Frequency of serological markers of rheumatoid arthritis in adult patients with active celiac disease

Abstract

Background

Celiac disease (CD) and rheumatoid arthritis (RA) are multisystem autoimmune diseases affecting 1% of general populationa. Both diseases share genetic and immunological features.

Aim

In this retrospective study, we aim to determine the frequency of auto‐antibodies of RA in adult patients with CD.

Materials and methods

Seventy seven adult patients with active CD were included in the present study. Ninety healthy blood donors (HBD) served as control group. Anti‐cyclic citrullinated peptides antibodies (CCP‐Ab) and rheumatoid factors (RF; IgA, IgG and IgM) were determined by enzyme linked immunosorbent assay (ELISA) for patients and control group. For statistical analysis, we used Chi‐square or Fisher's exact test.

Results

Our study included 77 adult patients with active celiac disease (57 female, 20 male). Twenty‐four (31.2%) active celiac patients and 7 (7.8%) blood donors had CCP‐Ab or RF (31.2% vs 7.8%, p < 10–4). Only two patients (2.6%) had both CCP‐Ab and RF. IgA was the predominant isotype of RF in celiac patients (n = 18; 23.4%) while none of healthy blood donors had RF‐IgA (23.4% vs 0.0%, p < 10–4).

Conclusion

The current study has shown that CD is associated with a high frequency of RF‐IgA suggesting that celiac patients could be at a higher risk of developing RA.

Seventy seven adult patients with active celiac disease (CD) and ninety healthy blood donors were tested for rheumatoid arthritis (RA) autoantibodies. The frequency of rheumatoid factors IgA (RF‐IgA) were significantly higher in celiac disease patients than in control group.

1. INTRODUCTION

Rheumatoid arthritis (RA) is the most frequent inflammatory arthritis. ¹ It causes progressive articular damage and can lead to irreversible osteoarticular destruction with functional loss. ² Rheumatoid factors (RF) and antibodies to cyclic citrullinated peptides (CCP‐Ab) are found in 80% and 95% of RA patients, respectively. ² Celiac disease (CD) is a multisystem autoimmune‐based disorder elicited by gluten and characterized by small intestinal enteropathy and systemic symptoms. ³ CD is a T‐cell‐mediated disease with T‐cells‐infiltrated lamina propria in response to gliadin‐derived peptides causing tissue damage and the release of cytokines such as Interferon‐gamma, Interleukin‐21, and Interleukin‐15. ⁴ , ⁵ , ⁶ The Interleukin‐15 has also been described to be implicated in the pathology of various autoimmune diseases such as type 1 diabetes, psoriasis, and RA. ⁷ On the contrary, the altered intestinal permeability in CD plays an important role in the progression of autoimmune diseases and so exposes patients to exogenous antigens, which could be a driver for autoimmune response in the joints of genetically predisposed subjects. ⁸ In fact, Podas et al. ⁹ suggested that RA is a disease that starts in the gut and Matei et al. ¹⁰ demonstrated that breakdown of gut‐barrier integrity contributes to arthritis development and proposed restoration of gut‐barrier homeostasis as a therapeutic approach of RA. In RA and in CD, it was reported quantitative alteration in intestinal microbiota. ¹¹ , ¹² , ¹³ In fact, the theory that the microorganism population within the gastrointestinal tract is a key factor in the pathogenesis of joint disease was reinforced by many researches. ¹² In addition, a study exploring the overlap of CD and RA susceptibility, loci identified association to six common regions. ¹⁴ , ¹⁵ Most of them further emphasize a role for adaptive and innate immunity in both diseases. Thirty‐one loci outside the HLA locus were confirmed to be shared between CD and RA. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ For the recognition of peptides by HLA molecules, post‐translational modifications were described for optimal peptide binding and the acquisition of auto‐antigenicity leading to tissue damage in CD and RA. ²⁰ , ²¹ , ²² Furthermore, patients with RA have frequent gastrointestinal complaints that sometimes are not fully clarified. ²³ Both CD and RA affect 1% of the general population. In the current study, we aimed to do a serological screening for RA immunological markers in adult patients with active CD.

2. PATIENTS AND METHODS

2.1. Study participants

In our retrospective study, sera of 80 adult patients (age ≥18 years) with active celiac disease (newly diagnosed or known having celiac disease) were included from the database of our Immunology laboratory. None of our patients was on gluten‐free diet. Sera were collected over a 24‐month period from four hospitals in the center of Tunisia. We included only patients who had anti‐endomysial IgA, anti‐transglutaminase 2 IgA, and anti‐gliadin deamidated peptide IgG and IgA antibodies. All patients underwent duodenal biopsy, which was in favor of CD. Patients with known RA were excluded from our study. Control sera were obtained from 90 healthy blood donors' (HBD) sex and age matched with CD patients.

All sera were screened for anti‐nuclear antibodies (ANA).

All sera were stored at −80°C until the usage. Ethical committee of our hospital gave approval for the study.

2.2. Methods

2.2.1. Celiac disease auto‐antibodies

Anti‐endomysial IgA antibodies were performed by indirect immunofluorescence using cryostat sections (4 µm thick) of human umbilical cord prepared in our laboratory as a substrate and fluorescein isothiocyanate‐labeled anti‐human IgA antibodies (Bio‐Rad^®). Sera were diluted to 1/10. A positive result was recorded if a connective tissue surrounding the muscle cells fluoresced brightly in a honey‐comb pattern.

Anti‐transglutaminase 2 IgA antibodies were determined by indirect ELISA (Orgentec^®). The cut‐off value is 25 RU/ml.

Anti‐gliadin deamidated peptide IgG and IgA were performed by indirect ELISA (Euroimmun^®). Cut‐off value was 25 UI/ml for both isotypes.

2.2.2. Rheumatoid arthritis auto‐antibodies

Serum samples were evaluated for IgG, IgA, and IgM rheumatoid factors (RF) using a commercial ELISA (Orgentec^®) as we described previously. ²⁴ Cut‐off values are 49.5, 45.5 and 31 U/ml for RF‐IgG, RF‐IgA, and RF‐IgM, respectively.

Anti‐cyclic citrullinated peptides antibodies (CCP‐Ab) were detected by using second‐generation ELISA (Euroimmun^®) kit as we described before. ²⁵

Results were expressed as arbitrary units. The cut‐off value is 5.5 RU/ml.

2.2.3. Antinuclear antibodies

ANA‐IgG were determined by indirect immunofluorescence using fluorescein isothiocyanate anti‐IgG conjugate (Bio‐Rad^®). Rat liver sections (4 μm thick) and HEp 20–10 (Euroimmun^®) cells were used as substrate.

Determination of the antigenic specificity of ANA (DNAn‐Ab, nucleosome‐Ab, histone‐Ab, RNP‐Ab, SSA‐Ab, SSA‐Ab, Scl70‐Ab, and Sm‐Ab) was done for patients with positive ANA and performed by indirect ELISA (Orgentec^®) using the instructions of the manufacturer.

2.3. Statistical analysis

The analysis of the results was performed by Epi Info version 3 using Chi‐squared or Fisher's exact test. p‐values less or equal to 0.05 were considered statistically significant; 95% of confidence interval (CI) was calculated.

3. RESULTS

High‐level ANA >1/160 were present in three patients for whom the determination of specificity of ANA showed a positivity for Sjogren's syndrome (SS) auto‐antibodies (SSA‐Ab and SSB‐Ab). Two patients out of three with SSA‐Ab and SSB‐Ab had RF‐IgG, IgA and IgM and one had only RF‐IgM. These patients were excluded from our study because RF positivity could be due to SS and not RA. So, out of 80 patients with CD, 77 were included in our statistical analysis.

Our study included 77 adult patients with active celiac disease (57 female, 20 male). Demographic and clinical characteristics were summarized in Table 1. In the 90 control subjects, 56 were female. The mean age of our control group was 37.6 years (range: 20–64).

TABLE 1.

Demographic and clinical characteristics of patient group

Patients	n = 77
Mean age (range)	38.6 (18–70)
Sex ratio (Female/Male)	57/20
Known celiac disease	22
Family history of celiac disease	4
Arthralgia	4
Cirrhosis	2
Weight loss	8
Abdominal pain	2
Diarrhea	22
Bloating	5
Osteomalacia	2
Hypocholesterolemia	4
Hypocalcemia	4
Leucopenia	1
Anemia	19
Cytolysis	1

Twenty‐four (31.2%) active celiac patients and 7 (7.8%) blood donors had CCP‐Ab or RF (31.2% vs. 7.8%, p < 10⁻⁴, CI 95% [21.9; 42.2], odds ratio = 6.2). Seven patients (9.1%) had CCP‐Ab with low titers ranged between 6.5 and 9.3 RU/ml. Only two patients (2.6%) had both CCP‐Ab and RF. IgA was the predominant isotype of RF in celiac patients (23.4%) while none of healthy blood donors had RF‐IgA, the difference was statistically significant (p < 10⁻⁴).

Rheumatoid factors and RF‐IgA were more frequent than CCP‐Ab (26.0% vs. 9.1%, p = 0.01 and 23.4% vs. 9.1% p = 0.02, respectively).

Results are gathered in Table 2 and Table 3.

TABLE 2.

Frequency of auto‐antibodies of rheumatoid arthritis in celiac patients and healthy controls

	Patients (n = 77) n (%)	Control group (n = 90); n (%)	p	CI
CCP‐Ab and/or RF	24 (31.2%)	7 (7.8%)	<10−4	[21.9; 42.2]
CCP‐Ab and RF	2 (2.6%)	0 (0.0)	NS	–
CCP‐Ab	7 (9.1%) a , b	3 (3.3%)	NS	–
CCP‐Ab only	5 (6.5%)	3 (3.3%)	NS	–
RF (IgG and/or IgA and/or IgM)	20 (26.0%) a	5 (5.5%)	<10−4	[17.4; 36.8]
RF‐IgG	2 (2.6%)	2 (2.2%)	NS	–
RF‐IgA	18 (23.4%) b	0 (0.0%)	<10−4	[15.2; 34.0]
RF‐IgM	4 (5.2%)	2 (2.2%)	NS	–
RF‐IgA only	14 (18.2)	0 (0.0%)	<10−4	[11.0; 28.4]

Abbreviations: CCP‐Ab, anti‐cyclic citrullinated peptides antibodies; CI, confidence interval; RF, rheumatoid factors.

^a RF versus CCP‐Ab p = 0.01.

^b RF‐IgA versus CCP‐Ab p = 0.02

TABLE 3.

Characteristics of patients with both antibodies to cyclic citrullinated peptides and rheumatoid factors

	Age	Sex	RF titers (U/ml)	CCP‐Ab levels (RU/ml)	Clinical manifestations
Patient 1	M	32	RF‐IgA 89.5 RF‐IgM negative RF‐IgG negative	9.3	No articular pain Diarrhea Celiac disease in first degree relative (brother)
Patient 2	F	18	RF‐IgA 229.5 RF‐IgM negative RF‐IgG negative	7.2	No joint pain Anorexia Weight loss

Abbreviations: CCP‐Ab, anti‐cyclic citrullinated peptides antibodies; F, female; M, male; RF, rheumatoid factors; RU, random unit.

4. DISCUSSION

From 1984, Parke et al. ²⁶ described three cases of adult patients with both CD and RA. Since then, only three studies have done the frequency of RA markers in CD patients. ²⁷ , ²⁸ , ²⁹ In the current study, we determined the frequency of antibodies of RA in adult patients with active CD.

We found that the frequency of RA markers was significantly higher in active CD patients than in HBD (31.2% vs. 7.8%; p < 10⁻⁴). These results are similar to those of Yang et al. ²⁸ who found RA markers in 20.4% of CD patients vs. 6% of controls (p < 0.05; Table 4). Out of three isotype of RF, only RF‐IgA was significantly more frequent in celiac patients than in healthy subjects. The frequency of RF‐IgA in our study (23.4%) is similar to that of Sjöker et al (18.8%). ²⁹ The frequency of CCP‐Ab in our study (9.1%) is similar to that of Yang et al. (10.2%). ²⁸ In both of our studies, this frequency was not statistically different between CD and control group.

TABLE 4.

Frequency of rheumatoid arthritis markers in adult patients with celiac disease

	Study population	RF n (%)	CCP‐Ab n (%)	CCP‐Ab and/or RF n (%)
Sökjer 1995 29	n = 22	IgM: 1 (4.5) IgG: 2 (9.1) IgA: 4 (18.1)	ND	–
Yang 2019 28	n = 49 age: 46±17 years Sex ratio (F/M): 35/14	IgM: 5 (10.2) IgG: ND IgA: ND	5 (10.2)	10 (20.4)
Fayyaz 2019 27	n = 45 mean Age: 26.28 years Sex ratio (F/M): 40/5	IgM: 15 (33.3) IgG: ND IgA: ND	2 (4.4)	16 (35.0)
Our study	n = 77 Mean age: 38.6 Sex ratio F/M:57/20	IgM: 4 (5.2) IgG: 2 (2.6) IgA: 18 (23.4)	7 (9.1)	24 (31.2)

Abbreviations: CCP‐Ab, anti‐cyclic citrullinated peptides antibodies; F, female; M, male; ND, not done; RF, rheumatoid factors.

The predominant isotype of RF in our study was IgA. Sökjer et al ²⁹ also found a selective increase of RF‐IgA in patients with CD suggesting that the RF production in patients with CD primarily results from activation of immune system in the gut mucosa. A study interested in the jejunal secretion of RF in patients with CD found an elevated jejuna fluid levels of RF in patients with active CD compared with those with inactive CD confirming that the major source of RF synthesis in those patients is located in the jejuna mucosal immune system. ³⁰

Rheumatoid factors are directed against the Fc portion of immunoglobulin IgG. RF involvement in the immune complex may lead to further complement activation and recruitment inflammatory cells resulting in tissue damage. ³¹ Our patients with positive RA serological markers did not suffer from joint involvement. This could be explained by the fact that RF are frequently detected in the serum of CD patients with non‐arthritic clinical disorders. ³⁰ We could also imagine that our patients with RF and/or CCP‐Ab will develop RA in future. In fact, it was demonstrated that development of auto‐antibodies can precede clinical onset of RA by several years. The frequency of RF in our study was higher than CCP‐Ab. All three RF isotypes could be detected earlier than CCP‐Ab in sera antedating symptoms of RA. ³² We also found that RF‐IgA was significantly more frequent than CCP‐Ab. It has been demonstrated that out of the three isotype of RF, IgA was the most sensitive and appeared earlier in patients' sera before any symptoms. ³² Indeed, RF‐IgA could appear 22 years before symptoms onset while CCP‐Ab appear before 9 years of clinical manifestations of RA.

On the contrary, several authors did a vice versa study by doing a serological screening of CD in patients with RA. ²³ , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ According to these studies, the prevalence of CD among RA patients range from 0.0% ³³ , ³⁶ , ³⁹ to 11.3% ³⁷ (Table 5). The frequency of CCP‐Ab in our CD patients was similar to that of anti‐transglutaminase 2 antibodies in RA patients of Moghtaderi (9.1% and 11.3%, respectively). ³⁷

TABLE 5.

Frequency of celiac disease in patients with rheumatoid arthritis in the literature

Reference	Serological markers used	% of positivity of serological markers
Paimela 1995 33	AGA	0
Francis 2002 34	EmA	0.6
Feighery 2003 35	EmA	1.8
Bizzaro 2003 36	TTG, EmA	0
Luft 2003 23	TTG, EmA	2
Moghtaderi 2006 37	TTG	11.3
Nisihara 2007 38	EmA	0
Koehne 2010 39	AGA, EmA	0
Koszarny 2015 40	AGA	4.1
Caio 2018 41	TTG, EmA, DPG	1.5

Abbreviations: AGA, anti‐gliadin antibodies; DPG, anti‐deaminated gliadin peptide antibodies; EmA, anti‐endomysial antibodies; TTG, anti‐transglutaminase 2 antibodies.

The current study has shown that CD is associated with a high frequency of RF‐IgA suggesting that celiac patients could be at a higher risk of developing RA. Serological and clinical follow‐up of our patients with positive RA auto‐antibodies over the years could confirm these findings.

Ghozzi M, Melayah S, Adaily N, Ghedira I. Frequency of serological markers of rheumatoid arthritis in adult patients with active celiac disease. J Clin Lab Anal. 2022;36:e24249. doi: 10.1002/jcla.24249

DATA AVAILABILITY STATEMENT

Data are not shared.