Journal of Dermatological Treatment

ISSN: 0954-6634 (Print) 1471-1753 (Online) Journal homepage: https://www.tandfonline.com/loi/ijdt20

Targeted therapies for patients with moderate-to-

severe psoriasis: a systematic review and network
meta-analysis of PASI response at 1 year

Najeeda Yasmeen, Laura M. Sawyer, Kinga Malottki, Lars-Åke Levin, Eydna

Didriksen Apol & Gregor B. Jemec

To cite this article: Najeeda Yasmeen, Laura M. Sawyer, Kinga Malottki, Lars-Åke Levin, Eydna
Didriksen Apol & Gregor B. Jemec (2020): Targeted therapies for patients with moderate-to-severe
psoriasis: a systematic review and network meta-analysis of PASI response at 1 year, Journal of
Dermatological Treatment, DOI: 10.1080/09546634.2020.1743811

To link to this article: https://doi.org/10.1080/09546634.2020.1743811

© 2020 The Author(s). Published with View supplementary material

license by Taylor & Francis Group, LLC

Accepted author version posted online: 23 Submit your article to this journal
Mar 2020.
Published online: 02 Apr 2020.

Article views: 1228 View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ijdt20
JOURNAL OF DERMATOLOGICAL TREATMENT
https://doi.org/10.1080/09546634.2020.1743811

ORIGINAL ARTICLE

Targeted therapies for patients with moderate-to-severe psoriasis: a systematic

review and network meta-analysis of PASI response at 1 year
Najeeda Yasmeena, Laura M. Sawyera, Kinga Malottkia, Lars-Åke Levinb, Eydna Didriksen Apolc and
Gregor B. Jemecd
a
Symmetron Limited, London, United Kingdom; bInstitute of Health and Medicine, University of Linko€ping, Link€oping, Sweden; cLEO
Pharma A/S, Ballerup, Denmark; dDepartment of Dermatology, Sjaellands Universitetshospital, Roskilde, Denmark

ABSTRACT ARTICLE HISTORY

Purpose: To compare PASI outcomes of approved biologics and apremilast after 1 year of treatment. Received 4 November 2019
Methods: A systematic review identified RCTs and long-term extensions reporting PASI 75, 90, and Accepted 7 February 2020
100 responses in adults with moderate-to-severe psoriasis. Data for analysis were modeled using a
Bayesian multinomial likelihood model with probit link. KEYWORDS
Psoriasis; biological therapy;
Results: Twenty-eight studies reporting PASI responses were included in the network meta-analysis. network meta-analysis;
Differences in study design led to a stepwise approach to synthesis, consisting of two analyses. The systematic literature review
primary analysis included nine RCTs investigating comparative efficacy at 1 year. Results indicated
risankizumab, brodalumab, and guselkumab were the most effective therapies, followed by ixekizu-
mab and secukinumab; all demonstrated superiority to ustekinumab and etanercept. The secondary
analysis extended the primary analysis with 19 further studies comparing active interventions to pla-
cebo outcomes extrapolated from induction. The interventions generating the highest PASI response
were the same as the primary analysis. These therapies were more effective than apremilast, ustekinu-
mab, adalimumab, certolizumab, etanercept, and infliximab.
Conclusions: This NMA demonstrated that evaluated IL-17 and IL-23 inhibitors outperformed other
biological therapies after 1 year. Risankizumab had a higher probability of achieving PASI outcomes
than all other biologics, except brodalumab and guselkumab, where no significant difference could
be concluded.

Introduction design compares continuous versus interrupted therapy, rather

Psoriasis is a chronic inflammatory systemic disorder which than the comparative efficacy of maintained interventions.
For clinical decision making, the most useful studies are
affects approximately 100 million people worldwide (1). Psoriasis
those that compare interventions head to head from baseline
symptoms can present on both skin and nails in varying severi-
through to a long-term endpoint, though there are not many of
ties and a quarter of patients suffer from moderate-to-severe
disease. The most characteristic skin lesions are localized, these, especially for older treatments. However, most clinical
demarcated, red papules, and plaques often covered with white studies focus on short-term induction periods (7).
In recent years, the number of treatments available for mod-
or silver scales (2).
Due to the chronic nature of the disease, most patients erate-to-severe psoriasis has increased, enabling dermatologists
require long-term therapy (3). In clinical trials for psoriasis there to optimize therapy for patients to achieve better outcomes (8);
are two defined treatment periods: a short 10- to 16-week however, choosing between therapies can be challenging.
induction phase, followed by maintenance phase therapy (4). Numerous network meta-analyses (NMAs) that have investigated
Studies evaluating treatments beyond induction use a variety of psoriasis therapies over the first 10 to 16 weeks of treatment
different study designs (5,6). In many trials, some or all patients have been published (9–14) and yet only two reviews have
from the induction phase placebo arm switch to the active ther- quantitatively assessed the long-term efficacy of psoriasis thera-
apy and can no longer provide a suitable comparison for the pies: one at 6 months (15) and another at 1 year (16). Since
patients receiving a licensed dose throughout the induction and these studies were published, four new treatments have been
maintenance phases. The maintenance phases of these trials are licensed for psoriasis – certolizumab pegol, guselkumab, tildraki-
usually non-randomized and may be considered single-arm zumab, and risankizumab. In addition, new long-term studies of
studies. In another group of trials, referred to as randomized previously approved biologics have been published (17). As new
withdrawal trials, treatment responders from one or more induc- therapies become licensed and new evidence emerges, it is
tion phase arms are re-randomized to continue active therapy important to evaluate the available therapies to inform treat-
or to receive placebo. Therefore, such a randomized withdrawal ment decisions.

CONTACT Laura M. Sawyer lsawyer@symmetron.net Symmetron Ltd, London, United Kingdom

Supplemental data for this article can be accessed here.
ß 2020 The Author(s). Published with license by Taylor & Francis Group, LLC
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/), which
permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.
2 N. YASMEEN ET AL.

The objective of this systematic literature review (SLR) and hierarchical Bayesian NMA of PASI responses using an ordered
NMA was to update the study by Sawyer et al. (16) with the probit model to estimate the probabilities of achieving each
inclusion of new evidence and recently licensed interventions. level of PASI response. PASI 75 responses of the reference arm
This study aimed to compare the efficacy of approved targeted in included studies were utilized to inform baseline event rates.
therapies (brodalumab, ixekizumab, secukinumab, guselkumab, Prior exposure to biological therapies varies across trials in psor-
risankizumab, tildrakizumab, ustekinumab, adalimumab, certoli- iasis and is thought to be a potential effect modifier (20,21);
zumab pegol, etanercept, infliximab, apremilast) for moderate- therefore, a sensitivity analysis was carried out excluding studies
to-severe psoriasis after around 1 year of therapy. Efficacy was which included <5% of patients with prior experience of biolog-
measured using the Psoriasis Area and Severity Index (PASI) ics or which did not report this information.
which captures both severity of psoriatic lesions and area cover- Due to differences in study design for maintenance phase
age. The PASI score ranges between 0 and 72, and response is data, a stepwise approach to synthesis was undertaken, with a
usually measured with percentage improvement in PASI score, primary and secondary analysis. The primary analysis (analysis 1)
with values of 50, 75, 90, and 100% frequently cited in the lit- included studies with comparative evidence at 52-weeks. In order
erature (18,19). to include more studies and interventions, a second analysis was
conducted (analysis 2). This analysis included long-term extension
Materials and methods data for active therapies and induction-phase PASI results for
patients receiving placebo who, in long-term extensions had
An SLR was conducted to identify randomized controlled trials switched from placebo to active treatment. In these studies, the
(RCTs) and long-term extensions that assessed the long-term induction phase data from the placebo arms were compared
efficacy of biologic therapies and apremilast in adult patients with the maintenance phase data from the active therapy arms,
with moderate-to-severe chronic plaque psoriasis. The SLR was assuming placebo responses during induction persist into main-
performed in accordance with a protocol developed prior to the tenance. In other words, placebo induction phase responses were
commencement of the review. carried forward to the end of maintenance.
MEDLINE, MEDLINE In-Process, Embase, and the Cochrane Both fixed and random-effects models were used to generate
Library were searched from the January 01, 2000 to September results. Model-fit statistics including deviance information criter-
23, 2019 for articles published in English. Search strategies ion (DIC) and the total residual deviance indicated that the ran-
included index and text terms for psoriasis, the relevant inter- dom-effects model was the best-fit for the data, therefore, only
ventions and RCTs. The full search strategies can be found in these results are presented in this study. Inconsistency between
the Supplementary Table 1. Supplementary searches were also direct and indirect estimates of effect was assessed for any
performed. These included checking the reference lists of loops in the evidence network using the two stage Bucher
included studies and searches in disease-specific and health method (22).
economic and outcomes research congresses. All identified titles WinBUGS version 1.2 statistical software was used to perform
and abstracts were assessed for inclusion by one reviewer, with all analyses, using non-informative priors. After an initial burn-in
a second reviewer independently performing a 50% check. Full of at least 20,000 simulations, convergence was confirmed
texts were then further reviewed in the same manner by both through visual inspection the Brook-Gelman-Rubin diagnostic and
reviewers. Discrepancies were resolved by discussion or, if history plots. Sampled parameters were then estimated using
necessary, by a third reviewer. 50,000 simulations on three chains. Results were calculated as the
Long-term extensions and RCTs comparing European absolute probabilities of response for each treatment and as risk
Medicines Agency (EMA)-approved interventions of interest at ratios (RRs) for every pairwise comparison at each level of PASI
their licensed doses – adalimumab, apremilast, brodalumab, response. Point estimates reflecting the median value are pre-
certolizumab pegol, etanercept, guselkumab, infliximab, ixekizu- sented, along with 95% credible intervals (95% CrIs), reflecting
mab, risankizumab, secukinumab, tildrakizumab, and ustekinu- the range of true effects with 95% probability. Statistically signifi-
mab – with any comparator including placebo, another biologic cant differences have been interpreted as a 95% CrI which does
therapy of interest or a non-biologic systemic therapy were eli- not include the null value, 1. Numbers needed to treat (NNT) are
gible for inclusion. Trials had to include an intervention regimen also presented for the achievement of complete clearance (PASI
that was administered at a licensed dose throughout the entire
100). A numeric presentation of each treatment’s rank distribu-
induction to maintenance phase. The outcome of interest was
tion, called the surface under the cumulative ranking (SUCRA)
improvement in PASI score (PASI 75, PASI 90, and PASI 100) at
curve, is also presented. The SUCRA takes into account the effect
or around 52 weeks. PASI 50 was not included as it was
sizes and accompanying uncertainty. If a treatment always ranks
reported in very few studies. The full set of inclusion criteria can
first, then the SUCRA is equal to 100%; if a treatment always
be found in the Supplementary Material S2.
ranks last, then the SUCRA equals 0%.
For studies meeting the eligibility criteria, characteristics
regarding the study design, patient population, interventions,
outcomes, and PASI results were extracted. The methodological Results
quality of studies was assessed using the Cochrane Risk of
Literature search results
Bias Tool.
Electronic database searches identified 7654 articles, and a fur-
ther 14 articles were identified through supplementary searches.
Analysis
A total of 5915 records were screened at title and abstract
The between-study heterogeneity was assessed by comparing stage, of these, 707 full text records were assessed. The SLR
features including treatment, outcome, study design, and included 88 articles consisting of both induction and mainten-
patient characteristics. Studies were combined using a ance phase data (Figure 1).
 JOURNAL OF DERMATOLOGICAL TREATMENT 3

Figure 1. PRISMA flow diagram of SLR and NMA.

Evidence network treatment effects. Therefore, a total of 28 studies were eligible

for quantitative synthesis. Figure 2 presents the network
Of the 88 studies included in the SLR, 44 included maintenance
diagram highlighting the available evidence for analysis 1 and
phase data between 40 and 64 weeks. Studies were compared
for potential sources of heterogeneity and the evidence consid- analysis 2.
ered relevant for synthesis further refined. Sixteen studies were Table 1 contains baseline characteristics of the included stud-
excluded from the NMA due to differences in study design and ies which were broadly similar. The mean age of the participants
the comparisons made. This included data from nine responder- ranged from 42 (43) to 51 (49) years, and 49% (50) to 91% (56,
enriched randomized withdrawal maintenance phase studies 60) were male. Previous conventional systemic therapy and
(23–31) and one responder-enriched re-randomized dose com- phototherapy was reported in 0% (49) to 95% (55) and 31% (23)
parison study of tildrakizumab (32). Three studies comparing dif- to 82% (60) patients, respectively. The proportion who had
ferent doses or dosing regimens of comparators listed in the received a previous biological therapy ranged from 0% (48,49)
eligibility criteria were excluded (33–35) as was a study compar- to 60% (46).
ing a biosimilar etanercept to the originator product (36). One
study was excluded due to protocol-driven dose intensifications
Risk of bias
(37) and another was excluded due to a very small sample size
(38). These exclusions were made to reduce heterogeneity in Most studies were rated as being low risk of bias, but there was
the network which could give rise to biased estimates of relative some heterogeneity between the included studies
4 N. YASMEEN ET AL.

Figure 2. Network diagram of PASI responses - analysis 1 and analysis 2.

ADA: adalimumab; APR: apremilast; BRO: brodalumab; CZP: certolizumab pegol; ETN: etanercept; GUS: guselkumab; INF: infliximab; IXE: ixekizumab; PBO: placebo;
SEC: secukinumab; RIS: risankizumab; UST WBD: ustekinumab weight-based dose.

(Supplementary Figure 1). Of the 28 included RCTs, four (14%) treatment with brodalumab or guselkumab is expected to result
reported an inadequate randomization method while 25 (89%) in slightly more PASI responders at all levels than treatment
supplied sufficient information to assess whether allocation con- with ixekizumab, though the differences are not statistically sig-
cealment was properly ensured. In four studies, the blinding of nificant. Ixekizumab and secukinumab were superior to usteki-
participants and personnel was insufficient as the long-term numab and adalimumab and all therapies were significantly
extension was open label (45,46,51,54). In all studies, the risk of better than etanercept. The relative treatment effects for all
attrition bias was low, as incomplete outcome data were suffi- interventions are presented in Table 2. The NNT for PASI 100
ciently addressed. The risk of reporting bias was low in most of compared to ustekinumab was four for risankizumab, five for
the studies. The risk of bias for each study is presented in brodalumab and guselkumab, seven for ixekizumab, and nine
Supplementary Figure 2. for secukinumab. The NNT for ustekinumab versus etanercept
was 9 and ustekinumab versus adalimumab was 17. The SUCRA
percentage indicated that risankizumab was the most efficacious
Primary analysis
therapy in 97.3% of Bayesian iterations with a median rank of 1
Twelve studies included head-to-head data on maintenance (CrI 1 3) and brodalumab was second with a SUCRA of 78.8%
phase outcomes. Nine of these studies (17,39–42,44,56) con- and rank of 2 (CrI 1 4). This was followed by guselkumab
nected to form an evidence network comparing eight interven- (SUCRA 78.2%, median rank 3 [CrI 1 4]), ixekizumab (57.1%, 4
tions (brodalumab, secukinumab, ixekizumab, risankizumab, [2 5]), secukinumab (45.7%, 5 [4 5]), ustekinumab (24.3%, 6
guselkumab, ustekinumab, adalimumab, and etanercept) for [6 7]), adalimumab (18.4%, 7 [6 7]), and etanercept (0.1%, 8
PASI 75, 90, and 100 at 48 to 52 weeks. The other three studies [8 8]). The upper and lower bounds of the 95% credible inter-
(47,50) did not share a common comparator with the nine stud- val for etanercept are the same which indicates a low level of
ies and were therefore disconnected from the network and uncertainty in its rank as the least effective therapy evaluated in
could not be synthesized. The nine connected studies, including this analysis.
5054 patients, were included in analysis 1.
Results indicate that treatment with risankizumab, brodalu-
Secondary analysis
mab, and guselkumab provide the highest levels of PASI 75, 90,
and 100 response (Figure 3). The probabilities of achieving com- Unlike the studies eligible for the primary analysis, the remain-
plete clearance ranged from 61.1% with risankizumab, 54.5% for ing 16 studies presented maintenance phase data for active
brodalumab, 54.2% for guselkumab, 46.1% for ixekizumab, therapies but included long-term extensions where patients
40.9% for secukinumab, 28.9% for ustekinumab, 26.4% for adali- switched from placebo to active treatment. To form a connected
mumab, and 16.8% for etanercept. Risankizumab was found to network, and to augment analysis 1, the induction phase data
be significantly superior to all therapies except for brodalumab from the placebo arms were compared with the maintenance
and guselkumab at achieving all levels of PASI response. phase data from the active therapy arms. This assumption
Brodalumab and guselkumab were both found to be signifi- allowed for the inclusion of data for adalimumab, apremilast,
cantly more efficacious than secukinumab, ustekinumab, adali- certolizumab pegol, guselkumab, and infliximab, as well as add-
mumab, and etanercept (Figure 4). Based on mean values, itional data for etanercept, secukinumab and ustekinumab (45
Table 1. Baseline characteristics of studies included in the primary analysis and the secondary analysis.
Disease duration Percentage psoriasis
Sample size Age (years) Male Weight (kg) PsA (years) Previous therapies affected BSA PASI score DLQI score
Phototherapy Systemic
Study, Ind. Main. (UVB or PUVA) non-biologic Biologic
publications Treatment n n Mean SD % Mean SD % Mean SD % % % Mean SD Mean SD Mean SD
Analysis 1 studies
AMAGINE-2 (39) Brodalumab 210mg 612 168 45 13 69% 91 23 19% 19 12 52% 77% 29% 26 16 20.3 8.3 14.7 7.1
Ustekinumab 300 289 45 13 68% 91 24 17% 19 13 50% 75% 28% 27 19 20 8.4 15.1 7.2
(WBD)
AMAGINE-3 (39) Brodalumab 210mg 624 171 45 13 69% 90 23 20% 18 12 40% 68% 25% 28 18 20.4 8.3 14.5 7.2
Ustekinumab 313 301 45 13 68% 90 22 20% 18 12 44% 70% 24% 28 18 20.1 8.4 14.6 7.4
(WBD)
IXORA-S (17) Ixekizumab 80 mg Q2W 136 136 42.7 12.7 66% 85.8 20.3 18 11.1 60% 93% 13% 27 17 19.9 8.2 11.1 7.2
Ustekinumab 166 166 44 13.3 68% 89.4 24.8 18.2 12 61% 92% 15% 28 17 19.8 9 12 7.3
(WBD)
ultIMMa-1 (40) Risankizumab 150mg 304 304 48.3 13.4 70% 87.8 22.9 28% 34% 26 15 20.6 7.7 13 7
Ustekinumab 100 100 46.5 13.4 70% 88.9 22.9 23% 30% 25 15 20.1 6.8 13.6 7.3
(WBD)
ultIMMa-2 (40) Risankizumab 150mg 294 294 46.2 13.7 69% 92.2 21.7 25% 40% 26 16 20.5 7.8 13.5 7.4
Ustekinumab 99 99 48.6 14.8 67% 91.9 21.4 27% 43% 21 12 18.2 5.9 11.7 6.6
(WBD)
CLEAR (41) Secukinumab 300mg 337 334 45.2 14 68% 87.4 20 21% 19.6 12.9 65% 14% 33 18 21.7 8.5
Ustekinumab 339 335 44.6 13.7 74% 87.2 22.1 16% 16.1 11.2 66% 13% 32 17 21.5 8.1
(WBD)
FIXTURE (42) Etanercept 50mg BIW 326 323 43.8 13 71% 84.6 20.5 14% 16.4 12 63% 14% 34 18 23.2 9.8 13.4
Secukinumab 300mg 327 323 44.5 13.2 69% 83 21.6 15% 15.8 12.3 60% 12% 34 19 23.9 9.9 13.3
VOYAGE 1 (43) Adalimumab 40mg Q2W 334 334 42.9 12.6 75% 19% 17 11.3 54% 64% 21% 29 17 22.4 9 14.4 7.3
Guselkumab 100 mg 329 329 43.9 12.7 73% 20% 17.9 12.3 57% 64% 22% 28 17 22.1 9.5 14 7.5
ECLIPSE (44) Guselkumab 100 mg 534 534 46.3 13.7 68% 18% 53% 52% 29% 20 7.4
Secukinumab 300mg 514 514 45.3 13.6 65% 15% 51% 56% 29% 20.1 7.6
Analysis 2 studies
Gordon 2006 (45) Placebo 52 52 43 20–70 65% 94 50–147 31% 19 1-40 28 7-75 16 5.5–40.4 12.2 8.1
Adalimumab 40mg Q2W 46 35 46 20–71 71% 93 63–159 33% 21 1-58 29 6-58 16.7 5.4–39.0 13.3 8.8
X-PLORE (46) Placebo 42 42 46.5 67% 93.6 22.6 29% 18 13.3 50% 50% 36% 28 19 21.8 10
Adalimumab 40mg Q2W 43 38 50 70% 91.6 19.9 26% 19.3 12.8 56% 40% 60% 27 17 20.2 7.6
Ohtsuki 2017 (47) Placebo 84 84 48.3 12 74% 68.5 13.8 12.4 9.4 26% 5% 28 15 19.9 8.9 7.5 5.3
Apremilast 30mg BID 85 85 51.7 12.7 84% 70.1 13 13.9 9.2 31% 2% 31 16 21.6 8.9 7.4 5.7
LIBERATE (48) Placebo 84 84 43.4 14.9 70% 89.5 23.1 16.6 12.1 83% 0% 27 16 19.4 6.8 11.4 6.3
Apremilast 30mg BID 83 74 46 13.6 59% 88.5 19.8 19.7 12.7 80% 0% 27 16 19.3 7 13.6 6.7
UNVEIL (49) Placebo 73 73 51.1 13.7 56% 89.6 19.1 13.9 12.6 0% 0% 7 2 8 3.2 11.1 6.5
Apremilast 30mg BID 148 148 48.6 15.4 50% 87.5 21.1 17.5 13.9 0% 0% 7 2 8.2 4 11 6.5
CIMPASI-1 (50) Placebo 51 51 47.9 12.8 69% 95.2 19.5 8% 18.5 12.9 29% 26 16 19.8 7.5 13.9 8.3
Certolizumab 200mg 95 95 44.5 13.1 71% 92.6 21 11% 16.6 12.3 32% 25 17 20.1 8.2 13.3 7.4
Certolizumab 400mg 88 88 43.6 12.1 68% 92.2 21.7 17% 18.4 12.9 33% 24 17 19.6 7.9 13.1 6.5
CIMPASI-2 (50) Placebo 49 49 43.3 14.5 53% 87.1 26.4 18% 15.4 12.2 29% 20 10 17.3 5.3 12.9 7.3
Certolizumab 200mg 91 91 46.7 13.3 64% 97.8 25.6 24% 18.8 13.5 35% 21 12 18.4 5.9 15.2 7.2
Certolizumab 400mg 87 87 46.4 13.5 49% 91.8 27.7 30% 18.6 12.4 35% 23 12 19.5 6.7 14.2 7.2
JOURNAL OF DERMATOLOGICAL TREATMENT

Tyring 2006 (51) Placebo 309 306 45.6 12.1 70% 91 33% 19.7 11.4 27 17 18.1 7.4 12.5 6.7
Etanercept 50mg BIW 311 304 45.8 12.8 65% 92.6 35% 20.1 12.3 27 18 18.3 7.6 12.1 6.7
(continued)
5
 6
Table 1. Continued.
Disease duration Percentage psoriasis
Sample size Age (years) Male Weight (kg) PsA (years) Previous therapies affected BSA PASI score DLQI score
Phototherapy Systemic
Study, Ind. Main. (UVB or PUVA) non-biologic Biologic
publications Treatment n n Mean SD % Mean SD % Mean SD % % % Mean SD Mean SD Mean SD
Ohtsuki 2018 (52) Placebo 64 64 48.3 10.6 84% 71.6 14 16% 13.7 10.2 42.20% 59% 16% 34 18 25.9 12.341 10.6 7.74
Guselkumab 100mg 63 63 47.8 11.07 75% 74.3 16 16% 14.4 9.2 47.60% 59% 18% 38 21 26.7 12.196 10.3 7.27
N. YASMEEN ET AL.

EXPRESS (53) Placebo 77 77 43.8 12.6 79% 89.3 18.7 29% 17.3 11.1 71% 46% 34 18 22.8 8.7 11.8 7.5
Infliximab 5mg/kg 301 281 42.6 11.7 69% 85.9 20.1 31% 19.1 11 65% 42% 34 19 22.9 9.3 12.7 7
EXPRESS II (54) Placebo 208 208 44.4 12.5 69% 91.1 22.6 26% 50% 13% 28 18 19.8 7.7 13.4 7.3
Infliximab 5mg/kg 314 134 44.5 13 65% 92.2 23.2 28% 55% 14% 29 16 20.4 7.5 13.1 7
Torii 2010 (55) Placebo 19 19 43.3 12.3 74% 69.7 8.9 37% 11.1 6.4 74% 95% 50 27 33.1 15.6 10.5 6.8
Infliximab 5mg/kg 35 32 46.9 13 63% 68.5 13.4 29% 14.2 8.9 63% 94% 46 21 31.9 12.8 12.7 6.8
UNCOVER-3 (23) Placebo 193 193 46 12 71% 91 21 18 13 31% 43% 17% 29 17 21 8 13 7
Ixekizumab 80mg Q2W 385 362 46 13 66% 90 23 18 12 39% 44% 15% 28 17 21 8 12 7
SUSTAImm (56) Placebo 58 54 50.9 11.2 78% 75.1 17.7 12% 24% 33 19 24 9.4 9.7 5.8
Risankizumab 150mg 55 54 53.3 11.9 91% 74.1 16.2 9% 29% 41 23 26.3 11.7 10.4 5.4
FEATURE (57) Placebo 59 59 46.5 14.1 66% 88.4 21.6 20.2 14.2 49% 44% 21.1 8.5
Secukinumab 300mg 59 58 45.1 12.6 64% 92.6 25.9 18 11.9 34% 39% 33 18 20.7 8
ERASURE (42) Placebo 248 246 45.4 12.6 69% 89.7 25 27% 17.3 12.4 44% 29% 30 16 21.4 9.1 12
Secukinumab 300mg 245 245 44.9 13.5 69% 88.8 24 23% 17.4 11.1 52% 29% 33 19 22.5 9.2 13.9
JUNCTURE (58) Placebo 61 61 43.7 12.7 62% 90.2 21.2 20% 19.9 12.2 48% 21% 26 15 19.4 6.7
Secukinumab 300mg 60 60 46.6 14.2 77% 91 23.1 23% 21 13.5 50% 25% 26 13 18.9 6.4
PHOENIX 1 (59) Placebo 255 255 44.8 11.3 72% 94.2 23.5 35% 20.4 11.7 59% 56% 50% 28 17 20.4 8.6 11.8 7.4
Ustekinumab 45mg 255 251 44.8 12.5 69% 93.7 23.8 29% 19.7 11.7 68% 55% 53% 27 18 20.5 8.6 11.1 7.1
Ustekinumab 90mg 256 246 46.2 11.3 68% 93.8 23.9 37% 19.6 11.1 67% 55% 51% 25 15 19.7 7.6 11.6 6.9
Igarashi 2012 (60) Placebo 32 31 49 84% 71.2 10.9 3% 16 11.2 63% 66% 0% 50 23 30.3 11.8 10.5 6.2
Ustekinumab 45mg 64 60 45 83% 73.2 15.4 9% 15.8 8.2 56% 73% 2% 47 24 30.1 12.9 11.4 6.5
Ustekinumab 90mg 62 56 44 76% 71.1 14 11% 17.3 10.7 82% 84% 0% 47 20 28.7 11.2 10.7 6.4
BID: twice daily; BIW: twice weekly; BSA: body surface area; DLQI: dermatology life quality index; Ind: induction; Maint: maintenance; PASI: psoriasis area and severity index; PsA: psoriatic arthritis; PUVA: psoralen
and ultraviolet A; Q2W: every 2 weeks; Q4W: every 4 weeks; SD: standard deviation; UVB: ultraviolet B; WBD: weight-based dose.
Where standard deviation is not reported, range is presented.
 JOURNAL OF DERMATOLOGICAL TREATMENT 7

Figure 3. Results of predicted percentage PASI 75, 90, and 100 responses for evaluated interventions (Analysis 1).
Q2W: every 2 weeks; Q4W: every 4 weeks; WBD: weight-based dose.

and 90 mg). This analysis also allowed for studies that could not highest SUCRAs, indicating best rank, are reported for risankizu-
be connected to the network in analysis 1 to be included via mab (98.2%) and guselkumab (89.6%), followed by brodalumab
their induction-phase placebo arm (43,50). Therefore, a total of (89.2%), ixekizumab (76.7%), and secukinumab (67.9%). In con-
28 studies, including 9940 patients, were included in analysis 2. trast, apremilast and etanercept were ranked the least effica-
Results of analysis 2 showed that all therapies were signifi- cious and had the lowest SUCRA scores of all the
cantly more effective than placebo (Table 3). Probabilities of active therapies.
achieving all levels of PASI response were highest for risankizu-
mab and brodalumab, followed by guselkumab, ixekizumab,
and secukinumab (Figure 5). The lowest probabilities were gen-
erated by apremilast and etanercept. Similar to analysis 1, risk Sensitivity analysis: exclusion of studies with less than 5% prior
ratios indicate treatment with risankizumab resulted in signifi- biologic exposure
cantly more PASI responders than any other intervention apart Eight studies (45,48,49,51–53,55,60) reported less than 5% of
from brodalumab and guselkumab. Brodalumab and guselku- patients with prior biologic exposure (including all apremilast
mab were found to be superior to apremilast, adalimumab, cer- trials) and were excluded in the sensitivity analysis. All of these
tolizumab pegol, etanercept, infliximab, secukinumab, and studies were included in analysis 2 only; therefore, this sensitiv-
ustekinumab, but similar in efficacy to ixekizumab (Table 3). ity analysis was only conducted for analysis 2. Overall, the rela-
Overall, results of analysis 2 were consistent with analysis 1. tive ranks and statistical significance of treatment effects were
The SUCRAs for each therapy are also presented in Figure 6 consistent with results of the base-case analysis 2
along with their median rank and associated uncertainty. The (Supplementary Table 4).
8 N. YASMEEN ET AL.

Figure 4. Analysis 1: Relative treatment effect for all interventions. Relative effects are plotted as the median difference in response on the probit scale, where
positive values indicate greater efficacy for the intervention and negative values indicate greater efficacy for the comparator.
ADA: adalimumab; BRO: brodalumab; ETN: etanercept; GUS: guselkumab; IXE: ixekizumab; SEC: secukinumab; RIS: risankizumab; UST: ustekinumab weight-
based dose.
 JOURNAL OF DERMATOLOGICAL TREATMENT 9

Table 2. NMA of 52-week active therapy RCTs (Analysis 1): results for PASI response.
Median risk ratio (95% credible interval)
Intervention vs. Comparator PASI 75 PASI 90 PASI 100
RIS BRO 1.03 (0.99, 1.12) 1.05 (0.98, 1.2) 1.11 (0.96, 1.38)
GUS 1.03 (0.98, 1.14) 1.06 (0.97, 1.24) 1.12 (0.94, 1.46)
IXE 1.08 (1.01, 1.28) 1.15 (1.02, 1.47) 1.31 (1.05, 1.91)
SEC 1.13 (1.04, 1.35) 1.23 (1.08, 1.59) 1.48 (1.18, 2.12)
UST 1.29 (1.1, 1.69) 1.52 (1.2, 2.15) 2.1 (1.5, 3.28)
ADA 1.33 (1.1, 1.95) 1.6 (1.2, 2.63) 2.28 (1.47, 4.44)
ETN 1.65 (1.22, 2.76) 2.18 (1.43, 4.14) 3.66 (2.03, 8.26)
BRO GUS 1 (0.94, 1.08) 1 (0.9, 1.14) 1.01 (0.82, 1.26)
IXE 1.05 (0.98, 1.21) 1.09 (0.97, 1.34) 1.17 (0.94, 1.63)
SEC 1.09 (1.02, 1.27) 1.16 (1.04, 1.44) 1.32 (1.09, 1.8)
UST 1.25 (1.09, 1.57) 1.43 (1.18, 1.92) 1.87 (1.42, 2.71)
ADA 1.29 (1.08, 1.83) 1.51 (1.17, 2.37) 2.04 (1.37, 3.76)
ETN 1.59 (1.2, 2.57) 2.05 (1.39, 3.72) 3.27 (1.9, 6.92)
GUS IXE 1.05 (0.97, 1.21) 1.08 (0.95, 1.35) 1.17 (0.9, 1.66)
SEC 1.09 (1.03, 1.23) 1.16 (1.06, 1.36) 1.32 (1.14, 1.64)
UST 1.24 (1.09, 1.58) 1.43 (1.17, 1.94) 1.85 (1.39, 2.78)
ADA 1.29 (1.09, 1.75) 1.51 (1.19, 2.22) 2.04 (1.44, 3.33)
ETN 1.59 (1.21, 2.51) 2.05 (1.4, 3.58) 3.26 (1.94, 6.49)
IXE SEC 1.04 (0.94, 1.18) 1.06 (0.91, 1.29) 1.12 (0.84, 1.53)
UST 1.18 (1.06, 1.44) 1.3 (1.11, 1.7) 1.57 (1.21, 2.27)
ADA 1.22 (1.05, 1.67) 1.37 (1.08, 2.09) 1.72 (1.16, 3.08)
ETN 1.51 (1.17, 2.34) 1.87 (1.32, 3.24) 2.75 (1.66, 5.62)
SEC UST 1.13 (1.05, 1.32) 1.22 (1.08, 1.49) 1.4 (1.16, 1.83)
ADA 1.18 (1.05, 1.5) 1.29 (1.09, 1.76) 1.53 (1.17, 2.33)
ETN 1.46 (1.17, 2.09) 1.76 (1.31, 2.71) 2.46 (1.66, 4.2)
UST ETN 1.04 (0.91, 1.26) 1.06 (0.86, 1.39) 1.09 (0.79, 1.64)
ADA 1.27 (1.08, 1.73) 1.43 (1.13, 2.09) 1.74 (1.23, 2.86)
ADA ETN 1.22 (1.04, 1.61) 1.34 (1.06, 1.92) 1.58 (1.09, 2.56)
ADA: adalimumab; BRO: brodalumab; ETN: etanercept; GUS: guselkumab; IXE: ixekizumab; PASI: psoriasis area and severity
index; SEC: secukinumab; RIS: risankizumab; UST: ustekinumab weight-based dose.

Discussion of PASI responders are similar from induction to mainten-

ance phase.
This SLR and NMA compared licensed systemic therapies for
The off-treatment durability of PASI response could be inves-
moderate-to-severe psoriasis. Tildrakizumab was included
tigated with randomized withdrawal trials. However, such trials,
among the comparators of interest, but because there was no
which may be considered responder-enriched, were excluded.
comparable maintenance data available at 1 year, it could not
Compared to maintenance phase studies without responder
be included in the NMA. The synthesis of evidence from nine
enrichment, this design may introduce bias in favor of the active
RCTs reporting PASI outcomes at week 48 to 52 in the primary
analysis, showed risankizumab, brodalumab, and guselkumab to intervention. Additionally, these studies varied in their definition
be associated with the highest likelihood of response, followed of responders and consequently on the eligibility for re-random-
by ixekizumab and secukinumab. Results indicated that risanki- ization and were therefore not comparable.
zumab was statistically superior to ixekizumab but not statistic- Another source of long-term evidence are registry studies
ally different from brodalumab or guselkumab. Brodalumab and which can provide evidence of therapies in the real world (61).
guselkumab produced greater PASI responders than ixekizumab, However, these observational studies are not always useful in
but this was not statistically significant. Analysis 1 showed all addressing questions of efficacy, as treatment choice is influ-
therapies except etanercept and adalimumab were found to be enced by factors such as patient characteristics, or healthcare
superior to ustekinumab, a result consistent with the findings of coverage (62). In addition, the data collection is not sufficiently
the head to head RCTs included in the analysis. In the second- consistent to allow the combination of data in an NMA (61).
ary analysis, studies from the first analysis were supplemented Such data may serve to supplement the efficacy assessment
with the inclusion of long-term extensions of placebo controlled presented here with real world evidence on drug survival and
RCTs reporting maintenance phase outcomes for other licensed tolerability.
therapies, including certolizumab pegol, infliximab and apremi-
last. Results were consistent across both analyses and suggested Strengths and limitations
the superiority of IL-17s and IL-23s over other systemic therapies
after 1 year of treatment. This study included a comprehensive and up to date SLR of
The results of this NMA are consistent with our original syn- English language studies, which allows recently approved thera-
thesis of long-term data (16), although further interventions are pies to be included. The study also presents the first long term
included in the current analysis (ixekizumab, risankizumab, indirect comparison with IL-23s, apart from tildrakizumab for
guselkumab, and certolizumab pegol). Additionally, our results which there is no comparable evidence.
are broadly in line with the results of previously published The number of studies reporting on the maintenance phase
NMAs (9–14) on induction phase data, which show proportions outcomes of interventions in psoriasis is limited, especially
10 N. YASMEEN ET AL.

Table 3. NMA of 52-week RCTs using induction phase placebo control (Analysis 2): results for PASI responses.
Median risk ratio (95% credible interval)
Intervention Comparator PASI 75 PASI 90 PASI 100
RIS BRO 1.02 (0.98, 1.11) 1.05 (0.97, 1.19) 1.1 (0.94, 1.36)
GUS 1.02 (0.98, 1.12) 1.05 (0.96, 1.21) 1.1 (0.92, 1.4)
IXE 1.07 (1.01, 1.24) 1.14 (1.02, 1.41) 1.28 (1.05, 1.76)
SEC 1.11 (1.03, 1.31) 1.22 (1.07, 1.54) 1.44 (1.16, 2)
INF 1.23 (1.05, 1.75) 1.43 (1.1, 2.34) 1.88 (1.22, 3.72)
UST 1.28 (1.1, 1.68) 1.54 (1.22, 2.17) 2.11 (1.51, 3.26)
CZP 400 1.28 (1.06, 1.96) 1.53 (1.12, 2.75) 2.1 (1.27, 4.69)
ADA 1.32 (1.1, 1.89) 1.62 (1.21, 2.61) 2.28 (1.49, 4.27)
CZP 200 1.45 (1.11, 2.47) 1.86 (1.24, 3.77) 2.82 (1.53, 7.2)
ETN 1.59 (1.2, 2.54) 2.13 (1.42, 3.86) 3.47 (1.98, 7.29)
APR 3.48 (1.78, 9.41) 6.29 (2.64, 20.56) 15.75 (5.2, 64.42)
PBO 15.76 (5.08, 66.67) 42.09 (10.9, 213.49) 179.82 (36.9, 1104.12)
BRO GUS 1 (0.94, 1.07) 1 (0.89, 1.13) 1 (0.82, 1.24)
IXE 1.04 (0.99, 1.18) 1.08 (0.97, 1.3) 1.16 (0.95, 1.54)
SEC 1.08 (1.02, 1.25) 1.16 (1.04, 1.42) 1.3 (1.08, 1.74)
INF 1.19 (1.03, 1.66) 1.36 (1.06, 2.16) 1.7 (1.13, 3.25)
UST 1.25 (1.09, 1.58) 1.46 (1.19, 1.97) 1.91 (1.43, 2.78)
CZP 400 1.24 (1.04, 1.86) 1.45 (1.09, 2.54) 1.9 (1.18, 4.1)
ADA 1.29 (1.09, 1.79) 1.53 (1.18, 2.38) 2.06 (1.39, 3.7)
CZP 200 1.41 (1.1, 2.35) 1.76 (1.2, 3.47) 2.55 (1.42, 6.29)
ETN 1.54 (1.19, 2.4) 2.02 (1.39, 3.52) 3.13 (1.87, 6.27)
APR 3.39 (1.77, 8.91) 5.97 (2.58, 18.8) 14.23 (4.9, 55.35)
PBO 15.34 (5.03, 62.62) 39.95 (10.67, 193.72) 162.5 (34.9, 941.62)
GUS IXE 1.04 (0.99, 1.17) 1.08 (0.97, 1.29) 1.16 (0.95, 1.53)
SEC 1.09 (1.03, 1.22) 1.16 (1.06, 1.36) 1.3 (1.13, 1.61)
INF 1.19 (1.03, 1.64) 1.36 (1.07, 2.12) 1.7 (1.15, 3.16)
UST 1.25 (1.09, 1.58) 1.46 (1.19, 1.99) 1.9 (1.42, 2.83)
CZP 400 1.25 (1.04, 1.84) 1.45 (1.09, 2.49) 1.9 (1.19, 3.97)
ADA 1.29 (1.09, 1.74) 1.54 (1.2, 2.25) 2.07 (1.46, 3.34)
CZP 200 1.41 (1.1, 2.33) 1.76 (1.21, 3.41) 2.55 (1.44, 6.12)
ETN 1.55 (1.19, 2.36) 2.02 (1.4, 3.42) 3.14 (1.91, 5.95)
APR 3.39 (1.77, 8.78) 5.97 (2.59, 18.41) 14.23 (4.98, 53.46)
PBO 15.34 (5.05, 62.07) 39.97 (10.73, 189.43) 162.79 (35.45, 905.8)
IXE SEC 1.03 (0.97, 1.15) 1.06 (0.95, 1.24) 1.12 (0.91, 1.42)
INF 1.14 (0.99, 1.52) 1.25 (0.99, 1.87) 1.46 (0.98, 2.6)
UST 1.19 (1.07, 1.44) 1.33 (1.13, 1.73) 1.62 (1.27, 2.29)
CZP 400 1.19 (1.01, 1.7) 1.33 (1.02, 2.2) 1.62 (1.03, 3.27)
ADA 1.23 (1.06, 1.63) 1.4 (1.12, 2.06) 1.76 (1.25, 2.94)
CZP 200 1.34 (1.07, 2.14) 1.61 (1.13, 3.01) 2.18 (1.26, 5.01)
ETN 1.47 (1.17, 2.17) 1.85 (1.33, 3.02) 2.67 (1.71, 4.93)
APR 3.23 (1.74, 8.04) 5.47 (2.48, 16.02) 12.14 (4.51, 43.15)
PBO 14.63 (4.96, 56.27) 36.66 (10.33, 162.71) 138.93 (32.36, 720.46)
SEC INF 1.1 (0.97, 1.42) 1.17 (0.94, 1.68) 1.31 (0.9, 2.2)
UST 1.15 (1.05, 1.33) 1.25 (1.1, 1.53) 1.45 (1.2, 1.88)
CZP 400 1.14 (0.98, 1.59) 1.25 (0.96, 1.98) 1.45 (0.94, 2.78)
ADA 1.18 (1.05, 1.49) 1.32 (1.11, 1.79) 1.57 (1.21, 2.36)
CZP 200 1.29 (1.05, 2) 1.51 (1.09, 2.71) 1.95 (1.16, 4.26)
ETN 1.42 (1.15, 1.98) 1.74 (1.31, 2.61) 2.4 (1.65, 3.9)
APR 3.11 (1.71, 7.42) 5.13 (2.4, 14.21) 10.85 (4.21, 35.91)
PBO 14.11 (4.9, 51.81) 34.36 (10.03, 144.13) 124.07 (30.49, 593.47)
INF UST 1.04 (0.85, 1.25) 1.06 (0.78, 1.41) 1.1 (0.69, 1.71)
CZP 400 1.04 (0.82, 1.41) 1.07 (0.73, 1.67) 1.11 (0.61, 2.18)
ADA 1.07 (0.88, 1.37) 1.11 (0.82, 1.6) 1.19 (0.74, 2.04)
CZP 200 1.17 (0.91, 1.75) 1.28 (0.87, 2.24) 1.48 (0.8, 3.25)
ETN 1.27 (1.05, 1.76) 1.45 (1.08, 2.25) 1.79 (1.14, 3.25)
APR 2.79 (1.62, 6.26) 4.29 (2.15, 11.23) 8.12 (3.35, 25.98)
PBO 12.68 (4.7, 42.63) 28.84 (9.22, 110.12) 92.94 (25.58, 407.83)
UST CZP 400 1 (0.82, 1.32) 1 (0.74, 1.5) 1.01 (0.62, 1.83)
ADA 1.03 (0.91, 1.23) 1.05 (0.87, 1.36) 1.09 (0.8, 1.57)
CZP 200 1.13 (0.92, 1.64) 1.21 (0.87, 2.02) 1.34 (0.8, 2.75)
ETN 1.23 (1.07, 1.59) 1.38 (1.12, 1.92) 1.64 (1.2, 2.52)
APR 2.7 (1.59, 5.94) 4.06 (2.08, 10.3) 7.42 (3.17, 22.47)
PBO 12.25 (4.6, 40.65) 27.22 (8.85, 101.5) 84.74 (23.74, 359)
CZP 400 ADA 1.03 (0.79, 1.32) 1.05 (0.7, 1.52) 1.08 (0.59, 1.91)
CZP 200 1.12 (1, 1.39) 1.2 (1, 1.59) 1.33 (1, 1.95)
ETN 1.22 (0.97, 1.68) 1.36 (0.96, 2.11) 1.61 (0.94, 2.99)
APR 2.66 (1.58, 5.86) 3.99 (2.03, 10.25) 7.25 (3.01, 22.87)
PBO 12.09 (4.6, 39.59) 26.75 (8.82, 99.6) 82.71 (23.47, 357.4)
(continued)
 JOURNAL OF DERMATOLOGICAL TREATMENT 11

Table 3. Continued.
Median risk ratio (95% credible interval)
Intervention Comparator PASI 75 PASI 90 PASI 100
ADA CZP 200 1.09 (0.86, 1.57) 1.15 (0.79, 1.89) 1.24 (0.69, 2.52)
ETN 1.19 (1.03, 1.51) 1.3 (1.05, 1.78) 1.5 (1.08, 2.29)
APR 2.6 (1.57, 5.54) 3.83 (2.01, 9.36) 6.78 (2.98, 19.74)
PBO 11.81 (4.55, 37.44) 25.72 (8.69, 90.91) 77.46 (22.82, 309.02)
CZP 200 ETN 1.08 (0.81, 1.45) 1.13 (0.74, 1.72) 1.21 (0.65, 2.22)
APR 2.34 (1.45, 4.89) 3.29 (1.75, 8) 5.38 (2.3, 16.13)
PBO 10.64 (4.33, 32.21) 21.97 (7.92, 75.13) 61.2 (19.11, 242.78)
ETN APR 2.17 (1.41, 4.16) 2.92 (1.68, 6.32) 4.49 (2.19, 11.38)
PBO 9.87 (4.19, 27.5) 19.6 (7.49, 59.56) 51.26 (17.54, 170.53)
APR PBO 4.43 (2.54, 8.99) 6.52 (3.43, 14.47) 11.09 (5.16, 27.72)
ADA, adalimumab; APR, apremilast; BRO, brodalumab; CZP, certolizumab pegol; ETN, etanercept; GUS, guselkumab; INF,
infliximab; IXE, ixekizumab; PASI, psoriasis area and severity index; PBO, placebo; SEC, secukinumab; RIS, risankizumab;
UST, ustekinumab weight-based dose.

Figure 5. Results of predicted percentage PASI 75, 90, and 100 responses for evaluated interventions (Analysis 2).
Q2W: every 2 weeks; Q4W: every 4 weeks; WBD: weight-based dose
12 N. YASMEEN ET AL.

Figure 6. SUCRA and ranking with error bars indicating the 95% credible interval (Analysis 2).
ADA: adalimumab; BRO: brodalumab; ETN: etanercept; GUS: guselkumab; IXE: ixekizumab; SEC: secukinumab; RIS: risankizumab; UST: ustekinumab weight-
based dose

considering the volume of RCT evidence evaluating induction week parallel RCTs, is the most robust comparison of biologic
phase efficacy and safety. In addition to their relative scarcity, therapies for moderate-to-severe psoriasis over the longer-term;
variations in their study design make quantitative comparisons however, it is limited by the number of comparisons it
challenging. The primary analysis, which relies on largely 52- can make.
 JOURNAL OF DERMATOLOGICAL TREATMENT 13

The secondary analysis aims to extend the comparisons that Funding

can be made, but it is limited by the uncertainty of an assumed
This study was funded by LEO Pharma A/S.
placebo response. Due to the treatment switching in these stud-
ies, the missing maintenance data were imputed using induc-
tion phase responses. A plateau in placebo response between
induction and 24 weeks has been noted in four RCTs (53,63–65), References
which lends some support to the assumption that placebo
responses do not improve with time. Carrying forward the pla- 1. Girolomoni G, Strohal R, Puig L, et al. The role of IL-23
cebo responses from induction fails to account for any losses of and the IL-23/TH 17 immune axis in the pathogenesis and
efficacy between 10 and 16 weeks and 1 year, thereby making treatment of psoriasis [Review]. J Eur Acad Dermatol
the assumption quite conservative. Venereol. 2017;31(10):1616–1626.
This study did not investigate adverse events or health- 2. World Health Organization. Global report on psoriasis.
related quality of life at 1 year due the heterogenous reporting World Health Organization; 2016 [Last accessed March
between trials. To enable a full evaluation of therapies for mod- 2020]. Available from: https://apps.who.int/iris/handle/
erate-to-severe plaque psoriasis, future long-term studies should 10665/204417
investigate these outcomes and report these in a more uniform 3. Griffiths CE, Barker JN. Pathogenesis and clinical features
way, enabling a formal comparison. of psoriasis. Lancet (London, England). 2007;370(9583):
Additionally, a potential limitation of analyses 1 and 2 is that 263–271.
4. Mrowietz U, Kragballe K, Reich K, et al. Definition of treat-
it may not be directly relevant to clinical practice as patients nor-
ment goals for moderate to severe psoriasis: a European
mally would discontinue treatment at the end of induction if the
consensus. Arch Dermatol Res. 2011;303(1):1–10.
response is insufficient or receive higher doses than recom-
5. Taylor GJ, Wainwright P. Open label extension studies:
mended. Since this study only included clinical trials and not
research or marketing? BMJ. 2005;331(7516):572–574.
real-world evidence, only licensed doses of interventions were
6. Buch MH, Aletaha D, Emery P, et al. Reporting of long-
included. Several studies have been conducted which report that
term extension studies: lack of consistency calls for con-
dose adjustments are used routinely in clinical practice (66,67).
sensus. Ann Rheum Dis. 2011;70(6):886–890.
7. Naldi L, Svensson A, Zenoni D, et al. Comparators, study
Conclusion duration, outcome measures and sponsorship in thera-
peutic trials of psoriasis: update of the EDEN Psoriasis
This NMA of maintenance phase RCT evidence indicates that
Survey 2001-2006. Br J Dermatol. 2010;162(2):384–389.
high levels of PASI can be achieved and maintained at around
8. Ronholt K, Iversen L. Old and new biological therapies for
1 year. Results demonstrated that the evaluated IL-17 and IL-23
psoriasis. Int J Mol Sci. 2017;18(11): 2297.
inhibitors outperformed other biological therapies in terms of
9. Hendrix N, Ollendorf DA, Chapman RH, et al. Cost-effect-
the proportion achieving all levels of PASI response after
iveness of targeted pharmacotherapy for moderate to
48–52 weeks of treatment. Risankizumab had a higher probabil-
severe plaque psoriasis. J Manag Care Spec Pharm. 2018;
ity of achieving PASI outcomes than all other biologics, except
24(12):1210–1217.
brodalumab and guselkumab, where no significant difference
10. Geng W, Zhao J, Fu J, et al. Efficacy of several biological
could be concluded. Further long-term trials with head to head
therapies for treating moderate to severe psoriasis: a net-
evidence in which patients receive a licensed therapy for the work meta-analysis. Exp Ther Med. 2018;16(6):5085–5095.
entire induction and maintenance periods are needed to valid- 11. Sawyer L, Fotheringham I, Wright E, et al. The comparative
ate these conclusions. Further real-world studies may also help efficacy of brodalumab in patients with moderate-to-
to validate the findings and provide evidence on the long-term severe psoriasis: a systematic literature review and net-
performance of these treatments in clinical practice. work meta-analysis. J Dermatolog Treat. 2018;29(6):
557–568.
12. Loos AM, Liu S, Segel C, et al. Comparative effectiveness
Acknowledgments
of targeted immunomodulators for the treatment of mod-
The authors thank Jana Tillotson for graphic design support dur- erate-to-severe plaque psoriasis: a systematic review and
ing the preparation of this article. network meta-analysis. J Am Acad Dermatol. 2018;79(1):
135–144.e7.
13. Lin PT, Wang SH, Chi CC. Drug survival of biologics in
Disclosure statement
treating psoriasis: a meta-analysis of real-world evidence.
Najeeda Yasmeen and Laura Sawyer are Symmetron employees, Sci Rep. 2018;8(1):16068.
as was Kinga Malottki when the work was undertaken; all were 14. Cameron C, Druchok C, Hutton B, et al. Guselkumab for
consultants to LEO Pharma for this study. Eydna Didriksen Apol the treatment of moderate-to-severe plaque psoriasis dur-
is a LEO Pharma employee. Lars-Åke Levin has received consult- ing induction phase: a systematic review and network
ing fees from LEO Pharma and Gregor B Jemec has received meta-analysis. J Psoriasis Psoriatic Arthritis. 2019;4(2):
honoraria from AbbVie, Coloplast, Galderma, Inflarx, LEO 81–92.
Pharma, MSD, Pierre-Fabre, UCB, and his department received 15. Nast A, Jacobs A, Rosumeck S, et al. Efficacy and safety of
grants from AbbVie, Actelion, Janssen-Cilag, LEO Pharma, systemic long-term treatments for moderate-to-severe
Novartis, Regeneron, Serono and UCB for his participation as an psoriasis: a systematic review and meta-analysis. J Invest
investigator. Dermatol. 2015;135(11):2641–2648.
14 N. YASMEEN ET AL.

16. Sawyer LM, Cornic L, Levin LA, et al. Long-term efficacy of III, double-blind, placebo- and active comparator-con-
novel therapies in moderate-to-severe plaque psoriasis: a trolled VOYAGE 2 trial. J Am Acad Dermatol. 2017;76(3):
systematic review and network meta-analysis of PASI 418–431.
response. J Eur Acad Dermatol Venereol. 2019;33(2): 30. Blauvelt A, Ferris LK, Yamauchi PS, et al. Extension of uste-
355–366. kinumab maintenance dosing interval in moderate-to-
17. Paul C, Griffiths CEM, van de Kerkhof PCM, et al. severe psoriasis: results of a phase IIIb, randomized,
Ixekizumab provides superior efficacy compared with uste- double-blinded, active-controlled, multicentre study
kinumab over 52 weeks of treatment: results from IXORA- (PSTELLAR). Br J Dermatol. 2017;177(6):1552–1561.
S, a phase 3 study. J Am Acad Dermatol. 2019;80(1): 31. Reich K, Gooderham M, Thaci D, et al. Risankizumab com-
70–79.e3. pared with adalimumab in patients with moderate-to-
18. Feldman SR, Krueger GG. Psoriasis assessment tools in severe plaque psoriasis (IMMvent): a randomised, double-
clinical trials. Ann Rheum Dis. 2005;64(suppl_2):ii65–ii68. blind, active-comparator-controlled phase 3 trial. Lancet
19. Shikiar R, Bresnahan BW, Stone SP, et al. Validity and reli- (London, England). 2019;394(10198):576–586.
ability of patient reported outcomes used in psoriasis: 32. Papp K, Thaci D, Reich K, et al. Tildrakizumab (MK-3222),
results from two randomized clinical trials [journal article. an anti-interleukin-23p19 monoclonal antibody, improves
Health Qual Life Outcomes. 2003;1(1):53. psoriasis in a phase IIb randomized placebo-controlled
20. Papp KA, Gordon KB, Langley RG, et al. Impact of previous trial. Br J Dermatol. 2015;173(4):930–939.
biologic use on the efficacy and safety of brodalumab and 33. Papp KA, Blauvelt A, Bukhalo M, et al. Risankizumab versus
ustekinumab in patients with moderate-to-severe plaque ustekinumab for moderate-to-severe plaque psoriasis. N
psoriasis: integrated analysis of the randomized controlled Engl J Med. 2017;376(16):1551–1560.
trials AMAGINE-2 and AMAGINE-3. Br J Dermatol. 2018; 34. Langley RG, Papp K. Efficacy and safety of continuous
179(2):320–328. every-2-week dosing of ixekizumab over 52 weeks in
21. Ruiz Salas V, Puig L, Alomar A. Ustekinumab in clinical patients with moderate-to-severe plaque psoriasis in a
practice: response depends on dose and previous treat- randomized phase III trial (IXORA-P). Br J Dermatol. 2018;
ment. J Eur Acad Dermatol Venereol. 2012;26(4):508–513. 178(6):1315–1323.
22. Bucher HC, Guyatt GH, Griffith LE, et al. The results of dir- 35. Khattri S, Goldblum O, Solotkin K, et al. Early onset of clin-
ect and indirect treatment comparisons in meta-analysis
ical improvement with ixekizumab in a randomized, open-
of randomized controlled trials. J Clin Epidemiol. 1997;
label study of patients with moderate-to-severe plaque
50(6):683–691.
psoriasis. J Clin Aesthet Dermatol. 2018;11(5):33–37.
23. Griffiths CE, Reich K, Lebwohl M, et al. Comparison of ixe-
36. Griffiths CEM, Thaci D, Gerdes S, et al. The EGALITY study:
kizumab with etanercept or placebo in moderate-to-
a confirmatory, randomized, double-blind study compar-
severe psoriasis (UNCOVER-2 and UNCOVER-3): results
ing the efficacy, safety and immunogenicity of GP2015, a
from two phase 3 randomised trials. Lancet (London,
proposed etanercept biosimilar, vs. the originator product
England). 2015;386(9993):541–551.
in patients with moderate-to-severe chronic plaque-type
24. Gordon KB, Blauvelt A, Papp KA, et al. Phase 3 trials of ixe-
psoriasis. Br J Dermatol. 2017;176(4):928–938.
kizumab in moderate-to-severe plaque psoriasis. N Engl J
37. Papp KA, Langley RG, Lebwohl M, et al. Efficacy and safety
Med. 2016;375(4):345–356.
of ustekinumab, a human interleukin-12/23 monoclonal
25. Papp K, Reich K, Leonardi CL, et al. Apremilast, an oral
phosphodiesterase 4 (PDE4) inhibitor, in patients with antibody, in patients with psoriasis: 52-week results from
moderate to severe plaque psoriasis: results of a phase III, a randomised, double-blind, placebo-controlled trial
randomized, controlled trial (Efficacy and Safety Trial (PHOENIX 2). Lancet (London, England). 2008;371(9625):
Evaluating the Effects of Apremilast in Psoriasis [ESTEEM] 1675–1684.
1). J Am Acad Dermatol. 2015;73(1):37–49. 38. de Vries AC, Thio HB, de Kort WJ, et al. A prospective
26. Papp KA, Reich K, Paul C, et al. A prospective phase III, randomized controlled trial comparing infliximab and eta-
randomized, double-blind, placebo-controlled study of nercept in patients with moderate-to-severe chronic pla-
brodalumab in patients with moderate-to-severe plaque que-type psoriasis: the Psoriasis Infliximab vs. Etanercept
psoriasis. Br J Dermatol. 2016;175(2):273–286. Comparison Evaluation (PIECE) study. Br J Dermatol. 2017;
27. Paul C, Cather J, Gooderham M, et al. Efficacy and safety 176(3):624–633.
of apremilast, an oral phosphodiesterase 4 inhibitor, in 39. Lebwohl M, Strober B, Menter A, et al. Phase 3 studies
patients with moderate-to-severe plaque psoriasis over 52 comparing brodalumab with ustekinumab in psoriasis.
weeks: a phase III, randomized controlled trial (ESTEEM 2). N Engl J Med. 2015;373(14):1318–1328.
Br J Dermatol. 2015;173(6):1387–1399. 40. Gordon KB, Strober B, Lebwohl M, et al. Efficacy and safety
28. Lebwohl M, Blauvelt A, Paul C, et al. Certolizumab pegol of risankizumab in moderate-to-severe plaque psoriasis
for the treatment of chronic plaque psoriasis: results (UltIMMa-1 and UltIMMa-2): results from two double-blind,
through 48 weeks of a phase 3, multicenter, randomized, randomised, placebo-controlled and ustekinumab-con-
double-blind, etanercept- and placebo-controlled study trolled phase 3 trials. Lancet (London, England). 2018;
(CIMPACT). J Am Acad Dermatol. 2018;79(2):266–276.e5. 392(10148):650–661.
29. Reich K, Armstrong AW, Foley P, et al. Efficacy and safety 41. Blauvelt A, Reich K, Tsai TF, et al. Secukinumab is superior
of guselkumab, an anti-interleukin-23 monoclonal anti- to ustekinumab in clearing skin of subjects with moder-
body, compared with adalimumab for the treatment of ate-to-severe plaque psoriasis up to 1 year: results from
patients with moderate to severe psoriasis with random- the CLEAR study. J Am Acad Dermatol. 2017;76(1):
ized withdrawal and retreatment: results from the phase 60–69.e9.
 JOURNAL OF DERMATOLOGICAL TREATMENT 15

42. Langley RG, Elewski BE, Lebwohl M, et al. Secukinumab in 54. Menter A, Feldman SR, Weinstein GD, et al. A randomized
plaque psoriasis–results of two phase 3 trials. N Engl J comparison of continuous vs. intermittent infliximab main-
Med. 2014;371(4):326–338. tenance regimens over 1 year in the treatment of moder-
43. Blauvelt A, Papp KA, Griffiths CE, et al. Efficacy and safety ate-to-severe plaque psoriasis. J Am Acad Dermatol. 2007;
of guselkumab, an anti-interleukin-23 monoclonal anti- 56(1):31.e1–15.
body, compared with adalimumab for the continuous 55. Torii H, Nakagawa H. Infliximab monotherapy in Japanese
treatment of patients with moderate to severe psoriasis: patients with moderate-to-severe plaque psoriasis and
results from the phase III, double-blinded, placebo- and psoriatic arthritis. A randomized, double-blind, placebo-
active comparator-controlled VOYAGE 1 trial. J Am Acad controlled multicenter trial. J Dermatol Sci. 2010;59(1):
Dermatol. 2017;76(3):405–417. 40–49.
44. Reich K, Armstrong AW, Langley RG, et al. Guselkumab 56. Ohtsuki M, Fujita H, Watanabe M, et al. Efficacy and safety
versus secukinumab for the treatment of moderate-to- of risankizumab in Japanese patients with moderate to
severe psoriasis (ECLIPSE): results from a phase 3, rando- severe plaque psoriasis: results from the SustaIMM phase
mised controlled trial. Lancet (London, England). 2019; 2/3 trial. J Dermatol. 2019;46(8):686–694.
394(10201):831–839. 57. Blauvelt A, Prinz JC, Gottlieb AB, et al. Secukinumab
45. Gordon KB, Langley RG, Leonardi C, et al. Clinical response administration by pre-filled syringe: efficacy, safety and
to adalimumab treatment in patients with moderate to usability results from a randomized controlled trial in
severe psoriasis: double-blind, randomized controlled trial psoriasis (FEATURE). Br J Dermatol. 2015;172(2):484–493.
and open-label extension study. J Am Acad Dermatol. 58. Lacour JP, Paul C, Jazayeri S, et al. Secukinumab adminis-
2006;55(4):598–606. tration by autoinjector maintains reduction of plaque
46. Gordon KB, Duffin KC, Bissonnette R, et al. A phase 2 trial psoriasis severity over 52 weeks: results of the randomized
of guselkumab versus adalimumab for plaque psoriasis. controlled JUNCTURE trial. J Eur Acad Dermatol Venereol.
N Engl J Med. 2015;373(2):136–144. 2017;31(5):847–856.
47. Ohtsuki M, Okubo Y, Komine M, et al. Apremilast, an oral 59. Leonardi CL, Kimball AB, Papp KA, et al. Efficacy and safety
phosphodiesterase 4 inhibitor, in the treatment of
of ustekinumab, a human interleukin-12/23 monoclonal
Japanese patients with moderate to severe plaque psoria-
antibody, in patients with psoriasis: 76-week results from
sis: efficacy, safety and tolerability results from a phase 2b
a randomised, double-blind, placebo-controlled trial
randomized controlled trial. J Dermatol. 2017;44(8):
(PHOENIX 1). Lancet (London, England). 2008;371(9625):
873–884.
1665–1674.
48. Reich K, Gooderham M, Green L, et al. The efficacy and
60. Igarashi A, Kato T, Kato M, et al. Efficacy and safety of
safety of apremilast, etanercept and placebo in patients
ustekinumab in Japanese patients with moderate-to-
with moderate-to-severe plaque psoriasis: 52-week results
severe plaque-type psoriasis: long-term results from a
from a phase IIIb, randomized, placebo-controlled trial
phase 2/3 clinical trial. J Dermatol. 2012;39(3):242–252.
(LIBERATE). J Eur Acad Dermatol Venereol. 2017;31(3):
507–517. 61. Eissing L, Rustenbach SJ, Krensel M, et al. Psoriasis regis-
49. Stein Gold L, Bagel J, Lebwohl M, et al. Efficacy and safety tries worldwide: systematic overview on registry publica-
of apremilast in systemic- and biologic-naive patients tions. J Eur Acad Dermatol Venereol. 2016;30(7):
with moderate plaque psoriasis: 52-week results of 1100–1106.
UNVEIL. J Drugs Dermatol. 2018;17(2):221–228. 62. Dreyer NA, Garner S. Registries for robust evidence. Jama.
50. Gottlieb AB, Blauvelt A, Thaci D, et al. Certolizumab pegol 2009;302(7):790–791.
for the treatment of chronic plaque psoriasis: results 63. Asahina A, Nakagawa H, Etoh T, et al. Adalimumab in
through 48 weeks from 2 phase 3, multicenter, random- Japanese patients with moderate to severe chronic plaque
ized, double-blinded, placebo-controlled studies (CIMPASI- psoriasis: efficacy and safety results from a Phase II/III
1 and CIMPASI-2). J Am Acad Dermatol. 2018;79(2): randomized controlled study. J Dermatol. 2010;37(4):
302–314.e6. 299–310.
51. Tyring S, Gottlieb A, Papp K, et al. Etanercept and clinical 64. Gottlieb AB, Matheson RT, Lowe N, et al. A randomized
outcomes, fatigue, and depression in psoriasis: double- trial of etanercept as monotherapy for psoriasis. Arch
blind placebo-controlled randomised phase III trial. Lancet Dermatol. 2003;139(12):1627–1632. discussion 1632.
(London, England). 2006;367(9504):29–35. 65. Krueger GG, Langley RG, Leonardi C, et al. A human inter-
52. Ohtsuki M, Kubo H. Guselkumab, an anti-interleukin-23 leukin-12/23 monoclonal antibody for the treatment of
monoclonal antibody, for the treatment of moderate to psoriasis. N Engl J Med. 2007;356(6):580–592.
severe plaque-type psoriasis in Japanese patients: efficacy 66. Feldman SR, Zhao Y, Zhou H, et al. Economic impact of
and safety results from a phase 3, randomized, double- above-label dosing with etanercept, adalimumab, or uste-
blind, placebo-controlled study. J Dermatol. 2018;45(9): kinumab in patients with psoriasis. J Manag Care Spec
1053–1062. Pharm. 2017;23(5):583–589.
53. Reich K, Nestle FO, Papp K, et al. Infliximab induction and 67. Esposito M, Gisondi P, Conti A, et al. Dose adjustment of
maintenance therapy for moderate-to-severe psoriasis: a biologic therapies for psoriasis in dermatological practice:
phase III, multicentre, double-blind trial. Lancet (London, a retrospective study. J Eur Acad Dermatol Venereol. 2017;
England). 2005;366(9494):1367–1374. 31(5):863–869.