Prevalence and diagnostic outcomes of children with duodenal lesions and negative celiac serology

1 Prevalence and diagnostic outcomes of children with duodenal lesions and negative celiac serology Ida Gustafsson, BM1*, Marleena Repo, MD1*, Alina Popp, MD1,2, Katri Kaukinen, MD3, Pauliina Hiltunen, MD1 Taina Arvola, MD4, Juha Taavela MD5, Martine Vornanen, MD6, Laura Kivelä, MD1,7, Kalle Kurppa, MD1,8 *The authors have contributed equally to this article Affiliations: 1Center for Child Health Research, Faculty of Medicine and Health Technology, Tampere University and Department of Pediatrics, Tampere University Hospital, Tampere, Finland; 2Carol Davila University of Medicine and Pharmacy, National Institute for Mother and Child Health, Bucharest, Romania; 3Department of Internal Medicine, Tampere University Hospital and Celiac Disease Research Center, Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland; 4Hämeenlinna Central Hospital, Hämeenlinna, Finland and Allergy Centre, Tampere University Hospital, Tampere, Finland; 5Central Finland Central Hospital, Jyväskylä, Finland; 6Department of Pathology, Tampere University Hospital, Tampere, Finland; 7University of Helsinki and Helsinki University Hospital, Children’s Hospital, and Pediatric Research Center, Helsinki, Finland; 8The University Consortium of Seinäjoki, Seinäjoki, Finland Correspondence: Kalle Kurppa, MD, PhD, Center for Child Health Research, Tampere University, Arvo Ylpön katu 34, 33520 Tampere, Finland. Email: [email protected]; Tel: +358503186255 This is the accepted manuscript of the article, which has been published in Digestive and Liver Disease , 2020, 52(3), 289-295. https://doi.org/10.1016/j.dld.2019.11.011 2 Conflict of Interest: None. Grant Support: The study was supported by the Competitive State Research Financing of the Expert Area of Tampere University Hospital and the Foundation for Pediatric Research. Electronic word count (excluding abstract, references, tables and figures): 2916 Prospective part of this study is registered at clinicaltrials.gov; trial identifying number NCT02072590. Abbreviations: AIE, autoimmune enteropathy; ARA, antireticulin antibody; EGD, esophagogastroduodenoscopy; EmA; endomysium antibody; HLA, human leucocyte antigen; IBD, inflammatory bowel disease; IEL, intraepithelial lymphocyte; PVA, partial villous atrophy; SVA, subtotal villous atrophy; tTGab, tissue transglutaminase antibody; TVA, total villous atrophy 3 Abstract Background: Celiac disease diagnostics begin by measuring autoantibodies, which may fail to identify seronegative patients. Duodenal lesion in the absence of antibodies is scarcely studied, especially in children. Aims: To investigate the prevalence and diagnostic outcomes of children with seronegative duodenal lesion in two countries with different disease profiles. Methods: Medical data, including the results of histology and transglutaminase (tTGab) and endomysium (EmA) antibody measurements were collected from 1,172 Finnish and 264 Romanian children with systematic duodenal sampling. Database of 509 Finnish children with celiac disease was examined to identify earlier seronegative patients. Results: Celiac disease was diagnosed in 307 Finnish and 83 Romanian children in the endoscopy cohorts. No seronegative patients were found among 899 celiac disease patients, although some were only tTGab or EmA positive. Non-celiac duodenal lesion was detected in eight Finnish and 32 Romanian children, their most common diagnoses being inflammatory bowel disease and infections, respectively. Six children with morphological lesion received no diagnosis. None of them developed celiac disease during a follow-up of 3–11 years. Conclusion: Pediatric seronegative celiac disease is exceptional in the era of modern autoantibodies. Other reasons for duodenal lesion should therefore be sought, bearing in mind possible differences across countries. Keywords: Antibodies; Biopsy; Seronegative; Duodenum; Lesion 4 Introduction The estimated prevalence of celiac disease is 1-3%, emphasizing the importance of its practical diagnosis.1 The traditional approach has been identification of duodenal lesion, but the revised European criteria allow non-invasive diagnosis in children.2 Recent studies support the accuracy of the serology-based guidelines,3,4 but there may be some caveats in the current diagnostic approach. Both European and US diagnostic algorithms2,5 start from the measurement of tissue transglutaminase antibodies (tTGab), which may fail to identify seronegative patients and leave them exposed to long-term complications. Known reasons for false-negative serology, such as IgA deficiency, can be taken into account during the evaluation,3 but at least adults may have a true seronegative celiac disease.6 The differential diagnostics of seronegative duodenal lesions is challenging, as has also been reported in non-celiac conditions,7-9 and the milder the lesion the more unspecific it is.10,11 Furthermore, early stage celiac disease may be associated with low and/or fluctuating tTGab levels, decreasing the utility of serology in these circumstances.12,13 An additional challenge is posed by the current approach to also take biopsies from the less specific proximal duodenum.14 At present, systematic studies on the frequency of seronegative celiac disease and other causes for duodenal lesion in children are lacking. This issue is complicated by the varying practices of endoscopic sampling and the non-standardization of tTGab assays. Our systematic duodenal sampling in each endoscopy, together with the long tradition of celiac disease research and use of validated histology,15-16 enabled us to explore the outcomes of seronegative lesions. We moreover managed to utilize a corresponding Romanian cohort to compare two countries with disparate disease profiles, and a large Finnish research database to gain historical perspective. 5 Materials and Methods Patients and study design The study was conducted at the Tampere University Hospital and at the National Institute for Mother and Child Health, Bucharest. Three cohorts were included: 1. The Finnish endoscopy cohort (Cohort 1) was composed by collecting the indications and findings of 2,395 consecutive gastrointestinal endoscopies carried out 2007-14 in Tampere. After exclusion of repeat endoscopies and cases without duodenal biopsies, the remaining 1,172 diagnostic esophagogastroduodenoscopies (EGD) were categorized on the basis of the presence of duodenal pathology (Figure 1). Children with pathologic findings were further divided into those with and without celiac disease and the latter group further into subjects with and without morphological duodenal lesion. From 2012 onwards >80% of children with suspected celiac disease have been prospectively enrolled. 2. The Romanian cohort (Cohort 2) was formed by evaluating results of the duodenal histology of 270 consecutive children referred to EGD in 2007-15 from the National Institute for Mother and Child Health. Of these, diagnostic EGD with biopsies was performed on 264 subjects, who were selected for corresponding analyses as in Cohort 1 (Figure 2). All children since 2012 have been invited to a prospective study. 3. Cohort 3 was extracted from a research database containing information on 1,070 Finnish children diagnosed with celiac disease. Patients diagnosed from 2007 onwards overlapped with Cohort 1 and were excluded, as were those with unclear diagnosis, leaving altogether 509 eligible children (Figure 3). 6 Demographic data and medical histories were collected from all children. The results of additional investigations, such as laboratory measurements, colonoscopy, magnetic resonance enterography and capsule endoscopy, were also recorded, along with the diagnosis. If no diagnosis was established, relevant follow-up data were collected until May 31, 2018. Celiac disease serology and genetics Previously used serum antigliadin antibodies were ignored due to their low specificity,3 as were the infrequently used deamidated gliadin antibodies. Serum antireticulin antibodies (ARA) were measured in the 1980/90s before being gradually replaced by measurement of endomysium antibodies (EmA). Tests for tTGab become available in 1998 and in the 2000s tTGab and EmA have been used almost invariably. In Finnish patients, ARA (Cohort 3) and EmA (Cohorts 1 and 3) have been analyzed by immunofluorescence utilizing rat liver, kidney and stomach tissue (ARA) or human umbilical cord (EmA) as a substrate.17,18A titre 1: ≥5 is considered positive. tTGab (Cohorts 1 and 3) have been measured either by conventional ELISA (Phadia®, Uppsala, Sweden) or, since 2011, by automatized EliA assay (Phadia®) considering values ≥7.0 U/l positive. Some Cohort 3 patients had been tested in the 1990s with other tTGab assays. Romanian patients (Cohort 2) have been tested for EmA by monkey esophagus-based assay (NOVA Lite®, INOVA Diagnostics, San Diego, CA, positive 1: ≥5) and for tTGab either by Phadia® ELISA or by QUANTA Lite® (INOVA; positive ≥20 U). Seronegative celiac disease denotes constant negativity of autoantibodies in a subject with a biopsy-proven diagnosis. Serological testing at the referral site was also reckoned if performed with equal test. Possible explanations for discrepancies between serology and Clinical data 7 histology, such as IgA deficiency, reduced gluten consumption and immunosuppression, were scrutinized, as well as was testing with IgG class antibodies in the case of IgA deficiency. The presence of human leucocyte antigen (HLA) DQ2/DQ8, if analyzed, was recorded. Lack of these haplotypes has a very high negative predictive value for celiac disease.19 Histology Almost all duodenal samples in Cohorts 1 and 2 were taken upon EGD, likewise samples after the mid-1980s in Cohort 3. Earlier samples in Cohort 3 and a few samples in Cohort 2 were taken by Watson capsule. In our routine, at least four biopsies are systematically obtained from distal duodenum during EGD and, since 2012, at least two samples also from the bulb.2 The biopsies are processed and interpreted by pathologists with expertise in the alimentary tract. Only representative and well-orientated biopsy cuttings are accepted for morphometric analyses made from at least three adjacent villous-crypt pairs. Duodenal abnormalities were categorized into morphological and non-morphological.20 The more detailed classification of histology, e.g. the frequency and distribution (regionally and along the villous-crypt axis) of intraepithelial lymphocytes (IELs) and other inflammatory cells, and presence of granulomas and abnormal number of plasma cells,8,9 was recorded, as was the categorization of villous atrophy to partial (PVA), subtotal (SVA) and total (TVA). Potential celiac disease was defined as positivity to tTGab/EmA without morphological changes of the duodenal mucosa. Besides the traditional histology, some of the biopsies from the prospective series were snapfrozen and used in the determination of small-bowel mucosal TG2 targeted IgA deposits6 8 and γδ+ IELs. The analysis was conducted using immunohistochemical double staining of the mucosal IgA and tTG with monoclonal antibodies.21 Statistics Quantitative variables are presented as number of patients, percentages or medians with upper and lower quartiles. All data were analyzed anonymously using SPSS version 20.0 (IBM, Armonk, NY, USA). Ethical aspects The Ethical Committees of the Pirkanmaa Hospital District (Code R11187, 21th Feb 2012) and of the University of Medicine and Pharmacy “Carol Davila” and the National Institute for Mother and Child Health “Alessandrescu-Rusescu” (Code 01242, 12th Sep 2011) approved the prospective studies. All families gave written informed consent. The Department of Pediatrics, Tampere University Hospital and the National Institute for Mother and Child Health approved collection of the retrospective data. Results Cohort 1 The median age of the 1,172 children in Cohort 1 was 8.7 years and 54.4% were girls. The most common diagnoses were celiac disease (Table 1) and inflammatory bowel disease (IBD, n=132). Altogether 366 children had abnormal duodenal histology, including all celiac disease patients (Figure 1). tTGab was measured from 272 (88.6%), EmA from 212 (69.1%), and 9 both from 210 (68.4%) of those with celiac disease. All of them were positive for either one or both antibodies (Figure 1). Two EmA positive (titers 1:50 and 1:5; PVA in both) patients had negative tTGab (values 5.1 U/l and 5.2 U/l) and seven EmA negative cases positive tTGab (8.8-21 U/l). Seven of these nine subjects had PVA, one TVA, and one non-specified morphological damage. Eight seronegative children had morphological duodenal lesion (Figure 1), none of whom had findings indicative of celiac disease (Table 2). Five children with IBD and the one with autoimmune enteropathy (AIE) responded to their disease-specific treatments and the rotavirus patient had a normal repeat biopsy (Table 2). The non-diagnosed child has not received any diagnosis during a follow-up of six years, the symptoms have diminished and she has remained seronegative. Fifty-one children had non-morphological duodenal abnormalities (Figure 1), including 14 with changes only in the bulb. Twelve cases had potential celiac disease, five of whom subsequently received a diagnosis. No other cases with celiac disease have so far been diagnosed. Cohort 2 The median age of the 264 Romanian children was 6.3 years and 51.1% were girls. Altogether 171 children had abnormal duodenal histology (Figure 2). Celiac disease was diagnosed in 83 children (Table 1), of whom all were seropositive, including one with IgA deficiency (Figure 2). Of the celiac patients, tTGab was measured from 82 (98.8%), EmA from 77 (92.8%), and both from 76 (91.6%). None of the EmA positive patients had negative tTGab, while one EmA negative case had positive tTG-ab. The latter child had normal villi 16 10. Lähdeaho ML, Kaukinen K, Collin P, et al. Celiac disease: from inflammation to atrophy: a long-term follow-up study. J Pediatr Gastroenterol Nutr 2005;41:44-8. 11. Shmidt E, Smyrk T, Faubion W, et al. Duodenal intraepithelial lymphocytosis with normal villous architecture in pediatric patients: Mayo Clinic experience, 2000-2009. J Pediatr Gastroenterol Nutr 2013;56:51-5. 12. Simell S, Hoppu S, Hekkala A, et al. Fate of five celiac disease-associated antibodies during normal diet in genetically risk children observed from birth in a natural history study. Am J Gastroenterol 2007;102:2026-35. 13. Auricchio R, Tosco A, Piccolo E, et al. Potential celiac children: 9-year follow-up on a gluten-containing diet. Am J Gastroenterol 2014;109:913-21. 14. Taavela J, Popp A, Korponay-Szabó IR, et al. A prospective study on the usefulness of duodenal bulb biopsies in celiac disease diagnosis in children: urging caution. Am J Gastroenterol 2016;111:124-33. 15. Kivelä L, Kaukinen K, Lähdeaho ML, et al. Presentation of celiac disease in Finnish children is no longer changing: A 50-year perspective. J Pediatr 2015;167: 1109-15. 16. Taavela J, Koskinen O, Huhtala H, et al. Validation of morphometric analyses of small-intestinal biopsy readouts in celiac disease. PLoS ONE 2013;8:e76163. 17. Ladinser B, Rossipal E, Pittchieler K, et al. Endomysium antibodies in coeliac disease: an improved method. Gut 1994;35;776-8. 18. Unsworth DJ. Serological diagnosis of gluten sensitive enteropathy. J Clin Pathol 1996;49:704-711. 19. Kaukinen K, Partanen J, Mäki M, et al. HLA-DQ typing in the diagnosis of celiac disease. Am J Gastro 2002;97:695-9. 17 20. Rostami K, Marsh MN, Johnson MW, et al. ROC-king onwards: intraepithelial lymphocyte counts, distribution & role in coeliac disease mucosal interpretation. Gut 2017;66:2080-6. 21. Järvinen TT, Kaukinen K, Laurila K, et al. Intraepithelial lymphocytes in celiac disease. Am J Gastro 2003;98:1132-7. 22. Agardh D, Lee HS, Kurppa K, et al. Clinical features of celiac disease: a prospective birth cohort. Pediatrics 2015;135:627-34. 23. Abrams JA, Diamond B, Rotterdam H, et al. Seronegative celiac disease: increased prevalence with lesser degrees of villous atrophy. Dig Dis Sci 2004;49;546-50. 24. Volta U, Gaio G, Boschetti E, et al. Seronegative celiac disease: shedding light on an obscure clinical entity. Dig Liv Dis 2016;48:1018-22. 25. Vivas S, Ruiz de Morales JG, Riestra S, et al. Duodenal biopsy may be avoided when high transglutaminase antibody titers are present. World J Gastroenterol 2009;15:477580. 26. Korponay-Szabo IR, Dahlbom I, Laurila K, et al. Elevation of IgG antibodies against tissue transglutaminase as a diagnostic tool for coeliac disease in selective IgA deficiency. Gut 2003;52:1567-71. 27. Lindfors K, Koskinen O, Kurppa K, et al. Serodiagnostic assays for celiac disease based on the open or closed conformation of the autoantigen, transglutaminase 2. J Clin Immunol 2011;31:436-42. 28. Kurppa K, Lindfors K, Collin P, et al. Antibodies against deamidated gliadin peptides in early-stage celiac disease. J Clin Gastroenterol 2011;45:673-8. 29. Koskinen O, Collin P, Lindfors K, et al. Usefulness of small-bowel mucosal transglutaminase-2 specific autoantibody deposits in the diagnosis and follow-up of celiac disease. J Clin Gastroenterol 2010;44: 483-8. 18 30. Karell K, Louka AS, Moodie SJ, et al. HLA types in celiac disease patients not carrying the DQA1*05-DQB1*02 (DQ2) heterodimer: results from the European Genetics Cluster on Celiac Disease. Hum Immunol 2013;64: 469-77. 31. Maglio M, Ziberna F, Aitoro R, et al. Intestinal production of anti-tissue transglutaminase 2 antibodies in patients with diagnosis other than celiac disease. Nutrients 2017;9:10. 32. Vegh Z, Burisch J, Pedersen N, et al. Incidence and initial disease course of inflammatory bowel diseases in 2011 in Europe and Australia: results of the 2011 ECCO-EpiCom inception cohort. J Crohns Colitis 2014;8:1506-15. 33. Jansson-Knodell C, Hujoel I, Rubio-Tapia A, et al. Not all that flattens villi is celiac disease: a review of enteropathies. Mayo Clin Proc 2018;93:509-17. 34. Hardee S, Alper A, Pashankar DS, et al. Histopathology of duodenal mucosal lesions in pediatric patients with inflammatory bowel disease: statistical analysis to identify distinctive features. Pediatr Dev Pathol 2014;17: 450-4. 35. Schiepatti A, Sanders D, Biagi F, et al. Seronegative coeliac disease: clearing the diagnostic dilemma. Curr Opin Gastroenterol 2018;34:154-8. 36. Tobin J, Sinha B, Ramani P, et al. Upper gastrointestinal mucosal disease in pediatric Crohn disease and ulcerative colitis: a blinded, controlled study. J Pediatr Gastroenterol Nutr 2001;32:443-8. 37. Järvinen TT, Collin P, Rasmussen M, et al. Villous tip intraepithelial lymphocytes as markers of early-stage coeliac disease. Scand J Gastroenterol 2004;39:428-33. 38. Hagopian W, Lee HS, Liu E, et al. Co-occurrence of type 1 diabetes and celiac disease autoimmunity. Pediatrics 2017;140:e20171305. 19 Figure legends Figure 1. Flowchart of the Finnish endoscopy cohort (Cohort 1). AIE, autoimmune enteropathy; CD, celiac disease; IBD, inflammatory bowel disease. aDuodenal samples lacking from 10 children; bIgG class autoantibodies positive; cNo diagnosis set during a follow-up of six years. Figure 2. Flowchart of the Romanian endoscopy cohort (Cohort 2). CD, celiac disease; CMA, cow´s milk allergy; IBD, inflammatory bowel disease. aIgG class autoantibodies positive Figure 3. Flowchart of the Finnish celiac disease cohort (Cohort 3). AGA, antigliadin antibodies; ARA, antireticulin antibodies; CD, celiac disease; EmA, endomysial antibodies; tTG, tissue transglutaminase antibodies. aARA, EmA and tTG 20 Appendix A. Supplementary data Supplementary Figure 1. Examples of misinterpreted histology (A-B) and differential diagnostics of non-celiac duodenal lesion (C). In panel A, incorrect cutting angle, demonstrated by the cross-sections of the mucosal crypts, gives a false impression of normal duodenal villi. Correct orientation with longitudinally cut crypts in panel B reveals clear villous shortening and crypt hyperplasia. Panel C shows a biopsy cutting of a 4-month-old child with autoimmune enteropathy (AIE). Histopathology can be challenging to distinguish from celiac disease, but characteristics for AIE are predominantly mononuclear inflammation in the lamina propria, crypt apoptosis, and loss of goblet cells. Here the patient had also never ingested gluten and demonstrated excellent clinical and histological response to tacrolimus treatment. Obtaining representative biopsy with correct cutting angle was hampered by the small size of the specimen taken with infant endoscope. 21 [Figure 1.] 22 [Figure 2.] 23 [Figure 3.] 24 Table 1. Characteristics of celiac disease patients in the three study cohorts Finnish endoscopy cohort Romanian endoscopy cohort Finnish research database N:o of patients 307 83 402 Age, median (quartiles),years 7.4 (4.6, 11.6) 5.3 (3.1, 6.4) 8.2 (5.0, 12.0) Girls, % 66.4 62.7 59.2 Main clinical presentation, % Gastrointestinal 73.3 55.6 45.7 Extraintestinala 8.3 24.1 30.4 Screen-detected 18.4 20.4 23.9 At-risk group for celiac disease, % First-degree relative of a patient 17.9 16.7 50.4 Type 1 diabetes 7.2 16.7 14.4 Autoimmune thyroidal disease 0.7 3.7 9.4 Severity of the diagnostic lesion, % Total villous atrophy 18.8 18.6 31.8 Subtotal villous atrophy 41.4 50 39.1 Partial villous atrophy 39.7 31.4 29.1 aE.g. poor growth, arthritis, neurological symptoms, rash. 25 Table 2. Main diagnostic findings and selected follow-up data of the Finnish children presenting with duodenal atrophy without celiac disease N:o Age, yr Sex Clinical presentation tTGab, U/la EmA, titer Duodenal histology Other relevant findings Diagnosis 1. 4.8 M Bloody stools, anemia abdominal pain ND 1: <5 PVA, inflammation in lamina propria; later biopsy normal Increased ESR, ANCA and F-calprotectin; pancolitis with granulomas; positive response to IBD treatment Crohn’s disease 2. 11.7 F Diarrhea, abdominal pain, poor growth 0.3 ND PVA, inflammation in lamina propria; later biopsy inflammation Increased ESR, ASCA and F-calprotectin; proctitis and granulomas; positive response to IBD treatment Crohn’s disease 3. 14.8 M Tiredness, anemia, diarrhea, poor growth 0.5 1:<5 PVA, epithelial inflammation; later biopsy inflammation Increased ESR, ASCA and F-calprotectin, fistulae and strictures; negative HLA DQ2/8; positive response to ileal resection Crohn’s disease 4. 15.2 F Anemia abdominal pain, poor growth 0.4 ND SVA, inflammation in lamina propria and epithelium Increased ESR and F-calprotectin; pancolitis and ulcerations in ileum; positive response to IBD treatment Crohn’s disease 5. 9.9 M Anemia, bloody stools 0.2 1: <5 PVA, inflammation in lamina propria and epithelium; later biopsy normal Increased ESR and ANCA; moderate pancolitis; positive response to IBD treatment Ulcerative colitis 6. 0.3 M Diarrhea, poor growth NDb NDb TVA, inflammation in lamina propria and epithelium, crypt apoptosis Negative anti-enterocyte antibodies; normal colonoscopy; positive response to tacrolimus Autoimmune enteropathy 7. 1.8 M Abdominal pain, anemia 0.1 1: <5 PVA, inflammation in lamina propria and epithelium; later biopsy normal Normal F-calprotectin and rectal biopsy; severe reflux; one month later positive stool rotavirus antigen Rotavirus infection 8. 4.1 F Abdominal pain, hematemesis, constipation 1.2 1: <5 PVA, no inflammation; later biopsy PVA in bulb Celiac disease serology repeatedly negative. No diagnosis during long-term follow-up No ANCA, ASCA, EmA and tTGAab, anti-neutrophil cytoplasmic-, anti-Saccharomyces cerevisiae, endomysium, and tissue transglutaminase antibodies respectively; ESR, erythrocyte sedimentation rate; HLA, human leukocyte antigen; IBD, inflammatory bowel disease; ND, no data; PVA, SVA and TVA, partial, subtotal, and total villous atrophy. aCelikey®, cut off <5.0 U/l; bHe had never consumed gluten