Psoriasis induced by anti–tumor necrosis factor therapy: A paradoxical adverse reaction



Administration of anti–tumor necrosis factor (anti-TNF) agents is beneficial in a variety of chronic inflammatory conditions, including psoriasis. We describe 5 patients in whom psoriasiform skin lesions developed 6–9 months after the initiation of anti-TNF therapy for longstanding, seropositive rheumatoid arthritis (etanercept or adalimumab), typical ankylosing spondylitis (infliximab), and Adamantiades-Behçet's disease (infliximab). In all 5 patients, the underlying disease had responded well to anti-TNF therapy. Four patients developed a striking pustular eruption on the palms and/or soles accompanied by plaque-type psoriasis at other skin sites, while 1 patient developed thick erythematous scaly plaques localized to the scalp. In 3 patients there was nail involvement with onycholysis, yellow discoloration, and subungual keratosis. Histologic findings from skin biopsies were consistent with psoriasis. None of these patients had a personal or family history of psoriasis. In all patients, skin lesions subsided either with topical treatment alone, or after discontinuation of the responsible anti-TNF agent. The interpretation of this paradoxical side effect of anti-TNF therapy remains unclear but may relate to altered immunity induced by the inhibition of TNF activity in predisposed individuals.

All anti–tumor necrosis factor (anti-TNF) agents currently in clinical use, namely the monoclonal antibodies infliximab and adalimumab, as well as the soluble TNF receptor etanercept, markedly decrease not only joint inflammation but also skin inflammation in patients with psoriatic arthritis (PsA), with a favorable safety profile (1). In addition, controlled studies have shown that etanercept and infliximab are efficacious in decreasing the clinical activity of skin lesions in patients with the severe form of psoriasis vulgaris (2). Psoriasis is a chronic skin disease affecting 1–3% of the population, and is characterized by hyperproliferation and abnormal differentiation of keratinocytes, as well as infiltration of activated T cells in the epidermis and papillary dermis. Although the mechanisms underlying the beneficial effects of anti-TNF treatment in patients with psoriatic skin lesions remain unclear, it is postulated that blocking the action of TNF decreases the production of other proinflammatory cytokines, such as interleukin-8 (IL-8), IL-6, and colony-stimulating factors, as well as the expression of adhesion molecules on endothelial cells and keratinocytes, leading to less inflammation and keratinocyte proliferation (1).

In recent years we have used infliximab, adalimumab, and etanercept to treat 100 patients with chronic arthritic conditions, including Adamantiades-Behçet's disease (ABD) (3), all of whom are being followed up regularly in our departments. To increase awareness of the potential adverse effects of anti-TNF therapy, we describe herein 5 of these patients in whom psoriasiform skin lesions developed during continuous administration of these agents.


The patients' characteristics, concomitant treatments, and clinical and histologic findings are summarized in Table 1. Both of the patients with rheumatoid arthritis (RA) had seropositive (rheumatoid factor titers more than 5 times the upper limit of normal), symmetric polyarthritis. Ankylosing spondylitis (AS) had been diagnosed on the basis of symmetric sacroiliitis and extensive spondylitis with typical syndesmophytes. Both of the patients with ABD had typical disease and relapsing panuveitis, and 1 of these patients had central nervous system involvement.

Table 1. Characteristics of 5 patients in whom psoriasis developed during anti-TNF treatment*
PatientAge/sexDiagnosis/ duration, yearsAnti-TNF agentMonths after treatment initiationConcomitant medicationType of eruptionNail involvementHistologic findings
  • *

    Anti-TNF = anti–tumor necrosis factor; AS = ankylosing spondylitis; INF = infliximab; RA = rheumatoid arthritis; ADA = adalimumab; LEF = leflunomide; SSZ = sulfasalazine; ABD = Adamantiades-Behçet's disease; AZA = azathioprine; ETA = etanercept; MTX = methotrexate.

133/FAS/13INF9Palmoplantar pustular lesions and scattered erythematous scaly plaques on arms, thighs, lateral trunkPsoriasiform hyperplasia, intraepidermal pustules
265/FRA/21ADA9LEF, SSZPalmoplantar pustular lesions, psoriasiform plaques on elbows, arms, thighsOnycholysis, “oil-spotting” (nail of right thumb)Intraepidermal pustule, parakeratosis, acanthosis, perivascular infiltrate
349/MABD/20INF6AZAExtensive erythematous plaques on the scalp with silvery white scaleParakeratosis, acanthosis, elongation of rete ridges, perivascular inflammatory infiltrate
443/MABD/21INF7AZAPalmoplantar pustular lesions, scaly plaques on elbows, knees, scalpOnycholysis, subungual keratosis in fingernails and toenailsNot done
548/FRA/14ETA7MTX, LEFPustules on soles, scaly plaques on elbows, tibiaSubungual hyperkeratosis and onychodystrophy of toenailsSmall intraepidermal pustule with neutrophils, psoriasiform hyperplasia, dilated capillaries in dermal papillae

None of the 5 patients had distal interphalangeal joint involvement, and none had a personal or family history of psoriasis. Typical features of plaque-type psoriasis were observed in all patients (formally evaluated by AS), with scattered desquamative erythematous plaques on the trunk, extremities, or scalp. Four patients developed pustular psoriasis of the palms and soles (Figure 1a), which resembled palmoplantar pustulosis. The psoriasis area and severity index did not exceed a score of 10 in any of the patients. The various anti-TNF agents for treatment of RA were administered at the dosages recommended by their respective manufacturers, while the patients with AS and ABD received infusions of 5 mg/kg infliximab every 6–8 weeks. With regard to the underlying rheumatic disease, all patients responded well to anti-TNF treatment.

Figure 1.

a, Patient 1. Psoriatic erythematous plaques of the midpalms with clusters of small (2–4-mm) pustules at different stages of evolution. Patient 3 b, Thick psoriatic plaques with silvery scale in the scalp and retroauricular area.

Patient 1. The patient, a 33-year-old woman with AS, was treated with infliximab. Nine months after the initiation of treatment, she developed a symmetric erythematous eruption on the palms and soles, with multiple pustules overlying erythematous pruritic skin (Figure 1a). A more widespread eruption composed of ill-defined erythematous scaly plaques developed on the arms, thighs, and lateral aspects of the trunk. A skin biopsy of the lesions on the soles showed psoriasiform epidermal hyperplasia with an intraepidermal pustule and a lymphohistiocytic perivascular infiltrate in the dermis. Due to the extent of the eruption, treatment was switched to etanercept, which resulted in partial resolution of the rash. The palmoplantar lesions continued to be present, but with a reduced and tolerable intensity.

Patient 2. Patient 2 was a 65-year-old woman with RA who was treated with adalimumab. During the ninth month of treatment, she developed a symmetric pustular eruption on her palms and soles, composed of multiple 1–2-mm pustules, with a background of erythematous, scaly skin. The skin eruption gradually progressed to involve other areas such as the elbows and the extensor surfaces of the arms and thighs, where well-defined erythematous, scaly plaques developed. In addition, she had new onset of onycholysis and “oil-spotting” on the fingernail of her right thumb, in addition to scaling of the scalp, consistent with seborrheic dermatitis. A skin biopsy from an indicative lesion of the left palm showed parakeratosis, acanthosis with an intraepidermal pustule, and a perivascular lymphocytic infiltrate in the dermis with neutrophil accumulation at the dermal papillae. Her skin condition responded well to treatment with topical steroids, and discontinuation of adalimumab was not required.

Patient 3. The patient, a 49-year-old man, was treated with infliximab for severe, relapsing ocular inflammation associated with ABD (3). Six months after the initiation of treatment, he developed an extensive scaly erythematous eruption, which was confined to the scalp and did not involve other areas of the skin (Figure 1b). A skin biopsy of the lesion was indicative of psoriasis, and findings included parakeratosis, epidermal hyperplasia, elongation of the rete ridges, dilated capillaries in the dermal papillae, and a lymphocytic and polymorphonuclear cell infiltrate in the dermis. Use of topical steroids resulted in a slight improvement of psoriasis, which subsided upon discontinuation of infliximab after 11 months of treatment. However, a recurrence of the eruption was observed a few months later, after infliximab had been reinstituted due to severe, relapsing ocular disease.

Patient 4. Patient 4 was a 43-year-old man who was treated with infliximab for severe, relapsing ocular inflammation associated with ABD (3). During the seventh month of continuous treatment, he developed a progressive eruption, with painful palmoplantar pustular lesions, scaly erythematous plaques on the elbows and knees, focal scaly plaques on the scalp, and onycholysis and subungual keratosis in several nails of the fingers and toes. Treatment with infliximab was discontinued, and psoriasis resolved with the addition of topical antiinflammatory treatment. However, because of frequent, severe ocular relapses during the next year, infliximab treatment was reinstated. A recurrence of psoriasis was observed during the fourth month after reinitiation of treatment, but infliximab was not discontinued because of the highly beneficial effect on his ocular condition.

Patient 5. The patient, a 48-year-old woman, was treated with infliximab for RA, but during the sixth infusion she developed an anaphylaxis-type reaction with hypotension and fever, and infliximab was discontinued. She was switched successfully to etanercept, and during the seventh month of treatment a pustular eruption developed on the soles, consisting of small discrete pustules on well-defined erythematous plaques. Clinical examination revealed psoriatic plaques with mild hyperkeratosis on the elbows and tibia. Subungual hyperkeratosis and onychodystrophy were also noted on the toenails. Histologic examination of the pustular lesions on the inner right foot showed psoriasiform hyperplasia with parakeratosis and a small intraepidermal pustule filled with neutrophils (Figure 2). The dermal papillae contained dilated capillaries, in addition to a dense perivascular inflammatory infiltrate, consisting of lymphocytes and neutrophils, in the dermis. The eruption was treated with a potent topical steroid, which resulted in subsequent improvement without discontinuing etanercept.

Figure 2.

Patient 5. Skin biopsy showing an intraepidermal pustule with neutrophils, acanthosis, minimal spongiosis, elongation of the rete ridges, and a dense inflammatory infiltrate in the upper dermis consisting of neutrophils and lymphocytes, typical of psoriasis. No eosinophils were detected (original magnification × 10).


We report the new onset of psoriasis in 5 patients who were receiving anti-TNF agents for chronic arthritic diseases. None of these patients had clinical evidence of psoriasis or psoriasiform skin/nail lesions either before or at the time of initiation of anti-TNF therapy. Examination of histologic samples from the palmoplantar pustular lesions showed typical features of localized pustular psoriasis (palmoplantar pustulosis), with intraepidermal pustules, epidermal hyperplasia, and a perivascular inflammatory infiltrate. Onycholysis, subungual hyperkeratosis, and yellowish discoloration were observed in 3 of the 5 patients. A close temporal association between the initiation of anti-TNF treatment and the onset of skin lesions was not evident, precluding an early allergic reaction. Moreover, no other psoriasis-inducing medications had been received by any patient in this study, and no clear exacerbating factors for psoriasis were reported. Moreover, readministration of the same anti-TNF agent in 2 of the 5 patients resulted in recurrence of the eruption.

These 5 patients were drawn from a cohort of 100 patients who had been treated in our centers with anti-TNF agents for >6 months. Infliximab, adalimumab, and etanercept were used to treat 47, 24, and 29 patients, respectively. Although the prevalence of psoriasis may be higher among patients with chronic inflammatory arthritic conditions, its occurrence in 5% of our patients who received anti-TNF treatment clearly exceeds the prevalence that might be expected by chance, especially within such a short period of observation. According to a method for estimating the probability of adverse drug reactions developed by Naranjo et al (4), the development of psoriasis following exposure to anti-TNF agents in patients 1, 2, and 5 is considered to be a probable adverse effect of treatment, while in both patients with ABD, in whom relapse occurred upon retreatment with infliximab, psoriasis is considered a definite adverse effect.

Alternative interpretations to explain the occurrence of a psoriasiform rash following anti-TNFα treatment seem unlikely. For example, the patients could have developed a local variant of acute generalized exanthematous pustulosis, which can mimic pustular psoriasis and may occur as an adverse drug event. In contrast to previously described cases of acute generalized exanthematous pustulosis (5), characterized by the acute appearance of a generalized pustular eruption, fever, and peripheral neutrophilia, the eruption in our patients was more localized and followed a more indolent course, which does not quite conform to the time frame of typical drug hypersensitivity.

Another potential explanation would be the possibility that the patients originally diagnosed as having RA or AS actually had psoriatic arthritis, and that their skin eruption was part of their underlying condition rather than an adverse effect of the anti-TNF treatment. PsA is known to be a quite heterogeneous disease, with clinical features that may occasionally resemble or overlap with those of other arthritic disorders, and ∼10% of patients with PsA present with arthritic symptoms before developing the cutaneous manifestations (6). However, the typical clinical, serologic, and radiologic findings in the patients described herein were more consistent with their original diagnosis of RA or AS. Furthermore, no concomitant arthritis flare was observed in any of the patients, which would have been expected given the similar immunologic mechanisms underlying the joint and skin involvement in PsA (1).

Several cases of a psoriasiform eruption induced by anti-TNF agents have been reported (7–11), including 2 patients who received infliximab for inflammatory bowel disease (7, 8), and 1 patient with RA who received adalimumab (9). Two of these patients developed a pustular eruption on the palms and soles (8, 9), accompanied by nail pitting in 1 patient (8). Also, 2 patients with seronegative RA developed psoriatic lesions that were typical from both the clinical and histologic perspectives, after treatment with infliximab and etanercept, respectively (10). Moreover, 3 of 127 patients with AS who had been treated with infliximab developed palmoplantar pustulosis at weeks 13, 34, and 23 of treatment. None of these patients had a personal or family history of psoriasis (11). Also, there have been anecdotal reports of 8 patients with RA and 4 patients with AS (12, 13) in whom new onset or aggravation of psoriasis developed following treatment with infliximab, adalimumab, or etanercept (4, 3, and 5 patients, respectively). Taken together, these data suggest that the development of psoriasiform lesions in previously unaffected individuals, particularly pustular lesions of the palms and soles, represents an adverse effect of anti-TNF treatment, which is a class effect rather than a drug-specific effect.

The underlying pathophysiologic mechanisms responsible for this phenomenon remain elusive. Several lines of evidence suggest that, under certain conditions, anti-TNF treatment promotes the activation of autoreactive T cells, leading to tissue damage via autoimmune mechanisms (14). This is exemplified by the common induction of antinuclear antibodies in some patients receiving anti-TNF agents and the occurrence of autoimmune syndromes, such as cutaneous lupus erythematosus or systemic lupus erythematosus–like syndromes (14). The psoriasiform lesions observed in our patients could be viewed in the context of altered immunity induced by the inhibition of TNF which, in turn, could promote an inflammatory “autoimmune” response in the skin of predisposed individuals. For example, TNF may have a suppressive regulatory role, possibly preventing the activation of psoriasis-specific autoreactive T cells, such as the cutaneous lymphocyte antigen (CLA)–expressing T cell subset (15). Moreover, anti-TNF treatment could increase the expression of chemokine receptors, such as CXCR3, which promote infiltration of autoreactive T cells to the skin that have been found to be up-regulated in psoriatic lesions (16). As recently reported, treatment of RA patients with infliximab and etanercept indeed results in increased numbers of peripheral T cells expressing CXCR3 (17); importantly, intraepidermal CLA+ T cells have been found to selectively express CXCR3 (18).

In turn, this hypothesis contradicts the documented therapeutic benefit of anti-TNF agents in psoriasis (1, 2). It should be remembered, however, that psoriasis is heterogeneous, and that certain variants of the disease, for example palmoplantar pustular psoriasis, may have immunologic backgrounds distinct from the typical plaque-type psoriasis (19). Therefore, the documented effect of anti-TNF therapy in plaque-type psoriasis does not imply a similar action in all psoriasis types. Further studies on similar cases are warranted not only to accurately interpret the clinical significance of this paradoxical reaction, but also to further explore the complex mechanisms underlying skin inflammation in these patients.