Autoimmune Diseases Are Linked to Type IIb Autoimmune Chronic Spontaneous Urticaria.

Chronic spontaneous urticaria (CSU) is defined as the spontaneous occurrence of wheals, angioedema or both for > 6 weeks due to known or unknown causes. 1 While exact pathomechanisms underlying CSU are, as of yet, not completely unraveled, there is strong evidence to support a link to autoimmunity. 2 3 Recently, 2 endotypes of CSU have been proposed: type I autoimmune CSU mediated by IgE autoantibodies (also known as autoallergic CSU) and type IIb autoimmune CSU (aiCSU), 4 5 6 7 in which IgG autoantibodies, and probably IgM and IgA are responsible for direct activation of mast cells by binding to high-affinity IgE receptors. 3 8 Positive autologous serum skin test (ASST), basophil histamine release assay (BHRA) and immunoassay for specific IgG autoantibodies against FcεRIα/IgE are the current gold standard for the diagnosis of aiCSU. 9 At least 8% of CSU patients have aiCSU and they are more severely affected with markedly lower total IgE levels, 4 and higher rates of eosinopenia and basopenia. 10 On the other hand, CSU patients are known to often exhibit comorbid autoimmune disorders, 4 11 12 13 14 mostly Hashimoto's thyroiditis (HT), vitiligo and rheumatoid arthritis. 11 12 Also, CSU has been described to occur in patients with common autoimmune diseases such as autoimmune thyroid disease (AITD) and relatively rare autoimmune disorders, e.g. autoimmune encephalitis. 11 15 16 17 CSU is thought to occur more frequently in patients with AITD, systemic lupus erythematosus (SLE), rheumatoid arthritis or celiac disease. 11 17 Three reasons underline the importance of research on the role and relevance of autoimmune diseases in CSU. First, the presence of autoimmune diseases in patients with CSU may predict the duration, activity and course of CSU and the response to treatment. Secondly, various features of CSU may be associated with a higher risk of having comorbid autoimmune diseases. Knowledge of these features can help facilitate early diagnosis of comorbid autoimmune diseases in patients with CSU. Thirdly, therapeutic interventions in patients with CSU and concomitant autoimmune diseases need to be aligned to assure maximum efficacy as well as minimum drug interactions and adverse effects. For example, CSU treatment, e.g. omalizumab or cyclosporine, may also be effective in some other autoimmune diseases, e.g. SLE, 18 bullous pemphigoid, 19 eosinophilic granulomatosis with polyangiitis 20 and autoimmune polyglandular syndromes. 21 Thus, the link between CSU and autoimmunity has been known for many years, although the specific distribution of autoimmune diseases in different CSU phenotypes and endotypes is less characterized. In this multicenter retrospective study with more than 1,000 CSU patients, we evaluated clinical and laboratory factors associated with a higher risk of having autoimmune diseases and we characterized this subpopulation of patients.

Of 1,199 CSU patients, 28% (n = 333) had one or more comorbid autoimmune disease ( Table 1 ). The most prevalent disease was AITD (25%, mostly HT, ≥ 21%). Other autoimmune comorbidities were seen in 3% of cases, mainly vitiligo (2%) and rheumatoid arthritis (1%). Of the 333 CSU patients with a comorbid autoimmune disease, 91% had one, most often AITD, 8% had 2, and less than 1% had 3 ( Table 2 ). About 1% of all CSU patients had HT and vitiligo, the most frequent combination of 2 comorbid autoimmune diseases ( Table 2 ). CSU, chronic spontaneous urticaria; AID, autoimmune diseases; F: female; M: male; ITP, idiopathic thrombocytopenic purpura; SLE, systemic lupus erythematosus. * References 25 53 ; † Some patients had more than one autoimmune disease; ‡ HT in 20.7% (248/1,199) of patients, Graves' disease in 1.2% (15/1,199) of patients, and diagnosis of autoimmune thyroid disease is not specified in 3.5% (42/1,199) of patients; § Pernicious anemia in 0.4% (3/751) of patients. CSU, chronic spontaneous urticaria; AITD, autoimmune thyroid disease; AID, autoimmune diseases. * Total = 333; † total = 1,199; ‡ Two patients had pernicious anemia. CSU patients who were 40 years or older had autoimmune comorbidities, especially AITD, RA and vitiligo, more often than younger patients ( Table 1 ). Autoimmune comorbidities were reported more frequently in female CSU patients (risk ratio [RR] = 2.3, P < 0.001) and patients with a family history of autoimmune diseases (RR = 2.0, P < 0.001; Table 3 , Supplementary Table S1 ). Patients with angioedema also had higher rates of comorbid autoimmune diseases (RR = 1.2, P = 0.05, Table 3 ). AID, autoimmune disease; RR, risk ratio; CI, confidence interval; F, female; M, male; TSH, thyroid-stimulating hormone; +, positive/present; −, negative/absent; ↑, increased; ↓, decreased; BAT, basophil activation test; BHRA, basophil histamine release assay; ASST, autologous serum skin test; CSU, chronic spontaneous urticaria. * Comparison between groups: The number of CSU patients with high normal/high or low TSH levels vs. patients with normal TSH levels: P < 0.001; † Comparison between groups: Type IIb autoimmune CSU (+) vs. partly type IIb (+/−) CSU: P = 0.01, type IIb autoimmune CSU vs. non type IIb autoimmune CSU (−): P = 0.01, partly type IIb autoimmune CSU vs. non type IIb autoimmune CSU: P = 0.402; ‡ In previous studies various cut-off values less than 50 kU/L were used; § Basopenia was defined as blood basophil numbers less than 0.01 × 10 9 /L. For eosinopenia/low blood eosinophil counts, we used a cutoff of less than 0.05 × 10 9 /L as described before. 37 The presence of autoimmune diseases in CSU patients tended to be associated with higher disease activity ( Table 3 , Fig. 1A and B ). AID, autoimmune disease; CSU, chronic spontaneous urticaria. The majority (91%) of CSU patients had normal thyroid-stimulating hormone (TSH) levels (0.27–3.9 mU/L). Twenty-eight (4%) patients showed low (below 0.27 mU/L) and 34 (5%) patients had high normal/high (above 3.9 mU/L) TSH levels. Of patients with high normal/high TSH levels, 68% (n = 23) had autoimmune diseases; of patients with low TSH levels, 54% (n = 15) had autoimmune diseases; of patients with normal TSH levels, only 24% (n = 143) had autoimmune diseases ( P < 0.001 for all, Table 3 ). No difference regarding autoimmune diseases was seen between patients with low and elevated TSH levels. Seventy-one percent (27/38) of patients with high normal/high or low TSH levels were diagnosed HT or had elevated IgG anti-TPO levels. One patient had Graves' disease and 7 patients had undefined AITD. Patients with any of the definition criteria for aiCSU such as a positive ASST, BAT and/or BHRA more often had autoimmune diseases as compared to patients without ( Table 3 ). Risk of having autoimmune diseases is 1.7, 2.9 and 3.3 times greater for CSU patients with positive ASST, BHRA and BAT as compared to CSU patients with negative results of ASST, BHRA and BAT, respectively. CSU patients with detectable IgG anti-FcεRIα/IgE showed similar rates of autoimmune diseases as patients without IgG anti-FcεRIα/IgE ( Table 3 ). The prevalence of autoimmune diseases was significantly higher in patients who had true ‘triple-positive’ aiCSU (60%) as compared to patients who were positive only for some but not all markers of aiCSU (28%) and patients who had ‘triple-negative’ CSU (17%, Table 3 ). Other features of aiCSU, i.e . lower blood eosinophil and basophil counts and lower total IgE levels, were also linked to presence of autoimmune diseases in CSU patients ( Fig. 1C-E , Table 3 ). In CSU patients, the prevalence of ANA and/or IgG anti-TPO positivity was significantly higher in non-responders to omalizumab treatment (71%, 5/7) as compared to complete responders (20%, 4/16, Table 4 ). ANA and/or IgG anti-TPO were negative in 80% of complete responders, 39% of partial responders and 29% of non-responders. TPO, thyroperoxidase; ANA, antinuclear antibodies; NR, non-responders; PR, partial responders; CR, complete responders; +, present; −, absent. * Comparison between groups: NR vs. PR: P = 0.68, NR vs. CR: P = 0.02, PR vs. CR: P = 0.019.

The results of our study confirm that around one-third of CSU patients have autoimmune comorbidities. Importantly, autoimmune comorbidity is strongly linked to type IIb aiCSU. These findings support the concept that aiCSU is a distinct endotype of CSU. They also help assess and monitor patients with CSU, especially those with aiCSU, for comorbid autoimmune diseases in routine clinical practice. How common are autoimmune comorbidities in the general population and in patients with CSU? Most AID are rare or very rare diseases, whereas others, such as thyroiditis or type I diabetes, are more common. A recent population-based study involving ~800,000 individuals in Catalonia (Spain) investigated the prevalence of 78 autoimmune diseases and identified AITD with 5% as the most prevalent one, followed by type I diabetes and rheumatoid arthritis (both < 1%). 25 In our population of CSU patients (only a few patients were also from Catalonia), 25% had comorbid AITD, HT or Graves' disease, indicating that AITD is more than 5 times as common in CSU as in the normal population. This is in line with the results of numerous previous studies, which reported rates of HT to range from 1% to 28% in CSU. 17 In our study, vitiligo is also more prevalent in CSU patients as compared to the general population (2% vs. < 1%). 25 Furthermore, the prevalence of rheumatoid arthritis and autoimmune gastritis is likely to be higher in CSU (1% in our study) than in the general population (< 1%, Table 1 ). In 2% of our patients, CSU was associated with 2 or 3 autoimmune diseases, mostly HT and vitiligo, autoimmune gastritis or rheumatoid arthritis, a so-called autoimmune polyglandular syndrome. This is in line with previous studies, where vitiligo and rheumatoid arthritis were also the most frequently reported autoimmune diseases in combination with AITD and CSU. 11 , 12 26 27 In CSU patients as well as in the general population, autoimmune diseases are associated with female sex, older age and having a family history of autoimmune diseases. 11 12 22 25 28 In 2 large retrospective studies, chronic urticaria in female patients was strongly associated with AITD, SLE, vitiligo, Henoch-Schönlein purpura, 29 rheumatoid arthritis, Sjögren syndrome, celiac disease and type I diabetes mellitus. 12 Autoimmune diseases that are more prevalent in children and young adults, e.g . celiac disease and diabetes mellitus type I, 22 28 are also more frequently observed in younger CSU patients of our cohort. Caminiti et al . 30 previously reported that the presence of celiac disease in children with chronic urticaria was significantly more frequent than in control children. However, an overall prevalence of autoimmune diseases is likely to be lower in young CSU patients than in middle-aged CSU patients. 17 31 This might be explained by the fact that autoimmunity, including autoimmune CSU, develops with age and that most autoimmune diseases appear 10 years after the onset of CSU. 12 What are the clinical implications of our findings? AITD presents with euthyroidism or hypothyroidism in HT and hyperthyroidism (thyrotoxicosis) in Graves’ disease. HT is linked to many other autoimmune diseases, for example, vitiligo. 26 32 IgG anti-TPO is considered a sensitive marker for both HT and Graves’ disease. IgG anti-TPO appears in 90% of HT patients compared to 25%–50% for IgG anti-thyroglobulin. 33 In patients with HT, euthyroidism, i.e . IgG anti-TPO positivity, normal TSH and thyroid hormone levels, and absence of symptoms, or subclinical hypothyroidism can progress to overt hypothyroidism. 34 Increased rates of this progression are associated with female sex, TSH ≥ 7 mU/L and increased IgG anti-TPO titers. 33 34 35 36 The frequency of IgG anti-TPO positivity also increases with age, with a peak at around 50 years. 33 Does this mean that adult CSU patients should be screened for IgG anti-TPO? Several independent lines of evidence support to do this and suggest that screening CSU patients for HT may improve outcomes of both HT and CSU. First, we and others detected an increased prevalence of HT diagnosed by serum levels of IgG anti-TPO in >10% of CSU patients. It was associated with female sex, older age and higher rates of hypothyroidism. 7 12 17 Hypothyroidism, including subclinical hypothyroidism, is known to be linked to an increased risk for cognitive impairment, cardiovascular events and mortality from coronary heart disease. 34 Levothyroxine treatment may improve the prognosis, decrease the risk of adverse cardiovascular events and prevent progression to overt hypothyroidism. 34 Secondly, in CSU patients, HT was shown to be associated with other organ-specific and/or systemic autoimmune diseases including autoimmune polyglandular syndromes. 12 33 Therefore, diagnosis of HT may help to suspect and reveal other autoimmune comorbidity. Thirdly, HT may hint to the presence of aiCSU, which is usually more severe and resistant to omalizumab treatment. 4 37 38 Furthermore, treatment of hypothyroidism with levothyroxine might be associated with improvement of CSU, 17 39 although some studies did not confirm this. 40 The international urticaria guideline recommends that physicians assess CSU patients for thyroid hormones and autoantibodies in the extended diagnostic workup if indicated by history. 1 In light of the evidence presented above as well as the findings of the present study, we recommend that all adult CSU patients 7 be screened for IgG anti-TPO. If elevated levels are detected, thyroid function should be checked ( e.g . TSH, free T4) and tests for aiCSU ( e.g. ASST, BAT/BHRA) can be performed. If thyroid function is not impaired and symptoms are absent, patients should be followed up periodically to monitor for symptoms of hypothyroidism and to detect any rise in their TSH and/or development autoimmune diseases other than HT ( Fig. 2 ). 34 41 CSU, chronic spontaneous urticaria; TPO, thyroperoxidase; TSH, thyroid-stimulating hormone; LT, levothyroxine; aiCSU, autoimmune chronic spontaneous urticaria; CHD, coronary heart disease. * Referral to an endocrinologist. Are comorbid autoimmune diseases in patients with CSU linked to type IIb autoimmunity? Our findings show that in CSU, the presence of comorbid autoimmune diseases and a positive family history of autoimmune diseases are associated with markers of aiCSU, e.g . ‘triple positivity’ and low total IgE, which is in line with previous findings. 4 12 42 This may be explained by the known co-occurrence of multiple autoimmune phenomena and autoimmune diseases. Our observations, on one hand, strengthen the evidence for the existence of a distinct autoimmune subtype of CSU. On the other hand, they indicate the need to include targeted questions in the history and/or further laboratory diagnostic workup of comorbid autoimmune diseases in a subgroup of CSU patients. This is especially true for patients who are positive to ASST, BHRA and/or BAT and have autoantibodies detected by immunoassay. Surprisingly, we did not see higher rates of autoimmune diseases in patients with presence of IgG anti-FcεRIα/IgE. This might be explained by the fact that these antibodies are nonfunctional in some patients and were also described in healthy people. 43 Patients with aiCSU are likely to have a worse and slow response to omalizumab 38 44 and a good response to cyclosporine 45 and inhibition of Bruton's tyrosine kinase by fenebrutinib (Metz et al ., personal communication). Whether the presence of comorbid autoimmune diseases can also predict response to treatment in CSU patients is unknown. We found that the presence of ANA and/or IgG anti-TPO, markers for autoimmunity and autoimmune diseases, was associated with non-response to omalizumab treatment. This is in line with recently reported data from a cohort of Turkish CSU patients, where poor response to omalizumab treatment was also linked to ANA-positivity. 46 In addition, in small studies and reports, omalizumab was less effective in patients with CSU and IgG anti-TPO. 47 48 In other studies, cyclosporine was more effective in CSU patients with ASST and/or BHRA positivity and raised thyroid autoantibodies. 49 50 The major strength of the current study is the inclusion of a large cohort of patients from 12 European centers and assessment of many parameters, including triple testing for aiCSU. The retrospective analysis is a limitation of our study. The data from the patients was extracted from different centers in different countries and it can be expected that the diagnosis, management and treatment of patients differ significantly among centers and countries. Moreover, we compared aiCSU patients with patients with non type IIb autoimmune CSU rather than those with type I autoimmune CSU. This is a limitation of this study and needs to be addressed by future studies. Importantly, such studies need to also compare patients with type I and type IIb autoimmune CSU with patients who have both type I and type IIb autoimmune CSU, a subpopulation of CSU patients that is currently ill characterized in terms of its rate and clinical characteristics. 51 In addition, the 3 markers for aiCSU, i.e. basophils tests, ASST and IgG anti-FcεRIα/IgE, were only analyzed, in combination, in the patients from the PURIST study. The therapeutic response to omalizumab was evaluated only in a small subgroup of patients. Finally, some of the routine analyses were not performed by a central laboratory, rather than the centers involved, which may have increased the heterogeneity of data. In conclusion, autoimmune diseases, especially HT and vitiligo, are common in CSU and linked to aiCSU and poor response to treatment with omalizumab. In CSU patients, autoimmune diseases are often overlooked, usually occur as CSU persists, can be treated and prevented by early treatment, and treatment can have positive effects on CSU. Therefore, we recommend that CSU patients be assessed for features of aiCSU and be investigated, at the onset of their urticaria and at regular intervals, for comorbid autoimmune diseases. In all adult CSU patients, we strongly recommend a) checking for symptoms of autoimmune diseases and screening for IgG anti-TPO with annual reassessment of thyroid function in positive asymptomatic cases; and b) a follow-up on a regular basis of patients with aiCSU, especially female patients with HT and/or a positive family history, to screen for signs and symptoms of other autoimmune diseases. 11 41 Further studies are needed and should use additional markers for aiCSU, for example total IgE 52 or the ratio of IgG anti-TPO to total IgE, a recently reported marker for aiCSU. 4

We analyzed data from 1,199 CSU patients from the urticaria specialty clinics of Charite-Universitätsmedizin Berlin, Germany (n = 791), I.M. Sechenov First Moscow State Medical University, Russia (n = 249) and 10 further collaborating urticaria centers (n = 159, the PURIST study, some data have been published before 4 ). Patient data were analyzed retrospectively. The study followed Good Clinical Practice guidelines and the Declaration of Helsinki, ethics approval was obtained by each of the cooperating centers as required, and all patients provided signed informed consent. CSU was defined as recurrent whealing, angioedema, or both for > 6 weeks due to known or unknown causes and without definite triggers. Patients with chronic inducible urticaria alone and patients with urticarial vasculitis were excluded. We analyzed the patients' records for a history or presence of autoimmune diseases including vitiligo, rheumatoid arthritis, celiac disease, type I diabetes mellitus, SLE, autoimmune gastritis/pernicious anemia, ankylosing spondylitis, alopecia areata, myasthenia gravis, antiphospholipid syndrome, scleroderma, idiopathic thrombocytopenic purpura, dermatomyositis, primary biliary cirrhosis and autoimmune hepatitis. CSU patients with several other conditions such as psoriasis, endometriosis, fibromyalgia, sensorineural hearing loss and thrombocytopenia, in which an autoimmune nature of the disease has been discussed, but not ultimately confirmed, were excluded from the analyses unless other autoimmune diseases were present. AITD was assessed if diagnosed by a physician and/or elevated levels of IgG anti-TPO with euthyroidism, hypothyroidism (HT) or hyperthyroidism (Graves' disease) were present. A few patients with CSU and hypothyroidism might have normal thyroid hormones levels because of the levothyroxine treatment. Blood was drawn and analyzed for routine patient care. From the patients' record, we extracted data on age, sex, duration of CSU, presence of concomitant angioedema, family history of autoimmune diseases, disease activity and ASST results, complete blood count, antinuclear antibodies (ANA), serum total IgE and IgG anti-TPO levels, when available. We analyzed rates of autoimmune diseases in 4 age groups of CSU patients ( 39 years) based on the previously reported data on the most common mean age-of-onset of autoimmune diseases of 20–29 and 40–50 years. 22 Disease activity was determined using the urticaria activity score (UAS), which was calculated as the sum of the itch (no = 0, mild = 1, moderate = 2, intense = 3) and the wheal score (no wheals = 0, 50 wheals/24 hr = 3) for 7 consecutive days (UAS7, minimum = 0, maximum = 42). CSU patients were divided into 3 groups according to CSU activity: low disease activity = 0–15 (urticaria-free, well-controlled CSU or mild CSU); moderate disease activity = 16–27 (moderate CSU); and high disease activity = 28–42 (severe CSU). 23 The aiCSU is defined according to an agreed set of 3 diagnostic criteria: (a) positive in vivo autoreactivity (positive ASST); (b) positive in vitro BHRA and/or basophil activation test (BAT) as evidence of functional serum histamine-releasing autoantibodies in patients' sera that are directed against IgE or high-affinity IgE receptors; and (c) a positive immunoassay for detection of IgG autoantibodies against FcεRIα and/or anti-IgE. The ASST is a screening test for aiCSU that evaluates the presence of histamine-releasing serum factors of any type, including IgG autoantibodies against FcεRIα or IgE. The ASST is regarded as positive when the serum wheal is ≥ 1.5 mm as compared to the negative control (saline). 24 In CSU patients from the PURIST study (n = 159), all three aiCSU-defining tests were performed as previously described. 4 Most analyses were performed in all centers, except BHRA/BAT, IgG anti-FcεRIα/IgE (data from the PURIST study) and UAS7 (data from the urticaria clinic of Berlin and the PURIST study). UAS7 was usually calculated seven days before the blood sampling and ASST. The association between the presence of markers of autoimmune diseases, namely ANA and IgG anti-TPO, and omalizumab 300 mg treatment was retrospectively analyzed in 45 CSU patients from the Berlin cohort. Of 45 patients, seven, 18, and 20 were non-responders (a reduction in UAS7 of less than 30%), partial responders (a reduction in UAS7 of 30% to 89%) and complete responders (a reduction of 90% or more from baseline in UAS7) to omalizumab, respectively. Pre-omalizumab treatment measurements of ANA and IgG anti-TPO were used for this analysis. Statistical analysis was performed using the Statistical Package for the Social Science (IBM SPSS version 25.0; IBM Corp, New York, NY, USA). Figures were made using GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA). Due to the methodology, there were missing data for some variables, and some analyses were performed in subgroups. Demographic characteristics of CSU patients were reported using descriptive statistics. Because data were not distributed normally, non-parametric tests were used. Groups were compared using the Mann-Whitney U test, the chi-square test of homogeneity and Fisher's exact test (if data violated the sample size adequacy assumption). P values < 0.05 were considered significant. Post hoc analysis involved pairwise comparisons using the z test of 2 proportions or multiple Fisher's exact tests with Bonferroni correction (statistical significance was set at P < 0.02).