Diagnosing coeliac disease: beyond the conventional histology

Abstract

Keliakiadiagnoosi perustuu nykyisten kriteerien mukaan ohutsuolikoepalaan, jossa tulee olla todettavissa ohutsuolen nukkakato eli villusatrofia sekä kuopakkeiden pidentyminen eli kryptahyperplasia. Keliakiavasta-aineet veressä tukevat diagnoosia. Keliakian ohutsuolimuutokset kehittyvät vähitellen: ensimmäinen muutos on lisääntynyt tulehdussolupitoisuus epiteelillä ja vasta lopuksi kehittyy villusatrofia. Diagnostisia ongelmia aiheuttaa tänä päivänä ohutusuolinäytteiden vaihteleva laatu, villusatrofian ajoittainen läiskittäisyys ja muut villusatrofiaa ja tulehdusmuutoksia aiheuttavat taudit.

Tämä tutkimus keskittyi keliakian diagnostiikan parantamiseen. Tutkimuksessa yhteensä 223 potilaalla oli hoitamaton ja 241 potilaalla hoidettu keliakia tai ihokeliakia. Lisäksi 66 potilaalla oli alkava keliakia ilman villusatrofiaa, ja 608 potilaalla keliakia poissuljettiin ohutsuolikoepalan perusteella epäilyistä huolimatta. Yhteensä 138 potilasta toimi ei-keliakiakontrollina. Hoitamattomassa keliakiassa ohutsuolen epiteelin tulehdussoluista CD3+ soluilla todettiin 85% herkkyys ja 69% tarkkuus tautiin, ja vastaavat luvut alfa/beta+ soluille olivat 80% ja 59% ja gamma/delta+ soluille 92% ja 81%. Lymfosyyttitiheyden määrittäminen pelkästään villusten kärjistä keliakiassa, jossa ohutsuolen suolinukka ei ole täysin tuhoutunut, todettiin vähintään yhtä hyväksi kuin gamma/delta+ solumääritys, joka vaatii jääleikkeen. Myöskin alkavassa keliakiassa ilman villusatrofiaa villusten kärkien lymfosyyttitiheyden ja gamma/delta+ solujen herkkyydet (85% ja 82%) olivat paremmat kuin CD3+ ja alfa/beta+ solujen (61% ja 58%).

Keliakiavasta-aineet (endomysium- tai retikuliinivasta-aineet) löysivät 84% keliaakikoista, joilla oli villusatrofia, eikä yhdelläkään ei-keliakikontrollilla ollut todettavissa näitä vasta-aineita. Vasta-ainenegatiivisilla keliaakikoilla oli enemmän oireita ja ohutsuolen lymfoomaa kuin vasta-ainepositiivisilla, vaikka suolivaurioaste todettiin ryhmissä samanlaiseksi. Tutkimus lisäksi osoitti, että keliakiavasta-aineet löysivät 70% potilaista, joilla oli alkuun normaali ohutsuolirakenne, mutta myöhemmin kehittyi keliakia. Verestä mitattavat keliakiavasta-aineet muodostuvat ohutsuolessa, ja tässä tutkimuksessa kaikilla hoitamattomilla keliaakikoilla, joilla oli villusatrofia, ja 93%:lla potilaista joilla oli alkava keliakia ilman villusatrofiaa, oli keliakialle tyypilliset vasta-ainekertymät todettavissa ohutsuolessa. Tutkimuksessa lisäksi osoitettiin, että keliaakoilla ohutsuolen vasta-ainekertymät kohdistuvat transglutaminaasi 2:ta vastaan.

Tutkimus osoitti, että keliakian diagnostiset kriteerit vaativat uudistamista. Yhteensä 66 potilaalla (23%:lla) 289:stä oli keliakia, jota ei pystytty toteamaan perinteisellä ohutsuolen suolinukan analysoinnilla. Kohonnut gamma/delta+ tai villusten kärkien lymfosyyttitiheys ja erityisesti veren keliakiavasta-aineet viittaavat keliakiaan tilanteissa, joissa ohutsuolen histologinen tutkimus on epävarma tai epäillään alkavaa keliakiaa. Tämän tutkimuksen perusteella kuitenkin kaikkein luotettavimmin keliakian löytää, jopa ennen villusatrofian kehittymistä, ohutsuolen transglutaminaasi 2:ta vastaan muodostuneet vasta-ainekertymät. Tämä menetelmä vaikuttaisi olevan erityisen luotettava ja käyttökelpoinen epävarmoissa diagnostisissa tilanteissa, kun veren keliakiavasta-aineet eivät tue diagnoosia tai ohutsuolen histologinen löydös on epäselvä.

According to the current criteria the diagnosis of coeliac disease should be based on the presence of small bowel mucosal villous atrophy and crypt hyperplasia; serology has a supportive role. However, the diagnosis is often complicated, since small bowel specimens may be of poor quality and villous atrophy patchy. Moreover, atrophy can also be present in other disorders, and early developing coeliac disease without villous atrophy is particularly difficult to diagnose, since early changes are unspecific. This study focused on improving the quality of coeliac disease diagnosis. The aim was to investigate the diagnostic significance of the determination of villous height-crypt depth ratio (Vh/CrD), the densities of CD3+, alfa/beta+, gamma/delta+ and villous tip intraepithelial lymphocytes (IELs) and serum and intestinal coeliac antibodies in untreated coeliac disease with villous atrophy, in early developing coeliac disease and in treated disease.

In studies I-IV altogether 223 patients had untreated and 241 treated coeliac disease or dermatitis herpetiformis. Further, 66 patients were considered to suffer from early developing coeliac disease, and in 608 patients coeliac disease was suspected but excluded because of normal small bowel mucosal villous architecture. Altogether 138 individuals served as non-coeliac controls. In untreated coeliac disease with villous atrophy CD3+ IELs had a 85% sensitivity and 69% specificity for the condition, the corresponding percentages for alfa/beta+ IELs being 80% and 59% and for gamma/delta+ IELs 92% and 81%, respectively. The villous tip IELs determined from haematoxylin- and eosin-stained samples cannot be investigated in untreated coeliac disease cases where total or subtotal villous atrophy is present; but in severe partial villous atrophy, this method nonetheless proved to be more accurate than the determination of CD3+ and alfa/beta+ cells and at least as good as the determination of gamma/delta+ cells, which requires frozen sections. In early developing coeliac disease the sensitivities of all IELs were lower than in coeliac disease with villous atrophy, but those of villous tip and gamma/delta+ IELs (85% and 82%) proved superior to CD3+ and alfa/beta+ IELs (61% and 58%, respectively). Serum immunoglobulin (Ig) A-class reticulin (ARA) and endomysial (EmA) antibodies proved reliable in disclosing coeliac disease with villous atrophy; the sensitivity of coeliac autoantibodies for the condition was 84%; specificity was 100%. In the second study (II), focusing on seronegative coeliac disease, altogether 22 IgA-competent untreated coeliac disease patients out of 177 were found to be serum EmA-negative. The majority of EmA-negative patients were male, and they were found to be on average older than EmA-positive patients. EmA-negative patients suffered from abdominal symptoms more often than EmA-positive, and three EmA-negative subjects were diagnosed with enteropathy-associated T-cell lymphoma at the time when coeliac disease diagnosis was established. The severity of histological lesion was similar in EmA-negative and ­positive patients, but the densities of gamma/delta+ IELs were statistically significantly higher in EmA-positive patients.

The follow-up study (IV) focusing on early developing coeliac disease showed that the majority of patients with positive IgA-class coeliac autoantibodies in the serum (ARA or EmA) having normal villous architecture subsequently develop coeliac disease; the sensitivity of coeliac autoantibodies in early developing coeliac disease was shown to be 70%. Thus patients having false-positive coeliac autoantibodies in the serum are at risk of developing overt coeliac disease and should be followed up.

Coeliac autoantibodies are produced in the small bowel mucosa, and in this study autoantibody deposits were also investigated from the intestine. All untreated coeliac disease patients (n = 35) with villous atrophy, both EmA-negative and ­positive, had intestinal coeliac-type IgA deposits, and these were shown to be gluten-dependent. None of the non-coeliac controls were found to have these intestinal IgA deposits. In the follow-up study (IV) focusing on coeliac disease without villous atrophy small bowel mucosal IgA deposits reached 93% sensitivity for early developing coeliac disease, and this method proved superior to earlier utilized approaches such as determination of IELs. IgA deposits detected in the intestine of untreated coeliac disease patients were found in co-localization with transglutaminse 2 (TG2) and the TG2-target specificity of the deposits was ascertained by showing that they had the ability to bind externally added recombinant human TG2. It emerged that the diagnostic criteria for coeliac disease need to be revised. In other words, it was shown that Vh/CrD determination as the gold standard in the diagnosis is nowadays inadequate; this method failed to recognize 66 (23%) out of 289 untreated coeliac disease patients. These 66 patients were suffering from early developing coeliac disease, where minor mucosal abnormalities may be present in the absence of villous atrophy. However, such changes are difficult and subjective in interpretation. Determination of IELs, especially of gamma/delta+ IELs, is helpful in obscure cases, but these cells are not invariably increased in untreated coeliac disease with villous atrophy. In severe partial villous atrophy the determination of villous tip IELs is at least as valuable as the determination of gamma/delta+ IELs. If early developing coeliac disease is suspected, patients evincing an increased density of gamma/delta+ or villous tip IELs in their small bowel mucosa shoul! d be followed up to detect forthcoming coeliac disease, but an increased density of CD3+ or alfa/beta+ IELs alone does not require routine surveillance. Serum coeliac autoantibodies (ARA/EmA) are highly specific for coeliac disease, and IgA-class serum coeliac autoantibodies with normal histology indicate early developing coeliac disease. However, the single most sensitive and specific means of establishing untreated coeliac disease with and without villous atrophy is the investigation of intestinal TG2-specific IgA deposits. The investigation of intestinal IgA deposits seems to be of special value in obscure cases when the serology does not support the diagnosis or conventional histology is ambiguous.