IgA vasculitis in patients with inflammatory bowel disease: new insights into the role of TNF-α blockers

Abstract

Introduction

Immunoglobulin A (IgA) vasculitis (IgAV), also known as Henoch-Schönlein purpura, is an immune complex-mediated small vessel vasculitis characterized by the presence of IgA1-dominant immune deposits within the vessels [1]. IgA vasculitis most often presents with a clinical triad comprising cutaneous involvement with vascular purpura, musculoskeletal involvement with arthralgia and/or arthritis, and gastrointestinal (GI) involvement ranging from abdominal pain to more severe disease with intestinal hemorrhage and/or perforation [2, 3]. Besides GI involvement, renal involvement presenting as glomerulonephritis affects the long-term prognosis [4, 5]. In contrast to the disease in children which is frequent and associated with benign course, IgAV is usually more severe in adults due to GI and renal involvement. Although symptomatic management is frequently sufficient in benign forms, severe cases may require more aggressive treatments, mainly glucocorticoids (GCs), sometimes in combination with immunosuppressants. IgAV pathogenesis is not fully elucidated but IgA is suspected to play a key role, especially with respect to its peculiar implication in innate immunity within the GI mucosal barrier, where various pathogens may interact with the host. Susceptibility genes have been also identified, involving the HLA system [6], which may promote the occurrence of the disease. Thus, environmental and genetic factors play a role in the pathophysiology of this multifactorial disease [7, 8].

IgA vasculitis frequently occurs in the absence of other underlying conditions. However, vasculitides may be associated with other diseases. For instance, association between vasculitis and IBD was reported [9], but concerned mainly large vessel vasculitis (i.e. giant cell arteritis and Takayasu arteritis) and ANCA-associated vasculitides. Besides, although IgA nephropathy is considered as a distinct entity from IgAV, a recent French cohort reported its association with IBD highlighting some interesting pathogenesis hypothesis [10]. Thus, IgA seems to play a role in both IgAV and IBD. Nevertheless, the association between IgAV and IBD was only described in a few case reports in the literature [11–16]. Therefore, additional data are required to better describe this poorly known association.

The aim of this study was to describe the association, phenotype and management of patients presenting with the association of IgAV and IBD.

Methods

Patients

We conducted an observational, multicentre, retrospective study in France, between 8 February 2019, and 16 February 2020, including adult patients at inclusion presenting with the association of IgAV and IBD, through the implication of physicians participating to the Groupe Français d'Etudes des Vascularites (FVSG) and the Groupe d'Etude Thérapeutique des Affections Inflammatoires du Tube Digestif (GETAID). These two networks were created respectively in 1981 for the FVSG and 2012 for the GETAID, with the aim of improving the knowledge and conducting studies on their topic of expertise. To be included in the study, patients had to have both IgAV and IBD, regardless of chronology between the two diseases. IBD was diagnosed according to the European Crohn's and Colitis Organisation (ECCO) guidelines [17] and the phenotype classification based on the Montreal classification [18]. Due to the lack of unicist classification for IgAV in adults, patients had to meet the ACR [19] and EULAR (European League Against Rheumatism) classification criteria [20] in the presence of purpura (mandatory criteria). IgA vasculitis patients were then categorized into definite or probable IgAV cases (Supplementary Table S1, available at Rheumatology online). A definite case was defined by the presence of IgA deposits on a biopsy performed in a patient suspected of vasculitis and showing either leukocytoclastic vasculitis on skin biopsy or glomerulonephritis on kidney biopsy. A probable case was defined in the absence of histological evidence of IgA deposition, if the purpura was highly suggestive of IgAV (i.e. extensive to the abdomen and/or the four limbs, and/or the presence of necrotic or bullous lesions), and was associated with at least one other organ involvement (i.e. arthralgia and/or arthritis, glomerulonephritis, GI involvement), and in the absence of any obvious differential diagnosis. All cases were reviewed independently by two physicians with expertise in vasculitis (C.R. and B.T.) using a standardized form, and in case of disagreement, the case was reviewed jointly prior to inclusion.

We excluded patients suspected for IgAV if they did not meet the criteria listed above, including the presence of a non-suggestive and isolated purpura with no other organ involvement or associated with another differential diagnosis. Also, other dermatoses associated with IgA deposits but not suggestive of vasculitis such as bullous or linear IgA dermatosis were excluded, as well as patients with cancer or liver disease (cirrhosis, autoimmune hepatitis or primary sclerosing cholangitis). In contrast, patients presenting with inflammatory conditions associated with IBD, i.e. spondyloarthritis, ocular inflammation or hidradenitis suppurativa, were included.

Ethics

This study was conducted in compliance with the Good Clinical Practice protocol and the Declaration of Helsinki principles and was approved by the local ethics committee Institutional Review Board from Cochin Hospital, Paris (CLEP No. AAA-2021–08002).

Patient and public involvement statement

There has been no patient and public involvement.

Data collection

We used a standardized case report form to collect data from each patient, from the start of the first disease diagnosed until last follow-up. For each patient, data related to IBD included gender, age at IBD diagnosis, familial history, smoking status, Montreal classification criteria [18] concerning location and behaviour in case of Crohn’s disease (CD) or extent and severity in case of ulcerative colitis (UC). Extraintestinal involvement and complications like abscess, stenosis, fistula, or history of surgery were also recorded. Finally, data concerning follow-up before, during and after IgAV were gathered. Data related to IgAV included age at IgAV diagnosis, clinical manifestations (i.e. cutaneous, musculoskeletal, GI and renal involvement, but also less frequent neurological, pulmonary and/or ocular involvement), routine laboratory results, biopsy performed, treatments received and follow-up outcomes (duration, relapses, complications and possible sequelae).

Subgroups

In this study, we have taken an interest in some specific situations that we have deemed relevant, including: the description of the overall cohort, the description of the situation of IgAV occurring after IBD, and the description of the situation of IgAV occurring after IBD, while treated with anti-TNF-α.

Statistical analysis

Patient’s characteristics were analysed using descriptive statistics. Categorical variables were expressed as number (percentage) and continuous variables as median with interquartile range (IQR).

Results

Patient’s characteristics

Sixty suspected cases from 17 centres were identified, among them 17 patients were excluded either because they did not meet inclusion criteria (n = 9), had a differential diagnosis (n = 7) or because of insufficient data (n = 1). Overall, 43 patients were finally included (Fig.  1). Characteristics of IBD are summarized in Table  1, and characteristics of IgAV in Table  2.

Fig. 1

Flow chart of the study

CD: Crohn’s disease; IgAV: IgA vasculitis; UC: ulcerative colitis.

Open in new tabDownload slide

Median age at IBD diagnosis was 26.0 (IQR 20.0–32.5) years and at IgAV diagnosis was 33.0 (IQR 26.0–46.5) years, including four patients with an age < 18 years, and 22 patients were men (51%).

IBD preceded the diagnosis of IgAV in 38 (88%) cases, all but one occurring in adults, with a median interval of 9.2 (IQR 5.4–15.4) years between the two diseases. In contrast, IgAV preceded IBD in five (12%) patients, with a median interval of 133 months (IQR 72.5–176.3). For these five patients, very few data concerning IgAV was available, especially for three patients who had a diagnosis of IgAV during childhood.

Description of inflammatory bowel diseases

Twenty-five patients (58%) had CD, and 18 (42%) had UC. Eleven (26%) patients had first or second-degree family history of IBD. Eighteen out of 38 (47.3%) patients were active or past smokers (data not available for five patients).

Crohn’s disease

Data on CD location, disease behaviour and complications were available for 24 out of 25 patients. Location was ileal, colonic, ileocolonic and upper digestive in five (21%), three (13%), 14 (58%) and two (8%) patients, respectively. Disease behaviour was inflammatory, stricturing and penetrating in 15 (62.5%), six (25%) and three (12.5%) patients, respectively. Sixteen patients had CD complications during follow-up, including abscess in eight (33%), fistulae in 10 (42%) and stenosis in 10 (42%). Fifteen CD patients (63%) had surgery, including intestinal surgeries such as ileocolic resection and/or colectomy in 10 cases and perianal surgery for fistula in three cases.

Ulcerative colitis

UC extent was rectal, left-sided and extensive in three (17%), six (33%) and nine (50%) patients, respectively. Three patients (17%) underwent surgery for colectomy.

Treatments used in inflammatory bowel diseases

IBD treatments are summarized in Table  3. IBD patients received topical or oral GCs in 30 (70%), AZA in 29 (67%), 5-aminosalicylic acid (5-ASA) in 25 (58%), anti-TNF-α therapy in 35 (81%), including adalimumab in 18, infliximab in 15 and golimumab in two, and less frequently other biological agents (ustekinumab in four and vedolizumab in four cases).

Characteristics of IgA vasculitis

Clinical and laboratory characteristics of IgAV are summarized in Table  2. Diagnosis of IgAV was definite in 30 (70%) and probable in 13 (30%) cases. Main clinical features of IgAV were available for 40 patients and showed vascular purpura in 40 (100%), joint involvement in 22 (55%), renal involvement in 17 (43%) and GI involvement in 12 (30%) patients. Tissue biopsy was performed in 40 patients, including skin biopsy in 28 and kidney biopsy in 13 patients, with IgA deposits identified in 30/40 (75%) of all cases and 12/12 (100%) of the kidney biopsies.

After a median follow-up of 17.5 (IQR 7.2–64.4) months, 14 (33%) patients had at least one relapse, presenting as purpura in 12, joint involvement in five, renal involvement in two and GI involvement in one. Rare manifestations, potentially related to both diseases included ocular inflammation (n = 4), including both uveitis (n = 2) and scleritis (n = 2), hidradenitis suppurativa (n = 3), ankylosing spondylitis (n = 2) and mononeuritis multiplex (n = 1).

Description of IgAV cases occurring after IBD diagnosis

IBD preceded the diagnosis of IgAV in 38 (88%) cases, with a median interval of 9.2 (IQR 5.4–15.4) years. Among these patients with known IBD at the time of IgAV diagnosis, five (13%) had an active IBD. Treatments in this group of patients are summarized in Table  3. Thirty-four (89%) patients received anti-TNF-α therapy for IBD during their evolution. IgAV occurred under treatment with anti-TNF-α therapy in 28/38 (74%), including adalimumab in 15 and infliximab in 13, with a median time after anti-TNF-α therapy initiation of 31.5 (IQR 19.0–55.5) months.

Management of IgAV included both specific treatment for vasculitis and modification of IBD treatment. Twenty-seven out of the 38 patients (71%) received IgAV specific treatments, including oral GCs in 25, colchicine in six, CYC in six, AZA in three, HCQ in three and dapsone in two.

For the 28 patients with IgAV occurring under TNF-α blockers, anti-TNF-α therapy were discontinued in 15 (54%) and continued in 13 (46%) patients. There was no difference regarding vasculitis features between these two situations. When discontinued, TNF-α blockers were switched to 5-ASA in three, AZA in five, ustekinumab in four and vedolizumab in three. No IgAV relapse was observed after discontinuation of TNF-α blockers, while IBD relapse or complications were noted in five (33%) patients, after a median time of 4 months (IQR 2–7.5). TNF-α blockers were reintroduced in six (40%) patients, using the same agent in four and another agent in two, with four (67%) experiencing a vasculitis relapse after a median time of 2.5 months (IQR 1.75–5.25). In these four patients, two relapsed after resuming the same anti-TNF-α, and two after switching to another. Among the 13 patients who continued anti-TNF-α therapy, three (23%) experienced IgAV relapse after a median time of 2 months (IQR 1.5–3), while IBD relapses and/or complications were noted in only one (8%) patient (Table  4).

Clinical features of patients under anti-TNF-α experiencing IgAV relapse, either because there were not discontinued (n = 3) or after reintroduction (n = 4), included cutaneous involvement purpura in all cases and joint involvement in two cases. No severe organ involvement was noted.

After a median follow-up of 17.3 (IQR 6.9–49.1) months, four patients presented chronic renal failure attributable to IgAV-related glomerulonephritis, including one patient with end-stage renal disease requiring renal transplantation. Two patients died: one from cardiogenic shock after renal transplantation and the other one from sepsis in the context of urothelial cancer.

Discussion

This study describes the association of IBD and IgAV and provides new insights into this rare but probably non-fortuitous association. The main conclusion drawn by our study is that IBD preceded IgAV in most cases and that anti-TNF-α therapy appeared to promote the onset of vasculitis. Also, discontinuation of anti-TNF-α was associated with the absence of IgAV relapse but high rate of relapses or complications of IBD, contrasting with an opposite situation when anti-TNF-α therapy were pursued.

Patients with both diseases displayed common demographic and clinical features compared with isolated IBD and IgAV. Both males and females were equally represented, while IgAV usually shows slight male predominance with a sex ratio of 1.5 [2, 21] and IBD affects both genders equally [22, 23]. The main IgAV characteristics were consistent with what was previously described in the literature [2, 4, 24], except for less frequent arthralgia and renal involvement. Moreover, IgAV prognosis was better than usually reported in adults, as chronic renal failure was observed in <10% in our study population. Also, a smaller case series on nine patients with the association of IgAV and IBD was recently published [25]. Median age was 40.8 years vs 33 years in our study, with the same male predominance. Among the nine patients, six were under anti-TNF-α at the time of IgAV onset, and the mean time between initiation of anti-TNF-α and IgAV was 3.3 years in this study compared with 2.6 years in our cohort. At vasculitis onset, three continued anti-TNF-α, two were switched to another anti-TNF-α and treatment was discontinued in one. At last follow-up, 3/5 still under anti-TNF-α achieved remission of vasculitis. Major organ involvement was comparable between the two studies, except for more frequent renal involvement (in 78% compared with 43% in our cohort). Moreover, a case-control analysis comparing cases with 18 patients with IgAV without IBD did not reveal any significant difference in demographic or clinical characteristics of IgAV.

Some evidence suggests that the association of IBD and IgAV, both multifactorial diseases, could be non-fortuitous [8]. First, in a retrospective review on native kidney biopsies from patients with IBD, IgA nephropathy was the most frequent kidney biopsy diagnosis in IBD and has a significantly higher diagnostic prevalence compared with all non-IBD kidney biopsy specimens [26]. Common genetic background could be shared between the two diseases, as supported by the description of a multiplex family with two children presenting IgAV and their mother having CD [27]. Nevertheless, genomic studies performed in these two diseases do not seem to find commonly shared variants [28, 29]. IBD and IgAV may also share common pathophysiological features. Indeed, IgA has the dual roles of maintaining homeostasis with the microbiome and protecting from intestinal infection [30]. Secretory IgA are produced by intestinal plasma cells, and Peyer’s patches, major organized lymphoid tissues of the small intestine, are the dominant source of IgA-producing cells [31]. IBD is associated with an immune activation in Peyer's patches and mucosal lymph nodes, with excessive stimulation of B cells. This could lead to an exaggerated production of IgA (especially IgA1) and its remote deposition within tissues, which could partly explain the onset of IgAV in the setting of IBD [32–34]. In addition to the excessive IgA production, another potentially important aspect would be its glycosylation state. As a matter of fact, defective IgA glycosylation seems to play a central role in IgA nephropathy and IgAV-associated nephropathy (histologically indistinguishable) [8]. Moreover, abnormalities in the O-glycan synthesis have been also linked to IBD, studies indicating that proper O-glycosylation of mucins may play a role in human susceptibility to IBD and decreased glycosylation was observed in the intestinal mucus of patients with IBD and correlated with disease activity [35, 36].

Most IgAV cases occurred under anti-TNF-α therapy prescribed to control IBD. We recently performed a pharmacovigilance study using the French pharmacovigilance database and the World Health Organization global individual case safety reports database (VigiBase), to identify the main drugs reported to induce IgAV. Drugs associated with a disproportionality in IgAV reporting included TNF-α blockers, especially when they were used for IBD [37]. In addition, vasculitis relapses after the rechallenge of the anti TNF-α therapy indicate a possible link between this treatment and the disease. Besides, relapse of IgAV occurred whether the same anti-TNF-α therapy was reintroduced or was switched to another, suggesting a therapeutic class effect rather than a specific molecule involvement. Although, we acknowledge one limitation being that the anti-TNF-α agent may have been introduced in a more severe subgroup of patients and a potential bias would be that the association is related to IBD severity rather than the treatment itself. Mechanisms by which anti-TNF-α could induce IgAV are not yet elucidated, but hypotheses from the literature include formation and deposition of anti-TNF/TNF-α immune complexes in capillaries, induction of shifts in T-cell responses in patients treated with anti-TNF-α therapy, and/or a direct drug toxicity [38, 39].

This association and its characteristics raise the question of its optimal management. Discontinuation of TNF-α blockers appeared to be beneficial for vasculitis with no relapse observed but not for IBD, as relapses or complications occurred in one-third of cases. Conversely, a significant proportion of IgAV relapses were noted when anti-TNF-α were continued or resumed. It should be noted that IgAV relapses were usually mild with mainly skin and joint involvement. Therefore, a multidisciplinary discussion should take place between the different specialists involved in the management of these patients to define the therapeutic strategy according to the benefit/risk ratio.

Finally, our study has some limitations because of its retrospective design. Some cases had missing data that may limit the detailed description of the long-term evolution. Also, we did not perform a case-control study including IBD without IgAV and IgAV without IBD. However, characteristics of our study population were comparable to features commonly reported in isolated IBD and IgAV. As strengths, we involved two French networks, i.e. the FVSG and GETAID, specialized in vasculitis and IBD, respectively, allowing the identification of an important number of patients for a rare association and thus making this study the largest cohort reporting IgAV associated with IBD. Besides, diagnosis of IBD was based on robust and validated criteria, while diagnosis of IgAV lacks solid criteria in adults. To overcome this, we applied very restrictive criteria that required evidence for IgA deposition within vessels and all cases were reviewed by two specialists with expertise in vasculitis.

Overall, in this large series, IBD patients who developed IgAV were mainly treated by anti-TNF-α therapy. In these patients, discontinuation of anti-TNF-α was associated with remission of vasculitis but a high rate of complications and/or relapses of IBD, whereas continuation of anti-TNF-α was associated with relapse of vasculitis but remission of IBD. This suggests a key role of anti-TNF-α agents in this condition.

Acknowledgements

We acknowledge the important contribution of the GFEV and GETAID networks in data collection. All individuals listed as authors should quality for authorship according to the following four ICMJE criteria: (i) substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND (ii) drafting the work or revising it critically for important intellectual content; AND (iii) final approval of the version to be published; AND (iv) agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: B.T. received some consulting fees and/or grants from Roche/Chugaï, AstraZeneca, GlaxoSmithKline, Bristol-Myers Squibb, Lilly, Vifor Pharma, LFB, Grifols and Terumo BCT.

Data availability statement

Data available upon request to the corresponding author.

Supplementary data

Supplementary data are available at Rheumatology online.

References