- Email: [email protected]
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal Antonella Meloni, MD, Chiara Mandas, MD, Rita De´sire´e Jores, MD, PhD, and Mauro Congia, MD Objective To study the prevalence of autoimmune thyroiditis (AT) in Sardinian children with celiac disease (CD) and the effects of a gluten-free diet (GFD) on thyroid function. Study design Children with biopsy-proven CD (n = 324; female:male 2:1; mean age, 6.6 years) followed from 1 to 15 years, were retrospectively evaluated for AT at onset of CD and during GFD. Serum thyroid peroxidase and thyroglobulin antibodies (AbTPO, AbTG), thyroid-stimulating hormone (TSH), free thyroxine (FT4), free triiodothyronine (FT3), and thyroid ultrasonography were considered. Age-matched Sardinian schoolchildren (n = 8040), previously evaluated for antithyroid antibodies and thyroid function, were used as controls. Results Thirty-four patients with CD (10.5%) developed AT (female:male 4,5:1; mean age, 10.5 years), 11 at onset of CD and 23 during GFD, with a higher prevalence than controls (P = 2.9 13). Twenty-eight patients were euthyroid and 6 hypothyroid. AbTPO and/or AbTG persisted elevated for 2 to 9 years despite the GFD in 9 of 11 patients with AT at onset of CD. Conclusions AT is strongly associated with CD in Sardinian children, has an age of onset of 10.5 years, and appears to be gluten-independent. In children with CD with AT, the female:male bias reported in adult AT is present before puberty. (J Pediatr 2009;155:51-5).
C
eliac disease (CD) is an autoimmune-mediated enteropathy triggered and maintained by the ingestion of gluten-containing cereals (wheat, rye, and barley) in genetically predisposed individuals.1 CD is considered to be a worldwide disorder, more common in Western countries. In Europe and in the United States, the disease has been found to affect about 1 in 100 people2; among Sardinians with CD, it has been reported with a similar incidence.3 Like type 1 diabetes, rheumatoid arthritis, and multiple sclerosis, CD has a strong genetic association with particular HLA class II alleles, namely the HLA genes encoding the class II DQ (a1*0501, b1*02) molecule.4 Sardinia is an ancient genetic isolate with a peculiar distribution of HLA class II alleles and haplotypes.5 The CD predisposing HLA class II DQ (a1*0501, b1*02) molecule is very frequent among Sardinians (43%), and this HLA molecule is almost always encoded in cis by the DRB1*0301, DQA1*0501, DQB1*0201 haplotype.6 An increased prevalence of autoimmune disorders, including type 1 diabetes and autoimmune thyroid diseases, has been repeatedly reported in patients with CD.7-9 Two hypotheses have been suggested to explain this association: (1) sharing of 1 or more genes responsible for CD and for the coexpressed autoimmune disease; or (2) continued gluten exposure in untreated CD might lead to loss of the intestinal barrier function and to alterations of the systemic immune response, ultimately helping to induce other autoimmune disorders.10 The first concept implies that the comorbidity associated with CD is part of the individual genetic background.11-13 On the contrary, the second concept may have important clinical implications because when the gluten induced immune reactivity and the interplay between autoimmune predisposing genes and trigger(s) is blocked, this comorbidity could be eliminated.14,15 In this retrospective study, a large cohort of 324 children with CD from Sardinia, followed for 1 to 15 years, was evaluated for the occurrence of autoimmune thyroiditis (AT). The aims of the study were to establish the prevalence of AT in CD and to verify if gluten exposure before the diagnosis of CD is correlated with development of thyroid autoimmunity, or vice versa if early gluten withdrawal is able to prevent the future development of AT.
AbTG AbTPO AT CD FT3 FT4 GFD TSH US
Thyroglobulin antibody Thyroid peroxidase antibody Autoimmune thyroiditis Celiac disease Free triiodothyronine Free thyroxine Gluten-free diet Thyroid-stimulating hormone Thyroid ultrasonography
From the Pediatric Clinic II, ‘‘Microcitemico’’ Hospital ASL 8, Department of Biological Sciences and Biotechnology, University of Cagliari, Sardinia, Italy Supported by a grant from Fritz-Thyssen-Stiftung, Germany (R.-D.J.). The authors declare no potential conflicts of interest, real or perceived. 0022-3476/$ - see front matter. Copyright Ó 2009 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2009.01.013
51
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Methods Patients Children with biopsy-proven CD (n = 324) from Sardinia (223 girls; mean age at diagnosis, 6.6 years; range, 10 months to 18 years), followed between 1992 and 2007 (mean followup period, 8 years) at the Pediatric Gastroenterological Unit in Cagliari, Italy, were included in the study. Clinical data were collected retrospectively; affected family members of probands, patients with preexisting autoimmune conditions (type 1 diabetes, autoimmune thyroid disorders, Addison’s disease), and patients with Down syndrome or Turner syndrome were excluded from the study. Screening for Celiac Disease The diagnosis of CD was based on the revised criteria of the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition,16 for example, clinical history, positivity of IgA antiendomysial antibody, typical histological features on small intestinal biopsy, and clinical response to the glutenfree diet (GFD) (Table I).17 All patients were typed for anti-gliadin antibody (AGA, both IgA and IgG) and antiendomysial antibody (EMA); most patients were also typed for anti-tissue transglutaminase antibody IgA (tTG-IgA) and anti-actin antibody IgA (AAA-IgA). The latter has been previously associated with the severity of intestinal damage.18 Mucosal atrophy has been graded according to the Marsh classification,19 modified by Oberhuber in type 3a, 3b, and 3c.20 After diagnosis of CD, all patients followed a GFD, and the dietary compliance was evaluated by assessment of AGA, EMA, and tTG every 12 months. Thyroid Function Assessment Screening for thyroid peroxidase antibody (AbTPO) and thyroid globulin antibody (AbTG) was performed in all 324 patients at the onset of CD and once per year after the institution of the GFD. When abnormal values were detected, a complete thyroid evaluation, including serum-free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), and thyroid ultrasonography, was performed. Thyroid autoantibody serum titers were detected using competitive radioimmunoassay (RIA Medical System; Genoa, Italy) with coated-tube technique thyroperoxidase antibodies (AbTPO, normal values <50 IU/mL), RIA-immunoradiometric assay (IRMA, Medical System; Genoa, Italy) with solid-phase technique antithyroglobulin (Ab TG, normal values <100 IU/mL). Serum FT3 and FT4 were detected by RIA (Medical System; Genoa, Italy) using commercial kits (normal values, 0.8 to 2 ng/mL for FT4; 1.40 to 4.2 pg/mL for FT3), whereas serum TSH was measured by a chemiluminescent method (Medical System; Genoa, Italy), with normal values ranging between 0.3 and 5 mU/mL. Ultrasonography of the thyroid gland was performed with a 7.5-MHz linear probe. Thyroid echographic images were classified and graded in 4 patterns according to Sostre 52
Vol. 155, No. 1
Table I. Clinical manifestations and serological markers of CD in 324 Sardinian children Patients (%) Clinical form Classic symptoms Nonclassic symptoms Iron deficiency anemia Failure to thrive Abdominal pain Pubertal delay Vomiting Diarrhea Chronic constipation Silent Serological markers AGA IgA AGA IgG tTG-IgA EMA AAA-IgA
39 56 22 18 5 4 3 2 2 5 73 88 95 99 64
et al21 and Marcocci et al.22 We restricted the diagnosis of AT to patients with high titers of AbTPO and/or AbTG associated with an abnormal thyroid echographic pattern. Therefore, patients with mild elevation of AbTPO and/or AbTG were eliminated from the statistical analysis to avoid possible overestimates. Thyroid function was classified according to the American Thyroid Association Guidelines.23 Control Subjects We compared the frequency of AT among patients with CD with the prevalence of antithyroid antibodies (ATA) in an age-matched Sardinian background population also evaluated for thyroid function.24 In this cohort of 8040 schoolchildren (4194 boys, 3846 girls, ages 6 to 15 years) ATA were detected in 235 children (2.92%). Statistical Analysis Data were evaluated by c2 test with Yates corrections, or by the Wilcoxon test, setting the level of significance at .05.
Results Clinical data of patients with CD with AT are summarized in Table II (available at www.jpeds.com). Overall AT was found in 34 (28 girls, 6 boys) of 324 patients with CD; 23 were on GFD when AT was diagnosed (Table III). Considered as a whole, a high prevalence of AT among children with CD (10.5%), compared with the Sardinian pediatric background population (2.92%), was found (P = 2.9 13). This prevalence is similar to or even lower than that reported in other studies conducted in adults and children.12,25 Euthyroidism was observed in 28 of 34 patients (82.4%), and 6 were hypothyroid. In 3, hypothyroidism preceded diagnosis of CD, and the other 3 patients developed hypothyroidism during GFD (Table II). The thyroid ultrasonography showed an abnormal pattern with varying severity, suggestive of an autoimmune infiltrate Meloni et al
ORIGINAL ARTICLES
July 2009
Table III. Prevalence of AT among patients with CD according to sex, age (in years), and treatment status No. of patients Female/male (ratio) Mean age SD Age range
CD
CD + AT
AT in untreated patients (GCD)
AT in treated patients (GFD)
324 223/101 (2.2:1) 6.6 4.5* 1-18
34/324 28/6 (4.5:1) 10.5 2.9* 5-16
11/34 10/1 (10:1) 9.5 2.9 5-15
23/34 18/5 (3.6:1) 11.0 2.8 7-16
*Statistically significant difference (P< 1.4 6) between the age of onset of CD alone and CD with AT.
(Table II). Two patients with isolated high levels of AbTPO and AbTG at CD diagnosis developed an overall AT with a grade 2 of Sostre after 1 year of GFD. A significant difference in the age of onset between patients with CD and patients with CD and AT was observed: 6.6 versus 10.5 years, respectively (P < 1.4 6; Table III). We were aware that a detection bias might have influenced the different ages of diseases onset because, by definition, AT was looked for only after the diagnosis of CD. On the other hand, no significant difference in the age of AT onset among patients at CD diagnosis (9.5 2.9) and patients on GFD (11.0 2.8) was found (Table III). The female/male ratio was higher in patients with CD and AT (F/M 4.5:1), compared with patients with CD alone (F/M 2.2:1; Table III). This finding was unexpected in our cohort of pediatric patients because a similar high F/M ratio in AT is commonly described in late adolescence and adulthood.26,27 Thus, to determine a possible role of sex hormones as an explanation for the observed F/M ratio, patients were subdivided into 2 groups: prepubertal (age #12 years) and pubertal (age >12 years) (Table IV). Unexpectedly, the F/ M ratio in prepubertal (12:1) was even higher than in pubertal patients (1:1), and the difference, when compared with the F/M ratio of prepubertal patients with CD alone (2:1), reached statistical significance (P < .0068; Table IV). The Figure shows the relationship of AT with the duration of gluten exposure. At onset of CD, 11 of 324 patients receiving a gluten-containing diet were affected by AT. After diagnosis of CD, an additional 23 patients placed on GFD developed AT after 1 to 11 years, and 3 of these shifted toward hypothyroidism. No difference was found in the serum titers of AbTPO, AbTG, and thyroid echographic pattern between patients with AT coexisting at the onset of CD and patients developing AT during GFD (Table II). In 9 of 11 patients with AT coexisting at the diagnosis of CD, very high titers of AbTG and/or AbTPO, suggestive of active AT, persisted despite 1 to 9 years of GFD (data not shown).
Discussion Our study documents the association between AT and CD in Sardinian children. Because it includes a large number of subjects, it provides the actual prevalence of AT in our pediatric population with CD. Previous studies on the prevalence of AT in Sardinian patients with CD, or vice versa on the occurrence of CD in AT, have been conducted only in adults.9,28 We found that the prevalence of AT in CD is about 4 times greater than that found in the age matched Sardinian schoolchildren background population, indicating that CD is a risk factor predisposing to AT. There are several possible explanations for the increased risk of developing AT in patients with CD: (1) the CD condition carries 1 or more genes (MHC or other autoimmunity genes) favoring AT; and (2) the timing of CD diagnosis and duration of gluten exposure might increase toxic and immunological effects of gluten, increasing the risk of AT.14,15,29 A lack of correlation between CD status and occurrence of AT, favoring the first hypothesis, is suggested by our findings. Indeed, AT developed in patients on a gluten-containing diet or in patients on GFD; in these latter patients, after many years (Table II; Figure) GFD did not improve the outcome of the disease, as demonstrated by very high titers of AbTG and/or AbTPO after 1 to 9 years of follow-up. There was no significant difference in the mean age between patients with AT coexisting at onset of CD and patients developing AT during the GFD: around 10.5 years in both groups, indicating that AT is independent of gluten exposure (Table III). The age of patients with CD with AT is the same observed in Sardinian children with isolated AT.30 Finally, the significant difference between the age of onset of CD (6.6 years) and CD with AT (10.5 years) suggests that CD and AT are two different autoimmune diseases, which association is likely explained by a shared genetic background. Our data, showing that the duration of gluten exposure is not of crucial importance in favoring the development of
Table IV. Comparison of female/male ratio according to age (prepubertal versus pubertal) in patients with CD and patients with CD and AT Patients
CD
F
M
F/M
CD + AT
F
M
F/M
P value
Total (n) Prepubertal (#12 y) Pubertal (>12 y)
290 254 36
195 169 26
95 85 10
2:1 2:1 2.6:1
34 25 8
28 24 4
6 2 4
4.6:1 12:1 1:1
NS <.007 NS
P value determined by c2 test. F, female; M, male; NS, P value not significant.
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal
53
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 155, No. 1
Figure. Vertical bars represent the number of patients with CD diagnosed with AT. The first bar represents patients with AT at CD diagnosis, the following represent patients developing AT during GFD.
AT, are in agreement with some authors11-13 and in contrast with others.14,15,29 Ventura et al14 found that the prevalence of AT correlated with the duration of gluten exposure and augmented with the age at CD diagnosis. In addition, thyroid autoantibodies had a tendency to disappear following the GFD. Possible explanations for these variable results could depend on different follow-up periods (15 years in our study and 6 months in the work of Ventura et al) and on age stratifications chosen by Ventura that did not take into account that the prevalence of certain autoimmune diseases raises with aging. A peculiar and previously unreported finding of our study concerns the F/M ratio of patients with CD and AT and the relationship between AT and puberty. The prevalence of girls in AT is well known, and the sex ratio favoring females is reported to be age-related, starting generally after puberty.26,27 This observation has supported the belief in a fundamental role of sexual hormones in favoring the development of AT. Accordingly, in the Sardinian schoolchildren population,24 a significant increase of thyroid autoimmunity was more evident in females than in males, in particular over 11 years (P < .01). Interestingly, in patients with CD with AT, the high F/M ratio was observed before puberty (Table IV), suggesting either a more complicated role of the sex hormones in conferring predisposition to AT, or, alternatively, an influence of the genetic background of CD in anticipating the high F/M ratio of AT. The high frequency of AT among patients with CD, even on GFD, suggests clinical usefulness for a longitudinal follow-up of thyroid function. n 54
The authors are grateful to Prof. A. Cao for critical reading of the manuscript. Submitted for publication Aug 29, 2008; last revision received Nov 24, 2008; accepted Jan 8, 2009. Reprint requests: Dr Antonella Meloni, Pediatric Clinic II, ‘‘Microcitemico’’ Hospital ASL 8, Department of Biological Sciences and Biotechnology, University of Cagliari, Sardinia, Italy. E-mail: [email protected].
References 1. Green PH, Cellier C. Celiac disease. N Engl J Med 2007;357:1731-43. 2. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology 2001;120:636-51. 3. Meloni G, Dore A, Fanciulli G, Tanda F, Bottazzo GF. Subclinical coeliac disease in schoolchildren from northern Sardinia. Lancet 1999;353:37. 4. Sollid LM. Coeliac disease: dissecting a complex inflammatory disorder. Nat Rev Immunol 2002;2:647-55. 5. Lampis R, Morelli L, De Virgiliis S, Congia M, Cucca F. The distribution of HLA class II haplotypes reveals that the Sardinian population is genetically differentiated from the other Caucasian populations. Tissue Antigens 2000;56:515-21. 6. Congia M, Frau F, Lampis R, Frau R, Mele R, Cucca F, et al. A high frequency of the A30, B18, DR3, DRw52, DQw2 extended haplotype in Sardinian celiac disease patients: further evidence that disease susceptibility is conferred by DQ A1*0501, B1*0201. Tissue Antigens 1992;39:78-83. 7. Collin P, Reunala T, Pukkala E, Laippala P, Keyrilainen O, Pasternack A. Coeliac disease–associated disorders and survival. Gut 1994;35:1215-8. 8. Counsell CE, Taha A, Ruddell WS. Coeliac disease and autoimmune thyroid disease. Gut 1994;35:844-6. 9. Velluzzi F, Caradonna A, Boy MF, Pinna MA, Cabula R, Lai MA, et al. Thyroid and celiac disease: clinical, serological, and echographic study. Am J Gastroenterol 1998;93:976-9. 10. Fasano A. Systemic autoimmune disorders in celiac disease. Curr Opin Gastroenterol 2006;22:674-9.
Meloni et al
July 2009 11. Sategna Guidetti C, Solerio E, Scaglione N, Aimo G, Mengozzi G. Duration of gluten exposure in adult coeliac disease does not correlate with the risk for autoimmune disorders. Gut 2001;49:502-5. 12. Ansaldi N, Palmas T, Corrias A, Barbato M, D’Altiglia MR, Campanozzi A, et al. Autoimmune thyroid disease and celiac disease in children. J Pediatr Gastroenterol Nutr 2003;37:63-6. 13. Viljamaa M, Kaukinen K, Huhtala H, Kyronpalo S, Rasmussen M, Collin P. Coeliac disease, autoimmune diseases and gluten exposure. Scand J Gastroenterol 2005;40:437-43. 14. Ventura A, Magazzu G, Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease: SIGEP Study Group for Autoimmune Disorders in Celiac Disease. Gastroenterology 1999;117:297-303. 15. Ventura A, Neri E, Ughi C, Leopaldi A, Citta A, Not T. Gluten-dependent diabetes-related and thyroid-related autoantibodies in patients with celiac disease. J Pediatr 2000;137:263-5. 16. Walker-Smith J, Guandalini S, Schmitz J, Shmerling DH, Visakorpi J. Revised criteria for diagnosis of coeliac disease: Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child 1990;65:909-11. 17. Fasano A. Clinical presentation of celiac disease in the pediatric population. Gastroenterology 2005;128:S68-73. 18. Clemente MG, Musu MP, Troncone R, Volta U, Congia M, Ciacci C, et al. Enterocyte actin autoantibody detection: a new diagnostic tool in celiac disease diagnosis: results of a multicenter study. Am J Gastroenterol 2004;99:1551-6. 19. Marsh MN, Crowe PT. Morphology of the mucosal lesion in gluten sensitivity. Baillieres Clin Gastroenterol 1995;9:273-93. 20. Oberhuber G. Histopathology of celiac disease. Biomed Pharmacother 2000;54:368-72. 21. Sostre S, Reyes MM. Sonographic diagnosis and grading of Hashimoto’s thyroiditis. J Endocrinol Invest 1991;14:115-21.
ORIGINAL ARTICLES 22. Marcocci C, Vitti P, Cetani F, Catalano F, Concetti R, Pinchera A. Thyroid ultrasonography helps to identify patients with diffuse lymphocytic thyroiditis who are prone to develop hypothyroidism. J Clin Endocrinol Metab 1991;72:209-13. 23. Surks MI, Chopra IJ, Mariash CN, Nicoloff JT, Solomon DH. American Thyroid Association guidelines for use of laboratory tests in thyroid disorders. JAMA 1990;263:1529-32. 24. Loviselli A, Velluzzi F, Mossa P, Cambosu MA, Secci G, Atzeni F, et al. The Sardinian Autoimmunity Study: 3, studies on circulating antithyroid antibodies in Sardinian schoolchildren: relationship to goiter prevalence and thyroid function. Thyroid 2001;11:849-57. 25. Sategna-Guidetti C, Volta U, Ciacci C, Usai P, Carlino A, De Franceschi L, et al. Prevalence of thyroid disorders in untreated adult celiac disease patients and effect of gluten withdrawal: an Italian multicenter study. Am J Gastroenterol 2001;96:751-7. 26. Chiovato L, Lapi P, Fiore E, Tonacchera M, Pinchera A. Thyroid autoimmunity and female gender. J Endocrinol Invest 1993;16:384-91. 27. Demirbilek H, Kandemir N, Gonc EN, Ozon A, Alikasifoglu A, Yordam N. Hashimoto’s thyroiditis in children and adolescents: a retrospective study on clinical, epidemiological and laboratory properties of the disease. J Pediatr Endocrinol Metab 2007;20:1199-205. 28. Meloni GF, Tomasi PA, Bertoncelli A, Fanciulli G, Delitala G, Meloni T. Prevalence of silent celiac disease in patients with autoimmune thyroiditis from Northern Sardinia. J Endocrinol Invest 2001;24:298-302. 29. Toscano V, Conti FG, Anastasi E, Mariani P, Tiberti C, Poggi M, et al. Importance of gluten in the induction of endocrine autoantibodies and organ dysfunction in adolescent celiac patients. Am J Gastroenterol 2000;95:1742-8. 30. Prinzis AGM, Cambuli VM, Concas D, Foschini ML, Marras V, Carta D, Loche S, Mariotti S. The female sex predominance in Hashimoto’s thyroiditis is a pubertal phenomenon. J Endocrinol Invest 2007;30:97.
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal
55
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 155, No. 1
Table II. Clinical data at disease onset of 34 pediatric patients with CD with AT CD AT AbTPO AbTg Patient Sex onset onset U/mL IU/mL Ultrasonography (Sostre grade) 1* 2† 3 4 5 6† 7 8† 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23* 24* 25 26 27 28 29 30 31 32 33 34
F F F F F F F F F F M F F F F M F F F F F F F F F M F M F F M M F F
5 7 7 7 8 10 11 11 12 12 15 2 2 6 5 4 5 7 8 3 4 6 7 1 1 6 6 12 10 6 13 5 7 15
5 7 7 7 8 10 11 11 12 12 15 7 7 7 8 9 9 9 9 10 10 10 10 11 12 12 12 13 14 14 14 15 16 16
1000 440 970 555 1000 250 900 300 140 4494 615 1000 100 1000 100 550 55 4000 260 1900 1926 420 5900 2600 1000 255 1000 3400 110 500 170 235 250 430
327 250 840 1500 800 600 600 280 160 314 215 370 288 100 1500 150 360 6400 380 350 340 600 225 230 730 67 50 1200 480 314 22 222 100 250
3 2 3 2 2 1 2 2 1 3 3 3 1 2 1 2 3 4 2 1 2 2 3 3 2 1 3 2 2 2 2 2 2 2
The first column shows the patients with CD with AT numbered from 1 to 34; the second shows the sex; the third and fourth show the age at CD and at AT onset in years; the fifth and sixth show the titers of circulating thyroid autoantibodies; the seventh shows the thyroid echographic pattern. Normal ranges: AbTPO <50 IU/mL, AbTG <100 IU/mL. *Patients 1, 23, 24: autoimmune hypothyroidism after 1, 7, and 11 years of GFD, respectively. †Patients 2, 6, and 8: autoimmune hypothyroidism at onset of CD.
55.e1
Meloni et al
C
eliac disease (CD) is an autoimmune-mediated enteropathy triggered and maintained by the ingestion of gluten-containing cereals (wheat, rye, and barley) in genetically predisposed individuals.1 CD is considered to be a worldwide disorder, more common in Western countries. In Europe and in the United States, the disease has been found to affect about 1 in 100 people2; among Sardinians with CD, it has been reported with a similar incidence.3 Like type 1 diabetes, rheumatoid arthritis, and multiple sclerosis, CD has a strong genetic association with particular HLA class II alleles, namely the HLA genes encoding the class II DQ (a1*0501, b1*02) molecule.4 Sardinia is an ancient genetic isolate with a peculiar distribution of HLA class II alleles and haplotypes.5 The CD predisposing HLA class II DQ (a1*0501, b1*02) molecule is very frequent among Sardinians (43%), and this HLA molecule is almost always encoded in cis by the DRB1*0301, DQA1*0501, DQB1*0201 haplotype.6 An increased prevalence of autoimmune disorders, including type 1 diabetes and autoimmune thyroid diseases, has been repeatedly reported in patients with CD.7-9 Two hypotheses have been suggested to explain this association: (1) sharing of 1 or more genes responsible for CD and for the coexpressed autoimmune disease; or (2) continued gluten exposure in untreated CD might lead to loss of the intestinal barrier function and to alterations of the systemic immune response, ultimately helping to induce other autoimmune disorders.10 The first concept implies that the comorbidity associated with CD is part of the individual genetic background.11-13 On the contrary, the second concept may have important clinical implications because when the gluten induced immune reactivity and the interplay between autoimmune predisposing genes and trigger(s) is blocked, this comorbidity could be eliminated.14,15 In this retrospective study, a large cohort of 324 children with CD from Sardinia, followed for 1 to 15 years, was evaluated for the occurrence of autoimmune thyroiditis (AT). The aims of the study were to establish the prevalence of AT in CD and to verify if gluten exposure before the diagnosis of CD is correlated with development of thyroid autoimmunity, or vice versa if early gluten withdrawal is able to prevent the future development of AT.
AbTG AbTPO AT CD FT3 FT4 GFD TSH US
Thyroglobulin antibody Thyroid peroxidase antibody Autoimmune thyroiditis Celiac disease Free triiodothyronine Free thyroxine Gluten-free diet Thyroid-stimulating hormone Thyroid ultrasonography
From the Pediatric Clinic II, ‘‘Microcitemico’’ Hospital ASL 8, Department of Biological Sciences and Biotechnology, University of Cagliari, Sardinia, Italy Supported by a grant from Fritz-Thyssen-Stiftung, Germany (R.-D.J.). The authors declare no potential conflicts of interest, real or perceived. 0022-3476/$ - see front matter. Copyright Ó 2009 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2009.01.013
51
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Methods Patients Children with biopsy-proven CD (n = 324) from Sardinia (223 girls; mean age at diagnosis, 6.6 years; range, 10 months to 18 years), followed between 1992 and 2007 (mean followup period, 8 years) at the Pediatric Gastroenterological Unit in Cagliari, Italy, were included in the study. Clinical data were collected retrospectively; affected family members of probands, patients with preexisting autoimmune conditions (type 1 diabetes, autoimmune thyroid disorders, Addison’s disease), and patients with Down syndrome or Turner syndrome were excluded from the study. Screening for Celiac Disease The diagnosis of CD was based on the revised criteria of the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition,16 for example, clinical history, positivity of IgA antiendomysial antibody, typical histological features on small intestinal biopsy, and clinical response to the glutenfree diet (GFD) (Table I).17 All patients were typed for anti-gliadin antibody (AGA, both IgA and IgG) and antiendomysial antibody (EMA); most patients were also typed for anti-tissue transglutaminase antibody IgA (tTG-IgA) and anti-actin antibody IgA (AAA-IgA). The latter has been previously associated with the severity of intestinal damage.18 Mucosal atrophy has been graded according to the Marsh classification,19 modified by Oberhuber in type 3a, 3b, and 3c.20 After diagnosis of CD, all patients followed a GFD, and the dietary compliance was evaluated by assessment of AGA, EMA, and tTG every 12 months. Thyroid Function Assessment Screening for thyroid peroxidase antibody (AbTPO) and thyroid globulin antibody (AbTG) was performed in all 324 patients at the onset of CD and once per year after the institution of the GFD. When abnormal values were detected, a complete thyroid evaluation, including serum-free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), and thyroid ultrasonography, was performed. Thyroid autoantibody serum titers were detected using competitive radioimmunoassay (RIA Medical System; Genoa, Italy) with coated-tube technique thyroperoxidase antibodies (AbTPO, normal values <50 IU/mL), RIA-immunoradiometric assay (IRMA, Medical System; Genoa, Italy) with solid-phase technique antithyroglobulin (Ab TG, normal values <100 IU/mL). Serum FT3 and FT4 were detected by RIA (Medical System; Genoa, Italy) using commercial kits (normal values, 0.8 to 2 ng/mL for FT4; 1.40 to 4.2 pg/mL for FT3), whereas serum TSH was measured by a chemiluminescent method (Medical System; Genoa, Italy), with normal values ranging between 0.3 and 5 mU/mL. Ultrasonography of the thyroid gland was performed with a 7.5-MHz linear probe. Thyroid echographic images were classified and graded in 4 patterns according to Sostre 52
Vol. 155, No. 1
Table I. Clinical manifestations and serological markers of CD in 324 Sardinian children Patients (%) Clinical form Classic symptoms Nonclassic symptoms Iron deficiency anemia Failure to thrive Abdominal pain Pubertal delay Vomiting Diarrhea Chronic constipation Silent Serological markers AGA IgA AGA IgG tTG-IgA EMA AAA-IgA
39 56 22 18 5 4 3 2 2 5 73 88 95 99 64
et al21 and Marcocci et al.22 We restricted the diagnosis of AT to patients with high titers of AbTPO and/or AbTG associated with an abnormal thyroid echographic pattern. Therefore, patients with mild elevation of AbTPO and/or AbTG were eliminated from the statistical analysis to avoid possible overestimates. Thyroid function was classified according to the American Thyroid Association Guidelines.23 Control Subjects We compared the frequency of AT among patients with CD with the prevalence of antithyroid antibodies (ATA) in an age-matched Sardinian background population also evaluated for thyroid function.24 In this cohort of 8040 schoolchildren (4194 boys, 3846 girls, ages 6 to 15 years) ATA were detected in 235 children (2.92%). Statistical Analysis Data were evaluated by c2 test with Yates corrections, or by the Wilcoxon test, setting the level of significance at .05.
Results Clinical data of patients with CD with AT are summarized in Table II (available at www.jpeds.com). Overall AT was found in 34 (28 girls, 6 boys) of 324 patients with CD; 23 were on GFD when AT was diagnosed (Table III). Considered as a whole, a high prevalence of AT among children with CD (10.5%), compared with the Sardinian pediatric background population (2.92%), was found (P = 2.9 13). This prevalence is similar to or even lower than that reported in other studies conducted in adults and children.12,25 Euthyroidism was observed in 28 of 34 patients (82.4%), and 6 were hypothyroid. In 3, hypothyroidism preceded diagnosis of CD, and the other 3 patients developed hypothyroidism during GFD (Table II). The thyroid ultrasonography showed an abnormal pattern with varying severity, suggestive of an autoimmune infiltrate Meloni et al
ORIGINAL ARTICLES
July 2009
Table III. Prevalence of AT among patients with CD according to sex, age (in years), and treatment status No. of patients Female/male (ratio) Mean age SD Age range
CD
CD + AT
AT in untreated patients (GCD)
AT in treated patients (GFD)
324 223/101 (2.2:1) 6.6 4.5* 1-18
34/324 28/6 (4.5:1) 10.5 2.9* 5-16
11/34 10/1 (10:1) 9.5 2.9 5-15
23/34 18/5 (3.6:1) 11.0 2.8 7-16
*Statistically significant difference (P< 1.4 6) between the age of onset of CD alone and CD with AT.
(Table II). Two patients with isolated high levels of AbTPO and AbTG at CD diagnosis developed an overall AT with a grade 2 of Sostre after 1 year of GFD. A significant difference in the age of onset between patients with CD and patients with CD and AT was observed: 6.6 versus 10.5 years, respectively (P < 1.4 6; Table III). We were aware that a detection bias might have influenced the different ages of diseases onset because, by definition, AT was looked for only after the diagnosis of CD. On the other hand, no significant difference in the age of AT onset among patients at CD diagnosis (9.5 2.9) and patients on GFD (11.0 2.8) was found (Table III). The female/male ratio was higher in patients with CD and AT (F/M 4.5:1), compared with patients with CD alone (F/M 2.2:1; Table III). This finding was unexpected in our cohort of pediatric patients because a similar high F/M ratio in AT is commonly described in late adolescence and adulthood.26,27 Thus, to determine a possible role of sex hormones as an explanation for the observed F/M ratio, patients were subdivided into 2 groups: prepubertal (age #12 years) and pubertal (age >12 years) (Table IV). Unexpectedly, the F/ M ratio in prepubertal (12:1) was even higher than in pubertal patients (1:1), and the difference, when compared with the F/M ratio of prepubertal patients with CD alone (2:1), reached statistical significance (P < .0068; Table IV). The Figure shows the relationship of AT with the duration of gluten exposure. At onset of CD, 11 of 324 patients receiving a gluten-containing diet were affected by AT. After diagnosis of CD, an additional 23 patients placed on GFD developed AT after 1 to 11 years, and 3 of these shifted toward hypothyroidism. No difference was found in the serum titers of AbTPO, AbTG, and thyroid echographic pattern between patients with AT coexisting at the onset of CD and patients developing AT during GFD (Table II). In 9 of 11 patients with AT coexisting at the diagnosis of CD, very high titers of AbTG and/or AbTPO, suggestive of active AT, persisted despite 1 to 9 years of GFD (data not shown).
Discussion Our study documents the association between AT and CD in Sardinian children. Because it includes a large number of subjects, it provides the actual prevalence of AT in our pediatric population with CD. Previous studies on the prevalence of AT in Sardinian patients with CD, or vice versa on the occurrence of CD in AT, have been conducted only in adults.9,28 We found that the prevalence of AT in CD is about 4 times greater than that found in the age matched Sardinian schoolchildren background population, indicating that CD is a risk factor predisposing to AT. There are several possible explanations for the increased risk of developing AT in patients with CD: (1) the CD condition carries 1 or more genes (MHC or other autoimmunity genes) favoring AT; and (2) the timing of CD diagnosis and duration of gluten exposure might increase toxic and immunological effects of gluten, increasing the risk of AT.14,15,29 A lack of correlation between CD status and occurrence of AT, favoring the first hypothesis, is suggested by our findings. Indeed, AT developed in patients on a gluten-containing diet or in patients on GFD; in these latter patients, after many years (Table II; Figure) GFD did not improve the outcome of the disease, as demonstrated by very high titers of AbTG and/or AbTPO after 1 to 9 years of follow-up. There was no significant difference in the mean age between patients with AT coexisting at onset of CD and patients developing AT during the GFD: around 10.5 years in both groups, indicating that AT is independent of gluten exposure (Table III). The age of patients with CD with AT is the same observed in Sardinian children with isolated AT.30 Finally, the significant difference between the age of onset of CD (6.6 years) and CD with AT (10.5 years) suggests that CD and AT are two different autoimmune diseases, which association is likely explained by a shared genetic background. Our data, showing that the duration of gluten exposure is not of crucial importance in favoring the development of
Table IV. Comparison of female/male ratio according to age (prepubertal versus pubertal) in patients with CD and patients with CD and AT Patients
CD
F
M
F/M
CD + AT
F
M
F/M
P value
Total (n) Prepubertal (#12 y) Pubertal (>12 y)
290 254 36
195 169 26
95 85 10
2:1 2:1 2.6:1
34 25 8
28 24 4
6 2 4
4.6:1 12:1 1:1
NS <.007 NS
P value determined by c2 test. F, female; M, male; NS, P value not significant.
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal
53
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 155, No. 1
Figure. Vertical bars represent the number of patients with CD diagnosed with AT. The first bar represents patients with AT at CD diagnosis, the following represent patients developing AT during GFD.
AT, are in agreement with some authors11-13 and in contrast with others.14,15,29 Ventura et al14 found that the prevalence of AT correlated with the duration of gluten exposure and augmented with the age at CD diagnosis. In addition, thyroid autoantibodies had a tendency to disappear following the GFD. Possible explanations for these variable results could depend on different follow-up periods (15 years in our study and 6 months in the work of Ventura et al) and on age stratifications chosen by Ventura that did not take into account that the prevalence of certain autoimmune diseases raises with aging. A peculiar and previously unreported finding of our study concerns the F/M ratio of patients with CD and AT and the relationship between AT and puberty. The prevalence of girls in AT is well known, and the sex ratio favoring females is reported to be age-related, starting generally after puberty.26,27 This observation has supported the belief in a fundamental role of sexual hormones in favoring the development of AT. Accordingly, in the Sardinian schoolchildren population,24 a significant increase of thyroid autoimmunity was more evident in females than in males, in particular over 11 years (P < .01). Interestingly, in patients with CD with AT, the high F/M ratio was observed before puberty (Table IV), suggesting either a more complicated role of the sex hormones in conferring predisposition to AT, or, alternatively, an influence of the genetic background of CD in anticipating the high F/M ratio of AT. The high frequency of AT among patients with CD, even on GFD, suggests clinical usefulness for a longitudinal follow-up of thyroid function. n 54
The authors are grateful to Prof. A. Cao for critical reading of the manuscript. Submitted for publication Aug 29, 2008; last revision received Nov 24, 2008; accepted Jan 8, 2009. Reprint requests: Dr Antonella Meloni, Pediatric Clinic II, ‘‘Microcitemico’’ Hospital ASL 8, Department of Biological Sciences and Biotechnology, University of Cagliari, Sardinia, Italy. E-mail: [email protected].
References 1. Green PH, Cellier C. Celiac disease. N Engl J Med 2007;357:1731-43. 2. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology 2001;120:636-51. 3. Meloni G, Dore A, Fanciulli G, Tanda F, Bottazzo GF. Subclinical coeliac disease in schoolchildren from northern Sardinia. Lancet 1999;353:37. 4. Sollid LM. Coeliac disease: dissecting a complex inflammatory disorder. Nat Rev Immunol 2002;2:647-55. 5. Lampis R, Morelli L, De Virgiliis S, Congia M, Cucca F. The distribution of HLA class II haplotypes reveals that the Sardinian population is genetically differentiated from the other Caucasian populations. Tissue Antigens 2000;56:515-21. 6. Congia M, Frau F, Lampis R, Frau R, Mele R, Cucca F, et al. A high frequency of the A30, B18, DR3, DRw52, DQw2 extended haplotype in Sardinian celiac disease patients: further evidence that disease susceptibility is conferred by DQ A1*0501, B1*0201. Tissue Antigens 1992;39:78-83. 7. Collin P, Reunala T, Pukkala E, Laippala P, Keyrilainen O, Pasternack A. Coeliac disease–associated disorders and survival. Gut 1994;35:1215-8. 8. Counsell CE, Taha A, Ruddell WS. Coeliac disease and autoimmune thyroid disease. Gut 1994;35:844-6. 9. Velluzzi F, Caradonna A, Boy MF, Pinna MA, Cabula R, Lai MA, et al. Thyroid and celiac disease: clinical, serological, and echographic study. Am J Gastroenterol 1998;93:976-9. 10. Fasano A. Systemic autoimmune disorders in celiac disease. Curr Opin Gastroenterol 2006;22:674-9.
Meloni et al
July 2009 11. Sategna Guidetti C, Solerio E, Scaglione N, Aimo G, Mengozzi G. Duration of gluten exposure in adult coeliac disease does not correlate with the risk for autoimmune disorders. Gut 2001;49:502-5. 12. Ansaldi N, Palmas T, Corrias A, Barbato M, D’Altiglia MR, Campanozzi A, et al. Autoimmune thyroid disease and celiac disease in children. J Pediatr Gastroenterol Nutr 2003;37:63-6. 13. Viljamaa M, Kaukinen K, Huhtala H, Kyronpalo S, Rasmussen M, Collin P. Coeliac disease, autoimmune diseases and gluten exposure. Scand J Gastroenterol 2005;40:437-43. 14. Ventura A, Magazzu G, Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease: SIGEP Study Group for Autoimmune Disorders in Celiac Disease. Gastroenterology 1999;117:297-303. 15. Ventura A, Neri E, Ughi C, Leopaldi A, Citta A, Not T. Gluten-dependent diabetes-related and thyroid-related autoantibodies in patients with celiac disease. J Pediatr 2000;137:263-5. 16. Walker-Smith J, Guandalini S, Schmitz J, Shmerling DH, Visakorpi J. Revised criteria for diagnosis of coeliac disease: Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch Dis Child 1990;65:909-11. 17. Fasano A. Clinical presentation of celiac disease in the pediatric population. Gastroenterology 2005;128:S68-73. 18. Clemente MG, Musu MP, Troncone R, Volta U, Congia M, Ciacci C, et al. Enterocyte actin autoantibody detection: a new diagnostic tool in celiac disease diagnosis: results of a multicenter study. Am J Gastroenterol 2004;99:1551-6. 19. Marsh MN, Crowe PT. Morphology of the mucosal lesion in gluten sensitivity. Baillieres Clin Gastroenterol 1995;9:273-93. 20. Oberhuber G. Histopathology of celiac disease. Biomed Pharmacother 2000;54:368-72. 21. Sostre S, Reyes MM. Sonographic diagnosis and grading of Hashimoto’s thyroiditis. J Endocrinol Invest 1991;14:115-21.
ORIGINAL ARTICLES 22. Marcocci C, Vitti P, Cetani F, Catalano F, Concetti R, Pinchera A. Thyroid ultrasonography helps to identify patients with diffuse lymphocytic thyroiditis who are prone to develop hypothyroidism. J Clin Endocrinol Metab 1991;72:209-13. 23. Surks MI, Chopra IJ, Mariash CN, Nicoloff JT, Solomon DH. American Thyroid Association guidelines for use of laboratory tests in thyroid disorders. JAMA 1990;263:1529-32. 24. Loviselli A, Velluzzi F, Mossa P, Cambosu MA, Secci G, Atzeni F, et al. The Sardinian Autoimmunity Study: 3, studies on circulating antithyroid antibodies in Sardinian schoolchildren: relationship to goiter prevalence and thyroid function. Thyroid 2001;11:849-57. 25. Sategna-Guidetti C, Volta U, Ciacci C, Usai P, Carlino A, De Franceschi L, et al. Prevalence of thyroid disorders in untreated adult celiac disease patients and effect of gluten withdrawal: an Italian multicenter study. Am J Gastroenterol 2001;96:751-7. 26. Chiovato L, Lapi P, Fiore E, Tonacchera M, Pinchera A. Thyroid autoimmunity and female gender. J Endocrinol Invest 1993;16:384-91. 27. Demirbilek H, Kandemir N, Gonc EN, Ozon A, Alikasifoglu A, Yordam N. Hashimoto’s thyroiditis in children and adolescents: a retrospective study on clinical, epidemiological and laboratory properties of the disease. J Pediatr Endocrinol Metab 2007;20:1199-205. 28. Meloni GF, Tomasi PA, Bertoncelli A, Fanciulli G, Delitala G, Meloni T. Prevalence of silent celiac disease in patients with autoimmune thyroiditis from Northern Sardinia. J Endocrinol Invest 2001;24:298-302. 29. Toscano V, Conti FG, Anastasi E, Mariani P, Tiberti C, Poggi M, et al. Importance of gluten in the induction of endocrine autoantibodies and organ dysfunction in adolescent celiac patients. Am J Gastroenterol 2000;95:1742-8. 30. Prinzis AGM, Cambuli VM, Concas D, Foschini ML, Marras V, Carta D, Loche S, Mariotti S. The female sex predominance in Hashimoto’s thyroiditis is a pubertal phenomenon. J Endocrinol Invest 2007;30:97.
Prevalence of Autoimmune Thyroiditis in Children with Celiac Disease and Effect of Gluten Withdrawal
55
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. 155, No. 1
Table II. Clinical data at disease onset of 34 pediatric patients with CD with AT CD AT AbTPO AbTg Patient Sex onset onset U/mL IU/mL Ultrasonography (Sostre grade) 1* 2† 3 4 5 6† 7 8† 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23* 24* 25 26 27 28 29 30 31 32 33 34
F F F F F F F F F F M F F F F M F F F F F F F F F M F M F F M M F F
5 7 7 7 8 10 11 11 12 12 15 2 2 6 5 4 5 7 8 3 4 6 7 1 1 6 6 12 10 6 13 5 7 15
5 7 7 7 8 10 11 11 12 12 15 7 7 7 8 9 9 9 9 10 10 10 10 11 12 12 12 13 14 14 14 15 16 16
1000 440 970 555 1000 250 900 300 140 4494 615 1000 100 1000 100 550 55 4000 260 1900 1926 420 5900 2600 1000 255 1000 3400 110 500 170 235 250 430
327 250 840 1500 800 600 600 280 160 314 215 370 288 100 1500 150 360 6400 380 350 340 600 225 230 730 67 50 1200 480 314 22 222 100 250
3 2 3 2 2 1 2 2 1 3 3 3 1 2 1 2 3 4 2 1 2 2 3 3 2 1 3 2 2 2 2 2 2 2
The first column shows the patients with CD with AT numbered from 1 to 34; the second shows the sex; the third and fourth show the age at CD and at AT onset in years; the fifth and sixth show the titers of circulating thyroid autoantibodies; the seventh shows the thyroid echographic pattern. Normal ranges: AbTPO <50 IU/mL, AbTG <100 IU/mL. *Patients 1, 23, 24: autoimmune hypothyroidism after 1, 7, and 11 years of GFD, respectively. †Patients 2, 6, and 8: autoimmune hypothyroidism at onset of CD.
55.e1
Meloni et al