C L I N I C A L A N D LA B O R A T O R Y I N V E S T I G A T I O N S BJD British Journal of Dermatology

Cytokine imbalance with increased production of

interferon-a in psoriasiform eruptions associated with
antitumour necrosis factor-a treatments
J. Seneschal, B. Milpied, B. Vergier,* S. Lepreux, T. Schaeverbeke and A. Taı̈eb
Department of Dermatology and Pediatric Dermatology, Saint André and Pellegrin Hospitals, Bordeaux University Hospitals, 33075 Bordeaux cedex, France
*Department of Pathology, Haut-Lévêque Hospital, Bordeaux University Hospitals, Bordeaux, France
Departments of Pathology and Rheumatology, Pellegrin Hospital, Bordeaux University Hospitals, Bordeaux, France

Summary

Correspondence Background Psoriasiform eruptions occur in association with antitumour necrosis

Alain Taı̈eb. factor (TNF)-a treatments in autoinflammatory diseases. The major reported clini-
E-mail: alain.taieb@chu-bordeaux.fr cal presentation is palmoplantar pustulosis, sometimes accompanied with plaque-
like psoriasis. In some reports, histological findings suggest psoriasis whereas
Accepted for publication
25 May 2009 others favour a lichenoid drug reaction. We present a case series with a compre-
hensive clinical, histopathological and immunohistochemical study.
Key words Objectives To investigate the mechanism involved in psoriasiform eruptions in
adverse drug reaction, interferon-a, psoriasiform patients receiving anti-TNF-a inhibitors.
eruption, tumour necrosis factor-a
Methods Between July 2004 and May 2008, 13 patients with psoriasiform erup-
tions arising under anti-TNF-a treatment were enrolled in the study. Punch bio-
Conflicts of interest
None declared. psy specimens of lesions were processed for standard and immunohistochemical
analyses using antibodies against CD3, CD4, CD8, CD20, CD1a, KP1, CXCR3,
DOI 10.1111/j.1365-2133.2009.09329.x CXCL9, Tia1 and MxA, which is specifically induced by type I interferons (IFNs).
Additionally, we analysed biopsies from lesional skin of patients with cutaneous
discoid lupus erythematosus, lichen planus and psoriasis. Control biopsies were
taken from unaffected skin.
Results All patients developed psoriasiform plaques on the body accompanied with
palmoplantar keratoderma or pustulosis in three patients. Histological and immuno-
histochemical findings showed a psoriasiform pattern with focal lichenoid and
spongiotic features. We detected strong production of the MxA protein in inflam-
matory cells, indicating involvement of type I IFNs, and the expression was
higher than in control psoriasis samples. Expression of MxA was closely associ-
ated with the recruitment of CXCR3+ lymphocytes in the skin bearing markers
of cytotoxic capacity.
Conclusions Results support the hypothesis that psoriasiform eruptions are a new
model of drug reaction characterized by an increased expression of type I IFNs
induced by anti-TNF-a.

Antitumour necrosis factor (TNF)-a agents are attractive
approaches for the treatment of a series of autoinflammatory
disorders including arthritis, inflammatory bowel disease and
psoriasis, with a favourable safety profile.1–3 Several studies
have shown that anti-TNF-a agents can have limited cutaneous
side-effects including urticaria, hyperhidrosis, dry skin, hyper-
keratosis, rosacea, hyperpigmentation, alopecia and delayed
hypersensivity reactions.4 The paradoxical induction of psoria-
sis-like eruption under anti-TNF-a treatments was first
reported by Verea et al. in 2004.5 Since that time, more than
170 cases of psoriasiform eruptions under anti-TNF-a agents
have been reported.6,7 The most common clinical presentation
is palmoplantar pustulosis accompanied or not by guttate or
plaque like-psoriasis. Histological findings were considered as
consistent with psoriasis in 40 cases. In other reports, skin
biopsies were not typical of psoriasis but rather showed a
lichenoid pattern. Nevertheless, psoriasiform eruptions have
not always been analysed histologically in a systematic manner
and the mechanisms underlying this paradoxical eruption are
not yet understood.

 2009 The Authors

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088 1081
1082 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al.
Interestingly, recent reports have shown that TNF-a inhibi-
tors increase the production of interferon (IFN)-a by plasma-
cytoid dendritic cells (pDCs).8 It is now admitted that IFN-a
plays a key role in the induction of cytotoxic skin reactions
such as found in cutaneous discoid lupus erythematosus
(CDLE),9 lichen planus (LP)10 or lichenoid eruptions. It has
been noted that type I IFNs (IFN-a ⁄b) induce the expression
of CXCR3 ligands such as CXCL9, CXCL10 and CXCL11 and
support a Th1-skewed inflammatory immune response, induc-
ing the recruitment of CXCR3+ cytotoxic T lymphocytes in
skin lesions.
This present study constitutes an additional and detailed clini-
cal, histological and immunohistochemical investigation to our
previous report.11 Thirteen patients with psoriasiform eruptions
occurring during the administration of anti-TNF-a agents are
now included, showing evidence of a ‘type I IFN signature’. To
investigate the mechanism of these eruptions, we analysed the
expression of the MxA protein which represents a specific
marker for type I IFN signalling. MxA is an abundant and
ubiquitous GTPase cytoplasmic protein with natural antiviral
activity and is inducible by IFN-a ⁄b.12 In previous studies, it
was shown that MxA expression, in type I IFN-dependent skin
infiltrates, is closely correlated with recruitment of chemokine
receptor (CXCR3)-positive lymphocytes characterizing the
induction of a Th1-biased cellular immune response.13
These Th1 cells show a cytotoxic profile analysed by the
expression of the cytotoxic protein Tia1. We performed
comparative immunohistochemical analyses to characterize the
phenotype of infiltrating inflammatory cells in these eruptions
compared with those in LP, CDLE and psoriasis.
Patients and methods
Between July 2004 and May 2008, patients who developed a
psoriasiform eruption after the beginning of anti-TNF-a treat-
ment were enrolled in this study. All patients were referred to
the Department of Rheumatology and Gastroenterology.
Lesional skin biopsies taken from 13 patients were investi-
gated. Details of patient characteristics are presented in Table 1.
All specimens were taken for diagnostic purposes from
untreated skin lesions. Control biopsies were taken from the
unaffected skin of three patients undergoing surgery for skin
tumours. Additionally, we analysed as positive controls
biopsies from lesional skin of patients with CDLE (n = 5), LP
(n = 5) and psoriasis (n = 6). Informed consent was obtained
from all donors.
Histology and immunohistology
Sections were prepared from formalin-fixed, paraffin-embed-
ded skin biopsies. Standard haematoxylin and eosin staining
was performed for diagnosis. The inflammatory infiltrate was
characterized by immunochemistry. Monoclonal antibodies
specific for CD3 (Dako, Glostrup, Denmark), CD4 (Menarini
Pharmaceuticals, Florence, Italy), CD8 (Dako), CD20 (Dako),
CD1a (Beckman Coulter, Paris, France), KP1 (Dako) and Tia1
Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
(Beckman Coulter) were used as primary antibodies, following
the manufacturer’s protocol. Labelling for MxA (M143, dilu-
tion 1 : 100; from Prof. O. Haller, University of Freiburg,
Germany), CXCR3 (1C6, dilution 1 : 100; PharMingen, San
Diego, CA, U.S.A.) and CXCL9 ⁄Mig (Mab392; R&D Systems,
Minneapolis, MN, U.S.A.) was performed on paraffin-embed-
ded tissue sections, as described previously.9 Secondary label-
ling was performed using the LSAB2 kit (Dako). Appropriate
isotype-matched controls were included. The stained slides
were analysed independently by two individuals in a blinded
manner. Cells were counted per three high power fields
(original magnification · 40) and the mean number was
calculated. The expression of CXCL9 was scored using a semi-
quantitative method: 0, no expression; +, weak expression;
++, fair expression; +++, strong expression. The expression
of MxA in inflammatory infiltrates was scored quantitatively
(percentage of strongly positive cells).
Statistical analysis
Pearson’s correlation analysis was used to measure correlation
between two parameters. The nonparametric Mann–Whitney
U-test was employed to compare the expression of CD3, CD4,
CD8, CD20, CD1a, KP1, CXCR3, CXCL9, Tia1 and MxA in dif-
ferent disease subsets and healthy controls in the skin.
P < 0Æ05 was considered to be significant (*), and P < 0Æ01
as highly significant (**).
Results
Case reports
The clinical and histological characteristics of patients who
developed psoriasis under anti-TNF-a treatment are sum-
marized in Table 1. Thirteen patients (seven men and six
women; median age 47 years, range 26–77) were enrolled.
Patients were followed for rheumatoid arthritis (five patients),
ankylosing spondylitis and psoriatic arthritis (four patients),
and inflammatory bowel disease (four patients). Seven patients
had received infliximab, three etanercept and three ada-
limumab. One patient had a personal history of psoriasis and
another had a family history of psoriasis.
The mean time elapsed from initiation of therapy to appear-
ance of skin lesions was variable: 15Æ8 months for infliximab
(range 1–26), 14Æ7 months for etanercept (range 2–24) and
4Æ8 months for adalimumab (range 1Æ5–7). Thirteen patients
developed a psoriasiform eruption consisting of small plaques,
associated with palmoplantar keratoderma or pustulosis in three
patients (Fig. 1a). One patient presented an inverted psoriasi-
form eruption. The lesions did not develop at injection sites.
Due to the highly beneficial effect on target diseases, psori-
asiform eruptions did not justify discontinuation of anti-TNF-
a treatments (except in patient 12). In three cases (patients 1,
3 and 4), treatment was stopped because of another adverse
reaction, e.g. immune keratitis (patient 1). Discontinuation
resulted in a complete clearing of lesions in two patients, and
 2009 The Authors
 Table 1 Descriptions of the 13 patients: patients 1–8 were described previously11

Personal ⁄ family
history ANA
Months after

 2009 The Authors

Age Diagnosis ⁄ duration Anti-TNF-a treatment Beginning
Patient (years) ⁄ sex (years) Psoriasis Eczema agent initiation Type of eruption Histological findings anti-TNF-a Eruption
1 35 ⁄ F RA seronegative ⁄ 3 ) ⁄) ) ⁄) ETA 2 Psoriasiform plaques legs, Psoriasiform pattern, focal lichenoid 1 ⁄ 100 1 ⁄ 500
trunk, scalp and spongiotic pattern
2 59 ⁄ M AS HLA-B27+ ⁄ 17 ) ⁄) ) ⁄) INF 18 Guttate psoriasiform lesions Spongiform pustulosis, minimal 1 ⁄ 50 1 ⁄ 100
scalp, legs, trunk, and spongiosis
palmoplantar pustulosis
3 45 ⁄ M RA seropositive ⁄ 4 ) ⁄) ) ⁄) ADA 7 Psoriasiform plaques scalp, ND – –
and palmoplantar pustulosis
4 26 ⁄ F AS HLA-B27) ⁄ 5 + ⁄ ) (2nd ) ⁄) INF 8 Plaques and guttate Psoriasiform pattern, apoptotic – 1 ⁄ 50
degree) psoriasiform eruption scalp, keratinocytes, focal lichenoid and
neck, umbilicus, legs spongiotic pattern
5 48 ⁄ M PsA HLA-B27+ ⁄ 25 + ⁄) ) ⁄) INF 26 Elbows: psoriasis plaques; Elbow: psoriasis; leg: psoriasiform – 1 ⁄ 500
legs: psoriasiform plaques pattern, spongiosis
6 61 ⁄ M AS HLA-B27+ ⁄ 30 ) ⁄) + ⁄) INF 26 Extensive erythematous Psoriasiform pattern 1 ⁄ 100 1 ⁄ 250
plaques on trunk, legs;
palmoplantar hyperkeratosis
and major skin fold lesions
7 30 ⁄ F RA seropositive ⁄ 13 ) ⁄) + ⁄) ETA 18 Psoriasiform lesions elbows, Hyperkeratosis, acanthosis, focal 1 ⁄ 1000 1 ⁄ 1000
knees, fingers lichenoid and spongiotic pattern
8 59 ⁄ M RA seropositive ⁄ 6 ) ⁄) ) ⁄) ETA 24 Erythematous plaques arms, ND 1 ⁄ 50 1 ⁄ 50
trunk
9 77 ⁄ F RA seronegative ⁄ 10 ) ⁄) ) ⁄) INF 24 Erythematous plaques arms, Hyperkeratosis, acanthosis, ? ?
trunk spongiosis
10 47 ⁄ M IBD ⁄ 10 ) ⁄) ) ⁄) INF 1 Palmoplantar keratoderma Psoriasiform pattern, focal lichenoid 1 ⁄ 50 ?
and spongiotic pattern
11 41 ⁄ F IBD ⁄ 15 ) ⁄) ) ⁄) INF 2 Erythematous plaques arms, Minimal parakeratosis, focal 1 ⁄ 50 1 ⁄ 100
trunk, umbilicus lichenoid pattern, apoptotic
keratinocytes
12 58 ⁄ F IBD ⁄ 20 ) ⁄) ) ⁄) ADA 1Æ5 Erythematous plaques trunk, Psoriasiform pattern ? ?

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
limbs
13 37 ⁄ M IBD ⁄ 8 ) ⁄) ) ⁄) ADA 6 Follicular and Parakeratosis, focal lichenoid pattern, ? ?
erythematosquamous apoptotic keratinocytes
plaques trunk, arms

ANA, antinuclear antibodies; TNF, tumour necrosis factor; RA, rheumatoid arthritis; AS, ankylosing spondylitis; PsA, psoriatic arthritis; IBD, inflammatory bowel disease; ETA, etanercept; INF, infliximab;
ADA, adalimumab; ND, not determined.
Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al. 1083
1084 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al.

(a)

(b)

Fig 1. Clinical and pathological features of psoriasiform eruptions. (a) Left panel, psoriasiform plaques on trunk (patient 4) and right panel,
palmar keratoderma and pustulosis (patient 3). (b) Left panel, histology shows a psoriasiform pattern with acanthosis and a focal lichenoid pattern
with necrotic keratinocytes (patient 4). Right panel, histology shows subcorneal pustulosis with a lymphocytic infiltrate (patient 3). (a,b) Left
panels are reproduced from our previous report.11

partial clearing in the third patient. In five patients the skin
disease was treated successfully with topical steroids, while
eight patients continued to have persistent skin disease despite
topical therapy and phototherapy. Antinuclear antibody titres
were evaluated before TNF-a-blocker treatment and during
psoriasiform eruptions. Four patients (patients 1, 2, 6 and 11)
had an increase in titre (Table 1).
Eleven patients had a skin biopsy of the psoriasiform erup-
tions (Fig. 1b). A psoriasiform pattern with parakeratosis,
hyperkeratosis and acanthosis was observed in five cases. Sur-
prisingly, three of these five biopsies showed an associated
focal lichenoid pattern characterized by epidermal basal-cell
damage and lymphocytic infiltrate in the upper dermis. A
focal lichenoid pattern was also shown in three other skin
samples. Apoptotic keratinocytes were present in patients 4,
11 and 13. Moreover, seven biopsies showed spongiosis with
epidermal oedema. A unilocular subcorneal pustule without
spongiform pustulation was found in patient 3. The inflam-
matory infiltrate was predominantly composed of mono-
nuclear cells associated with eosinophils in patients 2 and 7.
Skin infiltrating cells were mainly localized around vessels
but were also found in the upper dermis in a lichenoid pat-
tern. Interestingly, in patient 5 affected by psoriatic arthritis
we compared the pathological pattern of the skin biopsy per-
formed on a psoriasis lesion on the arm with a skin biopsy
taken from a psoriasiform lesion that appeared under anti-
TNF-a treatment on the leg. The psoriasis skin biopsy was
indistinguishable from psoriasis with parakeratosis, acanthosis,
elongation of rete ridges and vascular dilatation, while the
psoriasiform eruption biopsy showed a psoriasiform pattern
with moderate spongiosis and a striking perivascular mono-
nuclear cell infiltrate.

 2009 The Authors

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al. 1085

(a) Lupus Lichen Psoriasis Eruption (b) 900

Lupus
800 Lichen
CD3

Psoriasis
700
Eruption
Control
600

Cells per HPF

500
CD4

400

300

200
CD8

100

0
CD3 CD4 CD8 CD20 CD1a KP1

Fig 2. Characterization of the inflammatory infiltrate by immunohistochemistry (IHC) reveals CD8-dominated T-cell inflammation in psoriasiform
eruptions. IHC was performed using monoclonal antibodies specific for CD3, CD4, CD8, CD20, CD1a and KP1 to stain the lesional infiltrate.
Skin samples taken from patients with cutaneous discoid lupus erythematosus, lichen planus, psoriasis and healthy skin were included for control
purposes. (a) Representative findings of CD3, CD4 and CD8 expression (original magnification · 20). (b) The psoriasiform eruptions infiltrate is

(a) Lupus Lichen Psoriasis Eruption

(b) 100 **
90 *
80
% of strong positive cells

70
60
50
Fig 3. Enhanced MxA expression in psoriasiform eruption infiltrate.
40 (a) Representative findings of the specific type I interferon marker MxA in
30 cutaneous discoid lupus erythematosus, lichen planus, psoriasis and psoriasiform
20 eruptions (upper panel, original magnification · 20; lower panel, original
10
magnification · 40). (b) MxA expression on inflammatory cells was scored
0
Lupus Lichen Psoriasis Eruption Control quantitatively: percentage of strongly positive cells (*P < 0Æ05, **P < 0Æ01,
n=5 n=5 n=6 n = 10 n=3
nonparametric Mann–Whitney U-test).
MxA

 2009 The Authors

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
1086 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al.

than in psoriasis (P < 0Æ05) and healthy controls (P < 0Æ01),

T lymphocytes predominate in psoriasiform eruptions
The inflammatory infiltrate was characterized using antibodies
against CD3, CD4, CD8, CD20, CD1a and KP1 (Fig. 2a,b). T
cells were accompanied by several macrophages and a small
number of CD20+ B cells. The analyses revealed a clear major-
ity of CD3+ T cells among the infiltrating T cells. The number
of infiltrating immune cells was about 1Æ5 times lower than in
CDLE and LP but higher than in psoriasis, and the CD8 ⁄CD4
ratio (close to 1) was doubled in psoriasiform eruptions in
contrast to psoriasis samples.
MxA expression indicates production of type I interferons
in psoriasiform eruptions
To investigate the involvement of type I IFNs in psoriasiform
skin eruptions, we used a monoclonal antibody against MxA
(an antiviral protein which is specifically induced by type I
IFNs). Our analyses revealed MxA expression (Fig. 3a,b) in
the epidermis and dermal infiltrate. The expression in the epi-
dermis was comparable with that in psoriasis and lower than
in CDLE and LP (data not shown). However, MxA expression
in psoriasiform eruption infiltrates was significantly higher
but lower than in CDLE and LP.
The expression of the type I interferon-inducible ligand
CXCL9 is increased in psoriasiform eruptions
Type I IFNs induce lesional expression of several cytokines in
the skin, including the IFN-inducible protein CXCL9. This
cytokine is a ligand for the CXCR3 receptor and plays a role
in the migration of CXCR3+ lymphocytes into the skin.14,15
We found an increased expression of CXCL9 (Fig. 4a,b) in the
epidermis of psoriasiform eruptions vs. healthy control skin
(P < 0Æ01) and psoriasis (nonsignificant trend) (semiquantita-
tive analysis method).
CXCR3 and Tia1 are highly expressed in psoriasiform
eruption infiltrates
Type I IFNs are known to enhance a Th1-skewed immune
response like in cutaneous lupus erythematosus,16 LP10 or
lichenoid eruptions. We investigated the expression of the
chemokine receptor CXCR3 in the infiltrate which recog-
nizes Th1 lymphocytes. We found a strong increase of

(a) (b)
Strong **
Lupus Lichen Psoriasis Eruption

Fair
CXCL9

Weak

None
Lupus Lichen Psoriasis Eruption Control
n=5 n=5 n=6 n = 10 n=3
CXCL9
(c) 600
500
*
Cells per HPF

400 **
CXCR3

300

200

100

0
Lupus Lichen Psoriasis Eruption Control
n=5 n=5 n=6 n = 10 n=3
(d) CXCR3
120
*
100 **
Tia-1

Cells per HPF

80

60

40

20

0
Lupus Lichen Psoriasis Eruption Control
n=5 n=5 n=6 n = 10 n=3
Tia-1

Fig 4. Immunohistochemical analyses confirm significant ‘type I interferon signature’ in psoriasiform skin lesions. (a) Representative findings of
CXCL9, CXCR3 and Tia1 expression in cutaneous discoid lupus erythematosus, lichen planus, psoriasis and psoriasiform eruptions (original
magnification · 20). (b–d) Overview of the expression of CXCL9, CXCR3 and Tia1 on all investigated samples. Cells were counted per high
power field (HPF) (original magnification · 40). Data are given as mean ± SD (*P < 0Æ05, **P < 0Æ01, nonparametric Mann–Whitney U-test).

 2009 The Authors

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al. 1087

(a) 550 psoriasis lesions) presented skin lesions under TNF-a blockers
with clinical changes comparable with psoriasis, while histo-
500 r = 0·71, P < 0·05
CXCR3+ cells per HPF
450
pathology showed a psoriasiform pattern in association with
400 spongiosis in 10 cases and a lichenoid pattern in five cases.
350 Apoptotic keratinocytes were found in three cases. The inflam-
300 matory infiltrate was composed of a majority of mononuclear
250 cells with rarer eosinophilic infiltration. Immunohistochemical
200 analyses revealed that CD3+ T cells predominated. Skin total
150 CD3+ T-cell counts were found to be lower than in CDLE and
100 LP but higher than in psoriasis, and the CD8 ⁄CD4 ratio was
20 30 40 50 60 70 80 90
% of strong positive MxA cells increased like in CDLE and LP, suggesting a common patho-
genic mechanism between these three skin disorders.
(b) 550 It is now generally accepted that anti-TNF-a blockers
500 r = 0·64, P < 0·05 increase levels of type I IFNs (IFN-a ⁄b). IFN-a plays a major
CXCR3+ cells per HPF

450 role in cytotoxic skin reactions but also in psoriasis.8 A ‘type I

400
350
300
250
200
150
100
50
0 20 40 60 80 100 120 140
Tia1+ cells per HPF
Fig 5. Psoriasiform eruptions showed a correlation between (a) the
percentage of strongly positive MxA cells and (b) Tia1 expression,
with the CXCR3+ infiltrating cells. HPF, high power field.
CXCR3-expressing lymphocytes in psoriasiform eruptions as
compared with psoriasis (Fig. 4a,c). The CXCR3 receptor was
expressed at a level comparable with that in CDLE and LP. We
statistically correlated MxA expression in psoriasiform erup-
tion infiltrates and the number of Th1 (CXCR3-expressing)
lymphocytes (Fig. 5a) (r = 0Æ71, P < 0Æ05). Moreover, a large
number of cells in psoriasiform eruptions had a cytotoxic phe-
notype (Tia1+ staining) (Fig. 4a,d) when compared with
infiltrating cells in psoriasis (P < 0Æ05). We also found a cor-
relation between Tia1-positive cells and CXCR3-expressing
lymphocytes in psoriasiform eruption infiltrates (Fig. 5b)
(r = 0Æ64, P < 0Æ05).
Discussion
Since the first description of psoriasiform eruptions under
anti-TNF-a treatment in 2004, more than 170 cases have been
described in the international literature, but many reports
give insufficient descriptions and the pathogenic mechanism is
still not understood.6,7 Some authors suggested that these
eruptions show clinical and pathological changes indistin-
guishable from true psoriasis17 while others described a clini-
cal psoriasiform eruption with a lichenoid pattern on skin
biopsy.5 In the present study, we bring new pathological
and immunohistochemical evidence supporting the idea that
psoriasiform eruptions under anti-TNF-a treatment are dif-
ferent from true psoriasis. Indeed, our 13 patients (including
a patient affected by psoriatic arthritis with previous true
IFN signature’ has been shown to be associated with the
‘interface dermatitis’ found in several conditions including
cutaneous lupus erythematosus, LP and lichenoid drug
eruptions.18,19 In these conditions, the interaction between
the IFN-inducible chemokines CXCL9 and CXCL10 and their
common receptor CXCR3 appears to play a central pathogenic
role. According to our data, similar mechanisms are probably
involved in TNF-a-induced psoriasiform eruptions. Further,
we found a significantly increased expression of the MxA pro-
tein, indicating local type I IFN expression. The expression of
the MxA protein was clearly associated with the recruitment
of CXCR3+ T cells like in CDLE and LP. Moreover, these
CXCR3+ T cells expressed a cytotoxic Tia1 pattern. The IFN-
inducible chemokine CXCL9, which is enhanced in psoriasi-
form eruptions, probably plays an important proinflammatory
role by recruiting CXCR3+ T cells in the skin, as shown in
previous reports.20 De Gannes et al.21 have already reported an
increased immunohistochemical expression of the MxA pro-
tein in two cases of psoriasiform eruptions triggered by anti-
TNF-a treatment. However, their conclusion that psoriasiform
eruptions induced by TNF-a blockers are true psoriasis
remains debatable, as already discussed by us.22
Psoriasis is associated with a Th1 and a Th17 immune
response.23,24 The role of type I IFNs in the pathogenic mech-
anism of psoriasis has been well demonstrated,25 but accord-
ing to our data and other reports,20 the IFN-a signature
markers were clearly weaker than in psoriasiform eruptions
and classic ‘interface dermatitis’ lesions.
IFN-a is produced by pDCs that infiltrate the skin under
inflammatory conditions as was shown in ‘interface dermatitis’
lesions. Due to the rarity of the present adverse skin reaction
and the lack of available remaining sections, we were not able
to check the infiltration of pDCs. However, we can hypothesize
that pDCs, like in other ‘type I IFN’-dependent skin eruptions,
infiltrate the skin to produce large amounts of IFN-a. TNF-a
has been shown to regulate IFN-a production.8 TNF-a inhibits
the generation of pDCs from haematopoietic progenitors and
accelerates maturation of pDCs, at which stage pDCs cease to
produce IFN-a. So, blocking TNF-a could increase the produc-
tion of IFN-a by pDCs in the absence of other counterbalanc-
ing mechanisms, or enhance the production of IFN-a by pDCs

 2009 The Authors

Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
1088 Characteristics of psoriasiform eruptions following anti-TNF-a treatments, J. Seneschal et al.
following stimuli like viral reactivation.19 However, in the
present study we were not able to identify further triggers
besides anti-TNF-a treatment able to stimulate pDCs.
Taken together, the findings of the present study demon-
strate that psoriasiform eruptions arising during TNF-
a-blocker treatment are characterized by a ‘type I IFN
signature’ also found in CDLE or LP, which is characterized by
the cutaneous recruitment of Th1 cells. We suggest that psori-
asiform eruptions arising under anti-TNF-a agents constitute a
new model of adverse drug reaction. They differ from true
psoriasis. A cytokine imbalance with an increased production
of IFN-a by TNF-a inhibitors seems to play a key role. Our
data suggest a link between these eruptions and the broader
histopathological group of ‘interface dermatitis’, reflecting
increased in situ production of type I IFNs.
Acknowledgments
The study was supported by a grant from G.E.R.D.A.: Groupe
d’Etudes et de Recherche Dermato-Allergologique. Dr Seneschal
was the recipient of the 2007 Claude Benezra prize for this
project. We thank Dr P. Blanco for critically reading the
manuscript and for his helpful suggestions.
References
1 Weinblatt ME, Keystone EC, Furst DE et al. Adalimumab, a fully
human anti-tumor necrosis factor alpha monoclonal antibody,
for the treatment of rheumatoid arthritis in patients taking
concomitant methotrexate: the ARMADA trial. Arthritis Rheum 2003;
48:35–45.
2 Moreland LW, Schiff MH, Baumgartner SW et al. Etanercept therapy
in rheumatoid arthritis. A randomized, controlled trial. Ann Intern
Med 1999; 130:478–86.
3 Lipsky PE, van der Heijde DM, St Clair EW et al. Infliximab and
methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor
Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant
Therapy Study Group. N Engl J Med 2000; 343:1594–602.
4 Lee HH, Song IH, Friedrich M et al. Cutaneous side-effects in
patients with rheumatic diseases during application of tumour
necrosis factor-alpha antagonists. Br J Dermatol 2007; 156:486–91.
5 Verea MM, Del Pozo J, Yebra-Pimentel MT et al. Psoriasiform erup-
tion induced by infliximab. Ann Pharmacother 2004; 38:54–7.
6 Collamer AN, Guerrero KT, Henning JS et al. Psoriatic skin lesions
induced by tumor necrosis factor antagonist therapy: a literature
review and potential mechanisms of action. Arthritis Rheum 2008;
59:996–1001.
7 Wollina U, Hansel G, Koch A et al. Tumor necrosis factor-alpha
inhibitor-induced psoriasis or psoriasiform exanthemata: first 120
cases from the literature including a series of six new patients. Am
J Clin Dermatol 2008; 9:1–14.
8 Palucka AK, Blanck JP, Bennett L et al. Cross-regulation of TNF and
IFN-alpha in autoimmune diseases. Proc Natl Acad Sci USA 2005;
102:3372–7.
Journal Compilation  2009 British Association of Dermatologists • British Journal of Dermatology 2009 161, pp1081–1088
9 Wenzel J, Worenkamper E, Freutel S et al. Enhanced type I inter-
feron signalling promotes Th1-biased inflammation in cutaneous
lupus erythematosus. J Pathol 2005; 205:435–42.
10 Wenzel J, Scheler M, Proelss J et al. Type I interferon-associated
cytotoxic inflammation in lichen planus. J Cutan Pathol 2006;
33:672–8.
11 Seneschal J, Lepreux S, Bouyssou-Gauthier ML et al. Psoriasiform
drug eruptions under anti-TNF treatment of arthritis are not true
psoriasis. Acta Derm Venereol (Stockh) 2007; 87:77–80.
12 Haller O, Kochs G. Interferon-induced mx proteins: dynamin-like
GTPases with antiviral activity. Traffic 2002; 3:710–17.
13 Mohan K, Cordeiro E, Vaci M et al. CXCR3 is required for migra-
tion to dermal inflammation by normal and in vivo activated T cells:
differential requirements by CD4 and CD8 memory subsets. Eur J
Immunol 2005; 35:1702–11.
14 Tensen CP, Flier J, van der Raaij-Helmer EM et al. Human IP-9: a
keratinocyte-derived high affinity CXC-chemokine ligand for the
IP-10 ⁄ Mig receptor (CXCR3). J Invest Dermatol 1999; 112:716–
22.
15 Flier J, Boorsma DM, van Beek PJ et al. Differential expression of
CXCR3 targeting chemokines CXCL10, CXCL9, and CXCL11 in
different types of skin inflammation. J Pathol 2001; 194:398–
405.
16 Wenzel J, Tuting T. Identification of type I interferon-associated
inflammation in the pathogenesis of cutaneous lupus erythemato-
sus opens up options for novel therapeutic approaches. Exp Dermatol
2007; 16:454–63.
17 Kary S, Worm M, Audring H et al. New onset or exacerbation of
psoriatic skin lesions in patients with definite rheumatoid arthritis
receiving tumour necrosis factor alpha antagonists. Ann Rheum Dis
2006; 65:405–7.
18 Wenzel J, Tuting T. An IFN-associated cytotoxic cellular immune
response against viral, self-, or tumor antigens is a common
pathogenetic feature in ‘interface dermatitis’. J Invest Dermatol 2008;
128:2392–402.
19 Meller S, Gilliet M, Homey B. Chemokines in the pathogenesis
of lichenoid tissue reactions. J Invest Dermatol 2009; 129:315–
19.
20 Wenzel J, Peters B, Zahn S et al. Gene expression profiling of lichen
planus reflects CXCL9+-mediated inflammation and distinguishes
this disease from atopic dermatitis and psoriasis. J Invest Dermatol
2008; 128:67–78.
21 de Gannes GC, Ghoreishi M, Pope J et al. Psoriasis and pustular
dermatitis triggered by TNF-alpha inhibitors in patients with
rheumatologic conditions. Arch Dermatol 2007; 143:223–31.
22 Seneschal J, Lepreux S, Milpied B et al. Psoriasiform eruptions dur-
ing anti TNF-alpha treatment: psoriasis or not? Arch Dermatol 2007;
143:1593–5.
23 Griffiths CE, Barker JN. Pathogenesis and clinical features of psoria-
sis. Lancet 2007; 370:263–71.
24 Fitch E, Harper E, Skorcheva I et al. Pathophysiology of psoriasis:
recent advances on IL-23 and Th17 cytokines. Curr Rheumatol Rep
2007; 9:461–7.
25 Nestle FO, Conrad C, Tun-Kyi A et al. Plasmacytoid predendritic
cells initiate psoriasis through interferon-alpha production. J Exp
Med 2005; 202:135–43.
 2009 The Authors