What Is Known
- Serological diagnosis of celiac disease is highly sensitive, specific, less invasive, and more economical than biopsy-based pathway.
- Serology-based pathway for diagnosing celiac disease in selective symptomatic children with anti-tissue transglutaminase titer >10 times upper limit of normal has now become established clinical practice.
What Is New
- Study demonstrates serology-based diagnosis of celiac disease can be made reliably in high-risk asymptomatic children with anti-tissue transglutaminase titer >10 times upper limit of normal.
- Study provides further evidence supporting the proposal that all guidelines should be modified to recommend that asymptomatic children with anti-tissue transglutaminase titer >10 times upper limit of normal can be diagnosed as having celiac disease via serology-based pathway provided they have positive anti-endomysial antibody and HLA-DQ2/DQ8.
Celiac disease (CD) is defined as “an immune-mediated systemic disorder elicited by ingestion of gluten and related prolamines in genetically susceptible individuals. It is characterized by the presence of a variable combination of gluten-dependent clinical manifestations, CD-specific antibodies, human leucocyte antigen HLA-DQ2 or/and HLA-DQ8 haplotypes, and enteropathy” (1). It is caused by the ingestion of gluten, found in wheat, rye, and barley.
The prevalence of CD in children ages 7 years, based on a number of serological pediatric studies including the confidential Avon Longitudinal Study of Parents and Children (ALSPAC), is around 1% (2). Up to 90% of children with potential CD, however, may remain undiagnosed (3), as children are not routinely screened for this condition. CD can present with gastrointestinal and/or extraintestinal manifestations but can be silent (1,4,5). Initial serological screening involves measuring immunoglobulin-A (IgA) level and IgA-based anti-tissue transglutaminase (anti-tTG) titer (4,5). One percent of population has IgA deficiency, so it is important to check IgA level and if low request IgG-based anti-tTG and IgG-based anti-endomysial antibody (EMA) (5). Some children, however, from high-risk groups (first-degree relatives of patients with CD, type 1 diabetes mellitus [Ty-1 DM], Down syndrome, Turner syndrome, William syndrome, auto-immune thyroid disease, auto-immune liver disease, auto-immune ovarian failure) may be asymptomatic at diagnosis and are identified only following a positive serological screening with anti-tTG titer followed by small bowel biopsies.
The availability of more specific serological tests namely IgA-based anti-tTG and EMA, and evidence of their effective use in specific clinical settings has led to the European Society of Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) guidelines for diagnosing CD in children to be further revised in 2012 (1). Based on robust scientific evidence, these revised ESPGHAN guidelines recommend that in symptomatic children whose anti-tTG titer on screening is >10 times the upper limit of normal (>10× ULN), a diagnosis of CD can be made via serology-based pathway by a specialist without the need for small bowel biopsy provided they also have positive result for EMA and have a positive HLA-DQ2/DQ8 haplotype (1).
The ESPGHAN guidelines state that all other children, suspected of having CD, need a histological diagnosis based on endoscopic small bowel biopsies, taken while on a normal gluten-containing diet (1,4). This applies to the following group:
- All asymptomatic children who are found to be positive on screening for CD of high-risk groups or incidentally irrespective of their anti-tTG titer.
- Children symptomatic for CD but whose anti-tTG titer is <10× ULN.
- Children who have IgA deficiency and have symptoms suggestive of CD.
- Children symptomatic for CD and with anti-tTG >10× ULN but negative for HLA-DQ2/DQ8 genotype.
Bristol Royal Hospital for Children is the only tertiary referral center in the South West of England for pediatric gastroenterology and serves 12 secondary care district general hospitals (DGHs) providing pediatric care for a population of >5 million people. In 2013, the revised ESPGHAN guidelines were implemented in the South West of England via our regional pediatric gastroenterology clinical network. There have been multiple enquiries and requests from the DGH pediatricians, general practitioners, and parents as to whether asymptomatic children from high-risk groups whose anti-tTG is >10× ULN and who fulfill all other ESPGHAN criteria for a serology-based diagnosis of CD could be diagnosed via the nonbiopsy route. This clinical need forms the basis of this study.
AIMS AND OBJECTIVES
The aims and objectives of this study were:
- To examine the relationship between anti-tTG titer and histological grading in asymptomatic children diagnosed as having CD.
- To document whether the revised ESPGHAN guidelines for serology-based biopsy sparing diagnosis of CD could be reliably applied to asymptomatic children with anti-tTG >10× ULN who strictly meet the other ESPGHAN criteria for serology-based diagnostic pathway.
During a 10-year period (March 2007 to February 2017), prospective data on anti-tTG titer, age, sex, reason for screening, and histological findings have been collected at endoscopy on all children undergoing endoscopic assessment for suspected CD at Bristol Royal Hospital for Children. As diagnosis of CD on all these asymptomatic children was to be based on small bowel enteropathy detected on biopsy specimens, EMA testing and HLA typing were not indicated and from the onset was not part of the study.
All asymptomatic children referred to the gastroenterology team with raised anti-tTG titers (including those with anti-tTG >10× ULN) had upper gastrointestinal endoscopy with duodenal biopsies (4 from 2nd part of duodenum and 2 from duodenal bulb). Repeat anti-tTG titer was measured within previous 2 weeks or at the time of endoscopy using ELISA technique (normal level <10 IU/mL). All histological specimens were reported by a single specialist pediatric histopathologist. A cross-check with the database from the biochemistry laboratory for anti-tTG >10× ULN was done for both symptomatic and asymptomatic cases.
Marsh-Oberhuber (MO) classification was used to grade the histological lesions (6). The grading as suggested by MO classification was as follows: 0 = normal mucosa; MO-1 = increased number of intraepithelial lymphocytes; MO-2 = crypt hyperplasia; MO-3a = partial villous atrophy; MO-3b = subtotal villous atrophy; MO-3c = total villous atrophy (6).
Same data were also collected for all symptomatic children, but as serology-based pathway for diagnosing CD was introduced in our region in May 2013 following publication of the modified ESPGHAN guidelines in 2012, the data for symptomatic children are limited to our local population as symptomatic children with anti-tTG >10× ULN are now being diagnosed and followed up in their local DGHs. So comparison between asymptomatic and symptomatic children is limited to those between January 2007 and December 2012. Reasons for serological screening and performing subsequent small bowel biopsies on asymptomatic children with anti-tTG <10× ULN is shown in a Table 1, but has not been included in data analysis in this study.
Information was collected on cost of serology-based pathway (cost of confirmatory serology and confirmatory genetics) compared to biopsy-based diagnosis (cost of confirmatory serology and biopsy) from the laboratory and management office and cost–benefit of serology-based diagnosis calculated. As the costs for an initial appointment to arrange serological screening, a further appointment to communicate the diagnosis of CD (needed irrespective of the diagnostic pathway used), and a consultation with the pediatric dietician will be the same for both pathways, this has not been included in the cost–benefit analysis. The cost–benefit analysis is applicable only to the practice in the United Kingdom (UK) and is likely to have variations even within the different regions in the UK let alone other parts of the world.
This study was conducted as a service review with approval from the local audit department. As there were no identifiable patient data collected, ethical approval was not considered necessary.
During the 10-year period (March 2007–February 2017), there were a total of 157 asymptomatic children ages 0 to 17 years who were diagnosed as having CD following endoscopic small bowel biopsies and had histological changes concordant with a diagnosis of CD. Of these 157 patients, 73 of 157 had anti-tTG <10× ULN and were not studied in detail. Eighty-four of 157 had anti-tTG >10× ULN and formed the basis of this study. There were 34 boys and 50 girls. The mean age at diagnosis was 8.8 years (range 2.2–16.9 years). The mean age at diagnosis was 7.2 years (range 7 months to 16.2 years) for a similar cohort of patients with CD (n = 167) who were symptomatic at diagnosis. Figure 1 shows the number of children with anti-tTG >10× ULN (both symptomatic and asymptomatic) during a 6-year period from January 2007 to December 2012, who were referred for and underwent endoscopy and small bowel biopsies. When cross-checked with the biochemistry database, there were only 84 asymptomatic children with anti-tTG >10× ULN and all of them underwent small bowel endoscopy and biopsy and were subsequently diagnosed as having CD via the biopsy-based pathway.
Anti-tissue Transglutaminase Antibody Titer Less Than 10 Times Upper Limit of Normal and High-risk Group of Asymptomatic Children
Of 84 asymptomatic children with CD with anti-tTG >10× ULN, 89% (n = 75) belonged to high-risk groups. All the 84 asymptomatic children with anti-tTG titer of >10× ULN who had undergone endoscopic small bowel biopsies, had histological changes (MO 3a–3c) consistent with a definitive diagnosis of CD. Figure 2 lists the reasons that prompted serological screening in these asymptomatic children with CD (n = 84). The 2 most common reasons for screening these asymptomatic children were a preexisting diagnosis of Ty-1 DM in 36 of 84 (43%) and a positive family history of CD in first-degree relatives in 24 of 84 children (28%). The other high-risk groups (n = 15) included those with genetic associations (Down syndrome [n = 4], Turner syndrome [n = 2], William syndrome [n = 1]) and children with underlying autoimmune liver disease (n = 1), autoimmune ovarian failure (n = 1), arthropathy (n = 2), and cystic fibrosis (n = 4). The miscellaneous group included asymptomatic children in which screening with anti-tTG was done while investigating for other conditions such as food allergies (n = 1), unexplained proteinuria (n = 1), Enquiring About Tolerance (EAT) study (n = 1), and owing to family history of CD in second=degree relatives such as grandparents or cousins (n = 6).
Anti-tissue Transglutaminase Antibody Titer More Than 10 Times Upper Limit of Normal and High-risk Group of Asymptomatic Children
Asymptomatic children (73/157) had anti-tTG <10× ULN and these were not studied in details. Table 1 lists the reasons for their initial screening. Two main reasons for screening asymptomatic children were: family history of CD in first-degree relatives (n = 38) and Ty-1 DM (n = 13).
Anti-tissue Transglutaminase Antibody Titer More Than 10 Times Upper Limit of Normal and Severity of Small Bowel Enteropathy
Figure 3 shows the relationship between anti-tTG titers and severity of histological abnormality (enteropathy) based on the MO classification. There was a positive correlation between anti-tTG titer and severity of enteropathy. Total villous atrophy was present more often in those with anti-tTG of >200 IU/mL (29/53). The sensitivity and specificity of the anti-tTG titers of >10× ULN alone in correctly diagnosing CD in this cohort of asymptomatic patients (n = 84) predominantly from high-risk groups was 100%.
The cost of diagnosing CD via the biopsy-based pathway in our unit was £1340 per patient. The cost of the serology-based pathway was £65 (cost of anti-tTG: £5.86 [tested on 2 different samples], EMA: £14.30, and HLA-DQ2/DQ8 typing: £ 38.86). There is a cost difference of £1275 per patient if diagnosis can be robustly made by serology-based pathway.
This 10-year prospective study involved 84 asymptomatic children who on screening had all been found to have anti-tTG titer >10× ULN and subsequently underwent endoscopic small bowel biopsies. All 84 children had clear histological confirmation of CD with small bowel biopsies showing MO 3a–3c classification. Seventy-five of 84 belonged to groups at a high risk for developing CD. The sensitivity and specificity of anti-tTG >10× ULN in correctly diagnosing CD in these children was 100%. This study provides evidence that in asymptomatic children from high-risk groups, the anti-tTG >10× ULN is associated with small bowel enteropathy consistent with a diagnosis of CD in 100% of cases. As currently recommended in the ESPGHAN guidelines for symptomatic children, for a definitive diagnosis the asymptomatic children would also need to be positive for EMA and HLA-DQ2/DQ8 haplotype. HLA-DQ2/DQ8 haplotype was not tested in children for our study group, as they were all being diagnosed via the biopsy-based pathway. It is most likely that they would all have been positive, as high-risk groups are known to have a very high prevalence of HLA-DQ2/DQ8. For example, 92% to 94% of all children with diabetes are positive for HLA-DQ2/DQ8 haplotype (7,8).
There are known groups of asymptomatic children at high risk for developing CD wherein initial screening with anti-tTG titer and IgA level is recommended (1,5). These include Ty-1 DM, first-degree relatives with CD, Down syndrome, and a number of other syndromes and autoimmune conditions. In this study, the majority of asymptomatic children (75/84) with anti-tTG >10× ULN were screened for CD because they came from high-risk groups. The 2 main conditions associated with screening and subsequent diagnosis of CD in these asymptomatic children were Ty-1 DM (43%) and a family history of CD in first-degree relatives (28%).
The results of this study are in concordance with a smaller Italian study by Trovato et al (2015) (9) of 40 asymptomatic children with CD whose anti-tTG titers were >10× ULN. The authors concluded that if further similar evidence emerged, this should lead to the ESPGHAN guidelines for CD to be modified to recommend serological pathway for diagnosing CD in asymptomatic as well as symptomatic children with anti-tTG >10× ULN (9).
The most important aspect of managing CD is ensuring adherence to a strict gluten free diet (GFD). There have been theoretical concerns that a serology-based diagnosis may not be taken as seriously by the parents/children as would be the case following a histology-based diagnosis possibly leading to poor adherence to a GFD. A prospective Italian study involving 468 children which included 11% (n = 51) whose diagnosis of CD was made serologically, found no negative consequences related to clinical remission, adherence to GFD, and quality of life between the serology-based or histology-based diagnosis during a follow-up period of 1.9 years (10). A Finnish study by Kivelä et al (2017) (11) found that at-risk screened children with CD are comparable in dietary adherence to those diagnosed because of clinical manifestation.
A reliable serology-based diagnosis has many benefits. These include saving the child from having to go through an invasive upper gastrointestinal endoscopy and sedation/general anesthesia. It is also cheaper. Although the cost-saving will vary in different countries and different regions, the cost saving in the UK is around £1275 per patient.
This study from the UK together with the Italian study by Trovato et al (2015) provides evidence that in asymptomatic children from high-risk groups who are found on screening to have anti-tTG >10× ULN, diagnosis of CD can be made without the need for small bowel biopsies provided they have positive results for EMA and HLA-DQ2/DQ8 haplotype, that is, in a similar pathway to that recommended by the ESPGHAN and other guidelines for symptomatic children.
This large 10-year study of 84 asymptomatic children diagnosed as having CD following small bowel biopsy provides clear evidence that the diagnosis of CD can be made correctly and robustly in asymptomatic children from high-risk groups (n = 75) who are found on screening to have anti-tTG >10× ULN. This would be in the same way as is currently recommended by the ESPGHAN guidelines for diagnosis of CD in symptomatic children and requires that these children should also have positive results for EMA and HLA-DQ2/DQ8. Travato et al (2015) stated that there is a need for confirmation of the biopsy sparing protocol for asymptomatic children. This study provides evidence that in asymptomatic children from high-risk groups, finding of anti-tTG >10× ULN detected on screening for CD is associated with enteropathy consistent with a diagnosis of CD in 100% of cases. As already recommended by the ESPGHAN guidelines for symptomatic children, diagnosis will also require further consolidation with a positive EMA and HLA-DQ2/DQ8 haplotype. The ESPGHAN and other guidelines for diagnosis of CD in children should be modified in the light of this evidence.
1. Husby S, Koletzko S, Korponay-Szabó IR, et al. European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr
2. Bingley PJ, Williams AJ, Norcross AJ, et al. Undiagnosed coeliac disease at age seven: population based prospective birth cohort study. BMJ
3. Ravikumara M, Nootigattu VK, Sandhu BK. Ninety percent of celiac disease
is being missed. J Pediatr Gastroenterol Nutr
4. Murch S, Jenkins H, Auth M, et al. Joint BSPGHAN and Coeliac UK guidelines for the diagnosis and management of coeliac disease in children. Arch Dis Child
5. National Institute for Health and Care Excellence (September 2015). Coeliac disease: recognition, assessment and management. NICE guidelines 20. NICE, London. Available at https://www.nice.org.uk/guidance/ng20
6. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol
7. Binder E, Loinger M, Mühlbacher A, et al. Genotyping of coeliac-specific human leucocyte antigen in children with type 1 diabetes: does this screening method make sense? Arch Dis Child
8. Mitchell RT, Sun A, Mayo A, et al. Coeliac screening in a Scottish cohort of children with type 1 diabetes mellitus: is DQ typing the way forward? Arch Dis Child
9. Trovato CM, Montuori M, Anania C, et al. Are ESPGHAN “biopsy-sparing” guidelines for celiac disease
also suitable for asymptomatic patients? Am J Gastroenterol
10. Benelli E, Carrato V, Martelossi S, et al. Coeliac disease in the ERA of the new ESPGHAN and BSPGHAN guidelines: a prospective cohort study. Arch Dis Child
11. Kivelä L, Kaukinen K, Huhtala H, et al. At-risk screened children with celiac disease
are comparable in disease severity and dietary adherence to those found because of clinical suspicion: a large cohort study. J Pediatr
2017; 183:115–121. e2.