 1984; 86:104–113.

42 Cook MG, Dixon MF. An analysis of the reliability of detection anddiagnostic value of various pathological features in Crohn’s diseaseand ulcerative colitis. Gut 1973; 14:255–262.

43 Tanaka M, Saito H, Fukuda S, et al. Simple mucosal biopsy criteriadifferentiating among Crohn’s disease, ulcerative colitis and otherforms of colitis: measurement of validity. Scand J Gastroenterol 2000;35:281–286.

44 Geboes K, Ectors N, D’Haens G, et al. Is ileoscopy with biopsyworthwhile in patients presenting with symptoms of IBD. Am JGastroenterol 1998; 93:201–206.

45 Wright NA, Pike C, Elia G. Induction of a novel epidermal growth factorsecreting cell lineage by mucosal ulceration in human gastrointestinalstem cells. Nature 1990; 343:82–85.

46 Schmitz-Moormann P, Schäg M. Histology of the lower intestinal tractin Crohn’s disease of children and adolescents. Path Res Pract 1990;186: 479–484.

47 Gomes P, Du Boulay C, Smith CL, et al. Relationship between diseaseactivity and colonoscopic findings in patients with colonicinflammatory bowel disease. Gut 1986; 27:92–95.

48 Beatie RM, Schiffrin EJ, Donnet-Hughes A, et al. Polymeric nutrition asthe primary therapy in children with small bowel Crohn’s disease.Aliment Pharmacol Ther 1994; 8:609–615.

49 Baert F, D’Haens GR, Peeters M, et al. Tumor necrosis factor α antibody(infliximab) therapy profoundly downregulates the inflammation inCrohn’s ileocolitis. Gastroenterology 1999; 116:22–8.

50 Korelitz BJ, Sommers SC. Response to drug therapy in Crohn’s disease.Evaluation by rectal biopsy and mucosal cell counts. ClinGastroenterol 1984; 6:123–127.

51 Geboes K, Collins S. Structural abnormalities of the nervous system inCrohn’s disease and ulcerative colitis. Neurogastroenterol Mot 1998;10:189–202.

52 Dvorak AM, Osage JE, Monahan RA, et al. Crohn’s disease: transmissionelectron microscopic studies. III. Target tissues proliferation of andinjury to smooth muscle and the autonomic nervous system. HumPathol 1980; 11:620–634.

53 Belai A, Boulos PB, Robson T, et al. Neurochemical coding in the smallintestine of patients with Crohn’s disease. Gut 1997; 40:767–774.

54 Wakefield AJ, Sankey EA, Dhillon AP, et al. Granulomatous vasculitis inCrohn’s disease. Gastroenterology 1991; 100:1279–1287.

55 Mottet NK. Intestinal histopathology of regional enteritis:histopathologic spectrum of regional enteritis and ulcerative colitis.Philadelphia: WB Saunders; 1971.

56 Tanaka M, Riddell RH. The pathological diagnosis and differentialdiagnosis of Crohn’s disease. Hepatogastroenterol 1990; 37:18–31.

57 Graham MF, Diegelmann RF, Elson CO, et al. Collagen content andtypes in the intestinal strictures of Crohn’s disease. Gastroenterology1988; 94:257–265.

58 Gelbmann CM, Mestermann S, Gross V, et al. Strictures in Crohn’sdisease are characterized by an accumulation of mast cells colocalisedwith laminin but not with fibronectin or vitronectin. Gut 1999;45:210–217.

59 D’Haens G, Rutgeerts P, Geboes K, et al. The natural history ofesophageal Crohn’s disease: three patterns of evolution. GastrointestEndosc 1994; 40:296–300.

60 Oberhuber G, Puspok A, Oesterreicher C, et al. Focally enhancedgastritis: a frequent type of gastritis in patients with Crohn’s disease.Gastroenterology 1997; 112:698–706.

61 Ectors NL, Dixon MF, Geboes KJ, et al. Granulomatous gastritis: amorphological and diagnostic approach. Histopathol 1993; 23:55–61.

62 Gibson GR, Whitacre EB, Ricotti CA. Colitis induced by nonsteroidalanti-inflammatory drugs. Report of four cases and review of theliterature. Arch Intern Med 1992; 152:625–632.

R E F E R E N C E S

275

SECTION 3

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 275

63 Makapugay LM, Dean PJ. Diverticular-disease-associated chroniccolitis. Am J Surg Pathol 1996; 20:94–102.

64 Riddell RH, Goldman H, Ransohoff DF, et al. Dysplasia in inflammatorybowel disease: standardized classification with provisional clinicalapplications. Hum Pathol 1983; 14:931–968.

65 Odze RD, Brown CA, Hartmann CJ, et al. Genetic alterations in chroniculcerative colitis associated adenoma-like DALMS are similar to non-colitic sporadic adenomas. Am J Surg Pathol 2000; 24:1209–1216.

66 Sigel JE, Petras RE, Lashner BA, et al. Intestinal adenocarcinoma inCrohn’s disease. A report of 30 cases with a focus on coexistingdysplasia. Am J Surg Pathol 2000; 23:651–655.


EC

TIO

N 3

.D

IAG

NO

SIS

: A

CL

INIC

IAN

’S P

ER

SP

EC

TIV

E

276

IBD4E-18(255-276) 03/04/2003 10:36 AM Page 276

section 3 chapter 18 disease and ulcerative colitis and ulcerative colitis k. geboes goal to review...

Documents