A

O

KAT

a

ARAA

KC

1

grttisteourrg

A

1d

Digestive and Liver Disease 44 (2012) 375– 378

Contents lists available at SciVerse ScienceDirect

Digestive and Liver Disease

j our nal ho me page: www.elsev ier .com/ locate /d ld

limentary Tract

pen conformation tissue transglutaminase testing for celiac dietary assessment

umar Pallav1, Daniel A. Leffler ∗,1, Michael Bennett, Sohaib Tariq, Hua Xu, Toufic Kabbani,llan C. Moss, Melinda Dennis, Ciaran P. Kelly, Detlef Schuppan

he Celiac Center, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, United States

r t i c l e i n f o

rticle history:eceived 18 August 2011ccepted 12 December 2011vailable online 18 January 2012

eyword:oeliac disease

a b s t r a c t

Objectives: Anti tissue-transglutaminase antibody is the mainstay of celiac disease serologic testing.Whilst it has high sensitivity in patients on an unrestricted diet, sensitivity is poor for evaluation ofgluten free diet adherence.Aim: To assess the utility of a novel assay measuring Immunoglobulin-A antibodies to catalytically activeopen conformation tissue-transglutaminase in assessment of ongoing gluten exposure in celiac diseasepatients on an alleged gluten free diet.Methods: Through prospective assessment, 147 patients with celiac disease were divided into good andpoor adherence. Open and closed (conventional) tissue-transglutaminase titres were measured usingstandard enzyme linked immunosorbent assay. 50 patients with inflammatory bowel disease served asdisease controls.Results: Overall 128 patients had been on gluten free diet for more than six months and 19 were found tobe poorly adherent on dietary review. Within this group 13 (68.4%) and 10 (52.6%) patients respectively

were positive for the open conformation and conventional assay (p = 0.51). Two and one control patientstested positive for closed and open assays respectively.Conclusions: Compared to native assays open conformation tissue-transglutaminase may have higher sen-sitivity in the poor gluten free diet adherence group and higher specificity in the control population. Largerpopulation studies are warranted to assess whether the open conformation tissue-transglutaminase assaymay be superior to the conventional assay.

Gast

© 2012 Editrice

. Introduction

Celiac disease is increasingly recognised as the most commonastrointestinal disease with autoimmune features [1]. Increasedates of diagnosis worldwide have been in large part attributableo improvements in awareness and serologic testing. IgA anti-tissueransglutaminase (tTG), endomysial antibodies (EMA) and antibod-es to deamidated gliadin peptides (DGP) all have sensitivities andpecificities above 90% in most populations [2,3]. However, none ofhese serologic tests shows a high degree of responsiveness to mod-st changes in intestinal inflammation or limited amounts/durationf gluten re-exposure [3–7]. This is a major limitation in the eval-ation of patients with non responsive celiac disease (NRCD). Non

esponsive celiac disease is defined as ongoing symptoms or recur-ence of symptoms in celiac disease patients despite following aluten free diet (GFD) for at least 6 months. It has been reported that

∗ Corresponding author at: Beth Israel Deaconess Medical Center, 330, Brooklineve., Boston, MA 02215, United States. Tel.: +1 617 667 1272; fax: +1 617 667 8144.

E-mail address: dleffler@caregroup.harvard.edu (D.A. Leffler).1 Both authors contributed equally to this paper.

590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevieroi:10.1016/j.dld.2011.12.008

roenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

30–50% of NRCD cases are secondary to ongoing gluten exposure[8,9], but due to the inability of conventional tTG assays in detectinglow dose gluten exposure and limited availability of skilled celiacdieticians, this becomes a diagnosis of exclusion. Whilst in somepatients a dietary source of gluten can be identified through care-ful interview, in many a broad differential needs to be ruled outbefore arriving at that diagnosis.

In this study, we assessed the ability of a new test detectingthe antibody to the stabilised open (active) conformation tTG (O-tTG) in comparison to the conventional test which detects antibodyto the tTG of closed or undefined conformation (C-tTG) to predictgluten free diet adherence. We determined the dietary adherenceclinically by an expert dietician, and the specificity of both the testsin a control population with inflammatory bowel disease.

2. Methods

Serum was obtained from 147 individuals with biopsy proven

CD who had previously participated in a study evaluating glutenfree diet adherence and its relation to symptoms of celiac disease[9–11]. Additional tests were run on 50 patients with biopsy con-firmed inflammatory bowel disease (Tables 1 and 2).

Ltd. All rights reserved.

376 K. Pallav et al. / Digestive and Liver Disease 44 (2012) 375– 378

Table 1Characteristics of the celiac disease study population.

Study group (n = 147) Overall BIDMCb celiac population (n = 601) p value

Mean age at diagnosis (standard deviation) 44.8 (16) 43.7 (16.4) 45Females 113 (76.8%) 431 (71.7%) 23Ethnicity

White 145 (98.6%) 598 (99.5%)58Other 2 (1.4%) 3 (0.5%)

Other autoimmune disordersa 47 (30.5%) 163 (27%) 42

a Predominantly thyroid disease > type 1 diabetes mellitus � Raynaud’s phenomenon > inflammatory bowel disease > sarcoidosis > psoriasis.b Beth Israel Deaconess Medical Center.

Table 2Characteristics of the control cohort.

Ulcerative colitis(n = 26)

Crohn’s disease(n = 24)

Overall IBD(n = 50)

Celiac group(n = 147)

p value celiac vs.overall IBD

Mean age (years) at enrolment & SDa 43.9 (15) 35 (9.6) 39.5 (13.5) 50 (16) .0001Females 12 (46%) 13 (54%) 25 (50%) 113 (76.8%) .0006Active 3 (11.5%) 8 (33%) 11 (22%) NA NAOn immunomodulator 5 (19%) 18 (75%) 23 (46%) NA NA

D hn’s dc

tdfusswosfuoil

du(hdaO

dbZiodcaaitdc2Ict3

(BIDMC) (Table 1). 85% of the subjects studied under the celiac dis-ease group had ‘excellent’ or ‘good’ GFD adherence as measured bythe skilled dietician (Fig. 1).

esignated active if Crohn’s Disease Activity Index score ≥ 150 for patients with Croolitis.

a SD: standard deviation. IBD: inflammatory bowel disease.

Individuals in the celiac disease cohort underwent nutri-ional evaluation in a standardised fashion as we have previouslyescribed [7]. This included analysis of three-day food records, aood ingredient quiz, a dynamic interview and questionnaires eval-ating diet adherence, gastrointestinal and non-gastrointestinalymptoms and quality of life, as well as evaluation for gluten expo-ure by a highly skilled dietician with over 10 years of experienceorking with celiac disease. Global GFD adherence was recorded

n a 6 point Likert scale ranging from 1: ‘excellent adherence: con-uming gluten less than three times per year’ to 6: ‘not currentlyollowing a gluten free diet’ (see Table 1). For assessment of thetility of the serologic tests to monitor gluten free diet adherence,nly participants following the diet for at least six months werencluded. A subgroup analysis was performed on participants fol-owing the diet for at least 12 months as well.

The control population consisted of 24 patients with Crohn’sisease (Mean age 35 years St Dev 9.6 years) and 26 patients withlcerative colitis (Mean age 44 years St Dev 15 years). Twenty-four100%) of Crohn’s disease and 3 (11.5%) of ulcerative colitis patientsad active disease at the time of assessment. 18 (75%) of Crohn’sisease patients were on immunomodulator therapy at the time ofssessment as compared to 5 (19.23%) of ulcerative colitis patients.verall male:female ratio was 1:1.

Analysis of closed and open conformation anti-tTG IgA titres wasone by enzyme linked immunosorbent assay (ELISA) with recom-inant human antigen (Product No. E001, E002, E006 and E007,edira, Darmstadt, Germany). Recombinant tTG was produced innsect cells and purified to >95% purity and a specific activityf >2000 U/mg, measured by dansylcadaverine incorporation inimethylated casein (T036, Zedira). This tTG formulation (closedonformation) was then used to coat plates for ELISA-kits E001nd E002. For the production of open conformation tTG, the samentigen was used and incubated with the tTG specific irreversiblenhibitor ZED754 which locks the enzyme in the open conforma-ion. After removal of surplus inhibitor the remaining activity wasocumented at <10%. The open conformation tTG was proven byrystallisation and X-ray structure resolution (Lindemann et al.,0th International Symposium on Medicinal Chemistry; EFMC-

SMC 2008, Vienna, Austria, August 31–September 4, 2008). Theoatings for both the assays were performed under identical condi-ions. Serum at a dilution of 1:100 was incubated with the wells for0 min, followed by human IgA coupled to horseradish peroxidase

isease or Simple Clinical Colitis Activity Index score ≥ 5 for patients with ulcerative

for 30 min and colour development with 5,5′-tetramethylbenzidine(TMB) and hydrogen peroxide. The open and closed tTG-kits areidentical with the exception of the antigen utilised. The cut-off inthese assays has been set to 3 U/ml using analysis of blood donorsera.

Results were compared using Student’s t-test, ANOVA, Fisher’sexact test and Wilcoxon signed-rank test where appropriate.Informed consent was obtained from all study subjects to usebiological specimen. The study was approved by the Beth IsraelDeaconess Medical Center (BIDMC) Investigational Review Board.

3. Results

147 patients with biopsy proven celiac disease were tested.None of the participants had IgA deficiency. Time on gluten free dietranged from 3 to 576 (mean 56.5) months. The demographic char-acteristics of the study population were not significantly differentfrom that of overall celiac disease population seen at our hospital

Fig. 1. Distribution of standardised dietician evaluation scores. Standardised dieti-cian evaluation scores in 147 celiac disease patients. 1: excellent gluten free dietadherence, 2: good gluten free diet adherence, 3: fair gluten free diet adherence, 4:poor gluten free diet adherence; 5: very poor gluten free diet adherence, and 6: noton a gluten free diet.

K. Pallav et al. / Digestive and Liver

Frt

gafgoivw(hgtaftOmaii

ntpmatspalp

TT

ig. 2. Scatter plot of IgA anti-unfolded tTG vs. IgA anti-native tTG. Dashed linesepresent 95% confidence intervals for the correlation. Dotted lines mark cut-offs ofhe respective assays.

For assessment of the utility of the serologic tests to monitorluten free diet adherence, only participants following the diet fort least six months were included (n = 128). In this group, 19 wereound to have inadequate dietary adherence and 109 had eitherood or excellent adherence. The two groups were similar in termsf age at enrolment (50 years vs. 51 years p = 0.8), female predom-nance (76% vs. 68% p = 0.5) and time on gluten free diet 62 monthss. 70 months (p = 0.7). Within the poorly adherent group 10 (52.6%)ere positive for C-tTG and 13 (68.4%) tested positive for O-tTG

p = 0.5). Closed and open conformation IgA anti-tTG titres wereighly correlated (Fig. 2). In the 109 individuals with excellent orood gluten free diet adherence, 31 (28.4%) and 47 (43%) were C-TG and O-tTG positive, respectively (p = 0.034). Within the gooddherence group patients who were positive for either test wereound to be on the GFD for a significantly shorter duration thanhe patients with negative tests. (O-tTG positive, 41.8 months vs.-tTG negative, 77.5 months vs. p = 0.032 and C-tTG positive, 37.5onths vs. C-tTG negative, 71.9 months, p = 0.057). These findings

long with the dietetic survey findings supported the fact that pos-tive assays in the well adherent population were a marker for slowmmune recovery and not ongoing gluten exposure.

Within the poorly adherent group, patients with elevated andormal tTG titres were on the gluten free diet for similar lengths ofime (O-tTG positive; 76.1 months vs. O-tTG negative; 76.8 months,

= 0.98 and C-tTG positive; 91.8 months vs. C-tTG negative; 59.8onths, p = 0.38). Significantly more participants with poor GFD

dherence (13/19 68.4%) were positive for O-tTG compared to par-icipants with good GFD adherence (47/109 43.2%) (p = 0.048). Aimilar but lesser difference was seen with C-tTG (10/19 vs. 31/109,

= 0.059). No other differences were significant. See Table 3 for summary of the test characteristics of the two assays. We alsoooked at the subset following GFD for at least 12 months. Bothositive and negative predictive value increased slightly in this

able 3est characteristics of study assays for determination of poor dietary adherence.

Anti IgA tissue transglutaminase Sensitivity Specificity

Gluten free diet duration 6 months and moreClosed conformation 52.6% 71.5%

Open conformation 68.4% 56.8%

Gluten free diet duration 12 months and moreClosed conformation 56.2% 75%

Open conformation 68.7% 63.4%

Disease 44 (2012) 375– 378 377

selected group, however this did not reach statistical significance.See Table 3.

In the six individuals who fell outside of the 95% confidenceinterval for correlation between closed and open conformation IgAanti-tTG, four (67%) had been on the gluten free diet for less thanone year compared to 28% of the overall study population (seeFig. 2). However, this did not reach statistical significance (p = 0.06).No other demographic or clinical differences were noted betweenthese two groups.

Amongst the 50 patients with inflammatory bowel diseaseeither active or in remission (24 with Crohn’s disease, 26 withulcerative colitis) all but two individuals showed antibody titreswithin the normal range for both open and closed antibodies (rangeundetectable to 3.0). Of the individuals with positive titres, all wereless than two times the upper limit of normal. These included onepatient with ulcerative colitis with a C-tTG of 3.4 and an O-tTG of3.6, and a patient with Crohn’s disease with a C-tTG of 3.0.

4. Discussion

Whilst diagnostic methodologies for celiac disease are highlydeveloped, monitoring strategies remain limited and refinementsin serologic tests are a viable option. Analogous to the improvementmade to anti-gliadin testing by deamidated gliadin testing, openconformation tTG tests may offer benefits beyond conventional tTGassays.

The rationale for the novel assay is based on the fact that inhealthy tissue, tTG is largely confined to the intracellular com-partment and in an inactive closed conformation. However, ininflammation, increased amounts of tTG are secreted into the extra-cellular space where the concentration of ionised calcium is muchhigher [12]. The calcium binds to the beta-barrel domains of tTGcausing a significant change in conformation to the open formnecessary for catalytic activation [13,14]. Once in the open form,tTG is able to mediate deamidation of glutamine residues in cer-tain gluten peptides generating peptides with higher affinity for Tcell receptors and increased immunogenicity [15–17]. It is there-fore hypothesised that auto-antibodies to the open conformationof tTG may have a higher specificity for the inflammatory intestinalprocess active in celiac disease, and yield more otherwise crypticepitopes for auto-antibody formation, resulting in higher antibodytitres and thus higher sensitivity in patients with lesser degrees ofinflammation.

To date there are very limited data regarding the utility ofantibodies to open-conformation tTG as a test of dietary com-pliance in celiac disease [18]. Compared to closed-conformationantibodies, our data show an increased sensitivity of the open-conformation antibody test to detect dietary non-compliance. Alsosignificantly more participants with poor GFD adherence (13/1968.4%) were positive for O-tTG compared to participants withgood GFD adherence (47/109 43.2%) (p = 0.048). A similar but

non-significant difference was seen with C-tTG (10/19 vs. 31/109,p = 0.059). This may mean that the open conformation TTG assaymay be more effective at differentiating between good and pooradherence.

Positive predictive value Negative predictive value

24.4% 89.6%21.6% 91%

28% 91%24% 92%

3 Liver

(csa

aOf(iaclmd

vfaatfiatmes

ontcadao

adestaptio

F

CDc

[

[

[

[

[

[

[

[

[

[

78 K. Pallav et al. / Digestive and

As expected, neither antibody showed a very high sensitivity68.4% for O-TTG and 58.6% for C-tTG) to detect dietary non-ompliance. This is in line with past reports which suggest thaterologic tests alone are not accurate measures of gluten free dietdherence [7].

In the 109 individuals with excellent or good gluten free dietdherence, 31 (28.4%) and 47 (43%) had elevated titres of C-tTG and-tTG, respectively (p = 0.034). We also note that comparative titres

or the O-tTG were significantly higher than for the C-tTG assay10.3 vs. 6.2 p < 0.001), suggesting that O-tTG may be more sensitiven individuals with low titre celiac auto-antibodies. These findingsre in line with the hypothesis that auto-antibodies to the openonformation of tTG may have higher sensitivity in patients withesser degrees of inflammation or as in our case, resolving inflam-

ation, making it a valuable tool in workup of non responsive celiacisease.

It is also notable that in the disease control population, two indi-iduals were positive for C-tTG, whilst one individual was positiveor O-tTG. The combination of a higher percentage of patients withctive celiac disease who have a positive O-tTG, and lower percent-ge of disease controls with a positive titre in this assay is suggestivehat this novel test may truly have a higher sensitivity and speci-city than the commonly used C-tTG. Although the current C-tTGssays are highly accurate, on a population level positive predic-ive value is typically less than 80% [2,3]. For this reason, even

odest improvements in sensitivity and specificity of this cost-ffective and reproducible test would make celiac disease testingignificantly more efficient.

Whilst our study suggests a difference between the sensitivityf two tests in detecting dietary compliance some limitations areotable. First and the most important limitation of this study washat, adherence and serological status was not compared with con-urrent histological evaluation, although it is unclear whether, forssessment of GFD adherence, biopsy is superior to an expert celiacietician’s assessment [19,20]. Second, this study was performed at

single centre and limited sample size may have contributed to lackf statistical significance.

In conclusion, we present data on a novel open conformationnti-tTG assay. Because the test performance characteristics wereifferent between open and closed tTG, these tests may have differ-nt utility in different clinical settings and further study is neces-ary. Our data also suggest that the open tTG assay has the potentialo perform better than traditional closed conformation anti-tTGssays in detecting ongoing gluten exposure in celiac diseaseatients on a reported gluten free diet. Larger studies are neededo assess whether these differences in test characteristics translatento meaningful improvement in predictive value for assessmentf adherence to a gluten free diet and diagnosis of celiac disease.

inancial support

None.

onflict of interest statementaniel A. Leffler: Consultant to Alba Therapeutics, Alvine Pharma-euticals, Shire Pharmaceuticals and Prometheus Laboratories.

[

Disease 44 (2012) 375– 378

Alan C. Moss: Consultant to Prometheus Laboratories.Ciaran Kelly: received unrestricted research grants and acted as

a consultant for Alvine Pharmaceuticals, San Carlos, CA, and AlbaTherapeutics, Baltimore, MD.

Detlef Schuppan: Royalties from Eurospital, Trieste, Italy; con-sultant to Alba Therapeutics and Ferring, USA.

Acknowledgements

The open and closed conformation tTG assays were a kind gift ofDr. Ralf Pasternack and Martin Hils, Zedira, Darmstadt, Germany.

References

[1] Green PH, Cellier C. Celiac disease. N Engl J Med 2007;357:1731–43.[2] Hill ID. What are the sensitivity and specificity of serologic tests for celiac

disease? Do sensitivity and specificity vary in different populations? Gastroen-terology 2005;128(4 Suppl. 1):S25–32.

[3] Leffler D, Schuppan D. Update on Serologic Testing in Celiac Disease. Am JGastroenterol 2010;105:2520–4. Review.

[4] Vahedi K, Mascart F, Mary JY, et al. Reliability of antitransglutaminase antibod-ies as predictors of gluten-free diet compliance in adult celiac disease. Am JGastroenterol 2003;98:1079–87.

[5] Catassi C, Fabiani E, Iacono G, et al. A prospective, double-blind, placebo-controlled trial to establish a safe gluten threshold for patients with celiacdisease. Am J Clin Nutr 2007;85:160–6.

[6] Pyle GG, Paaso B, Anderson BE, et al. Low-dose gluten challenge in celiacsprue: malabsorptive and antibody responses. Clin Gastroenterol Hepatol2005;3:679–86.

[7] Leffler DA, Edwards George JB, Dennis M, et al. A prospective comparative studyof five measures of gluten-free diet adherence in adults with coeliac disease.Aliment Pharmacol Ther 2007;26:1227–35.

[8] Abdulkarim AS, Burgart LJ, See J, et al. Etiology of nonresponsive celiacdisease: results of a systematic approach. Am J Gastroenterol 2002;97:2016–21.

[9] Leffler DA, Dennis M, Hyett B, et al. Etiologies and predictors of diag-nosis in nonresponsive celiac disease. Clin Gastroenterol Hepatol 2007;5:445–50.

10] Leffler DA, Edwards-George J, Dennis M, et al. Factors that influence adher-ence to a gluten-free diet in adults with celiac disease. Dig Dis Sci 2008;53:1573–81.

11] Leffler DA, Dennis M, Edwards George JB, et al. A simple validated gluten-freediet adherence survey for adults with celiac disease. Clin Gastroenterol Hepatol2009;7. pp. 530–6, 536 e1–2.

12] Elli L, Bergamini CM, Bardella MT, et al. Transglutaminases in inflammation andfibrosis of the gastrointestinal tract and the liver. Dig Liver Dis 2009.

13] Pinkas DM, Strop P, Brunger AT, et al. Transglutaminase 2 undergoes a largeconformational change upon activation. PLoS Biol 2007;5:pe327.

14] Agardh D, Roth B, Lernmark A, et al. Calcium activation of tissue transglutam-inase in radioligand binding and enzyme-linked autoantibody immunoassaysin childhood celiac disease. Clin Chim Acta 2005;358:95–103.

15] van de Wal Y, Kooy Y, van Veelen P, et al. Selective deamidation by tissuetransglutaminase strongly enhances gliadin-specific T cell reactivity. J Immunol1998;161:1585–8.

16] Schuppan D, Junker Y, Barisani D. Celiac disease: from pathogenesis to noveltherapies. Gastroenterology 2009;137:1912–33.

17] Molberg O, Mcadam SN, Korner R, et al. Tissue transglutaminase selectivelymodifies gliadin peptides that are recognized by gut-derived T cells in celiacdisease. Nat Med 1998;4:713–7.

18] Lindfors K, Koskinen O, Kurppa K, et al. Serodiagnostic assays for celiac diseasebased on the open or closed conformation of the autoantigen, transglutaminase2. J Clin Immunol 2011;31:436–42.

19] Ciacci C, Cirillo M, Cavallaro R, et al. Long-term follow-up of celiac adults on

gluten-free diet: prevalence and correlates of intestinal damage. Digestion2002;66:178–85.

20] Lee SK, Lo W, Memeo L, et al. Duodenal histology in patients with celiac dis-ease after treatment with a gluten-free diet. Gastrointest Endosc 2003;57:187–91.