Psoriatic skin lesions induced by tumor necrosis factor antagonist therapy: A literature review and potential mechanisms of action

Authors


  • The opinions or assertions contained herein are those of the authors and are not to be construed as official policy of the Department of the Army, Department of the Air Force, or the Department of Defense.

Abstract

Objective

Numerous reports of the induction or worsening of psoriasis in patients treated with tumor necrosis factor (TNF) antagonists indicate that this is not a rare phenomenon. The etiology of this paradoxical clinical response remains unclear. The aim of this study was to describe similar cases, conduct a comprehensive analysis of the literature, explore possible immunologic mechanisms of action of this perplexing reaction, and recommend management options.

Methods

A systematic literature review was performed using the PubMed and Medline databases (1996 to September 2007) searching the index terms “infliximab,” “etanercept,” “adalimumab,” “tumor necrosis factor alpha inhibitor,” and “TNF inhibitor,” combined with the terms “psoriasis,” “pustular,” “skin,” “rash,” and “palmoplantar.” All relevant articles in English were reviewed. Pertinent secondary references were also analyzed.

Results

According to the literature, new-onset psoriasis may occur any time after initiation of TNF antagonist therapy, is often of an uncommon morphology, may respond to psoriasis treatments, and usually resolves with TNF discontinuation. The pathogenesis of this response appears to involve a disruption in cytokine balance following TNF inhibition, resulting in the up-regulation of plasmacytoid dendritic cells and the subsequent production of unopposed interferon-α, following a triggering event in predisposed individuals.

Conclusion

TNF antagonist–induced psoriasis is a newly recognized adverse effect of these medications that typically does not require therapy cessation. We recommend aggressive treatment of the skin disease and consideration of a change in the TNF antagonist if the lesions are unresponsive to conventional psoriasis treatment.

INTRODUCTION

Tumor necrosis factor (TNF) antagonists are successfully utilized in the treatment of a wide variety of chronic autoimmune diseases and inflammatory conditions, including psoriasis. The complete mechanism of action of these agents remains elusive; however, they have proven to be safe and effective in most patients, despite initial fears of untoward immune system dysregulation (1–4). One unexpected side effect of TNF antagonists increasingly reported in the literature is the new onset or worsening of psoriatic skin lesions (5–37).

Currently there are more than 110 reports of patients who developed psoriasis for the first time while receiving a TNF inhibitor, or who had paradoxical worsening of their previous skin disease during treatment (5–37). Having seen 4 patients in the course of 1 year in our clinics, we suspect the number of unreported or unrecognized cases is much higher. Several small reviews have attempted to analyze these cases; however, none have been comprehensive, nor have they provided a treatment strategy for approaching this challenging clinical problem. Of the mechanisms of action suggested as explanations for this perplexing response, none can fully explain the etiopathogenesis.

In this report, we attempt to fully review the current literature regarding TNF antagonist–induced psoriasis, to propose the most probable mechanism of action, and to provide a practical approach for treatment.

PATIENTS AND METHODS

A systematic literature review was performed utilizing the PubMed and Medline databases (1996 to September 2007) searching the index terms “infliximab,” “etanercept,” “adalimumab,” “tumor necrosis factor alpha inhibitor,” and “TNF inhibitor,” combined with the terms “psoriasis,” “pustular,” “skin,” “rash,” and “palmoplantar.” All relevant articles in English were reviewed. Pertinent secondary references, including meeting abstracts, were also analyzed. All patients in our institution during the previous year who experienced new onset or worsening of psoriasis while being treated with a TNF antagonist were included. An effort was made to include only those patients who appeared to meet accepted classification criteria for their respective disease, who had enough detailed information present to allow conclusions to be drawn, and who did not have a more probable explanation for their skin findings.

Clinical vignettes.

The following cases were seen in our institution from January 2006 to August 2007 and are representative of cases reported in the literature.

Case 1.

A 51-year-old woman with a 30-year history of ankylosing spondylitis began etanercept therapy in 2003 with satisfactory relief of her inflammatory back pain. In 2007, a pustular eruption was noted on her right hand. There was no history of psoriasis, recent infection, or trauma. The patient was referred for dermatology consult, and biopsy samples of the lesions demonstrated psoriasiform epidermal hyperplasia with hyperkeratosis and parakeratosis and intraepidermal pustulosis, consistent with pustular psoriasis. Etanercept was discontinued and the eruption was treated with topical steroids with partial resolution (Figure 1). The patient experienced a flare of her axial symptoms and chose to resume TNF antagonist therapy, despite the persistence of pustular palmar psoriasis. After 3 months, the eruption spontaneously resolved.

Figure 1.

Erythematous 2-mm papulopustular lesions on the palm of patient 1, following tumor necrosis factor antagonist discontinuation and treatment with topical corticosteroids. Scaling, hyperkeratosis, and desquamation are also present.

Case 2.

A 54-year-old man with a 23-year history of plaque psoriasis, which had been in remission without skin lesions for 4 years, began etanercept for axial involvement of psoriatic arthritis in 2007. He noted dramatic improvement in fatigue and systemic symptoms; however, following the fourth injection, the patient experienced a severe outbreak of erythematous pustular lesions on his palms and soles as well as guttate papules and scaly plaques on his thorax and lower extremities. The lesions were diagnosed by our dermatologists as prototypical palmoplantar and guttate psoriasis. Etanercept was stopped and the lesions resolved. Adalimumab was subsequently initiated without return of the pustular lesions.

Case 3.

A 53-year-old man with a long history of plaque psoriasis and a 3-year history of psoriatic arthritis, treated for 1 year with adalimumab as well as methotrexate and azathioprine, presented with severe palmoplantar pustular lesions on the hands and feet. He had no previous history of pustular lesions and denied recent illness or trauma. Palmoplantar psoriasis was diagnosed by his dermatologists, and narrow band ultraviolet phototherapy and topical steroids were started with subsequent improvement of the skin lesions, while continuing TNF antagonist therapy.

Case 4.

A 24-year-old woman with a history of Crohn's disease, treated with infliximab for 2 years, presented to the dermatology department with a 4-week history of guttate erythematous plaques over her upper and lower extremities and trunk. She did not have a history of psoriasis, previous infection, or recent trauma. A skin biopsy result was consistent with psoriasis. Methotrexate 7.5 mg weekly, topical steroids, and calcipotriene were started. The patient noted improvement in her psoriatic skin lesions within 6 weeks and clearance in 9 weeks. She was maintained on 5 mg of methotrexate weekly and infliximab without recurrence.

RESULTS

A total of 114 cases were identified by comprehensive literature review, using the methods detailed above (5–37). Of these, 104 patients were included in the final analysis. As detailed in Table 1, nearly 50% of patients were treated for rheumatoid arthritis, 22% were treated for a seronegative spondylarthritis, and 16% were treated for an inflammatory bowel disease (Crohn's disease or ulcerative colitis). Seventy percent of the patients were female. Overall, 52% developed pustular psoriasis, 49% plaque psoriasis (often appearing in unusual areas of the body such as the groin or pubic regions), and 15% experienced guttate lesions. Some patients developed lesions of more than 1 morphology. Infliximab was utilized in 53% of cases, etanercept in 29%, and adalimumab in 18%. More than 50% had documented histologic confirmation of psoriasis. Excluding the 13 patients being treated for psoriasis or psoriatic arthritis, 8 patients had a history of psoriasiform lesions (often remote). No causative infections, new drug introductions, recent trauma, or significant life stressors were identified.

Table 1. Collective data of 104 patients who developed psoriasis during TNF antagonist treatment*
 Total (n = 104)RA (n = 49)SN (n = 23)IBD (n = 17)Psoriasis (n = 9)Other (n = 6)
  • *

    Values are the number (percentage) unless otherwise indicated. TNF = tumor necrosis factor; RA = rheumatoid arthritis; SN = seronegative spondylarthritis (including ankylosing spondylitis, reactive arthritis, and 4 patients with psoriatic arthritis); IBD = inflammatory bowel disease (including Crohn's disease and ulcerative colitis); NA = not applicable.

  • Psoriasis with skin involvement only.

  • Includes 2 patients with Behçet's disease, 3 with juvenile idiopathic arthritis, and 1 with Shulman's fasciitis.

Men29 (30)5 (11)12 (52)6 (37.5)4 (57)3 (50)
Women69 (70)41 (89)11 (48)10 (62.5)3 (43)3 (50)
Sex unknown, no.630120
Age, mean ± SD years45.4 ± 13.952.8 ± 11.541.9 ± 10.632.1 ± 10.148.7 ± 8.134.0 ± 14.5
Age unknown, no.630120
Infliximab55 (53)18 (37)16 (70)17 (100)0 (0)4 (67)
Etanercept30 (29)15 (31)6 (26)0 (0)8 (89)1 (16)
Adalimumab19 (18)16 (32)1 (4)0 (0)1 (11)1 (16)
Pustular54 (52)27 (55)16 (70)7 (41)0 (0)4 (67)
Guttate16 (15)6 (12)2 (9)1 (6)7 (78)0 (0)
Plaque51 (49)24 (48)9 (39)12 (71)2 (22)4 (67)
New onset83 (87)43 (88)17 (74)17 (100)NA6 (100)
Exacerbation12 (13)6 (12)6 (26)0 (0)NA0 (0)
Resolved off anti-TNF14 (13)5 (10)2 (9)3 (18)3 (33)1 (16)
Partial or no resolution off anti-TNF4 (4)1 (2)1 (4)1 (6)0 (0)1 (16)
Resolved on anti-TNF31 (30)15 (31)10 (43)5 (29)1 (11)0 (0)
Partial resolution on anti-TNF32 (31)13 (27)8 (35)6 (35)5 (56)0 (0)
No resolution on anti-TNF2 (2)2 (84)0 (0)0 (0)0 (0)0 (0)
Resolved off anti-TNF, recurred with reintroduction4 (4)3 (6)1 (4)0 (0)0 (0)2 (33)
Resolved off anti-TNF, recurred with reintroduction of different anti-TNF8 (8)6 (12)1 (4)1 (6)0 (0)0 (0)
No recurrence with change of anti-TNF5 (5)4 (8)0 (0)0 (0)0 (0)1 (16)
Outcome unknown4 (4)2 (4)0 (0)1 (6)0 (0)1 (16)

Subgroup analysis was generally similar to analysis of the group as a whole. Nine patients with classic psoriasis vulgaris were reported to have worsening of their baseline disease and development of psoriasis of a new morphology (78% guttate) not previously experienced. Two patients with psoriatic arthritis and a history of pustular psoriasis also experienced new guttate lesions, in contrast to 12% of patients with rheumatoid arthritis and 15% for the overall group. Seventy percent of patients with ankylosing spondylitis or another seronegative spondylarthritis developed pustular outbreaks versus 52% for the group overall.

DISCUSSION

It is now generally accepted that the development or worsening of psoriasis during treatment with TNF antagonists can occur at any time from days to years after drug initiation (5–37). No predilection for age or sex has been statistically proven. This phenomenon has been reported in nearly all diseases treated with TNF antagonists, including psoriasis, rheumatoid arthritis, ankylosing spondylitis, and other seronegative spondylarthritides, including psoriatic arthritis, Behçet's disease, inflammatory bowel disease, and juvenile idiopathic arthritis. Because TNF antagonists are predominantly utilized in systemic inflammatory rheumatic diseases, the preponderance of patients affected were undergoing treatment for a rheumatic disorder (75% in this series). The majority of patients were reported to have an excellent therapeutic response to the drug, with the exception of the unexpected development of psoriatic skin lesions (5–37). Bacterial infection, especially staphylococcal infection, is an established trigger of psoriasis in persons predisposed to the condition (38, 39) and several authors have attempted to rule out bacterial or viral causes of this reaction (5, 7, 8, 21, 28, 34); however, no confirmed cases of infection-induced psoriasis or patients with clinical indicators of infection have been reported.

As the use of TNF antagonists has increased, it has been increasingly apparent that a wide variety of undesirable cutaneous reactions, other than psoriasis, can be associated with this therapy (40). These include eczematous eruptions, cutaneous lymphomas, herpes simplex, bacterial infections, lichenoid eruptions, erythema multiforme, lupus erythematosus, and acute generalized exanthematous pustulosis. Biopsy samples of TNF antagonist–induced psoriatic skin lesions are histologically identical to those of patients with idiopathic psoriasis (in contrast to the histology seen in patients with psoriasis induced by other medications) (41), implying that the mechanism of action of the 2 processes is similar or identical. Lesions of 3 distinct psoriatic morphologies may develop and patients often present with more than 1 type of lesion. In this analysis, patients with a history of plaque psoriasis tended to present with lesions of a new morphology in previously unaffected skin areas. Those with seronegative spondylarthritides developed pustular lesions more frequently than the group overall; however, the small sample size may impact the significance of this result.

Literature analysis revealed that the majority of patients (79%) were able to continue TNF antagonist therapy, either their original agent or a substitute from the same class (Table 1). Because some patients discontinued TNF therapy when psoriasis developed, it is possible that some of these patients may have been able to tolerate the drug. Of those who discontinued therapy, only 4% failed to have full resolution of the skin disease (although improvement may have occurred after case publication). Nearly 25% had full resolution while continuing TNF therapy. Despite the ability of all TNF antagonists to induce psoriasiform skin lesions, we can only speculate regarding the exact mechanism by which they do so, and because the majority of patients do not relapse following TNF antagonist switch, the pathophysiology appears to vary between the 3 agents.

There appeared to be a trend toward better control in patients who were treated aggressively by dermatologists with strong topical corticosteroids, keratolytics, vitamin D analogs, ultraviolet therapy, increases in baseline methotrexate doses, and the occasional use of cyclosporine when necessary. One-third of patients continued TNF antagonist therapy with only partial resolution of the psoriasis, with varying degrees of treatment administered. Eleven patients chose to stop the TNF antagonist, with subsequent resolution of the psoriasis, but later restarted a TNF agent secondary to flare of the primary disease and experienced recurrent skin lesions, which were sometimes then successfully treated while continuing biologic therapy (5–37).

Review of the literature revealed several closely associated immunologic pathways that may induce this paradoxical reaction in a predisposed subset of patients (9, 39–48). Plasmacytoid dendritic cells (PDCs) and their production of interferon-α (IFNα) appear to be the key factors in psoriasis induction in these patients (42–48). TNFα normally inhibits PDC maturation from hematopoietic progenitors as well as IFNα production (9), and the inhibition of TNFα may allow unlimited and unregulated production of IFNα by PDCs (39). TNFα inhibition is also thought to mimic infection or injury, further stimulating aberrant IFNα expression at the tissue level in predisposed individuals (9).

PDCs are found in early psoriatic lesions and even in the normal-appearing skin of psoriatic patients, prior to the appearance of the psoriatic plaque. PDCs have also been discovered in the skin of patients with autoimmune disease, whereas they are absent in healthy controls (42–44). Increased IFNα expression has been demonstrated in the lesional dermal vasculature and in the perivascular lymphocytic infiltrate of patients who develop psoriatic lesions while receiving TNF antagonist therapy (9), implicating this cytokine in the pathogenesis of skin dysregulation (9, 44). In the setting of TNFα inhibition, the cytokine profile favors an increase in IFNα production originating from PDCs (9).

TNFα inhibition has also been shown to decrease trafficking of Th1 lymphocytes to the sites of joint destruction in autoimmune inflammatory disorders, resulting in an increase in the pooling and sequestration of these cells in the peripheral circulation (46). This dormant population of T cells has the potential for activation and mobilization to cutaneous sites. In patients with rheumatoid arthritis, IFNα has been shown to induce the expression of the chemokine receptor CXCR3 on T cells, facilitating homing to the skin (48). PDC-derived IFNα also promotes the maturation of myeloid dendritic cells and the stimulation, survival, and expression of pathogenic Th1 cells via IFNα induction of interleukin-15 (45). Thus, the inhibition of TNFα, resulting in increased IFNα expression and homing of Th1 cells to the skin, is a probable mechanism for the development of paradoxical psoriasis in patients undergoing TNF antagonist therapy. Further research and investigation is necessary to gain a clearer understanding of the complicated and delicate relationship between TNFα inhibition, IFNα induction, and the factors that predispose a subset of patients to develop psoriatic lesions from 1 or more of the various TNF antagonists.

The development of psoriasiform lesions is a vexing problem for both the patient and clinician. Patients, particularly those with any remote history of psoriasis, should be counseled regarding the possibility of this side effect on medication initiation. The majority of patients will have resolution of their skin lesions following TNF discontinuation. Based on literature review and treatment of our patient cohort, we propose the following treatment algorithm (Figure 2).

Figure 2.

Algorithm for treatment of psoriasis induced by tumor necrosis factor (TNF) antagonist therapy.

Patients presenting with skin lesions resembling psoriasis should be evaluated for possible viral or bacterial infection with blood, urine, and stool cultures and as clinically indicated. Possible modifiable mechanical stressors, or trauma (including emotional) that can trigger psoriasis, should be identified. All patients should be referred to a dermatologist for evaluation and biopsy to confirm the diagnosis of psoriasis and to rule out acute exanthematous pustulosis and other psoriasis mimickers.

Patients with confirmed psoriasis should be evaluated based on the extent and tolerability of the lesions. Patients with severe, intolerable lesions and those who do not wish to proceed with therapy should discontinue the TNF antagonist and should receive aggressive treatment of the psoriasis under the guidance of a dermatologist. Patients who experience severe and increasing lesions with each subsequent infusion or injection are unlikely to be able to tolerate the drug and should discontinue the agent immediately.

Patients who feel their psoriasis lesions are tolerable and prefer to continue TNF antagonist therapy, despite the skin disease, should be offered an alternative TNF antagonist. Patients with psoriasis covering <5% of body surface area should be treated aggressively with topical treatments (corticosteroids, keratolytics, and vitamin D analogs); if there is no significant resolution in 2–3 weeks, ultraviolet phototherapy and the addition of methotrexate should be considered. Those with palmoplantar pustular disease and lesions covering >5% of body surface area should be given topical therapies with occlusive therapy to the palms and soles, as well as ultraviolet phototherapy. If skin lesions remain uncontrolled, utilization of methotrexate, acitretin, and cyclosporine, under the guidance of a dermatologist, should be considered. For those who continue to have recalcitrant skin disease, we recommend consideration of modalities of therapy other than TNF antagonists for the patient's primary autoimmune disorder.

The introduction of TNF antagonists has changed the prognosis of rheumatic disease and provided a level of disease control not previously attainable. However, with relief from the inflammatory milieu has come an unexpected disruption in cytokine balance, resulting in the predisposed individual developing an unpleasant secondary autoimmune condition. These challenging cases serve as a reminder to the clinician that our understanding of the immunoregulatory system is imperfect and much remains undiscovered. As the use of TNF antagonists continues to increase, we must remain vigilant in the monitoring and reporting of adverse and unexpected reactions, with hope of increasing our knowledge of their pathophysiology.

AUTHOR CONTRIBUTIONS

Dr. Collamer had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Collamer, Battafarano.

Acquisition of data. Collamer, Henning, Battafarano.

Analysis and interpretation of data. Collamer, Guerrero, Henning, Battafarano.

Manuscript preparation. Collamer, Guerrero, Henning, Battafarano.

Statistical analysis. Collamer, Battafarano.

Ancillary