Review

Paradoxical effects of anti-TNF-a agents in inflammatory diseases Expert Rev. Clin. Immunol. 10(1), 159–169 (2014)

Daniel Wendling* and Cle´ment Prati CHRU de Besanc¸on (University Teaching Hospital) and Universite´ de FrancheComte´, Boulevard Fleming, F-25030 Besanc¸on, France *Author for correspondence: [email protected]

Anti-TNF agents represent a major breakthrough in the management of inflammatory diseases. Among the side effects of these agents are the so-called paradoxical effects described in this review. They represent new onset or exacerbation of a condition (symptom/disease), usually improved with TNF blockers. These paradoxical effects are mainly psoriasiform skin reactions, uveitis and granulomatous diseases (such as sarcoidosis and Crohn’s disease). Infrequent and probably underreported, they should be discussed from the viewpoint of spontaneous features of the underlying disease (e.g., uveitis or psoriasis in a case of spondyloarthritis). The causal mechanism of occurrence is still a matter of debate, but may implicate an imbalance of cytokines toward interferons, chemokines and probably IL-17. These reactions may raise differential diagnosis problems. Symptoms resolve, most of the time, due to the discontinuation of the anti-TNF agent or sometimes a switch to another TNF blocker; but in some cases, it is a class effect that could lead to the withdrawal of all anti-TNF agents. KEYWORDS: anti-TNF agents • Crohn’s disease • paradoxical effect • psoriasis • sarcoidosis • TNF-a • uveitis • vasculitis

Anti-TNF agents represent a major breakthrough in the therapeutic armamentarium of inflammatory diseases. These agents have proved efficacious in several conditions with a safety profile and guidelines for current use [1]. Besides these adverse events, the so-called ‘paradoxical effects’ have been described with the use of anti-TNF agents. These are defined as new onset or exacerbation of a condition (symptom/disease) usually improved with TNF blockers. TNF-a: a major cytokine in inflammation

This cytokine is produced by many cell types (T cells, macrophages, fibroblasts, keratinocytes) after stimulation (bacteria, virus, immune complexes, cytokines [IL-1, IL-17, GM-CSF, INF-g], complement factors, microcrystals, tumor cells, ischemia, trauma etc.) Its transmembrane precursor is cleaved in a soluble form by a TNF-a converting enzyme (TACE). Both transmembrane and soluble forms are ligands for the TNF receptors 1 and 2 and transmembrane TNF, which may act as a ligand too [2]. TNF exerts many biological actions via several target cells (BOX 1) [3,4]. www.expert-reviews.com

10.1586/1744666X.2014.866038

TNF overexpression has been demonstrated in sites of interest (inflammatory diseases): synovial membrane in rheumatoid arthritis (RA), sacroiliac joints from ankylosing spondylitis (AS) patients, psoriatic skin or bowel mucosa from Crohn’s disease (CD). The anti-TNF agents

There are currently five anti-TNF agents available and bio-similars in development [5–7]. These are four anti-TNF monoclonal antibodies and one soluble p75 TNF receptor (TABLE 1). Although these agents act as TNF inhibitors, bind soluble and membrane TNF; they have different structures, half-life and immunological behavior. For example, etanercept has the capacity to bind to lymphotoxin; the modification of the Fc region (truncated in etanercept, absent in certolizumab) may explain the differences in cell lysis and granuloma formation for these several anti-TNF compounds. Paradoxical effects of anti-TNF treatment

Many and various paradoxical effects have been reported under anti-TNF treatment. They may be grouped under several headings (BOX 2).

 2014 Informa UK Ltd

ISSN 1744-666X

159

Review

Wendling & Prati

Box 1. Main biologic actions of TNF-a in inflammation. • Activation of transcription of different proinflammatory cytokines (TNF-a, IL-1, IL-6) and growth factors (GM-CSF) • Hepatic synthesis of acute phase reactants of inflammation (CRP) through IL-6 production • CNS thermoregulation (fever) • Cellular recruitment in inflammatory sites related to overexpression of endothelial and cellular adhesion molecules (ICAM-1, VCAM-1, selectins) and chemokines (MCP-1, IL-8, and so on) • Monocyte and macrophage activation • Release of immediate defense molecules (free radicals, nitric oxide, metalloproteinases, and so on) by neutrophils, chondrocytes, epithelial cells, and so on • Arachidonic acid release and prostaglandin secretion • Apoptosis induction of infected cells • Mitogenic activity on fibroblasts • Adaptative immune system activation. TNF acts as Treg activator, inducing their proliferation and membrane expression of TNF receptors

These correspond to the occurrence of an unexpected symptom or disease, which is a potential indication of TNF blockade. They affect various organs or tissues and are not considered as an evolution of the underlying disease treated by the anti-TNF agent. Psoriasis & psoriasiform lesions

These are the most common and the first reported paradoxical effects of TNF blockers. Anti-TNF agents (infliximab, etanercept and adalimumab) are licensed for the treatment of the severe forms of skin and nail psoriasis reactions. Evidence [8,9] showed onset or exacerbation of cutaneous psoriasis under antiTNF treatment (TABLE 2).

Harrison et al. [10] investigated the occurrence of psoriasis skin lesions in RA patients receiving anti-TNF treatment from the registry of the British Society for Rheumatology. They found 36 cases (25 were analyzed) out of 23,996 person-years of follow-up receiving anti-TNF and none in 5207 personyears receiving conventional disease-modifying anti-rheumatic drugs (DMARDs). This allowed a rate calculated using a hypothetical case of psoriasis per 1000 person-years (95% CI: 0– 0.71) for controls; 1.04 (0.67–1.54) for anti-TNF; 0.59 (0.22– 1.28) for etanercept; 0.88 (0.32–1.93) for infliximab and 1.84 (0.98–3.15) for adalimumab. Tillack et al. [11] reported that among the 434 anti-TNFtreated patients with inflammatory bowel diseases (IBD), 21 (4.8%) developed psoriasiform skin lesions. In multiple logistic regressions, it was revealed that smoking (odds ratio [OR]: 4.24; 95% CI: 1.55–13.60) and an increased body mass index (OR: 1.12, 95% CI: 1.01–1.24) are main predictors of these lesions. These data demonstrate psoriasis occurrence irrespective of the disease treated, even if some may be developed or associated with psoriasis such as spondyloarthritis (SpA) or IBD with each of the TNF blockers (monoclonal antibodies and soluble receptor). This event has also been described with certolizumab [12]. Age and sex ratio vary from one study to the other, depending on the type of disease (RA is mostly detected in women over 50 years old). Skin psoriasis reaction occurs after 5 months (extremes 1–24) of exposition to TNF blockers. Eighty-five percent of the cases from the literature survey by Collamer et al. [13] and Collamer and Battafarano [14] resulted in this same side effect, which is a new onset of psoriasis and a relapse in the remaining cases. This side effect occurs with the several anti-TNF agents, as well as in monoclonal antibodies as the soluble receptor; distribution may vary from study to study.

Table 1. Main characteristics of the several anti-TNF agents. Anti-TNF agent

Structure

Infliximab

Chimeric anti-TNF monoclonal anti body (mouse variable region, human IgG1 constant region)

Particularities

Indications RA, AS, PsA CD, UC Pso JIA

Adalimumab

Human anti-TNF monoclonal antibody (human variable region, human IgG1 constant region)

RA, AS, PsA CD, UC Pso JIA

Golimumab

Human anti-TNF monoclonal anti body (human variable region, Human IgG1 constant region)

RA, AS, PsA

Certolizumab

Humanized anti-TNF Fab, fusion with PEG

No Fc region: no complement activation

RA CD

Etanercept

Fusion protein of human TNF R2 and human IgG1 truncated constant region

Binds to lymphotoxin Does not bind to complement (due to truncated Fc), no complement dependent cytotoxicity triggered by binding of tm TNF

RA, AS, PsA Pso JIA

AS: Ankylosing spondylitis; CD: Crohn’s disease; JIA: Juvenile idiopathic arthritis; PsA: Psoriatic arthritis; Pso: Psoriasis; RA: Rheumatoid arthritis; tm: Transmembrane; UC: Ulcerative colitis.

160

Expert Rev. Clin. Immunol. 10(1), (2014)

Paradoxical effects of anti-TNF-a agents in inflammatory diseases

Skin lesion plaque psoriasis in 50% of the cases were reported by Collamer et al. [13] and Collamer and Battafarano [14], though an increased rate of pustular psoriasis was mentioned [8], sometimes, the lesions are more of psoriasiform lesions [15]. Several possible outcomes were also reported. Resolution with the discontinuation of anti-TNF therapy is reported in 24% of the 207 cases reviewed [14], whereas 26% was resolved and 25% was partially resolved on anti-TNF and 5% exhibited partial or no resolution off anti-TNF. Six percent had no recurrence with change of anti-TNF, 2% resolved off anti-TNF and recurred with reintroduction of the same anti-TNF and 6% after switching to another anti-TNF therapy; no resolution on anti-TNF mentioned in 1% and no outcome in 5%. This may illustrate the possibility of a class effect related to TNF blockade with the absence of a unique mechanism, in view of the cases of regression of psoriasis while continuing the antiTNF therapy. In major cases, psoriasis occurs, whereas the TNF blocker controls the underlying disease. Hypothesis of mechanism

Hypersensitivity reaction such as acute generalized exanthematous pustulosis or bacterial infection secondary to TNF inhibition should be ruled out as they may necessitate some specific treatments, but these hypotheses are not in accordance with the outcome described earlier. One hypothesis gives the central part to the type 1 IFN-a [10], which plays a central role in psoriasis; TNF downregulates the production of IFNs by the plasmacytoid dendritic cells of the derma, thus inhibiting TNF would enhance IFN production and favor psoriasis development. Another mechanism may be the increased expression of chemokines and receptors (such as CXCL9, CXCR3) leading to increased homing of lymphocytes in the skin [16]; IFN-a is known to induce expression of chemokine receptor CXCR3 on T cells [14]. Finally, the importance of the IL-12/IL-23 pathway and activation of the Th17 pathway have been underlined [17] with increased skin expression of IL-1b, IL-17, IL-21 and IL-22, increased serum levels of IL-17A and IL-22, hyperactivation of the Th17 and Th1 pathways and reduction of Treg activity. Skin lesions are histologically characterized by infiltrates of IL-17A/IL-22 secreting Th17 cells and IFN-g secreting Th1 cells and IFN-a expressing cells [11]. To sum up these hypotheses, it appears that the imbalance between cytokines secondary to TNF blockade (leading to dendritic cell activation) represents the major common mechanism leading to psoriasis onset or reactivation [18]. Differential diagnosis

Occurrence of skin lesions under anti-TNF treatment may represent a challenge for the clinician (BOX 3), which sometimes need a skin biopsy to rule out some other diagnosis [19,20]. Data from the Spanish Registry of biologic therapies in rheumatic diseases (Biobadaser) [21] revealed 17,329 person-years of www.expert-reviews.com

Review

Box 2. The paradoxical effects of anti-TNF agents. • • • • • • •

Psoriasis and psoriasiform lesions Uveitis and other eye diseases Inflammatory bowel diseases Sarcoidosis Other granulomatous disorders (rheumatoid nodules) Vasculitis Miscellaneous auto-immune

exposure to anti-TNF agents incidence rates for cutaneous adverse events of 53 per 1000 person-years; infection 28/1000 person-years; infusion reactions 15; autoimmune skin disease 5 and skin malignancy 3/1000 person-years. Management proposal

In most of the cases, skin lesion improves after withdrawal of the biologic agent. Treatment of psoriasis may allow the maintenance of anti-TNF treatment for the control of the underlying inflammatory disease, thus avoiding a possible flare of it in case of discontinuation of the biologic agent; this may be a suitable option in more than two-third of the cases [13]. Collamer et al. [13] proposed a treatment algorithm for psoriasiform skin eruptions under TNF antagonists. Patients with lesion should be referred to a dermatologist for clinical and potential histological confirmation. In case of severe intolerable lesion or patient preference, anti-TNF therapy should be discontinued. In case of tolerable lesion less than 5% body surface area, anti-TNF therapy may be continued with topical treatment for psoriasis; if psoriasis is recalcitrant, a switch to another anti-TNF agent or a discontinuation should be proposed. In case of tolerable lesion more than 5% body surface area or palmo plantar psoriasis, an alternative anti-TNF should be considered and skin lesions treated topically by PUVA or generally, if psoriasis is recalcitrant, a switch to another antiTNF agent or a discontinuation should be proposed. In some severe cases, anti-IL-12/IL-23 monoclonal antibody (ustekinumab) therapy demonstrates efficacy upon skin lesions induced by anti-TNF [11,22], in accordance with the mentioned cytokine imbalance. Uveitis

Uveitis represents an inflammation of the uveal tract of the eye and may be classified by location as anterior, posterior or intermediate; panuveitis refers to an involvement of two or more segments. The clinical course allows a classification as acute (3 months) or recurrent, with flare after complete recovery of the first episode. Anti-TNF agents demonstrate a reduction of incidence of uveitis (new onset or flares) [23,24], especially in SpA, a condition associated with acute recurrent uveitis [25–29]. On the other hand, the occurrence of uveitis (flare or new onset) has been reported under anti-TNF treatment [30] as single cases or retrospective cohort studies [31]. Lim et al. [30] in the US database of spontaneous reports, found 43 cases under etanercept, 14 with infliximab and 2 with adalimumab. After adjustment in the 161

162 207; M: 35%; 45

Retrospective literature

Retrospective Registry BSBR

Retrospective literature

Retrospective monocentric

Collamer and Battafarano (2010)

Harrison et al. (2009)

Denadai et al. (2013)

Fouache et al. (2009)

296

9826

IFX: 60% Eta: 20% Ada: 20%

IFX: 69%

Crohn 78% IBD: 22%

SpA

IFX: 25% Eta: 25% Ada: 50% Of newonset psoriasis

Rheumatoid arthritis16/ 18 EULAR responders

6 (1–24)

Psoriasiform lesions: 56%

PPP: 6/19

29%

9/36

Complete remission of skin lesions 64% without discontinuation of anti-TNF

8/25 stopped antiTNF due to psoriasis, 6 of them improved after stop

[32]

[15]

[10]

[14]

Continuation 66% (57% resolution) Switch 25/207 (53% no skin relapse RA: 10% SpA: 30% IBD: 5%

Plaque: 50% PP: 56% Guttate: 12%

IFX: 59% Eta: 19% Ada: 22%

SpA: 26% RA: 43% IBD: 20% JIA, Pso: 11%

[8]

Continuation 7/12 (1 improvement) Discontinuation 5/12 (4 skin resolution)

50%

Plaque: 50% PP: 40% Other: 10%

4.1 (1–15)

IFX: 20% Eta: 50% Ada: 30%

SpA: 50% RA: 30% PsA: 20%

NA

Ref.

Outcome

Previous history of psoriasis

Type of psoriasis reaction

Duration of exposition to anti-TNF (months; range)

Type of anti-TNF

Indication for antiTNF

Exposed patients (n)

Ada: Adalimumab; Eta: Etanercept; IBD: Inflammatory bowel diseases; IFX: Infliximab; M: Male; PP: Pustulotic psoriasis; PsA: Psoriatic arthritis; SpA: Spondyloarthritis; RA: Rheumatoid arthritis.

5

222; M: 52%; 26

36; M: 4/25; 60

12; M: 6; 45

Retrospective monocentric

Wendling et al. (2008)

Cases (n); sex; age (years)

Type of study

Study (year)

Table 2. Psoriasis reactivation or occurrence under anti-TNF treatment

Review Wendling & Prati

Expert Rev. Clin. Immunol. 10(1), (2014)

Paradoxical effects of anti-TNF-a agents in inflammatory diseases

Box 3. Main skin lesions associated with treatment with TNF antagonists. Skin lesions related to the administration of treatment • Erythema, urticaria, eczema, rash Skin infections • Bacterial infections: cellulitis • Viral infections: herpes zoster Skin neoplasia • Non-melanoma skin tumors: basocellular and spinocellular carcinoma • Melanocytic tumors: benign melanocytic naevus • Malignant melanoma • Skin lymphomas: T cells, fungoid mycosis, Sezary syndrome Immune-mediated diseases • ‘De novo’ psoriasis and exacerbation of prior psoriasis • Cutaneous lupus • Alopecia areata • Skin vasculitis • Vitiligo • Relapsing polychondritis • Polymyositis/dermatomyositis • Localized scleroderma (morphea) • Granuloma anularis • Lichen or lichenoid reaction • Pemphigus

number of patients treated with each agent, etanercept is found to be associated with a greater number of uveitis than antibodies. Rudwaleit et al. [29] also found 2 cases out of 1250 AS patients under adalimumab treatment, with duration of 20 weeks. In a systematic review of the studies with etanercept in AS (RCT + OLE), Sieper et al. [28] reported 21 cases among the 1074 exposed patients with a follow-up of 12–264 weeks, leading to an incidence rate of 1.8/100 patient-years. Fouache et al. [32] reported an incidence rate of 1/100 patientyears in patients with SpA treated with etanercept and the same incidence with conventional treatment and no uveitis found under anti-TNF monoclonal antibodies. A French national survey gathered 31 cases of the new onset of uveitis under anti-TNF therapy [31]. This study analyzed 121 other cases in the literature and summarized the main data in TABLE 3. National survey recorded 19 men and 12 women, with a mean age of 43 years (extremes 5–70), with a mean disease duration of 15 years (extremes 1.5–35) and a mean duration of global anti-TNF exposure of 27 months (4–96) and 19 months

Review

for the last anti-TNF therapy without duration difference between the three anti-TNF therapy. In the US registry [30], the mean duration of exposure was approximately 13 months. The rheumatologic condition responded well to anti-TNF at the time of uveitis (acute anterior) in 27/31 cases. These data argue, after adjustment for the estimated number of patients treated with each anti-TNF, for an increased risk of uveitis occurrence for etanercept and during treatment of SpA. Nevertheless, uveitis may occur in several conditions, not exclusively rheumatologic disorders and with each of the anti-TNF, witnessing a potential class effect; some patients were previously treated with another TNF blocker without uveitis episode. Many data argued the implication of the IL-23/IL-17 pathway in uveitis [25,33] and as discussed earlier in this review, a cytokine imbalance under anti-TNF therapy may increase Th17 activity. Outcome and management are variables. In most cases of anterior uveitis, local treatment was sufficient while systemic steroid treatment was required in 4/31 cases in the French survey. Complications seem rarely reported; in the French survey, 2 cases of anterior uveitis developed retinal involvement. Discontinuation of treatment was necessary in 7 out of 31 patients, although some positive re-challenge were reported in the literature. The switch to another anti-TNF (mainly from etanercept to a monoclonal antibody) is associated with an absence of uveitis recurrence in most cases. The possibility of the uveitis resolution without relapse while continuing the same anti-TNF was also reported. Besides uveitis, some other cases of the different types of ocular inflammation have been reported. For example, the case for scleritis as described in the literature on patients treated with etanercept for RA [34], with a good response of the rheumatologic condition to etanercept; ocular inflammation went into remission after discontinuation of etanercept and a positive rechallenge test has been reported in one case. Inflammatory bowel diseases

IBD, especially CD, represent another licensed used for antiTNF agents, at least monoclonal antibodies, since etanercept did not demonstrate efficacy in CD. In this situation, development of de novo, or exacerbation of a previously diagnosed IBD corresponds to a paradoxical effect. Braun et al. [35] reviewed the data from nine trials (seven placebo-controlled) of patients with AS treated with anti-TNF agents. The data were available for 419 AS patients treated with etanercept (625 patient-years), 366 treated with infliximab

Table 3. Cases of new onset of uveitis in the French national survey and from the literature review. New-onset uveitis under anti-TNF

AS

RA

PsA

JIA

Etanercept

Infliximab

Adalimumab

French national survey (n = 31; %)

61

19

12

6

74

16

10

Literature review (n = 121; %)

72

10

6

11

84.3

12.4

3.3

AS: Ankylosing spondylitis; JIA: Juvenile idiopathic arthritis; RA: Rheumatoid arthritis; PsA: Psoriatic arthritis. Data taken from [31].

www.expert-reviews.com

163

Review

Wendling & Prati

Table 4. New onset of inflammatory bowel diseases occurring under anti-TNF treatment. Study, subjects (n)

Disease treated with anti-TNF

Type of anti-TNF

Duration of anti-TNF exposition (months)

Type of IBD

Outcome

French national survey, 16

AS: 12 RA: 1 PsA: 1 JIA: 2

ETA: 14 IFX: 2

29 ± 20

CD: 8 CD like: 6 UC: 1 Indeterminate: 1

Anti-TNF discontinuation and switch to another anti-TNF Mab Favorable outcome, no flare

Literature review, 20

AS: 7 RA: 0 PsA: 2 JIA: 11

ETA: 18 IFX: 2

6–78

CD

Maintenance of same anti-TNF + mesalazine [2] Discontinuation of ETA and switch [17] NA [1]

AS: Ankylosing spondylitis; CD: Crohn’s disease; ETA: Etanercept; IBD: Inflammatory bowel diseases IFX: Infliximab; Mab: Monoclonal antibody; NA: Not available; RA: Rheumatoid arthritis; PsA: Psoriatic arthritis; JIA: Juvenile idiopathic arthritis; UC: Ulcerative colitis. Data taken from [36].

(618 patient-years), 295 with adalimumab (132 patient-years) and 434 placebo patients (150 patient-years). The results show an IBD flare OR of 18.0 (95% CI: 2–154) while taking etanercept and 4.2 (95% CI: 0.4–44) while taking adalimumab, in comparison with infliximab, but the incidence of the new onset of IBD was statistically indifferent from placebo for all anti-TNF agents. The rate of flare or new onset of IBD was evaluated at 2.2/100 patient-years with etanercept, 0.2/100 patient-years with infliximab, 2.3/100 patient-years with adalimumab and 1.3 with placebo. In juvenile idiopathic arthritis (JIA), new-onset IBD occurrence was estimated at 1.9/100 patient-years [36]. A French retrospective multicenter study retrieved 16 cases and found 20 other cases [36], the 3 cases previously reported by Fouache et al. [32] were included in the national series (TABLE 4). In the series of Fouache et al., the extrapolated incidence rate of CD in AS patients treated with anti-TNF was 1/100 patient-years for etanercept, 0/284 patient-years for infliximab and 0/62 patient-years for adalimumab. It appears clearly from these data that these cases of new onset of IBD occur mainly under etanercept and in the context of SpA (AS and psoriatic arthritis). This type of event is also possible with infliximab and adalimumab [37]. CDtype IBD is the most frequent pathological feature in this situation. IBD may be considered as severe in about 40% of the patients [36]. Anti-Saccharomyces cerevisiae antibodies were found in two out of the nine patients from the French survey. Mechanism

Few data are available for underlying predisposing factors in these patients that might have precipitated the IBD occurrence upon anti-TNF influence such as NOD2/CARD15 mutation found only in one of the two patients in the French cohort. Some structural and functional differences in anti-TNF agents may account for the greater incidence of this type of event with etanercept as compared with monoclonal antibodies and absence of efficacy of this soluble receptor in pre-existing IBD [38]. These include differences in soluble and membrane TNF binding and in apoptosis and cell lysis. 164

Another explanation may be the implication of the IL-23/ IL-17 axis in the development of CD [39]. Outcome & management

In some rare cases, the anti-TNF may be maintained without change, pending additional treatment of IBD with mesalazine; but in the majority of the cases, the initial TNF blocker is discontinued with favorable outcome for IBD and secondary switch to a monoclonal antibody. This represents the best strategy of management of this kind of event. Sarcoidosis

Anti-TNF agents may be effective in sarcoidosis, as emphasized in a recent review [40]; infliximab has some positive effects upon pulmonary disease in sarcoidosis. Along the years, several studies reported occurrence under anti-TNF therapy, for various indications, of histologically proven sarcoidosis [41–43] or sarcoid-like lesions [44]. Till date, more than 50 observations have been published. Etanercept is the supposed culprit in about two-third of the cases [43]; the remaining cases are equally dispatched between infliximab and adalimumab. Sarcoidosis symptoms appear after an exposition of several months (1–69); the underlying disease is RA in about 60% of the cases, AS in 20% and the remaining psoriatic arthritis, JIA and IBD. Therefore, this event is described in the whole spectrum of the diseases treated with TNF blockers. Clinical presentation is systemic in three-fourth of the cases, whereas the underlying disease is well controlled by the anti-TNF treatment. The symptoms are respiratory features such as dyspnea, chest pain, cough, with hilar and mediastinal adenopathy and sometimes pulmonary infiltrates [42]. Other sarcoid involvements have been reported with skin lesions (erythema nodosum, pigmented scars, nodular lesions) [45], neuro-ocular manifestations (and possibility of uveitis) and parotid gland enlargement. Differential diagnosis is an important issue and a challenge of pulmonary lesions in patients treated with anti-TNF agent. Infection (pyogenic or opportunistic), as well as tumoral development (lymphoma) should be ruled out and histological confirmation of sarcoidosis (well-formed non-necrotizing granuloma) is also required in this situation. Expert Rev. Clin. Immunol. 10(1), (2014)

Paradoxical effects of anti-TNF-a agents in inflammatory diseases

Review

Table 5. Vasculitis occurring under anti-TNF therapy. Series, subjects (n)

Underlying disease (%)

Type of anti-TNF (%)

Clinical picture

Type of vasculitis

Outcome

US FDA, 35

RA: 74 CD: 14 JIA: 6 AS: 3 PsA: 3

Eta: 57 IFX: 43

Skin involvement: 100%

Leukocytoclastic: 100%

Improvement after discontinuation: 62.5% Rechallenge: – positive: 6 cases – negative: 4 cases

French national survey, 39

RA: 87.5 JIA: 5 AS: 5 PsA: 2.5

ETA: 56 IFX: 39 ADA: 5

Skin involvement: 82%

Non-necrotizing: 44% Necrotizing: 26% Other/unknown: 30%

Improvement after discontinuation: 82% Improvement without discontinuation: 15% Rechallenge: – positive: 1 case – negative: 1 case Switch without relapse: 23% Death: 1 case

Literature, 118

RA: 84 CD: 7 Other: 9

ETA: 50 IFX: 43 ADA: 4 NA: 3

Skin involvement: 86%

Leukocytoclastic: most frequent

Improvement after discontinuation: 81% Death: 2% Unknown: 17%

Mayo Clinic, 8

RA: 50 CD: 12.5 UC: 37.5

ETA: 25 IFX: 62.5 ADA: 12.5

Skin involvement: 63%

Cutaneous small vessel: 50%

Improvement after discontinuation: 88% Death: 1/8

ADA: Adalimumab; AS: Ankylosing spondylitis; CD: Crohn’s disease; ETA: Etanercept; IFX: Infliximab; JIA: Juvenile idiopathic arthritis; PsA: Psoriatic arthritis; RA: Rheumatoid arthritis; UC: Ulcerative colitis. Data taken from the literature.

Pathogenesis

The role of anti-TNF on granuloma [46] and the different behavior of antibodies and soluble receptor in granulomatous diseases, with a greater tendency of etanercept to induce granulomatous reactions, may account for the occurrence of lesions. Monoclonal antibodies (and not etanercept) may induce apoptosis in activated monocytes and T cells. But sarcoidosis-like lesion may also appear under monoclonal antibodies, raising the question of other mechanisms: anti-TNF therapy increases the expression of IL-17 and IFN-g; and induces an imbalance in Th17 and Treg in favor of Th17. These two factors (IFN-g, Th17) may play a role in sarcoidosis development [43]. Outcome & management

Outcome is favorable in major cases, with resolution of clinical and radiological signs after withdrawal of the TNF blocker, alone or in association with steroids (in about 40% of the cases) [43]. In a fourth of the patients, another anti-TNF agent was re-introduced (mostly adalimumab, after etanercept) due to the activity of the underlying disease without recurrence of sarcoidosis. A re-challenge was positive in two cases [43]. Other aseptic granulomatosis were reported during anti-TNF treatment, including the case of pulmonary lesions, a collaborative series of 11 cases revealing at the histopathological level typical rheumatoid nodules in four cases and non-caseating granulomatous lesions in four cases, with a favorable outcome in all patients, with or without discontinuation of the www.expert-reviews.com

anti-TNF agent [47]. Skin granulomatous lesions may also occur or increase when the patient is treated with anti-TNF agents. These may be rheumatoid nodules [19], several dermatologic entities such as interstitial granulomatous dermatitis [48] or lichen planus [49] with about 20 cases reported in the literature. Vasculitis

Vasculitis is another type of paradoxical event described with anti-TNF treatment [50]. Vasculitis has been described as case reports, but also in series (US FDA [51], French [52], analyzed in several reviews of Ramos-Casals et al. [53]. With the series of the Mayo clinic [54], there are more than 140 cases reported in the literature. They represent almost half of the autoimmune diseases developing after TNF-targeted therapy found by Ramos-Casals in 2007 (TABLE 5) [53]. Vasculitis occurs predominantly in patients with RA treated mainly with etanercept or infliximab, in a variable duration of exposition ranging from 0.5 to 80 months. A female predominance of 75% and a mean age of 50 years are in accordance with a RA population. The underlying disease is usually controlled by the anti-TNF therapy. The clinical presentation is related to cutaneous involvement present in more than 80% of the cases. Purpura is the most frequent cutaneous lesion. The skin involvement allows a biopsy and histological diagnosis. Leukocytoclastic vasculitis is the most frequent type. Extra cutaneous involvement is present in one-quarter of the cases of vasculitis [53] involving central or peripheral nervous system or kidney. 165

Review

Wendling & Prati

Pathogenesis

The pathogenesis of vasculitis associated with anti-TNF treatment is not completely explained. An immune complexmediated hypersensitivity vasculitis may be related to the development of antibodies against anti-TNF agents, but in this hypothesis, the cases with etanercept should be less frequent since the treatment with soluble receptor induces less antibody formation than monoclonal antibodies. Overexpression of type I interferon secondary to imbalance between Th1 and Th2 cytokine production under TNF inhibition may favor induction of autoimmune disorders such as vasculitis. Outcome & management

Outcome is favorable in major cases (2/3), with resolution of symptoms after discontinuation of the anti-TNF agent in a mean time of 6.9 months [54], an adjunction of systemic steroids and sometimes immunosuppressive therapy in 50% of them. Re-challenge test was positive in seven cases of the US FDA and French surveys (six Eta, one IFX) and negative in five cases (three ETA, two IFX); in the French cohort, switching to another anti-TNF in nine cases was not associated with a relapse of vasculitis. In 6 out of 39 cases from the French survey, anti-TNF was maintained with improvement of vasculitis [52]. Death was reported in few cases (2% in the review of Ramos-Casals et al.) [53]. Miscellaneous

Favorable impact of anti-TNF therapy in cases of inflammatory muscle diseases (polymyositis, dermatomyositis) [55] and cases of the development of primary myositis during TNF antagonist therapy was previously reported and published with several myositis (such as polymyositis, orbital myositis, anti Jo-1 myositis) [55,56] mainly in the background of RA. Paradoxical joint inflammation in patients with IBD treated with anti-TNF agents has been described [57]. Twenty-one cases of disabling polyarthralgia occurred in a series of 1300 patients treated with anti-TNF monoclonal antibodies for IBD. Only two patients had a concomitant axial SpA. Arthralgia (pain and morning stiffness) without synovitis, involving hands and wrists, appeared after a mean duration of 12 months of antiTNF therapy. In about half of these patients, positive antinuclear antibodies in high titer were found. No correlation was found between this articular manifestation development and presence of anti infliximab antibodies. The evolution of these rheumatologic disorders was improved by anti-TNF discontinuation in 10 patients, adjunction of corticosteroids in 4 patients and immune modulators (azathioprine, methotrexate) in 4 patients, whereas increase of dosage or switching of the antiTNF agent were not effective. Such paradoxical flares of arthritis may occur in RA patients; although, the flare of arthritis being interpreted as lack of response to the anti-TNF agent was underreported. Another issue in paradoxical effects may be pyoderma gangrenosum (PG), a rare inflammatory disease of unknown etiology and poorly understood pathogenesis. In a recent systematic 166

review of IBD-associated PG, Agarwal and Andrews [58] identified 60 cases, 36 of these cases received an anti-TNF agent and 33 (92%) responded well to the therapy. On the other hand, there are case reports on the development of PG during antiTNF therapy [58] for IBD [59] or RA [60]. This may represent another situation of paradoxical effect of anti-TNF agents. But we have to keep in mind another evidence that PG may occur without any therapy in the course of inflammatory diseases. Expert commentary

All the conditions analyzed in this review represent situations in which anti-TNF treatment are usually effective and potentially, the paradoxical effects of treatment with a new kind of adverse event under biological treatment [61]. Data from the literature document this event and reveal a temporal relationship between anti-TNF treatment and occurrence of these effects with a potential resolution in major cases after the discontinuation of the anti-TNF agent. In many situations, there are positive re-challenge tests. Moreover, these socalled paradoxical effects have been described with different anti-TNF agents. The underlying disease is usually well controlled by anti-TNF treatment, giving no argument for an eventual flare of the systemic condition to explain the occurrence of the new clinical feature. All these situations may be underreported. In first analysis, a potential paradoxical effect from a feature that may be related to the disease treated with anti-TNF agents need to be distinguished. For example, the occurrence of uveitis, psoriasis or CD in a background of SpA, or vasculitis in cases of RA is not very surprising. Are these events side effects or events related to their underlying diseases? Does the anti-TNF agent play a role in induction, revealing or precipitating this event? In many cases, there is no increase of occurrence of the event under anti-TNF treatment compared with the controls (e.g., IBD). Moreover, these effects are not restricted to anti-TNF agents; an example is psoriasis onset, which has been described under the treatment with abatacept, rituximab or anti IL-6 (tocilizumab) treatments [62,63] and uveitis, which was described in patients treated with tocilizumab or rituximab [62,64]. This kind of event may be added to the autoimmune or immune-mediated effects of anti-TNF agents [65], including SLE, demylinating diseases or auto-antibodies occurrence. In fact, these so-called paradoxical effects may reflect an imbalance of cytokines under TNF blockade that may represent the major common immunopathological factor of these effects. Five-year view

These events need to be better evaluated and studied. Of interest would be a better analysis of immunological modifications at time of ‘paradoxical’ effect and individualization of prognosis or predictive factors for the occurrence of these effects. Nationwide registries or collaborative studies may catch these cases and may allow a more accurate calculation of incidence rate, comparing patients with or without these paradoxical reactions under anti-TNF agents. Expert Rev. Clin. Immunol. 10(1), (2014)

Paradoxical effects of anti-TNF-a agents in inflammatory diseases

Beyond these considerations, a better analysis of these patients would help to increase our knowledge about the diseases treated and the immunological mechanism of action of these anti-TNF agents. Financial & competing interests disclosure

D Wendling has received speaking fees and/or consultancy fees from Abbott, MSD, Pfizer, Roche Chugai, BMS, Amgen, Nordic Pharma and Swedish Orphan Biovitrum, and grants from Abbott, MSD, Pfizer, Roche

Review

Chugai and BMS. D Wendling is also National co-ordinator of the study, CAIN 457F2305 (secukinumab in AS, Novartis) and National coordinator and investigator of the study ALIGN (sarilumab in AS, SanofiAventis). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.

Key issues • Paradoxical effects of anti-TNF agents are defined as manifestations occurring under anti-TNF agents. • Mainly occur while the underlying disease motivating TNF blockers is controlled. • Temporal relationship of resolution after discontinuation of the TNF blockers and positive rechallenge test are arguments for a direct responsibility of anti-TNF agents. • The possibility of a feature of the underlying disease and the occurrence of this kind of effect under other treatments may represent arguments against a paradoxical effect.

References

9

Papers of special note have been highlighted as: • of interest 1

Pham T, Bachelez H, Berthelot JM et al. TNF alpha antagonist therapy and safety monitoring. Joint Bone Spine. 78(Suppl. 1), 15–185 (2011).

2

Bradley JR. TNF-mediated inflammatory disease. J. Pathol. 214(2), 149–160 (2008).

3

Tracey D, Klareskog L, Sasso EH et al. Tumor necrosis factor antagonist mechanisms of action: a comprehensive review. Pharmacol. Ther. 117(2), 244–279 (2008).

4

Biton J, Boissier MC, Bessis N. TNFa: activator or inhibitor of regulatory T cells? Joint Bone Spine 79(2), 119–123 (2012).

5

Benucci M, Saviola G, Manfredi M et al. Tumor necrosis factors blocking agents: analogies and differences. Acta Biomed. 83(1), 72–80 (2012).

6

Thalayasingam N, Isaacs JD. Anti-TNF therapy. Best Pract. Res. Clin. Rheumatol. 25(4), 549–567 (2011).

7

8

Park W, Hrycaj P, Jeka S et al. A randomised, double-blind, multicentre, parallel-group, prospective study comparing the pharmacokinetics, safety, and efficacy of CT-P13 and innovator infliximab inpatients with ankylosing spondylitis: the PLANETAS study. Ann. Rheum. Dis. 72(10), 1605–1612 (2013) Wendling D, Balblanc JC, Brianc¸on D et al. Onset or exacerbation of cutaneous psoriasis during TNF alpha antagonist therapy. Joint Bone Spine 75(3), 315–318 (2008).

www.expert-reviews.com

Ko JM, Gottlieb AB, Kerbleski JF. Induction and exacerbation of psoriasis with TNF-blockade therapy: a review and analysis of 127 cases. J. Dermatolog. Treat. 20(2), 100–108 (2009).

10

Harrison MJ, Dixon WG, Watson KD et al. Rates of new-onset psoriasis in patients with rheumatoid arthritis receiving anti-tumour necrosis factor alpha therapy: results from the British Society for Rheumatology Biologics Register. Ann. Rheum. Dis. 68(2), 209–215 (2009).

•

A retrospective analysis from a nationwide registry of new-onset psoriasis occurrence under anti-TNF therapy, compared with non-exposed patients, with calculation of incidence ratio.

11

Tillack C, Ehmann LM, Friedrich M et al. Anti-TNF antibody-induced psoriasiform skin lesions in patients with inflammatory bowel disease are characterised by interferon-g-expressing Th1 cells and IL17A/IL-22-expressing Th17 cells and respond to anti-IL-12/IL-23 antibody treatment. Gut doi:10.1136/gutjnl-2012– 302853 (2013) (Epub ahead of print).

12

Klein RQ, Spivack J, Choate KA. Psoriatic skin lesions induced by certolizumab pegol. Arch. Dermatol. 146, 1055–1056 (2010).

13

Collamer AN, Guerrero KT, Henning JS, Battafarano DF. Psoriatic skin lesions induced by tumor necrosis factor antagonist therapy: a literature review and potential mechanisms of action. Arthritis Rheum. 59(7), 996–1001 (2008).

14

Collamer AN, Battafarano DF. Psoriatic skin lesions induced by tumor necrosis

factor antagonist therapy: clinical features and possible immunopathogenesis. Semin. Arthritis Rheum. 40(3), 233–240 (2010). •

A large analysis of the literature about psoriasis onset under anti-TNF treatments.

15

Denadai R, Teixeira FV, Steinwurz F et al. Induction or exacerbation of psoriatic lesions during anti-TNF-a therapy for inflammatory bowel disease: a systematic literature review based on 222 cases. J. Crohns Colitis. 7(7), 517–524 (2013).

16

Seneschal J, Milpied B, Vergier B et al. Cytokine imbalance with increased production of interferon-alpha in psoriasiform eruptions associated with antitumour necrosis factor-alpha treatments. Br. J. Dermatol. 161(5), 1081–1088 (2009).

17

Ma HL, Napierata L, Stedman N et al. Tumor necrosis factor alpha blockade exacerbates murine psoriasis-like disease by enhancing Th17 function and decreasing expansion of Treg cells. Arthritis Rheum. 62(2), 430–440 (2010).

18

Joyau C, Veyrac G, Dixneuf V, Jolliet P. Anti-tumour necrosis factor alpha therapy and increased risk of de novo psoriasis: is it really a paradoxical side effect? Clin. Exp. Rheumatol. 30(5), 700–706 (2012).

19

Viguier M, Richette P, Bachelez H et al. Paradoxical adverse effects of anti-TNF-alpha treatment: onset or exacerbation of cutaneous disorders. Expert Rev. Clin. Immunol. 5(4), 421–431 (2009).

20

Flendrie M, Vissers WH, Creemers MC et al. Dermatological conditions during TNF-alpha-blocking therapy in patients with rheumatoid arthritis: a prospective

167

Review

Wendling & Prati

study. Arthritis Res. Ther. 7(3), R666–R676 (2005). 21

Herna´ndez MV, Sanmartı´ R, Can˜ete JD et al. Cutaneous adverse events during treatment of chronic inflammatory rheumatic conditions with tumour necrosis factor antagonists. Study based on the Spanish Registry of Adverse Events of Biological Therapies in Rheumatic Diseases. Arthritis Care Res (Hoboken) doi:10.1002/ acr.22096 (2013) (Epub ahead of print).

•

Updated registry report of cutaneous adverse events under anti-TNF therapy.

22

Puig L, Morales-Mu´nera CE, Lo´pez-Ferrer A, Geli C. Ustekinumab treatment of TNF antagonist-induced paradoxical psoriasis flare in a patient with psoriatic arthritis: case report and review. Dermatology 225(1), 14–17 (2012).

23

24

25

26

27

Cordero-Coma M, Yilmaz T, Onal S. Systematic review of anti-tumor necrosis factor-alpha therapy for treatment of immune-mediated uveitis. Ocul. Immunol. Inflamm. 21(1), 19–27 (2013). Rifkin LM, Birnbaum AD, Goldstein DA. TNF Inhibition for ophthalmic indications: current status and outlook. Bio. Drugs 27(4), 347–357 (2013).

•

Literature analysis and nationwide survey of uveitis occurring under anti-TNF treatment.

32

Fouache D, Goe¨b V, Massy-Guillemant N et al. Paradoxical adverse events of anti-tumour necrosis factor therapy for spondyloarthropathies: a retrospective study. Rheumatology (Oxford) 48(7), 761–764 (2009).

•

Guignard S, Gossec L, Salliot C et al. Efficacy of tumour necrosis factor blockers in reducing uveitis flares inpatients with spondylarthropathy: a retrospective study. Ann. Rheum. Dis. 65(12), 1631–1634 (2006). Braun J, Baraliakos X, Listing J, Sieper J. Decreased incidence of anterior uveitis in patients with ankylosing spondylitis treated with the anti-tumor necrosis factor agents infliximab and etanercept. Arthritis Rheum. 52(8), 2447–2451 (2005). Sieper J, Koenig A, Baumgartner S et al. Analysis of uveitis rates across all etanercept ankylosing spondylitis clinical trials. Ann. Rheum. Dis. 69(1), 226–229 (2010).

29

Rudwaleit M, Rødevand E, Holck P et al. Adalimumab effectively reduces the rate of anterior uveitis flares in patients with active ankylosing spondylitis: results of a prospective open-label study. Ann. Rheum. Dis. 68(5), 696–701 (2009).

30

Lim LL, Fraunfelder FW, Rosenbaum JT. Do tumor necrosis factor inhibitors cause uveitis? A registry-based study. Arthritis Rheum. 56(10), 3248–3252 (2007). Wendling D, Paccou J, Berthelot JM et al. New onset of uveitis during anti-tumor

168

A monocentric study of paradoxical events under anti-TNF treatments for spondyloarthritis (SpA), compared with patients not exposed to TNF blockers.

33

Horai R, Caspi RR. Cytokines in autoimmune uveitis. J. Interferon. Cytokine Res. 31(10), 733–744 (2011).

34

Gaujoux-Viala C, Giampietro C, Gaujoux T et al. Scleritis: a paradoxical effect of etanercept? Etanercept-associated inflammatory eye disease. J. Rheumatol. 39(2), 233–239 (2012)

35

Braun J, Baraliakos X, Listing J et al. Differences in the incidence of flares or new onset of inflammatory bowel diseases in patients with ankylosing spondylitis exposed to therapy with anti-tumor necrosis factor alpha agents. Arthritis Rheum. 57(4), 639–647 (2007). Toussirot E´, Houvenagel E´, Goe¨b V et al.

Wendling D. Uveitis in seronegative arthritis. Curr. Rheumatol. Rep.14(5), 402–408 (2012).

28

31

necrosis factor treatment for rheumatic diseases. Semin. Arthritis Rheum. 41(3), 503–510 (2011).

36

Development of inflammatory bowel disease during anti-TNF-a therapy for inflammatory rheumatic disease: a nationwide series. Joint Bone Spine 79(5), 457–463 (2012). •

The largest retrospective study of inflammatory bowel disease (IBD) onset during anti-TNF therapy, with review of the literature.

37

Ferna´ndez Salazar L, Barrio Andre´s J, Gonza´lez Herna´ndez JM. Crohn’s disease diagnosis during adalimumab treatment. Scand. J. Gastroenterol. 48(2), 254–255 (2013).

38

Haraoui B, Krelenbaum M. Emergence of Crohn’s disease during treatment with the anti-tumor necrosis factor agent etanercept for ankylosing spondylitis: possible mechanisms of action. Semin. Arthritis Rheum. 39(3), 176–181 (2009).

39

Siakavellas SI, Bamias G. Role of the IL-23/ IL-17 axis in Crohn’s disease. Discov. Med. 14(77), 253–262 (2012).

40

Maneiro JR, Salgado E, Gomez-Reino JJ et al. Efficacy and safety of TNF antagonists in sarcoidosis: data from the Spanish

registry of biologics BIOBADASER and a systematic review. Semin. Arthritis Rheum. 42(1), 89–103 (2012). 41

Toussirot E, Pertuiset E, Kantelip B, Wendling D. Sarcoidosis occurring during anti-TNF-alpha treatment for inflammatory rheumatic diseases: report of two cases. Clin. Exp. Rheumatol. 26(3), 471–475 (2008).

42

Massara A, Cavazzini L, La Corte R, Trotta F. Sarcoidosis appearing during anti-tumor necrosis factor alpha therapy: a new “class effect” paradoxical phenomenon. Two case reports and literature review. Semin. Arthritis Rheum. 39(4), 313–319 (2010).

43

Vigne C, Tebib JG, Pacheco Y, Coury F. Sarcoidosis: an underestimated and potentially severe side effect of anti-TNF-alpha therapy. Joint Bone Spine 80(1), 104–107 (2013).

44

Daı¨en CI, Monnier A, Claudepierre P et al. Sarcoid-like granulomatosis in patients treated with tumor necrosis factor blockers: 10 cases. Rheumatology (Oxford) 48(8), 883–886 (2009).

45

Fok KC, Ng WW, Henderson CJ, Connor SJ. Cutaneous sarcoidosis in a patient with ulcerative colitis on infliximab. J. Crohns Colitis. 6(6), 708–712 (2012).

46

Furst DE, Wallis R, Broder M, Beenhouwer DO. Tumor necrosis factor antagonists: different kinetics and/or mechanisms of action may explain differences in the risk for developing granulomatous infection. Semin. Arthritis Rheum. 36(3), 159–67 (2006).

47

Toussirot E, Berthelot JM, Pertuiset E et al. Pulmonary nodulosis and aseptic granulomatous lung disease occurring in patients with rheumatoid arthritis receiving tumor necrosis factor-alpha-blocking agent: a case series. J. Rheumatol. 36(11), 2421–2427 (2009).

48

Deng A, Harvey V, Sina B et al. Interstitial granulomatous dermatitis associated with the use of tumor necrosis factor alpha inhibitors. Arch. Dermatol. 142(2), 198–202 (2006).

49

Wendling D, Biver-Dalle C, Vidon C et al. Lichen planus under anti TNF therapy for ankylosing spondylitis. Joint Bone Spine 80(2), 227–228 (2013).

50

Wendling D, Streit G, Lehue´de´ G et al. Cutaneous lymphocytic vasculitis during TNF-alpha antagonist therapy for polyarthritis. Joint Bone Spine 73(2), 215–216 (2006).

51

Mohan N, Edwards ET, Cupps TR et al. Leukocytoclastic vasculitis associated with

Expert Rev. Clin. Immunol. 10(1), (2014)

Paradoxical effects of anti-TNF-a agents in inflammatory diseases

review of the literature. Clin. Rheumatol. 29(5), 563–566 (2010).

tumor necrosis factor-alpha blocking agents. J. Rheumatol. 31(10), 1955–1958 (2004). 52

Saint Marcoux B, De Bandt M; CRI (Club Rhumatismes et Inflammation). Vasculitides induced by TNFalpha antagonists: a study in 39 patients in France. Joint Bone Spine 73(6), 710–713 (2006).

57

Fiorino G, Danese S, Pariente B, Allez M. Paradoxical immune-mediated inflammation in inflammatory bowel disease patients receiving anti-TNF-a agents. Autoimmun. Rev. 13(1), 15–19 (2013).

53

Ramos-Casals M, Brito-Zero´n P, Mun˜oz S et al. Autoimmune diseases induced by TNF-targeted therapies: analysis of 233 cases. Medicine (Baltimore) 86(4), 242–251 (2007).

58

Agarwal A, Andrews JM. Systematic review: IBD-associated pyodermagangrenosum in the biologic era, the response to therapy. Aliment Pharmacol. Ther. 38(6), 563–572 (2013)

54

Sokumbi O, Wetter DA, Makol A, Warrington KJ. Vasculitis associated with tumor necrosis factor-a inhibitors. Mayo Clin. Proc. 87(8), 739–745 (2012).

59

55

Wendling D. Biologics in the treatment of primary inflammatory myositis. Joint Bone Spine 74(4), 316–318 (2007).

56

Ishikawa Y, Yukawa N, Ohmura K et al. Etanercept-induced anti-Jo-1-antibodypositive polymyositis in a patient with rheumatoid arthritis: a case report and

www.expert-reviews.com

Review

biological agents. J. Crohns Colitis. 7(4), 257–262 (2013). 62

Wendling D, Dougados M, Berenbaum F et al. Rituximab treatment for spondyloarthritis. A nationwide series: data from the AIR registry of the French Society of Rheumatology. J. Rheumatol. 39(12), 2327–2331 (2012).

63

Wendling D, Letho-Gyselinck H, Guillot X, Prati C. Psoriasis onset with tocilizumab treatment for rheumatoid arthritis. J. Rheumatol. 39(3), 657 (2012).

Jaimes-Lo´pez N, Molina V, Arroyave JE et al. Development of pyodermagangrenosum during therapy with infliximab. J. Dermatol. Case Rep. 3(2), 20–23 (2009)

64

Wendling D, Dernis E, Prati C et al. Onset of inflammatory eye disease under tocilizumab treatment for rheumatologic conditions: a paradoxical effect? J. Rheumatol. 38(10), 2284 (2011).

60

Vandevyvere K, Luyten FP, Verschueren P et al. Pyodermagangrenosum developing during therapy with TNF-alpha antagonists in a patient with rheumatoid arthritis. Clin. Rheumatol. 26(12), 2205–2206 (2007)

65

Perez-Alvarez R, Pe´rez-de-Lis M, Ramos-Casals M; BIOGEAS study group. Biologics-induced autoimmune diseases. Curr. Opin. Rheumatol. 25(1), 56–64 (2013).

61

Aubin F, Carbonnel F, Wendling D. The complexity of adverse side-effects to

169