Professional Documents
Culture Documents
Vasiliki Nikolaou, MD, Vincent Sibaud, MD, Davide Fattore, MD, Pietro Sollena, MD,
Ariadna Ortiz-Brugués, MD, Damien Giacchero, MD, Maria Concetta Romano, MD,
Julia Riganti, MD, Konstantinos Lallas, MD, Ketty Peris, MD, Dimitra Voudouri, MD,
Aimilios Lallas, MD, Gabriella Fabbrocini, MD, Elisabeth Lazaridou, MD, Cristina
Carrera, MD, Maria Carmela Annunziata, MD, Ernesto Rossi, MD, Angela Patri, MD,
Dimitrios Rigopoulos, MD, Alexander J. Stratigos, MD, Zoe Apalla, MD
PII: S0190-9622(20)33115-7
DOI: https://doi.org/10.1016/j.jaad.2020.08.137
Reference: YMJD 15460
Please cite this article as: Nikolaou V, Sibaud V, Fattore D, Sollena P, Ortiz-Brugués A, Giacchero D,
Romano MC, Riganti J, Lallas K, Peris K, Voudouri D, Lallas A, Fabbrocini G, Lazaridou E, Carrera C,
Annunziata MC, Rossi E, Patri A, Rigopoulos D, Stratigos AJ, Apalla Z, Immune checkpoint-mediated
psoriasis: a multicentric European study of 115 patients from European Network for Cutaneous ADverse
Event to Oncologic drugs (ENCADO) group., Journal of the American Academy of Dermatology (2021),
doi: https://doi.org/10.1016/j.jaad.2020.08.137.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier on behalf of the American Academy of Dermatology, Inc.
1
of
13
14
ro
15 Affiliations:
1
16 1st Department of Dermatology, “Andreas Sygros” Hospital for skin diseases,
-p
17 National & Kapodestrian University of Athens, Medical School, Athens, Greece.
2
18 Institut Universitaire du cancer, Toulouse Oncopole, Toulouse, France
re
3
19 Section of Dermatology, Department of Clinical Medicine and Surgery, University
20 of Naples Federico II, Naples, Italy
lP
4
21 Department of Dermatology, Università Cattolica del Sacro Cuore, Rome, Italy;
22 Fondazione Policlinico “A. Gemelli” IRCCS, Rome, Italy
5
23 Hospital Clinic Barcelona, Spain
na
6
24 Centre Antoine Lacassagne, Nice, France
7
25 San Camillo Forlanini Hospital, Rome, Italy
ur
8
26 Hospital Italiano of Buenos Aires, Argentina
9
27 First Dermatology Department, Medical School, Aristotle University of
Jo
28 Thessaloniki, Greece
10
29 Second Dermatology Department, Medical School, Aristotle University of
30 Thessaloniki, Greece
11
31 Medical Oncology, Fondazione Policlinico “A. Gemelli” IRCCS, Rome, Italy
32
33
34 Corresponding author: Dr Vasiliki Nikolaou
35 Cutaneous Toxicities Clinic, OncoDermatology Department
36 “Andreas Sygros” Hospital for skin diseases
37 5 I.Dragoumi str
38 12161 Athens , Greece
39 e-mail: drviknik@yahoo.com
40 tel: 0030-210-7265153
41
42
2
of
53 Reprint requests: Vasiliki Nikolaou
ro
54 Keywords: immune checkpoint inhibitors, psoriasis, adverse events, nivolumab,
55 pembrolizumab, immunotherapy, skin toxicity
56
-p
re
57
58
lP
59
60
na
61
ur
62
63
Jo
64
65
66
67
68
69
70
71
72
73
74
75
3
76
77 Abstract
80 Objective: To report data on ICI-mediated psoriasis, emerging from the largest so far
of
82 Μethods: The medical records of all patients with ICI-mediated psoriasis were
ro
83 retrospectively reviewed across nine institutions.
-p
84 Results: We included a cohort of 115 individuals. Grade-1, 2 and 3 disease severity
re
85 was reported in 60/105 (57.1%, 10 missing data), 34/105 (32.4%) and 11/105
lP
86 (10.5%), respectively. The ratio between de novo and exacerbation cases of psoriasis
na
87 was 21/90 (23.3%). The most common systemic therapy was acitretin (23 patients,
ur
90 patients (25.9%) interrupted and 20/111 (18%) permanently discontinued ICIs due to
91 psoriasis. BSA>10% at baseline had a 3.6 increased risk for ICIs treatment
94 grade2/3 disease were significant positive predictors for antitumor response of ICI
96
98
4
99 Conclusion: Acitretin, apremilast and methotrexate are safe and effective modalities
100 for ICI-mediated psoriasis. In most cases, ICI can be completed unhindered. A
102
103
104
of
105
ro
106
107
-p
re
108
lP
109
na
110
ur
111
Jo
112
113
114
115
116
117
118
5
119 Introduction
120 Immune checkpoint inhibitors (ICIs) are monoclonal antibodies targeting cytotoxic T
122 or programmed death ligand-1 (PD-L1). Their addition in our armamentarium has
123 radically transformed the therapeutic arena in oncology, providing the potential of
125 The introduction of ICIs has resulted in the recognition of a new spectrum of
126 immune-related adverse events’ (irAEs) that may occasionally limit their use [1]. The
of
127 nature of irAEs has not been completely elucidated, however it is meaningful that
ro
128 they are mostly immune-mediated, resulting by the T-cell activation of cytotoxic
129
-p
CD4+/CD8+ [2]. Skin toxicities are the most prevalent irAEs related to
re
130 immunotherapy. Lichenoid reactions, maculopapular rashes, vitiligo and other
lP
131 autoimmune skin diseases, including bullous disorders, have been reported in the
na
133 Considering that the use of anti-PD-1 anti-PD-L1 antibodies is relatively new, our
ur
134 experience with the management of immune-triggered skin toxicity remains empirical
Jo
135 and occasionally exploratory. In the current manuscript we present our experience
138
139
140
141
142
143
6
144
147 conducted in the name of ENCADO (European Network for Cutaneous ADverse
148 Event to Oncologic drugs) group. For the aims of the study we used the databases of 9
149 oncodermatology units from Greece, France, Italy, Spain and Argentina and searched
150 for psoriasis cases, developing through the treatment course with ICI until the end of
151 December 2019. Inclusion criteria were patients developing psoriasis after anti-PD-1
of
152 or anti-PDL1 treatment with available data on sex, age, psoriasis type(s), type of
ro
153 immunotherapy, primary cancer and the number of ICI doses until the event. We also
154
-p
recorded and analyzed more parameters including personal/family history of the
re
155 disease, psoriasis grading and pruritus , the need for interruption or ICI
lP
157 recorded epidemiologic data of the patients including sex, age, primary cancer,
158 psoriasis subtype(s) and personal/family history of the disease. The severity of
ur
159 psoriasis as well as pruritus were classified using the Common Terminology Criteria
Jo
160 for Adverse Events (CTCAE). Disease severity was classified as grade-1 (<10% Body
161 Surface Area [BSA], mild), grade-2 (10-30% BSA, moderate) and grade-3 (>30%
162 BSA, severe). The type of immunotherapy, the number of ICI doses until the event,
163 the need for interruption or ICI discontinuation, the therapeutic interventions, as well
164 as the treatment outcomes were also recorded. The therapeutic responses to psoriasis
165 treatment were evaluated based on the reduction of BSA and it was graded as no
166 response (<30% BSA reduction of the initial BSA involved), partial response (30%-
167 80% BSA reduction) and excellent response (>80% BSA reduction). Best oncologic
7
171 For continuous variables, we conducted a descriptive analysis calculating the mean
172 and standard deviation and we checked for normality using Kolmogorov–Smirnov
173 test. Also, Mann–Whitney T-test and ANOVA Kruskal–Wallis were used for the
174 comparisons between them. Pairwise comparisons were also conducted. For
175 dichotomous and categorical variables, Pearson X2–test was used for the association
of
176 between the variables. Crude and adjusted odds ratio and corresponding 95%
ro
177 Confidence Intervals (CI) were calculated from univariate and multivariate
178
-p
conditional logistic regression, respectively. A level of a=0.10 was used as a cutoff
re
179 for variable removal in the automated model selection for multivariate logistic
lP
180 regression. A survival analysis with Kaplan – Meier method and Log – Rank test for
na
181 the comparison of time to response to psoriatic treatment were also conducted. All the
182 statistical tests were two–sided and the level of significance was set at a=0.05. Data
ur
183 analysis was carried out using SPSS Statistics for Windows, version 25.0 (IBM-
Jo
184 Corp).
185
186
187
188
189
190
191
8
192
193
194 Results
195 ENCADO members provided data of 115 patients with anti-PD-1/PD-L1 induced
196 psoriasis that were included in the study. The mean (SD) follow-up period after
of
198 Demographic characteristics of the study group are listed in Table-1. Thirty-three out
ro
199 of 115 (30.8%) patients had a personal history of psoriasis; in 20 out of 33 of them the
200 -p
disease was active and clinically apparent upon ICI initiation. Data about family
re
201 history of psoriasis were available in only 93 patients and was positive in 32 of them
lP
202 (34.4%).
203
na
The mean number (SD) of drug dosages until psoriasis onset or exacerbation was 11.2
204 (14.9). In patients with clinically present psoriasis upon ICI initiation, deterioration of
ur
205 the disease was recorded sooner compared to those with no active psoriatic at baseline
Jo
206 (mean number of infusions 5.4 vs 12.2, p<0.05) (Table-2). No dosage differences
207 were noted among anti-PD-1 and anti-PDL1 agents or in regards with any other
208 specific drug of this group. Patients with non small cell lung cancer (NSCLC)
209 developed psoriasis sooner compared to those with melanoma (9.7 vs. 20.8 mean
210 number of infusions), but this was not statistically significant (Figure-1a). The
213 Thirty patients (30/107, 28%) developed additional cutaneous irAEs, namely macular
214 rash (8 patients), vitiligo (7 patients) and lichen planus (5 patients). Another 20.8%
217 Plaque psoriasis was the most commonly diagnosed clinical form (49/115, 42.6%),
219 (26.1%) patients developed simultaneously or subsequently more than one clinical
of
220 subtypes of psoriasis. Nail involvement was recorded in 37/115 cases (32.7%),
ro
221 whereas 8.1% reported symptoms of psoriatic arthritis. Pruritus was present in 67/115
222
-p
(58.3%) patients including 38 (33%) Grade 1, 21 (18.3%) grade 2 and 8 (7%) grade 3
re
223 reactions. No differences were found regarding number of infusions and type of
lP
224 psoriatic lesion (Figure-1b). The mean (SD) BSA affected at visit one was 15.9
225 (17.22) and the worst mean BSA during follow-up was 16.3 (15.9). Figure 2
na
227
Jo
229 Sixty-eight patients (68/115, 59.1%) were solely treated with topical agents, mainly
230 topical steroids (37 patients), followed by calcipotriol plus betamethasone (26
231 patients). Four patients underwent narrow band UVB phototherapy combined with
232 topical steroids (2 patients) or with calcipotriol plus betamethasone (2 patients) and
233 one patient with topical retinoids (tazarotene gel) plus topical steroids
234 Forty-seven patients (40.9%) were treated with both systemic and topical agents
235 (Suppl Table-1). The most common systemic therapy was acitretin monotherapy (21
10
238 20mg/week, 4.3%) and biologics (4 patients, 3.5%). Systemic steroids were
239 prescribed at an initial dose of 25-50 mg/day with a gradual decrease in a mean period
240 of 5.8 weeks. Acitretin was used in combination with topical corticosteroids or/and
242 patients, 23.8%). Four patients were treated with biologic agents, including two with
243 anti-TNFa (adalimumab and infliximab), one with ustekinumab and one with the
of
244 combination of acitretin and guselkumab. Eighteen out of 23 patients treated with
ro
245 acitretin showed either excellent response of psoriasis with complete clearance of the
246
-p
lesions (6/23, 26%) or partial response (12/23, 52.1%). A significant clinical
re
247 improvement (excellent response or partial response) was also reported in all patients
lP
248 treated with apremilast. Interestingly, all patients treated with systemic steroids, also
na
249 showed a positive clinical response. Regarding biologics, partial response was
250 observed in one patient in infliximab and in one patient in ustekinumab, whilst two
ur
251 individuals did not response to adalimumab and guselkumab. The first non responder
Jo
252 had intolerable grade-2 disease presenting with plaques, inverse and guttate lesions
253 whereas palms, soles and nails were also involved. The second non responder had
254 grade-3 plaque psoriasis with nail and palmoplantar involvement also. The Median
255 time (CI) to response to psoriasis treatment was 8 (7.02 – 8.98) months for acitretin,
256 14 (1.16 – 26.8) for apremilast , 6 (4.30 – 7.69) for methotrexate , 3 (1.17 – 4.28) for
257 steroids. Patients treated with systemic steroids showed quicker response to psoriasis
258 treatment and this was the only statistically significant comparison (steroids vs other
261 Overall, 29/112 patients (25.9%) interrupted ICI treatment due to psoriasis, whereas
263 differences regarding the need for oncological treatment discontinuation were
264 recorded between patients treated exclusively with topical regimens and patients
265 treated with systemic therapies (8 patients/12.5% with topical treatment only vs. 12
266 patients/26.7% with systemic treatment, p=0.128). Patients with psoriasis affecting
267 more than 10% of the BSA had 3.6 increased risk for treatment modification
268 (OR=3.64, CI 1.27-10.45, p=0.03) and 6.4 increased risk for treatment interruption
of
269 (OR=6.41, CI 2.40-17.11, p<0.001). The presence of pustular psoriasis also increased
ro
270 the risk of treatment interruption (odds ratio: 4.9, CI 1.41-16.96, p<0.012).
-p
re
271 Antitumoral effect of ICI
lP
272 Overall, 67.7% of the treated patients improved after immunotherapy, by either
273 complete (19/99, 19.2%) or partial (48/99, 48.5%) response. From univariate
na
274 regression, Guttate psoriasis and disease affecting >10 of BSA were found to be
ur
275 positive predictors leading to a 2.73-fold (p=0.05, C.I 0.98 – 7.55) and 2.55-fold
Jo
276 (p=0.03, C.I 1.05 – 6.22) increased probability for response, respectively.
277 Contrariwise, patients with pruritus posed a decreased probability for response and
278 the relationship was statistical significant (OR=0.38, p=0.03, C.I 0.16 – 0.91). From
279 multivariate logistic regression, only severity (positive predictor for response to ICI,
280 OR=3.15, C.I. 1.18-8.41) and pruritus (negative predictor for response to ICI, OR=
282
283
284
12
285
286
287
288
289 Discussion
290 Psoriatic lesions have been reported in various ICI-treated case series
of
291 [7,8,9,10,11,12]. Several studies report that the majority of these patients are
ro
292 complicated with psoriasis on the setting of exacerbation of a preexisting condition
293
-p
[10,13,14,15]. We present the largest to date series of cancer patients treated with
re
294 ICIs, complicated with psoriasis. According to our analysis 70% of the cases were
lP
295 affected by psoriasis de novo with unique disease features. Based on our data, as well
na
298 Our study supports that development of psoriasis occurs later than other skin
299 toxicities. According to recent literature, in patients treated with nivolumab, skin
300 irAEs presented after a median of 5 weeks, while similar results (6.4 weeks) were
301 reported in lung cancer studies [19,20]. Phillips et al, in a large study of 285 patients
302 with cutaneous adverse reactions also reported that psoriasiform rashes presented in a
303 median time of 61 days for 21/285 patients with recorded skin toxicities [21]. In our
304 population, the minimum period until psoriasis diagnosis was significantly longer
305 than the aforementioned (12 weeks). Nevertheless, there was a noticeable difference
306 between the group of patients with a prior psoriasis history vs. the de novo group,
13
307 with the former being affected significantly earlier. For both groups, close skin
308 surveillance and early expert consultation should be part of standard care.
309 Regarding response to immunotherapy, our rates were quite positive reaching overall
310 response rates of 67.7%. Previous studies on melanoma and lung cancer patient have
311 supported that skin reactions are analogically related to therapeutic efficacy
312 [22,23]. Our study, mainly based on melanoma and lung cancer patients as well, is in
313 line with latter observation. Moreover, in our study the presence of guttate lesions, as
of
314 well as psoriasis affecting >10% of body surface area were both associated with better
ro
315 response rates to immunotherapy in comparison with other types of psoriasis and mild
316 -p
symptomatology, respectively. On the contrary, the presence of pruritus was a
317
re
negative predictor. However, larger prospective studies are needed in order to safely
319 Apart from the epidemiological and disease characteristics, in our cohort we
na
320 attempted to focus on therapeutic aspects of this novel type of irAE. The most
ur
321 common systemic agent administered was acitretin. Acitretin, opposed to many other
Jo
322 anti-psoriatic drugs, does not harbor immunosuppressive properties, and in this
323 context it does not interfere with ICI treatment [24]. In our sample, response rates
324 were satisfactory and the treatment was well tolerated, with no unpredicted adverse
325 events and therefore, we recommend acitretin as one of the first-line options in ICI-
327 Our study included 7 individuals with moderate to severe psoriasis successfully
328 managed with apremilast. All patients responded well, with 2 out of 7 achieving
330 apremilast is that, based on the summary of product characteristics (SmPC) of the
14
333 overlooking the weaknesses of a small sample, we believe that apremilast may
335 Methotrexate was administered in only 5 patients in our group with all of them
336 achieving excellent and partial responses. Recent psoriasis guidelines from the French
337 group reviewed the risk of cancer associated with systemic therapies and reported that
of
338 there does not seem to be an increased risk for cancer in patients treated with
ro
339 methotrexate except for a possible increase of non-melanoma skin cancer [8].
340 -p
However, further studies are needed, to obtain solid evidence.
re
341 The role of systemic steroids in the treatment of psoriasis is controversial. Ιn our daily
lP
342 clinical practice we avoid systemic steroids, since it is well-known that rapid
343 withdrawal of systemic steroids may lead to rebound phenomenon. To date, only one
na
344 case report has been published, reporting a patient with pembrolizumab induced
ur
345 psoriatic dermatitis that was successfully treated with systemic steroids, without
Jo
346 exacerbation after discontinuation [27]. In our cohort, 8 individuals were managed
347 with systemic steroids, resulting in satisfactory response. However, since systemic
348 steroids might intervene with ICI treatment, in 5 out of 8 cases immunotherapy was
349 discontinued. Despite our small sample, and based on the relatively large number of
352 Our experience with biologics was limited, since these agents are not recommended in
353 patients with active malignancies. However, a recent meta-analysis of nine studies,
354 assessing the risk of carcinogenesis or cancer recurrence in 11679 patients exposed
15
356 detect any increased risk for the former [28]. Moreover the “prophylactic”
358 capable of resolving several shortcomings of the latter, while retaining its anti-tumor
359 efficacy [29]. Phillips et al. reported that IL-23a and IL-12/23 inhibition might be
360 beneficial in patients with steroid refractory irAEs, whereas antitumoral response was
361 maintained in two patients treated with guselkumab and ustekinumab respectively
362 [21]. On the other hand, Esfahani K et al, supported loss of anti-tumor efficacy in a
of
363 patient with pembrolizumab-induced psoriasis treated with secukinumab [30]. In our
ro
364 study, four patients were treated with biological agents, of which two responded well
365
-p
and the other two had to discontinue immunotherapy because of side effects and
re
366 despite clinical improvement. Given the increased disease severity, response rates
lP
367 were satisfying, however the decision for biological therapy should be always
na
368 individualized.
ur
370 Recent studies, in groups with pre-existing autoimmune disease, have shown that the
371 median progression free survival (PFS) was shorter in patients receiving
372 immunosuppressive therapy at ICI initiation [16]. Although there were differences
373 with our study, related to the initiation time of immunomodulators our results indicate
374 that coadministration of systemic treatment for psoriasis and ICI did not effect the
375 overall therapeutic effect of the latter. This indicates that treatment of psoriasis is of
376 great clinical value even by the use of systemic therapies, which do not seem to
378 Some of the major limitations of our study are due to its retrospective nature.
379 Significant parameters, like PASI scores, cannot therefore be evaluated, resulting in
16
380 missing data for a more accurate for patient’s categorization. Future prospective
381 studies, including pathology data, will advocate further analysis and better
383 In conclusion, we present the largest –to our knowledge– series of patients
384 complicated with psoriasis during anti-PD1/L1 treatment. The clinical characteristics
385 of this entity are highly diverse. Early diagnosis and adequate management with
386 agents that do not interfere with immunotherapy, or with the underlying malignancy,
of
387 is crucial. In this context, we introduce a practical therapeutic algorithm that we
ro
388 consider useful for both, dermatologists and oncologists. In most cases, ICI treatment
389 -p
can be completed, as long as strict dermatological surveillance is present throughout
re
390 treatment.
lP
391
na
392
ur
393
Jo
394
395
396
397
398
399
400
401
17
402
403
404
405
406
407
of
408
ro
409 Abstract
-p
410 Background: Immune checkpoint inhibitors (ICIs)-mediated psoriasis poses
re
411 significant diagnostic and therapeutic challenges.
lP
412
na
Objective: To report data on ICI-mediated psoriasis, emerging from the largest so far
414 Μethods: The medical records of all patients with ICI-mediated psoriasis were
416 Results: We included a cohort of 115 individuals. Grade-1, 2 and 3 disease severity
417 was reported in 60/105 (57.1%, 10 missing data), 34/105 (32.4%) and 11/105
418 (10.5%), respectively. The ratio between de novo and exacerbation cases of psoriasis
419 was 21/90 (23.3%). The most common systemic therapy was acitretin (23 patients,
420 20.1%), followed by systemic steroids (8 patients, 7%), apremilast (7 patients, 6.1%),
421 methotrexate (5 patients, 4.3%) and biologics (4 patients, 3.6%). Overall, 29/112
422 patients (25.9%) interrupted and 20/111 (18%) permanently discontinued ICIs due to
18
423 psoriasis. BSA>10% at baseline had a 3.6 increased risk for ICIs treatment
424 modification (OR=3.64, CI 1.27-10.45, p=0.03) and a 6.4 increased risk for
426 grade2/3 disease were significant positive predictors for antitumor response of ICI
428
430
of
431 Conclusion: Acitretin, apremilast and methotrexate are safe and effective modalities
ro
432 for ICI-mediated psoriasis. In most cases, ICI can be completed unhindered. A
435
na
436
437
ur
438
Jo
439
440
441
442
443
444
445
446
447
19
448
449
450
451
452
453
454
455
of
456 Introduction
ro
457 Immune checkpoint inhibitors (ICIs) are monoclonal antibodies targeting cytotoxic T
458
-p
lymphocyte-associated antigen-4 (CTLA-4), programmed cell death protein-1 (PD-1)
re
459 or programmed death ligand-1 (PD-L1). Their addition in our armamentarium has
lP
460 radically transformed the therapeutic arena in oncology, providing the potential of
na
462 The introduction of ICIs has resulted in the recognition of a new spectrum of
ur
463 immune-related adverse events’ (irAEs) that may occasionally limit their use [1]. The
Jo
464 nature of irAEs has not been completely elucidated, however it is meaningful that
465 they are mostly immune-mediated, resulting by the T-cell activation of cytotoxic
466 CD4+/CD8+ [2]. Skin toxicities are the most prevalent irAEs related to
468 autoimmune skin diseases, including bullous disorders, have been reported in the
470 Considering that the use of anti-PD-1 anti-PD-L1 antibodies is relatively new, our
471 experience with the management of immune-triggered skin toxicity remains empirical
472 and occasionally exploratory. In the current manuscript we present our experience
20
475
476
477
478
479
480
of
481
ro
482 Patients and methods
483
-p
This was a multi-centric retrospective study of psoriasis related to immunotherapy,
re
484 conducted in the name of ENCADO (European Network for Cutaneous ADverse
lP
485 Event to Oncologic drugs) group. For the aims of the study we used the databases of 9
na
486 oncodermatology units from Greece, France, Italy, Spain and Argentina and searched
487 for psoriasis cases, developing through the treatment course with ICI until the end of
ur
488 December 2019. Inclusion criteria were patients developing psoriasis after anti-PD-1
Jo
489 or anti-PDL1 treatment with available data on sex, age, psoriasis type(s), type of
490 immunotherapy, primary cancer and the number of ICI doses until the event. We also
491 recorded and analyzed more parameters, including personal/family history of the
492 disease, psoriasis grading and pruritus , the need for interruption or ICI
493 discontinuation, the therapeutic interventions, as well as the treatment outcomes. The
494 severity of psoriasis as well as pruritus were classified using the Common
495 Terminology Criteria for Adverse Events (CTCAE). Disease severity was classified
496 as grade-1 (<10% Body Surface Area [BSA], mild), grade-2 (10-30% BSA,
497 moderate) and grade-3 (>30% BSA, severe). The type of immunotherapy, the number
21
498 of ICI doses until the event, the need for interruption or ICI discontinuation, the
499 therapeutic interventions, as well as the treatment outcomes were also recorded. The
500 therapeutic responses to psoriasis treatment were evaluated based on the reduction of
501 BSA and it was graded as no response (<30% BSA reduction of the initial BSA
502 involved), partial response (30%-80% BSA reduction) and excellent response (>80%
503 BSA reduction). Best oncologic response to immunotherapy was recorded and graded
504 according to Response Evaluation Criteria in Solid Tumors (RECIST) criteria [6].
505
of
506 Statistical analysis
ro
507 For continuous variables, we conducted a descriptive analysis calculating the mean
508
-p
and standard deviation and we checked for normality using Kolmogorov–Smirnov
re
509 test. Also, Mann–Whitney T-test and ANOVA Kruskal–Wallis were used for the
lP
510 comparisons between them. Pairwise comparisons were also conducted. For
na
511 dichotomous and categorical variables, Pearson X2–test was used for the association
512 between the variables. Crude and adjusted odds ratio and corresponding 95%
ur
513 Confidence Intervals (CI) were calculated from univariate and multivariate
Jo
514 conditional logistic regression, respectively. A level of a=0.10 was used as a cutoff
515 for variable removal in the automated model selection for multivariate logistic
516 regression. A survival analysis with Kaplan – Meier method and Log – Rank test for
517 the comparison of time to response to psoriatic treatment were also conducted. All the
518 statistical tests were two–sided and the level of significance was set at a=0.05. Data
519 analysis was carried out using SPSS Statistics for Windows, version 25.0 (IBM-
520 Corp).
521
522
22
523
524
525
526
527
528
of
529
ro
530
-p
re
531 Results
lP
532 ENCADO members provided data of 115 patients with anti-PD-1/PD-L1 induced
na
533 psoriasis that were included in the study. The mean (SD) follow-up period after
535 Demographic characteristics of the study group are listed in Table-1. Thirty-three out
536 of 115 (28.7%) patients had a personal history of psoriasis; in 20 out of 33 of them the
537 disease was active and clinically apparent upon ICI initiation. Data about family
538 history of psoriasis were available in only 93 patients and was positive in 32 of them
539 (34.4%).
540 The mean number (SD) )of drug dosages until psoriasis onset or exacerbation was
541 11.2 (14.9). In patients with clinically present psoriasis upon ICI initiation,
542 deterioration of the disease was recorded sooner compared to those with no active
543 psoriatic at baseline (mean number of infusions 5.4 vs 12.2, p=0.01 (Table-2). No
544 dosage differences were noted among anti-PD-1 and anti-PDL1 agents or in regards
23
545 with any other specific drug of this group. Patients with non small cell lung cancer
546 (NSCLC) developed psoriasis sooner compared to those with melanoma (9.7 vs. 20.8
547 mean number of infusions), but this was not statistically significant (Figure-1a). The
550 Thirty patients (30/107, 28%) developed additional cutaneous irAEs, namely macular
551 rash (8 patients), vitiligo (7 patients) and lichen planus (5 patients). Another 20.8%
of
552 (20/96) of cases were complicated with systemic irAEs.
ro
553 Clinical characteristics
554
-p
Plaque psoriasis was the most commonly diagnosed clinical form (49/115, 42.6%),
re
555 followed by the palmoplantar (14/115, 12.2%) (Table-1). Interestingly, 30/115
lP
556 (26.1%) patients developed simultaneously or subsequently more than one clinical
na
557 subtypes of psoriasis. Nail involvement was recorded in 37/115 cases (32.7%),
558 whereas 8.1% reported symptoms of psoriatic arthritis. Pruritus was present in 67/115
ur
559 (58.3%) patients including 38 (33%) Grade 1, 21 (18.3%) grade 2 and 8 (7%) grade 3
Jo
560 reactions. No differences were found regarding number of infusions and type of
561 psoriatic lesion (Figure-1b). The mean (SD) BSA affected at visit one was 15.9 (7.22)
562 and the worst mean (SD) BSA during follow-up was 16.3 (15.9). Figure 2 illustrates
564
566 Sixty-eight patients (68/115, 59.1%) were solely treated with topical agents, mainly
567 topical steroids (37 patients), followed by calcipotriol plus betamethasone (26
24
568 patients). Four patients underwent narrow band UVB phototherapy combined with
569 topical steroids (2 patients) or with calcipotriol plus betamethasone (2 patients) and
570 one patient with topical retinoids (tazarotene gel) plus topical steroids.
571 Forty-seven patients (40.9%) were treated with both systemic and topical agents
572 (Suppl Table-1). The most common systemic therapy was acitretin monotherapy (21
of
575 20mg/week, 4.3%) and biologics (4 patients, 3.5%). Systemic steroids were
ro
576 prescribed at an initial dose of 25-50 mg/day with a gradual decrease in a mean period
577 -p
of 5.8 weeks. Acitretin was used in combination with topical corticosteroids or/and
578
re
calcipotriol (16 patients, 76.2%), or in combination with phototherapy (UVB-NB, 5
579 patients, 23.8%). Four patients were treated with biologic agents, including two with
lP
580 anti-TNFa (adalimumab and infliximab), one with ustekinumab and one with the
na
581 combination of acitretin and guselkumab. Eighteen out of 23 patients treated with
ur
582 acitretin showed either excellent response of psoriasis with complete clearance of the
Jo
583 lesions (6/23, 26%) or partial response (12/23, 52.1%). A significant clinical
584 improvement (excellent response or partial response) was also reported in all patients
585 treated with apremilast. Interestingly, all patients treated with systemic steroids, also
586 showed a positive clinical response. Regarding biologics, partial response was
587 observed in one patient in infliximab and in one patient in ustekinumab, whilst two
588 individuals did not response to adalimumab and guselkumab. The first non responder
589 had intolerable grade-2 disease presenting with plaques, inverse and guttate lesions
590 whereas palms, soles and nails were also involved. The second non responder had
591 grade-3 plaque psoriasis with nail and palmoplantar involvement also. The median
592 time (CI) to response to psoriasis treatment was 8 (7.02 – 8.98) months for acitretin,
25
593 14 (1.16 – 26.8) for apremilast, 6 (4.30 – 7.69) for methotrexate, 3 (1.17 – 4.28) for
594 steroids. Patients treated with systemic steroids showed quicker response to psoriasis
595 treatment and this was the only statistically significant comparison (steroids vs other
598 Overall, 29/112 patients (25.9%) interrupted ICI treatment due to psoriasis, whereas
of
600 differences regarding the need for oncological treatment discontinuation were
ro
601 recorded between patients treated exclusively with topical regimens and patients
602
-p
treated with systemic therapies (8 patients/12.5% with topical treatment only vs. 12
re
603 patients/26.7% with systemic treatment, p=0.128). Patients with psoriasis affecting
lP
604 more than 10% of the BSA had 3.6 increased risk for treatment modification
605 (OR=3.64, CI 1.27-10.45, p=0.03) and 6.4 increased risk for treatment interruption
na
606 (OR=6.41, CI 2.40-17.11, p<0.001). The presence of pustular psoriasis also increased
ur
607 the risk of treatment interruption (odds ratio: 4.9, CI 1.41-16.96, p<0.012).
Jo
609 Overall, 67.7% of the treated patients improved after immunotherapy, by either
610 complete (19/99, 19.2%) or partial (48/99, 48.5%) response. From univariate
611 regression, Guttate psoriasis and disease affecting >10 of BSA were found to be
612 positive predictors leading to a 2.73-fold (p=0.05, C.I 0.98 – 7.55) and 2.55-fold
613 (p=0.03, C.I 1.05 – 6.22) increased probability for response, respectively.
614 Contrariwise, patients with pruritus posed a decreased probability for response and
615 the relationship was statistical significant (OR=0.38, p=0.03 , C.I 0.16 – 0.91). From
616 multivariate logistic regression, only severity (positive predictor for response to ICI,
26
617 OR=3.15, C.I. 1.18-8.41) and pruritus (negative predictor for response to ICI, OR=
619
620
621
622
623
of
ro
624
625
-p
re
lP
626
na
627 Discussion
ur
628 Psoriatic lesions have been reported in various ICI-treated case series
Jo
629 [7,8,9,10,11,12]. Several studies report that the majority of these patients are
631 [10,13,14,15]. We present the largest to date series of cancer patients treated with
632 ICIs, complicated with psoriasis. According to our analysis 70% of the cases were
633 affected by psoriasis de novo with unique disease features. Based on our data, as well
636 Psoriasis exacerbation and psoriasiform rashes have been reported in various case
637 series [7,8,9,10,11,12]. However, the precise clinical characteristics of this novel
638 toxicity and its effect on the overall course of immunotherapy are vastly unknown.
27
641 setting, based upon our experience, as well as current European guidelines [13,14,15].
642 The safety of ICIs in individuals with pre-existing autoimmune diseases has gained
643 great attention. Previous studies report that the majority of patients treated with ICIs
645 condition [10,16,17,18]. Our data significantly differ since approximately 70% of the
of
646 studied population was affected by psoriasis de novo. Given the retrospective nature
ro
647 of our study this fact should be evaluated accordingly, however, the increasing ratio
648 -p
of de novo cases, in combination with the unique disease phenotype, suggests that a
re
649 multidisciplinary approach towards optimization of cutaneous toxicities management
lP
651 dermatologist.
na
652 Our study supports that development of psoriasis occurs later than other skin
ur
653 toxicities. According to recent literature, in patients treated with nivolumab, skin
Jo
654 irAEs presented after a median of 5 weeks, while similar results (6.4 weeks) were
655 reported in lung cancer studies [19,20]. Phillips et al, in a large study of 285 patients
656 with cutaneous adverse reactions also reported that psoriasiform rashes presented in a
657 median time of 61 days for 21/285 patients with recorded skin toxicities [21]. In our
658 population, the minimum period until psoriasis diagnosis was significantly longer
659 than the aforementioned (12 weeks). Nevertheless, there was a noticeable difference
660 between the group of patients with a prior psoriasis history vs. the de novo group,
661 with the former being affected significantly earlier. For both groups, close skin
662 surveillance and early expert consultation should be part of standard care.
28
663 Regarding response to immunotherapy, our rates were quite positive reaching overall
664 response rates of 67.7%. Previous studies on melanoma and lung cancer patient have
665 supported that skin reactions are analogically related to therapeutic efficacy
666 [22,23]. Our study, mainly based on melanoma and lung cancer patients as well, is in
667 line with latter observation. Moreover, in our study the presence of guttate lesions, as
668 well as psoriasis affecting >10% of body surface area were both associated with better
669 response rates to immunotherapy in comparison with other types of psoriasis and mild
of
671 negative predictor. However, larger prospective studies are needed in order to safely
ro
672 confirm if these clinical signs could be viable prognostic factors.
-p
673
re
Apart from the epidemiological and disease characteristics, in our cohort we
674 attempted to focus on therapeutic aspects of this novel type of irAE. The most
lP
675 common systemic agent administered was acitretin. Acitretin, opposed to many other
na
676 anti-psoriatic drugs, does not harbor immunosuppressive properties, and in this
ur
677 context it does not interfere with ICI treatment [24]. In our sample, response rates
Jo
678 were satisfactory and the treatment was well tolerated, with no unpredicted adverse
679 events and therefore, we recommend acitretin as one of the first-line options in ICI-
681 Our study included 7 individuals with moderate to severe psoriasis successfully
682 managed with apremilast. All patients responded well, with 2 out of 7 achieving
684 apremilast is that, based on the summary of product characteristics (SmPC) of the
687 overlooking the weaknesses of a small sample, we believe that apremilast may
689 Methotrexate was administered in only 5 patients in our group with all of them
690 achieving excellent and partial responses. Recent psoriasis guidelines from the French
691 group reviewed the risk of cancer associated with systemic therapies and reported that
692 there does not seem to be an increased risk for cancer in patients treated with
693 methotrexate except for a possible increase of non-melanoma skin cancer [8].
of
694 However, further studies are needed, to obtain solid evidence.
ro
695 The role of systemic steroids in the treatment of psoriasis is controversial. Ιn our daily
696
-p
clinical practice we avoid systemic steroids, since it is well-known that rapid
re
697 withdrawal of systemic steroids may lead to rebound phenomenon. To date, only one
lP
698 case report has been published, reporting a patient with pembrolizumab induced
699 psoriatic dermatitis that was successfully treated with systemic steroids, without
na
700 exacerbation after discontinuation [27]. In our cohort, 8 individuals were managed
ur
701 with systemic steroids, resulting in satisfactory response. However, since systemic
Jo
702 steroids might intervene with ICI treatment, in 5 out of 8 cases immunotherapy was
703 discontinued. Despite our small sample, and based on the relatively large number of
706 Our experience with biologics was limited, since these agents are not recommended in
707 patients with active malignancies. However, a recent meta-analysis of nine studies,
708 assessing the risk of carcinogenesis or cancer recurrence in 11679 patients exposed
710 detect any increased risk for the former [28]. Moreover the “prophylactic”
30
712 capable of resolving several shortcomings of the latter, while retaining its anti-tumor
713 efficacy [29]. Phillips et al. reported that IL-23a and IL-12/23 inhibition might be
714 beneficial in patients with steroid refractory irAEs, whereas antitumoral response was
715 maintained in two patients treated with guselkumab and ustekinumab respectively
716 [21]. On the other hand, Esfahani K et al, supported loss of anti-tumor efficacy in a
717 patient with pembrolizumab-induced psoriasis treated with secukinumab [30]. In our
718 study, four patients were treated with biological agents, of which two responded well
of
719 and the other two had to discontinue immunotherapy because of side effects and
ro
720 despite clinical improvement. Given the increased disease severity, response rates
721
-p
were satisfying, however the decision for biological therapy should be always
re
722 individualized.
lP
724 Recent studies, in groups with pre-existing autoimmune disease, have shown that the
ur
725 median progression free survival (PFS) was shorter in patients receiving
Jo
726 immunosuppressive therapy at ICI initiation [16]. Although there were differences
727 with our study, related to the initiation time of immunomodulators our results indicate
728 that coadministration of systemic treatment for psoriasis and ICI did not effect the
729 overall therapeutic effect of the latter. This indicates that treatment of psoriasis is of
730 great clinical value even by the use of systemic therapies, which do not seem to
732 Some of the major limitations of our study are due to its retrospective nature.
733 Significant parameters, like PASI scores, cannot therefore be evaluated, resulting in
734 missing data for a more accurate for patient’s categorization. Future prospective
31
735 studies, including pathology data, will advocate further analysis and better
737 In conclusion, we present the largest –to our knowledge– series of patients
738 complicated with psoriasis during anti-PD1/L1 treatment. The clinical characteristics
739 of this entity are highly diverse. Early diagnosis and adequate management with
740 agents that do not interfere with immunotherapy, or with the underlying malignancy,
of
742 consider useful for both, dermatologists and oncologists. In most cases, ICI treatment
ro
743 can be completed, as long as strict dermatological surveillance is present throughout
744 treatment. -p
re
745
lP
746
na
747
749
Jo
768
769
770
771
772
773
774
775
776
777
778
779
780
781
of
782
783
ro
784
785
786
787
-p
re
788
789
lP
790
791
792
na
793
794 #
Patients treated for melanoma are excluded
ur
795
796 * Patients with active psoriasis under systemic treatment can start ICI under close follow-up:
797 -Patients under Acitretin, Apremilast or Methotrexate can start ICI under close surveillance
Jo
798 -Patients under cyclosporine should be referred to dermatology expert for treatment modification
799 -Patients under biologics should be discussed case by case in a multidisciplinary approach with oncologists
800
801
802
803 **Special indications for systemic treatment
804 - Involvement of sensitive areas such as face, hands, genitals, nails, scalp
805 - Disseminated lesions
806 - Mild to moderate psoriatic arthritis
807 - DLQI > 10
808 - Severe symptoms (eg itch, burning)
809 - Intolerable Grade 2
810
811 *** Special indications for ICI cessation or withholidng
812 Grade 3 (BSA >30)
813 Erythrodermic psoriasis: Consider introducing systemic corticosteroids (prednisolone 0.2-0.5 mg/kgr)
814 Generalized Pustular psoriasis: Consider introducing systemic corticosteroids (prednisolone 0.2-0.5 mg/kgr)
815 Severe symptoms (eg itch, burning)
816
817 **** Topical treatments
818 - Topical steroids
819 - Calcipotriol
820 - Tazarotene
821 - UVB-NB
822
823
824 ***** Discuss with cancer specialists on a case by case basis
33
825
826
827
828
829
830
831
832
833
834
835
836
837
838
of
839
840
ro
841
842
843
844
-p
re
845
846
lP
847
848
849
na
850
851
ur
852 Table 1: Epidemiologic and clinical characteristics at baseline and therapeutic intervention
853
Jo
854
Total number (n)=115 (%)
Sex
Male 88 (76.5)
Female 27 (23.5)
Mean age (SD)
Male 68.5 (9.3)
Female 63.7 (11.8)
Primary cancer
NSCLC 69 (60)
Melanoma 17 (14.8)
Head & Neck SCC 6 (5.2)
Renal Cell Carcinoma 6 (5.2)
Urothelial Carcinoma 6 (5.2)
Hodgkin’s Lymphoma 2 (1.7)
Merkel Cell Carcinoma 1 (0.9)
Hepatocellular Carcinoma 3 (2.6)
Gastric Cancer 2 (1.7)
Mesothelioma 1 (0.9)
Ovarian Cancer 1 (0.9)
Pulmonary Neuroendocrine Cancer 1 (0.9)
Immune Checkpoint Inhibitor
Anti-PD1: 99 (86.1)
34
Nivolumab 68 (59.2)
Pembrolizumab 30 (26.1)
Spartalizumab 1 (0.8)
Anti-PDL1: 16 (13.9)
Durvalumab 6 (5.2)
Atezolizumab 9 (7.9)
Avelumab 1 (0.8)
Anti-PD-1/PD-L1 Monotherapy
Yes 100
No: 15
Ipilimumab 2
Cabozantinib 1
Pazopanib 1
Bevacizumab 1
Campetinib 1
Anti-LAG-3 1
Chemotherapy (Platinum base) 6
Dabrafenib 2
of
Psoriasis type
Plaque psoriasis 49 (42.6)
ro
Pustular psoriasis 8 (7)
Palmoplantar psoriasis 14 (12.2)
Guttate psoriasis 8 (7)
Nail psoriasis
Inverse psoriasis
-p2 (1.7)
1 (0.9)
re
Erythrodermic psoriasis 3 (2.6)
> 1 type: 30 (26.1)
Plaque & palmoplantar psoriasis 11
lP
865
866
867
868
869
870
871
872
873
874
875
876
877
878
of
879
880
ro
881
882
883
884
-p
re
885
886
lP
887
888
889
na
890
891 Table 2: Number of infusions until psoriasis
ur
892
Number of infusions (Mean, SD) P- value ₸
Sex 0.025
Jo
of
896
897
ro
898
899
900
901
-p
re
902
903
lP
904
905
906
na
907
908
909
ur
919
920
921
922
923
924
925
of
926
ro
927
928
-p
re
929
lP
930
na
931
ur
938 ENCADO: European Network for Cutaneous ADverse Event to Oncologic drugs
947
948
949
950
of
951
ro
952
953
-p
re
954
lP
955
na
956
957 References
ur
Jo
1
O'Reilly A, Hughes P, Mann J, et al. An immunotherapy survivor population:
health-related quality of life and toxicity in patients with metastatic melanoma treated
effects with shared cancer and tissue antigens in non-small cell lung cancer. JAMA
5
Sibaud V. Dermatologic reactions to immune checkpoint inhibitors: skin toxicity
6
Schwartz LH, Litiere S, De Vries E et al. RECIST 1.1- Update and clarification:
of
Probl Cnacer 2017; 41:407-412.
ro
9
Ruiz-Banobre J, Perez-Pampin E, Garcia-Gonzalea J et al. Development of psoriatic
-p
arthritis during nivolumab therapy for metastatic non-small cell lung cancer, clinical
re
outcome analysis and review of the literature. Lung Cancer 2017;108: 217-221.
lP
10
Guven DC, Kilickap S, Guner G, Taban H, Dizdar O. Development of de novo
na
psoriasis during nivolumab therapy in a patient with small cell lung cancer. J Oncol
11
Chujo S, Asahina A, Itoh Y, et al. New onset psoriasis during nivolumab treatment
Jo
during treatment with a PD-1 checkpoint inhibitor: case report and literature review.
14
Scarfi F, Lacava R, Patrizi A, et al. Follicular psoriasis induced by pembrolizumab
in a patient with advanced non-small-cell lung cancer. Int J Dermatol 2019; 58:e151-
e152.
15
Matsumura N, Ohtsuka M, Kikuchi N, et al. Exacerbation of psoriasis during
nivolumab therapy for metastatic melanoma. Acta Derm Venereol 2016; 96:259-260.
16
Nast A, Gisondi P, Ormerod AD, et al. European S3-Guidelines on the systemic
of
EADV and IPC. J Eur Acad Dermatol 2015; 29:2277-2294.
ro
17
Amatore F, Villani AP, Tauber M et al. French guidelines on the use of systemic
-p
treatments for moderate-to-severe psoriasis in adults. J Eur Acad Dermatol Venereol
re
2019; 33:464-483.
lP
18
Gisondi P, Altomare G, Ayala F et al. Italian guidelines on the systemic treatments
790.
ur
19
Weber JS, Hodi FS, Wolchock JD et al. Safety profile of nivolumab monotherapy:
Jo
a pooled analysis of patients with advanced melanoma. J Clin Oncol 2017; 35:785-
792.
20
Aso M, Toi Y, Sugisaka J et al. Association between skin reaction and clinical
23
Chan L, Hwang SJE, Byth K, et al Survival and prognosis of individuals receiving
programmed cell death 1 inhibitor with and without immunologv cutaneous adverse
of
110:107-109.
ro
26
Apalla Z, Psarakis E, Lallas A, et al. Psoriasis in Patients With Active Lung
-p
Cancer: Is Apremilast a Safe Option? Dermatol Pract Concept 2019; 9: 300-1.
re
27
Suzuki M, Matsumoto S, Takeda Y, Sugiyama H. Systemic psoriasiform dermatitis
lP
appeared after the administration of pembrolizumab. Intern Med. 2019; Nov 29 [Epub
ahead of print]
na
28
Micic D, Komaki Y, Alavanja A, et al. Risk of cancer recurrence among
ur
dual PD-1 and CTLA-4 immunotherapy efficacy and toxicity. Cell Stress 2019;
3:236-239.
30
Esfahani K, Miller Jr WH. Reversal of autoimmune toxicity and loss of tumor
of
ro
-p
re
lP
Figure legends
na
Figure 1 a. Boxplot of Number of doses to Primary cancer. There was not a statistical
ur
significant difference between the number of doses for adverse event and the primary
Jo
doses to Type of psoriasis. There was not a statistical significant difference between
the number of doses for adverse event and the type of psoriasis (ANOVA p=0.09,
of
ro
-p
re
lP
na
ur
Jo