BJD

S Y S TE M A T IC R E V IE W British Journal of Dermatology

Short-term efficacy and safety of new biological agents

targeting the interleukin-23–T helper 17 pathway for
moderate-to-severe plaque psoriasis: a systematic review
and network meta-analysis

mez-Garc
ıa,1,2 D. Epstein,3 B. Isla-Tejera,2,4 A. Lorente,1,2 A. Ve
F. Go
lez Garc
ıa-Nieto1,2 and J. Ruano1,2
1
Departments of Dermatology and 4Pharmacy, Hospital Universitario Reina Sofıa, Menendez Pidal Ave, 14004 Cordoba, Spain
2
Instituto Maimonides de Investigacion Biomedica de Cordoba (IMIBIC)/Hospital Universitario Reina Sofıa/Universidad de Cordoba, Cordoba, Spain
3
Department of Applied Economics, School of Economics, University of Granada, Granada, Spain

Summary

Correspondence A new generation of biologics targeting the interleukin-23–T helper 17 pathway

Juan Ruano. has been developed. This study aimed to assess the short-term effectiveness and
E-mail: [email protected]
iD http://orcid.org/0000-0002-0286-4107 safety of these new agents using a network meta-analysis. Twenty-seven ran-
Accepted for publication domized clinical trials (10 629 patients) were identified by a comprehensive sys-
23 May 2016 tematic literature review (PROSPERO 2015: CRD42015025472). Quality of evi-
dence was assessed following Cochrane-compliant rules and the Grading of
Funding sources
This work has been funded in part by project
Recommendations, Assessment, Development and Evaluations approach. Efficacy
ICI1400136 to J.R., integrated into the National and safety outcomes at weeks 10–16 were compared using a random-effects net-
Plan of R+D+I 2008–2011 and cofinanced by work meta-analysis within a frequentist framework to estimate pooled odds ratios
the ISCIII-Subdireccion General de Evaluacion and (ORs) of direct and indirect comparisons among the therapeutic options. There
European Regional Development Fund (ERDF); and were six direct drug-to-drug comparisons in the network, with a high degree of
by grant PP13/009 of Plan Propio de movilidad
consistency between the direct and indirect evidence. From the available evi-
para investigadores del Instituto Maimonides de
Investigacion Biomedica de Cordoba (IMIBIC). No dence, infliximab 5 mg kg 1 every 8 weeks [OR 118
89, 95% confidence inter-
funding was received from any pharmaceutical val (CI) 60
91–232
04] and secukinumab 300 mg every 4 weeks (OR 87
07,
companies. 95% CI 55
01–137
82) are shown to be among the most effective short-term
treatments, but are ranked as the biologics most likely to produce any adverse
Conflicts of interest event or an infectious adverse event, respectively. Ustekinumab 90 mg every
J.R. has received honoraria for lecturing and grants
12 weeks, the third most efficacious treatment (OR 73
67, 95% CI 46
97–
for research from Pfizer, honoraria for lecturing
from Janssen-Cilag and Novartis, and other finan- 115
56), was the only agent that did not show increased risk of adverse events
cial benefits from AbbVie and Novartis; A.V.G.N. compared with placebo. Treatment recommendations should also consider long-
has received honoraria for lecturing from Pfizer, term outcomes and costs.
Novartis, AbbVie and Janssen-Cilag, and other
financial benefits from AbbVie, Novartis and Jans-
sen-Cilag; F.G.G. has received honoraria for
What’s already known about this topic?
research from Pfizer, and for lecturing from AbbVie
• Old antitumour necrosis factor-a agents and recently licensed drugs blocking the
and Janssen-Cilag. A.L. has received honoraria for interleukin-23–T helper 17 pathway are more efficacious than placebo for the
lecturing from AbbVie and other financial benefit
treatment of moderate-to-severe plaque psoriasis.
•
from Novartis. D.E. has received funding from
Traditionally, meta-analyses have usually compared only two interventions at a
Boehringer Ingelheim and Roche. B.I.T. has no
time, but the need to summarize a comprehensive and coherent set of comparisons
disclosures.
based on all of the available evidence has led more recently to synthetic methods
DOI 10.1111/bjd.14814 that address multiple interventions.

594 British Journal of Dermatology (2017) 176, pp594–603 © 2016 British Association of Dermatologists

mez-Garc
ıa et al. 595
Efficacy and safety of new biologics in psoriasis, F. Go

What does this study add?

• To the best of our knowledge, this is the first network meta-analysis to study the
comparative short-term efficacy and safety of these agents using the new 2015
PRISMA statement for network meta-analyses.
• Infliximab and secukinumab ranked as the most efficacious drugs, but with the
highest risk for any adverse event and associated infections, respectively.
• Ustekinumab was the agent with the best efficacy–safety profile.

Psoriasis is a chronic inflammatory skin disease mediated by review and meta-analysis performed by Nast et al. evaluated
the cells and components of both the innate and adaptive direct evidence of the long-term efficacy and safety of some
immune systems, affecting 1–3% of the general population.1 of these drugs.13 Only the last two were performed following
Tumour necrosis factor (TNF)-a antagonists have been at the the PRISMA-P statement, but none used the recently published
centre of treatment for patients with moderate-to-severe pla- PRISMA NMA checklist.15
que psoriasis who are unresponsive to, or intolerant of, non- This study aims to extend the existing NMAs to assess the
biological systemic agents.2,3 However, our understanding of direct and indirect evidence for the short-term efficacy and
this disease has progressed greatly, and new drugs targeting safety of new biologics targeting the IL-23–T helper 17 path-
the cytokines involved in the interleukin (IL)-23–T helper 17 way in comparison with anti-TNF-a drugs for the treatment
pathway have emerged.4 Several studies have shown that a of moderate-to-severe plaque psoriasis.
new generation of monoclonal antibodies that block IL-12/
23p40, IL-23p19, IL-17A or IL-17RA can reverse the clinical,
Materials and methods
histological and molecular features of psoriasis in approxi-
mately 70–80% of patients, compared with 45–50% in the
Systematic review
case of TNF-a antagonists.5–7
As patients with psoriasis require lifelong treatment, reliable This systematic review and meta-analysis was conducted in
evidence of the comparative benefits and harms of interventions accordance with the modified 32-item PRISMA extension
is needed to make clinical decisions regarding their use. Meta- statement for NMA of 2015.15 The selection of databases, eli-
analyses are conducted to assess the strength of recommenda- gibility criteria, outcomes of the review and analytical meth-
tions and quality of evidence available for a disease and multiple ods were defined a priori in an internal protocol and registered
treatment alternatives, improving the precision of estimates of on PROSPERO under the code CRD42015025472 (Table S1;
effect and answering questions not posed by the individual see Supporting Information). Interventional studies were eligi-
studies. In network meta-analyses (NMAs), several treatments ble for inclusion if they were randomized, placebo-controlled
can be compared by connecting evidence from clinical trials that or head-to-head trials published in English, of infliximab
have investigated two or more treatments. The resulting trial [5 mg kg 1 at weeks 0, 2 and 6 then every 8 weeks (Q8W)],
network may allow estimation of the relative effects of all pairs etanercept [50 mg twice weekly (BIW) for 12 weeks, then
of treatments, taking direct and indirect evidence into account.8 25 mg BIW or 50 mg Q1W], adalimumab (80 mg at week 0,
For this reason, they are gaining popularity among clinicians, 40 mg at week 1, then 40 mg Q2W), ustekinumab (45 mg
guideline developers and health technology agencies as evidence or 90 mg at weeks 0 and 4, then Q12W) or secukinumab
on new interventions. Systematic review authors and assessors (300 mg at weeks 0, 1, 2 and 3, then Q4W) as monotherapy
are now strongly encouraged to make use of the Preferred for the treatment of plaque psoriasis in adult patients.
Reporting Items for Systematic review and Meta-analysis Proto-
col (PRISMA-P) when drafting and appraising review protocols.
Databases searched and study identification
Five NMAs have compared the short-term efficacy of treat-
ments for moderate-to-severe psoriasis. Woolacott et al. and Details regarding the databases searched and study identifica-
Bansback et al. compared the efficacy of systemic treatments tion for this review are provided in Appendix S1 (see Support-
for moderate-to-severe plaque psoriasis including anti-TNF-a ing Information).
agents such as etanercept, adalimumab or infliximab with that
of other biologics and nonbiologics.9,10 Recently, three studies
Data extraction
included ustekinumab, an anti-IL-12/23p40 agent, in the
NMA.11–13 Signorovitch et al. used NMA to compare the effi- Treatment effects were evaluated based on the intention-to-
cacy of biological treatments for moderate-to-severe psoriasis, treat efficacy rates (PASI 75 and PASI 90; number of patients
adjusting the model for placebo responses.14 A systematic with five-point Investigator’s Global Assessment, Physician’s

© 2016 British Association of Dermatologists British Journal of Dermatology (2017) 176, pp594–603

mez-Garc
ıa et al.
596 Efficacy and safety of new biologics in psoriasis, F. Go

Global Assessment or static Physician’s Global Assessment of 0 duration was 18
3 years (range 10–23) and the disease sever-
or 1 at weeks 10–16; number of patients with Dermatology ity was evaluated across the trials with a baseline PASI score of
Life Quality Index of 0 or 1 at weeks 10–16) and safety 20
2 (range 15–27) and a body surface area of 28
1% (range
parameters [number of patients with at least one adverse event 20–45). Psoriatic arthritis was diagnosed in 27
3% of cases
(AE), number of patients with at least one serious AE (SAE), (range 14–78). Trials included 67% men (range 54–79). The
number of patients with at least one infectious AE, and num- mean age was 45 years (range 35–51) and the mean weight
ber of patients withdrawing owing to AE] reported in the ran- 90 kg (range 79–99). The majority of the patients were
domized trials identified during the systematic review. PASI 75 white. Detailed information for all of the included studies is
and PASI 90 represent ≥ 75% and ≥ 90% reduction, respec- presented in Table S3. A network graph summarizing the
tively, in the PASI score with respect to baseline. For more comparisons is provided in Figure 1.
details, see the PROSPERO register file and PRISMA NMA 2015
checklist (Tables S1, S2; see Supporting Information).
Risk of bias
The risk of bias among the included studies was rated as ‘low
Quality of evidence and risk-of-bias assessment
risk’ or ‘unclear risk’ (Fig. S2; see Supporting Information).
Details regarding the quality of evidence and risk-of-bias Of the 27 included clinical trials, 21 (78%) reported an ade-
assessment for this review are provided in Appendix S1 (see quate randomization method, and allocation concealment was
Supporting Information).16 properly ensured in 18 studies (67%). In all studies, the
blinding of participants and personnel was sufficient. In 21
trials (78%), the risk of attrition bias was low, as incomplete
Meta-analysis of direct treatment effects
outcome data were sufficiently addressed. The risk of report-
Data extracted from trials were combined by a random-effects ing bias was low in most of the studies (78%).
model, with effect sizes expressed as the odds ratio (OR) of
achieving each outcome at weeks 10–16 in the treatment arm
Quality of evidence
vs. the control arm. Total effect size was calculated by the Man-
tel–Haenszel method. Heterogeneity was evaluated with I2 cal- When considering efficacy parameters, trials were of generally
culations. Statistical analysis was performed with RevMan 5
3 moderate quality according to GRADEpro assessment
(http://tech.cochrane.org/revman), with two-tailed P-values (Table S4; see Supporting Information). However, there was a
< 0
05 considered significant. Forest plots and funnel plots wide range of values related to quality assessment of safety
were obtained for each outcome analysed at 10–16 weeks. variables. Studies comparing agents head to head showed bet-
ter quality of evidence than those reporting placebo as the
comparator (Fig. S3; see Supporting Information). This was
Network meta-analysis
less evident when the comparator was another biologic
NMA was used to make mixed comparisons among the thera- belonging to a different pharmaceutical company.
peutic options and to rank the treatments, using the package
mvmeta in Stata version 12
0 (Stata Corp, College Station, TX,
Efficacy and safety of direct comparisons of monotherapy
U.S.A.) (Appendix S2; see Supporting Information). Inconsis-
vs. placebo
tency between direct and indirect evidence in the network
was analysed using the ratio of odds ratios. A value ≥ 2 indi- All biologics showed superior efficacy over placebo with
cates that the direct and indirect treatment comparisons may respect to all efficacy outcomes (Table 1; Fig. 2; Fig. S4
2,
be inconsistent.17 4
4, 4
6, 4
8–12; see Supporting Information). Secukinumab
300 mg Q4W and infliximab 5 mg kg 1 Q8W were the most
effective biologics, as demonstrated by their PASI 75 and PASI
Results
90 responses.
All treatments showed a higher OR for ‘at least one AE’,
Results of the search
although the ORs for adalimumab and for all doses of ustek-
During the abstract review phase, 2025 records were identi- inumab were not statistically significant (Fig. S4
9). Compared
fied (Fig. S1; see Supporting Information). Twenty-seven stud- with placebo, no significant differences in the risk of ‘at least
ies assessed for eligibility met the inclusion criteria and were one infectious AE’ were shown for etanercept 25 mg BIW,
included in the qualitative and quantitative analysis (10 629 ustekinumab 45 mg Q12W or ustekinumab 90 mg Q12W
randomized patients: 6540 to biologics and 4089 to conven- (Fig. S4
10). After short-term treatment, no significant differ-
tional treatments) (Table S3; see Supporting Information).18–42 ences were observed in the risks of ‘at least one SAE’ for any
There are only three included head-to-head trials reporting agent (Fig. S4
11). When considering ‘withdrawal due to AE’,
efficacy and safety data. The study sample size varied from 33 no significant risk differences were found, except for ustek-
to 1230. For the majority of trials, the double-blind period inumab 90 mg Q12W, which showed a lower risk (OR 0
14,
comprised 12 weeks (range 10–16). The mean prior disease 95% CI 0
03–0
63) (Fig. S4
12).

British Journal of Dermatology (2017) 176, pp594–603 © 2016 British Association of Dermatologists

mez-Garc
ıa et al. 597
Efficacy and safety of new biologics in psoriasis, F. Go

Fig 1. Network graph. Treatments are represented by nodes, and head-to-head studies between treatments are represented by lines. The area of
the node circle is proportional to the number of studies including that treatment, and the width of the line is proportional to the average
effectiveness in the placebo arms of the studies. There were 15 direct comparisons in the network: nine of biologics vs. placebo and six of
biologics against other biologics. The graph contains three loops and three closed loops. Both anti-interleukin-23–T helper 17 and antitumour
necrosis factor-a agents are represented in them. Etanercept 50 mg twice a week (BIW) is the node with the highest connectivity (excluding
placebo). Infliximab and adalimumab, two antitumour necrosis factor-a agents, are not present in any direct comparison with other biologic.
Q2W, every 2 weeks.

Table 1 Results of pooled odds ratios of the direct comparisons of each biologic included in the network vs. placebo

Number of trials
Treatment Comparator PASI 75 PASI 90 making direct comparison
Infliximab 5 mg kg 1 Q8W Placebo 118
9 (60
9–232
0) 84
1 (31
0–228
5) 5
Secukinumab 300 mg Q4W Placebo 87
1 (55
0–137
8) 96
0 (48
8–188
6) 4
Ustekinumab 90 mg Q12W Placebo 73
7 (47
0–115
6) 61
3 (13
1–287) 2
Ustekinumab 45 mg Q12W Placebo 56
2 (36
0–87
8) 56
0 (20
6–152
2) 2
Adalimumab 40 mg Q2W Placebo 30
7 (21
5–43
9) 22
1 (8
2–60
0) 4
Etanercept 50 mg BIW Placebo 17
9 (14
0–22
9) 16
5 (9
8–28
0) 9
Etanercept 25 mg BIW Placebo 16
1 (9
2–23
0) 15
1 (5
1–44
8) 4

Values are the odds ratio (95% confidence interval). PASI 75, ≥ 75% improvement in Psoriasis Area and Severity Index; BIW, twice a week;
Q2W, every 2 weeks.

differences between the treatments in any AE (Fig. S5; see

Mixed treatment comparisons
Supporting Information). However, as already noted, use of
When considering PASI 75 response, adalimumab 40 mg secukinumab 300 mg Q4W carries a significantly greater risk
Q2W is 2
5-fold more effective than etanercept 25 mg BIW of infectious AEs than etanercept 50 mg BIW (OR 1
64, 95%
(OR 2
52, 95% CI 1
46–4
35), while infliximab 5 mg kg 1 CI 1
04–2
59) and ustekinumab 45 mg or 90 mg Q12W (OR
Q8W (OR 0
26, 95% CI 0
12–0
56) and secukinumab 2
56, 95% CI 1
23–5
26).
300 mg Q4W (OR 0
33, 95% CI 0
20–0
55) are about four-
and threefold more effective than adalimumab 40 mg Q2W,
Ranking of treatments by efficacy
respectively (Table 2).
Infliximab 5 mg kg 1 Q8W and secukinumab 300 mg Q4W
are ranked the most effective (PASI 75 and PASI 90, respec-
Inconsistency analysis
tively), followed by ustekinumab 90 mg Q12W (Table 4).
The indirect comparison of etanercept 50 mg BIW vs. ustek- The probability that secukinumab 300 mg Q4W will be the
inumab 45 mg Q12W or ustekinumab 90 mg Q12W showed most effective treatment option (PASI 90) for particular cases
a treatment effect about twice as large as the direct evidence is almost 46
9%, compared with infliximab 5 mg kg 1 Q8W
(Table 3). The direct evidence (comparing one biologic with (44
6%) and ustekinumab 90 mg Q12W (4
3%) (Fig S6; see
another head to head) does not show any significant Supporting Information).

© 2016 British Association of Dermatologists British Journal of Dermatology (2017) 176, pp594–603

mez-Garc
ıa et al.
598 Efficacy and safety of new biologics in psoriasis, F. Go

Fig 2. Scatter plot of surface under the cumulative ranking (SUCRA) values for efficacy vs. safety outcomes at weeks 10–16 according to
treatments and targeting pathways. This graph displays four separate sets of data using two y-axes for SUCRA of ≥ 75% improvement in
Psoriasis Area and Severity Index (PASI 75) and PASI 90, and two x-axes for SUCRA of patients with at least one adverse event and patients
with at least one infectious adverse event, based on the mixed treatment comparisons. Each treatment is represented by a geometric figure
created by crossing the projection of these SUCRA values. Blue dotted lines represent SUCRA for PASI 90 values. Yellow dashed lines represent
SUCRA for patients with at least one infectious adverse event. Points closer to the bottom left of the figure are relatively less effective and have
fewer adverse events, while points closer to the top right are relatively more effective and relatively more at risk of adverse events. ADA,
adalimumab; BIW, twice a week; ETN, etanercept; IL, interleukin; INFLIX, infliximab; SECUK, secukinumab; TNF, tumour necrosis factor; USK,
ustekinumab.

Ranking of treatments by safety

Discussion
This review is the first study to evaluate the short-term effects
Etanercept 25 mg BIW ranked most likely to produce any AE
of secukinumab in patients with psoriasis, in addition to an
(Table 4). This arises principally from a single trial,26 which
agent that blocks IL-12/23p40 (ustekinumab) and the classical
estimated an OR of 4
59 (95% CI 1
88–11
20) compared with
anti-TNF-a drugs (adalimumab, etanercept and infliximab).
placebo (Figs S4
9, S7; see Supporting Information). It seems
Our results suggest a division of treatments in terms of short-
clinically implausible that a lower dose would produce consid-
term effectiveness, with infliximab and secukinumab as the
erably more AEs than a higher dose, and this result may be
most effective agents and adalimumab and etanercept as the
due to a different patient population or methods of outcome
least effective options. The efficacy of ustekinumab is posi-
measurement. Among the other treatments, infliximab
tioned at an intermediate point between those drugs, but with
5 mg kg 1 Q8W is ranked the second highest for risk of any
a higher overall quality of evidence and better safety profile
AE. Secukinumab 300 mg Q4W is ranked as the treatment
than infliximab and secukinumab.
most likely to produce an infectious AE (OR 2
05, 95% CI
Given that head-to-head comparisons between biologics are
1
57–2
67) (Fig. S4
11).
scarce, one of the strengths of this study was the use of NMA.
In the absence of trials involving a direct comparison of treat-
Efficacy vs. safety at weeks 10–16 according to ments of interest, an indirect comparison can provide useful
treatments and targeting pathways evidence for the difference in treatment effects between com-
peting interventions (which otherwise would be lacking) and
Infliximab 5 mg kg 1 Q8W is shown to be among the most
for judiciously selecting the best choice(s) of treatment. For
effective treatments in terms of PASI 75 response, but is
example, in our study we were able indirectly to compare
ranked as the biologic most likely to produce an AE. However,
infliximab 8 mg Q8W and adalimumab 40 mg Q2W against
infliximab 5 mg kg 1 Q8W is ranked as one of the treatments
any other treatment, comparisons that have never been made
least likely to produce any infectious AEs. On the other hand,
in any randomized controlled trial.
secukinumab 300 mg Q4W is also among the most effective
Several factors typically need to be taken into account
treatments in terms of PASI 90 response, but is ranked most
when recommending an intervention, not only its efficacy.
likely to produce an infectious AE.

British Journal of Dermatology (2017) 176, pp594–603 © 2016 British Association of Dermatologists

mez-Garc
ıa et al. 599
Efficacy and safety of new biologics in psoriasis, F. Go

Table 2 Mixed treatment comparisons estimated using random-effects network meta-analysis for ≥ 75% and ≥ 90% improvement in Psoriasis Area
and Severity Index (PASI 75 and PASI 90) at weeks 10–16

Treatment 1 Treatment 2 PASI 75 PASI 90

Adalimumab 40 mg Q2W Etanercept 25 mg BIW 2
52 (1
46–4
35) 2
49 (0
95–6
52)
Etanercept 50 mg BIW 1
44 (0
92–2
26) 1
23 (0
55–2
73)
Infliximab 5 mg kg 1 Q8W 0
26 (0
12–0
56) 0
26 (0
07–0
92)
Secukinumab 300 mg Q4W 0
33 (0
20–0
55) 0
25 (0
10–0
65)
Ustekinumab 45 mg Q12W 0
74 (0
43–1
26) 0
53 (0
21–1
31)
Ustekinumab 45 mg per 90 mg Q12W 0
90 (0
50–1
62) 0
70 (0
22–2
26)
Ustekinumab 90 mg Q12W 0
57 (0
34–0
98) 0
44 (0
18–1
10)
Etanercept 25 mg BIW Etanercept 50 mg BIW 0
57 (0
41–0
79) 0
49 (0
28–0
87)
Infliximab 5 mg kg 1 Q8W 0
10 (0
05–0
22) 0
10 (0
03–0
37)
Secukinumab 300 mg Q4W 0
13 (0
08–0
21) 0
10 (0
05–0
23)
Ustekinumab 45 mg Q12W 0
29 (0
19–0
46) 0
21 (0
10–0
46)
Ustekinumab 45 mg per 90 mg Q12W 0
36 (0
21–0
62) 0
28 (0
10–0
82)
Ustekinumab 90 mg Q12W 0
23 (0
15–0
35) 0
18 (0
08–0
39)
Etanercept 50 mg BIW Infliximab 5 mg kg 1 Q8W 0
18 (0
09–0
36) 0
21 (0
07–0
67)
Secukinumab 300 mg Q4W 0
23 (0
17–0
32) 0
21 (0
11–0
38)
Ustekinumab 45 mg Q12W 0
51 (0
37–0
71) 0
43 (0
25–0
74)
Ustekinumab 45 mg per 90 mg Q12W 0
63 (0
40–0
99) 0
57 (0
23–1
43)
Ustekinumab 90 mg Q12W 0
40 (0
29–0
55) 0
36 (0
21–0
62)
1
Infliximab 5 mg kg Q8W Secukinumab 300 mg Q4W 1
30 (0
61–2
75) 0
98 (0
28–3
39)
Ustekinumab 45 mg Q12W 2
89 (1
34–6
22) 2
02 (0
60–6
85)
Ustekinumab 45 mg per 90 mg Q12W 3
54 (1
59–7
90) 2
71 (0
65–11
21)
Ustekinumab 90 mg Q12W 2
25 (1
05–4
83) 1
71 (0
51–5
77)
Secukinumab 300 mg Q4W Ustekinumab 45 mg Q12W 2
23 (1
44–3
47) 2
07 (0
95–4
54)
Ustekinumab 45 mg per 90 mg Q12W 2
73 (1
79–4
18) 2
77 (1
40–5
51)
Ustekinumab 90 mg Q12W 1
74 (1
13–2
68) 1
75 (0
80–3
82)
Ustekinumab 45 mg Q12W Ustekinumab 45 mg per 90 mg Q12W 1
22 (0
72–2
10) 1
34 (0
47–3
79)
Ustekinumab 90 mg Q12W 0
78 (0
62–0
99) 0
84 (0
57–1
25)
Ustekinumab 45 mg per 90 mg Q12W Ustekinumab 90 mg Q12W 0
64 (0
38–1
08) 0
63 (0
22–1
78)

Values are the odds ratio (95% confidence interval). A mixed treatment comparison is calculated from both the direct and indirect evidence,
assuming a common between-study variance across all the comparisons in the network. Q2W, every 2 weeks; BIW, twice a week.

Many systematic reviews therefore examine measures of both statement has been published for the reporting of systematic
efficacy and safety, and the ranking of competing treatments reviews incorporating NMAs.15 This extension adds five new
for these two outcomes might differ considerably. In our items that authors should consider when reporting an NMA,
study we decided to take into account both sources of infor- as well as 11 modifications to the existing PRISMA items. To
mation together to display the relative position of every the best of our knowledge no meta-analysis has previously
treatment according to these variables. ORs for infectious AEs evaluated the short-term efficacy of all of the approved drugs
were higher for secukinumab 300 mg Q4W, adalimumab for psoriasis treatment, including secukinumab, based on these
40 mg Q2W and etanercept 50 mg BIW than with placebo. reporting guidelines.
In the majority of cases, infectious AEs were moderate or Our study is limited to using PASI 75 and PASI 90 as the
mild. Indeed, severe cases of infection were probably primary end points of the trials. Nevertheless, the lack of PASI
accounted for as SAEs, and are thus difficult to identify in 90 values in older clinical trials makes it difficult to compare
most RCTs. them with new ones. Reich et al. handled the missing data by
In the case of safety, the quality of the long-term studies jointly modelling PASI 50, PASI 75 and PASI 90 achievement
evaluated by Nast et al. was considered moderate or low, with in a Bayesian hierarchical framework.12 Signorovitch et al. tried
the quality of evidence strongly limited.13 This situation is to address this problem by modelling the relative risk of PASI
very similar to that which we found in relation to short-term 90 score adjusted by the response rate in the placebo group in
safety. This highlights the low quality of evidence of AE- each trial, but this adjustment did not substantially change the
related published data provided by authors, which is in con- results of the NMA.14 We believe the more traditional analysis
trast to the efficacy of these drugs. shown here, although restricted to PASI 75 and PASI 90, is
Despite this limitation, the use of the PRISMA statement more transparent and more robust as it requires fewer mod-
when drafting and appraising review protocols has improved elling assumptions.
the quality of the evidence from systematic reviews and Another limitation of our study is that response to treat-
meta-analyses. Recently, a modified 32-item PRISMA extension ment over such short time periods may not be representative

© 2016 British Association of Dermatologists British Journal of Dermatology (2017) 176, pp594–603

mez-Garc
ıa et al.
600 Efficacy and safety of new biologics in psoriasis, F. Go

Table 3 Pooled odds ratios of the indirect, direct and mixed treatment comparisons of each head-to-head analysis included in the network. This
table summarizes the direct treatment comparisons for efficacy (≥ 75% and ≥ 90% improvement in Psoriasis Area and Severity Index, PASI 75 and
PASI 90) and safety (patients with at least one adverse event or patients with at least one infectious adverse event), and shows the indirect and
mixed comparisons calculated by the network meta-analysis

Indirect treatment Direct treatment Mixed treatment

Treatment 1 Treatment 2 comparisonsa comparisonsb comparisonsc RoR P-value nd
PASI 75
Secukinumab 300 mg Q4W Etanercept 50 mg BIW 4
74 (2
92–7
70) 4
28 (3
05–6
02) 4
37 (3
16–6
05) 1
10 0
85 1
Ustekinumab 90 mg Q12W Etanercept 50 mg BIW 4
12 (2
46–6
87) 2
15 (1
55–2
94) 2
50 (1
82–3
44) 1
92 0
034* 1
Ustekinumab 45 mg Q12W Etanercept 50 mg BIW 3
14 (1
88–5
23) 1
57 (1
10–2
26) 1
95 (1
40–2
73) 1
99 0
030* 1
Ustekinumab 90 mg Q12W Ustekinumab 45 mg Q12W 1
31 (0
69–2
47) 1
28 (0
97–1
69) 1
28 (1
01–1
62) 1
02 0
81 1
Ustekinumab 45 mg Secukinumab 300 mg Q4W 0
40 (0
20–0
79) 0
35 (0
21–0
58) 0
36 (0
24–0
56) 1
15 0
92 1
per 90 mg Q12W
Etanercept 50 mg BIW Etanercept 25 mg BIW 1
11 (0
60–2
09) 1
89 (1
40–2
56) 1
75 (1
26–2
42) 1
63 0
11 2
PASI 90
Secukinumab 300 mg Q4W Etanercept 50 mg BIW 6
09 (2
37–15
68) 4
41 (3
19–6
39) 4
83 (2
59–8
98) 1
35 0
59 1
Ustekinumab 90 mg Q12W Etanercept 50 mg BIW 3
46 (1
26–9
44) 2
69 (1
94–3
73) 2
75 (1
58–4
80) 1
28 0
78 1
Ustekinumab 45 mg Q12W Etanercept 50 mg BIW 3
36 (1
23–9
17) 1
91 (1
31–2
78) 2
33 (1
32–4
10) 1
76 0
44 1
Ustekinumab 90 mg Q12W Ustekinumab 45 mg Q12W 1
02 (0
30–3
50) 1
18 (0
84–1
67) 1
18 (0
78–1
77) 1
15 0
61 1
Ustekinumab 45 mg Secukinumab 300 mg Q4W – 0
36 (0
25–0
50) 0
36 (0
17–0
73) – – 1
per 90 mg Q12W
Etanercept 50 mg BIW Etanercept 25 mg BIW 1
08 (0
31–3
75) 2
14 (1
41–3
24) 2
03 (1
15–3
58) 1
98 0
18 2
Any adverse event
Secukinumab 300 mg Q4W Etanercept 50 mg BIW 1
01 (0
75–1
37) 0
96 (0
70–1
31) 1
02 (0
83–1
26) 1
05 0
85 1
Ustekinumab 90 mg Q12W Etanercept 50 mg BIW 0
75 (0
55–1
03) 0
96 (0
69–1
33) 0
80 (0
65–0
99) 1
28 0
23 1
Ustekinumab 45 mg Q12W Etanercept 50 mg BIW 0
95 (0
69–1
30) 0
83 (0
58–1
20) 0
91 (0
73–1
13) 1
14 0
61 1
Ustekinumab 90 mg Q12W Ustekinumab 45 mg Q12W 0
80 (0
54–1
17) 0
88 (0
70–1
11) 0
88 (0
73–1
07) 1
1 0
13 1
Ustekinumab 45 mg Secukinumab 300 mg Q4W 0
92 (0
57–1
46) 0
78 (0
57–1
07) 0
84 (0
65–1
07) 1
09 0
50 1
per 90 mg Q12W
Etanercept 50 mg BIW Etanercept 25 mg BIW 0
58 (0
37–0
90) 1
05 (0
70–1
57) 0
73 (0
51–1
03) 1
81 0
10 2
Infectious adverse event
Secukinumab 300 mg Q4W Etanercept 50 mg BIW 1
64 (1
04–2
59) 1
15 (0
81–1
64) 1
38 (0
98–1
96) 1
42 0
30 1
Ustekinumab 90 mg Q12W Etanercept 50 mg BIW 0
83 (0
51–1
35) 1
03 (0
74–1
43) 0
90 (0
65–1
26) 1
24 0
67 1
Ustekinumab 45 mg Q12W Etanercept 50 mg BIW 0
99 (0
62–1
60) 1
08 (0
74–1
56) 1
03 (0
73–1
44) 1
09 0
90 1
Ustekinumab 90 mg Q12W Ustekinumab 45 mg Q12W 0
84 (0
48–1
45) 0
88 (0
66–1
17) 0
88 (0
65–1
18) 1
05 0
66 1
Ustekinumab 45 mg Secukinumab 300 mg Q4W 0
39 (0
19–0
81) 0
82 (0
58–1
15) 0
65 (0
43–0
99) 2
10 0
029* 1
per 90 mg Q12W
Etanercept 50 mg BIW Etanercept 25 mg BIW 1
04 (0
59–1
82) 0
97 (0
54–1
74) 0
99 (0
62–1
59) 1
07 0
93 2

Values are the odds ratio (95% confidence interval). RoR, ratio of odds ratios; Q2W, every 2 weeks; BIW, twice a week. aEstimated using
placebo as a common comparator. bEstimated directly from the trial using random-effects meta-analysis. cEstimated using random-effects net-
work meta-analysis. dNumber of clinical trials making a direct comparison. *Statistically significant.

of the long-term effects of treatment, and the side-effect pro- of-treatment comparisons reflect a predominant use of placebo
file of individual treatments may be an important determinant as a comparator, resulting in a lack of direct comparison of
of long-term success. In most of the long-term studies, the biologics, which represented only 40% of the direct compar-
placebo groups are discontinued after 10–16-weeks of induc- isons in the network and 11% of the 27 RCTs reviewed in this
tion, making it difficult to obtain indirect evidence by means study. Estellat et al. found similar results (27
8% and 9
5%,
of NMA beyond this period. Recently, Nast et al. found inflix- respectively) when they analysed randomized controlled trials
imab, secukinumab and ustekinumab, followed by adali- of biologics for rheumatoid arthritis.43
mumab and etanercept, to be the most efficacious long-term Another potential limitation is the inconsistency in some
treatments in patients with moderate-to-severe psoriasis,13 comparisons. In our study, we found significant inconsistency
which is in line with the results reported herein. They between indirect and direct PASI 75 estimates for etanercept
employed an imputation approach to obtain long-term efficacy 50 mg BIW vs. ustekinumab 45 mg Q12W or ustekinumab
data. This approach led to ‘serious’ or ‘very serious’ risk of 90 mg Q12W comparisons. Inconsistency in the treatment
bias and an overall quality-of-evidence score of ‘low’ for most effect might arise from heterogeneity in the underlying sever-
of the efficacy outcomes. ity of disease in the populations across different studies. How-
One of the key elements of comparative treatment effective- ever, we did not find any differences in variables such as
ness research is head-to-head trials. In our review, network- weight or ethnicity among agents or studies.

British Journal of Dermatology (2017) 176, pp594–603 © 2016 British Association of Dermatologists

mez-Garc
ıa et al. 601
Efficacy and safety of new biologics in psoriasis, F. Go

Table 4 Ranking by the effectiveness and safety of the treatments included in the network

PASI 75 at weeks PASI 90 at weeks Patients with at least Patients with at least
10–16 10–16 one AE one infectious AE
Mean Mean PrHighest Mean PrHighest Mean
Treatment SUCRA PrBest rank SUCRA PrBest rank SUCRA risk rank SUCRA risk rank
Infliximab 5 mg kg 1 Q8W 95
9 70
8 1
3 86
7 46
2 2
1 90 42
1 1
8 29
2 1
3 6
7
Secukinumab 300 mg Q4W 90
9 28
9 1
7 91 46
7 1
7 64
9 1
3 3
8 87
9 52
2 2
Ustekinumab 90 mg Q12W 73
6 0
3 3
1 71
7 4
3 3
3 21
2 0 7
3 39
1 1
1 5
9
Ustekinumab 45 mg Q12W 58
3 0 4
3 63
4 1
9 3
9 43
8 0
1 4
5 58
8 6
7 4
3
Ustekinumab 45 mg 48
9 0 5
1 52
9 0
5 4
8 33
1 0 6
4 38 2
5 6
per 90 mg Q12W
Adalimumab 40 mg Q2W 43
2 0 5
5 39
7 0
4 5
8 37
4 0
3 6 71
8 22
3 3
3
Etanercept 50 mg BIW 26
6 0 6
9 30
6 0 6
6 61
4 0
4 4
1 52
6 3
9 4
8
Etanercept 25 mg BIW 12
6 0 8 14
1 0 7
9 92
4 55
8 1
6 52
5 10 4
8
Placebo 0 0 9 0 0 9 5
7 0 8
5 20 0 7
4

PASI 75, ≥ 75% improvement in Psoriasis Area and Severity Index; AE, adverse event; SUCRA, surface under the cumulative ranking; PrBest,
probability of best treatment; PrHighest risk, probability of the highest risk of AEs. Q2W, every 2 weeks; BIW, twice a week.

In our study, visual inspection of funnel plots revealed pos- and indirect comparison of the efficacy and safety of psoriasis
sible publication bias in many of the active vs. placebo com- biological agents after long-term follow-up.
parisons. However, an asymmetrical funnel plot should not be
equated with publication bias. In our case, many trials with
Acknowledgments
different drugs were plotted in the same funnel plot. Taking
into account all plotted studies as if they were related to the The authors would like to thank Editage (www.editage.com)
same drug is not appropriate, and it may be the reason for for English-language editing.
this visual bias. Nevertheless, there seems to be publication
bias against null results for PASI 75 and PASI 90 outcomes for
References
secukinumab trials compared with other agents.
Finally, this meta-analysis includes only all currently 1 Lowes MA, Su
arez-Fari~ nas M, Krueger JG. Immunology of psoria-
licensed psoriasis biologics. Some published phase II and sis. Annu Rev Immunol 2014; 32:227–55.
2 Rustin MH. Long-term safety of biologics in the treatment of
phase III clinical trials have studied other biologics that block
moderate-to-severe plaque psoriasis: review of current data. Br J
IL-23p19 (guselkumab,44 tildrakizumab),45 IL-17A (ixek- Dermatol 2012; 167(Suppl. 3):3–11.
izumab)29 or IL-17RA (brodalumab)42,46 and investigated the 3 Pathirana D, Ormerod AD, Saiag P et al. European S3-guidelines on
dose–response relationship for the drug. Biologics that are the systemic treatment of psoriasis vulgaris. J Eur Acad Dermatol Vener-
pending licensing were not included in this study for two rea- eol 2009; 23(Suppl. 2):1–70.
sons. Firstly, only approved drugs provide estimates of com- 4 Di Cesare A, Di Meglio P, Nestle FO. The IL-23/Th17 axis in the
parative effectiveness that will be potentially useful to current immunopathogenesis of psoriasis. J Invest Dermatol 2009;
129:1339–50.
decision makers. Secondly, addressing the PROSPERO rules
5 Croxtall JD. Ustekinumab: a review of its use in the manage-
requires submitting the study protocol to the database at least ment of moderate to severe plaque psoriasis. Drugs 2011;
6 months before the anticipated completion date, and this will 71:1733–53.
undoubtedly help to reduce unplanned duplication and 6 Levin AA, Gottlieb AB. Specific targeting of interleukin-23p19 as
increase transparency, helping safeguard against selective effective treatment for psoriasis. J Am Acad Dermatol 2014; 70:
reporting. 555–61.
In conclusion, from the available evidence, infliximab, 7 Chiricozzi A, Krueger JG. IL-17 targeted therapies for psoriasis.
Expert Opin Investig Drugs 2013; 22:993–1005.
secukinumab and ustekinumab were found to be the most
8 Bafeta A, Trinquart L, Seror R, Ravaud P. Reporting of results from
efficacious short-term treatments for moderate-to-severe pla- network meta-analyses: methodological systematic review. BMJ
que psoriasis. However, infliximab and secukinumab showed 2014; 348:g1741.
the highest risk for any AEs and associated mild-to-moderate 9 Woolacott N, Hawkins N, Mason A et al. Etanercept and efal-
infections, respectively, while ustekinumab was the agent with izumab for the treatment of psoriasis: a systematic review. Health
the best efficacy–safety profile. Our results could potentially Technol Assess 2006; 10:1–233.
aid the future assessment of the incremental cost-effectiveness 10 Bansback N, Sizto S, Sun H et al. Efficacy of systemic treatments
for moderate to severe plaque psoriasis: systematic review and
of alternative treatments, and may provide a useful basis for
meta-analysis. Dermatology 2009; 219:209–18.
the preparation of treatment guidelines for the use of a new 11 Lin VW, Ringold S, Devine EB. Comparison of ustekinumab with
generation of biological therapies in moderate-to-severe pla- other biological agents for the treatment of moderate to severe
que psoriasis. Further research is warranted to enable direct

© 2016 British Association of Dermatologists British Journal of Dermatology (2017) 176, pp594–603

mez-Garc
ıa et al.
602 Efficacy and safety of new biologics in psoriasis, F. Go

plaque psoriasis: a Bayesian network meta-analysis. Arch Dermatol moderate to severe chronic plaque psoriasis. Br J Dermatol 2011;
2012; 148:1403–10. 165:652–60.
12 Reich K, Burden AD, Eaton JN, Hawkins NS. Efficacy of biologics 29 Griffiths CE, Reich K, Lebwohl M et al. Comparison of ixekizumab
in the treatment of moderate to severe psoriasis: a network meta- with etanercept or placebo in moderate-to-severe psoriasis
analysis of randomized controlled trials. Br J Dermatol 2012; (UNCOVER-2 and UNCOVER-3): results from two phase 3 ran-
166:179–88. domised trials. Lancet 2015; 386:541–51.
13 Nast A, Jacobs A, Rosumeck S, Werner RN. Efficacy and safety of 30 Bachelez H, van de Kerkhof PC, Strohal R et al. Tofacitinib versus
systemic long-term treatments for moderate-to-severe psoriasis – a etanercept or placebo in moderate-to-severe chronic plaque psoriasis:
systematic review and meta-analysis. J Invest Dermatol 2015; a phase 3 randomised non-inferiority trial. Lancet 2015; 386:552–61.
135:2641–8. 31 Gordon KB, Langley RG, Gottlieb AB et al. A phase III, randomized,
14 Signorovitch JE, Betts KA, Yan YS et al. Comparative efficacy of controlled trial of the fully human IL-12/23 mAb briakinumab in
biological treatments for moderate-to-severe psoriasis: a network moderate-to-severe psoriasis. J Invest Dermatol 2012; 132:304–14.
meta-analysis adjusting for cross-trial differences in reference arm 32 Menter A, Tyring SK, Gordon K et al. Adalimumab therapy for
response. Br J Dermatol 2015; 172:504–12. moderate to severe psoriasis: a randomized, controlled phase III
15 Hutton B, Salanti G, Caldwell DM et al. The PRISMA extension trial. J Am Acad Dermatol 2008; 58:106–15.
statement for reporting of systematic reviews incorporating net- 33 Saurat JH, Stingl G, Dubertret L et al. Efficacy and safety results
work meta-analyses of health care interventions: checklist and from the randomized controlled comparative study of adalimumab
explanations. Ann Intern Med 2015; 162:777–84. vs. methotrexate vs. placebo in patients with psoriasis (CHAM-
16 The Cochrane Collaboration. Assessing risk of bias in included PION). Br J Dermatol 2008; 158:558–66.
studies. Available at: http://methods.cochrane.org/bias/assessing- 34 Gordon KB, Duffin KC, Bissonnette R et al. A phase 2 trial of
risk-bias-included-studies (last accessed 25 August 2016). guselkumab versus adalimumab for plaque psoriasis. N Engl J Med
17 Salanti G, Ades AE, Ioannidis JPA. Graphical methods and numeri- 2015; 373:136–44.
cal summaries for presenting results from multiple-treatment 35 Leonardi CL, Kimball AB, Papp KA et al. Efficacy and safety of
meta-analysis: an overview and tutorial. J Clin Epidemiol 2011; ustekinumab, a human interleukin-12/23 monoclonal antibody, in
64:163–71. patients with psoriasis: 76-week results from a randomised, dou-
18 Chaudhari U, Romano P, Mulcahy LD et al. Efficacy and safety of ble-blind, placebo-controlled trial (PHOENIX 1). Lancet 2008;
infliximab monotherapy for plaque-type psoriasis: a randomised 371:1665–74.
trial. Lancet 2001; 357:1842–7. 36 Papp KA, Langley RG, Lebwohl M et al. Efficacy and safety of
19 Gottlieb AB, Evans R, Li S et al. Infliximab induction therapy for ustekinumab, a human interleukin-12/23 monoclonal antibody, in
patients with severe plaque-type psoriasis: a randomized, double- patients with psoriasis: 52-week results from a randomised, dou-
blind, placebo-controlled trial. J Am Acad Dermatol 2004; 51: ble-blind, placebo-controlled trial (PHOENIX 2). Lancet 2008;
534–42. 371:1675–84.
20 Reich K, Nestle FO, Papp K et al. Infliximab induction and mainte- 37 Griffiths CE, Strober BE, van de Kerkhof P et al. Comparison of
nance therapy for moderate-to-severe psoriasis: a phase III, multi- ustekinumab and etanercept for moderate-to-severe psoriasis. N
centre, double-blind trial. Lancet 2005; 366:1367–74. Engl J Med 2010; 362:118–28.
21 Menter A, Feldman SR, Weinstein GD et al. A randomized compar- 38 Langley RG, Elewski BE, Lebwohl M et al. Secukinumab in plaque
ison of continuous vs. intermittent infliximab maintenance regi- psoriasis – results of two phase 3 trials. N Engl J Med 2014;
mens over 1 year in the treatment of moderate-to-severe plaque 371:326–38.
psoriasis. J Am Acad Dermatol 2007; 56(31):e1–15. 39 Blauvelt A, Prinz JC, Gottlieb AB et al. Secukinumab administration
22 Gottlieb AB, Matheson RT, Lowe N et al. A randomized trial of by pre-filled syringe: efficacy, safety and usability results from a
etanercept as monotherapy for psoriasis. Arch Dermatol 2003; randomized controlled trial in psoriasis (FEATURE). Br J Dermatol
139:1627–32. 2015; 172:484–93.
23 Leonardi CL, Powers JL, Matheson RT et al. Etanercept as 40 Paul C, Lacour JP, Tedremets L et al. Efficacy, safety and usability
monotherapy in patients with psoriasis. N Engl J Med 2003; of secukinumab administration by autoinjector/pen in psoriasis: a
349:2014–22. randomized, controlled trial (JUNCTURE). J Eur Acad Dermatology
24 Papp KA, Tyring S, Lahfa M et al. A global phase III randomized Venereol 2015; 29:1082–90.
controlled trial of etanercept in psoriasis: safety, efficacy, and 41 Thacßi D, Blauvelt A, Reich K et al. Secukinumab is superior to
effect of dose reduction. Br J Dermatol 2005; 152:1304–12. ustekinumab in clearing skin of subjects with moderate to severe
25 Tyring S, Gottlieb A, Papp K et al. Etanercept and clinical out- plaque psoriasis: CLEAR, a randomized controlled trial. J Am Acad
comes, fatigue, and depression in psoriasis: double-blind placebo- Dermatol 2015; 73:400–9.
controlled randomised phase III trial. Lancet 2006; 367:29–35. 42 Lebwohl M, Strober B, Menter A et al. Phase 3 studies comparing
26 van de Kerkhof PC, Segaert S, Lahfa M et al. Once weekly adminis- brodalumab with ustekinumab in psoriasis. N Engl J Med 2015;
tration of etanercept 50 mg is efficacious and well tolerated in 373:1318–28.
patients with moderate-to-severe plaque psoriasis: a randomized 43 Estellat C, Ravaud P. Lack of head-to-head trials and fair control
controlled trial with open-label extension. Br J Dermatol 2008; arms: randomized controlled trials of biologic treatment for
159:1177–85. rheumatoid arthritis. Arch Intern Med 2012; 172:237–44.
27 Strober BE, Crowley JJ, Yamauchi PS et al. Efficacy and safety 44 Sofen H, Smith S, Matheson RT et al. Guselkumab (an IL-23-speci-
results from a phase III, randomized controlled trial comparing the fic mAb) demonstrates clinical and molecular response in patients
safety and efficacy of briakinumab with etanercept and placebo in with moderate-to-severe psoriasis. J Allergy Clin Immunol 2014;
patients with moderate to severe chronic plaque psoriasis. Br J Der- 133:1032–40.
matol 2011; 165:661–8. 45 Papp K, Thacßi D, Reich K et al. Tildrakizumab (MK-3222), an anti-
28 Gottlieb AB, Leonardi C, Kerdel F et al. Efficacy and safety of IL-23p19 monoclonal antibody, improves psoriasis in a phase 2b
briakinumab vs. etanercept and placebo in patients with randomized placebo-controlled trial. Br J Dermatol 2015; 173:930–9.

British Journal of Dermatology (2017) 176, pp594–603 © 2016 British Association of Dermatologists

mez-Garc
ıa et al. 603
Efficacy and safety of new biologics in psoriasis, F. Go

46 Papp KA, Leonardi C, Menter A et al. Brodalumab, an anti-interleu- Fig S1. Sample search strategy.
kin- 17-receptor antibody for psoriasis. N Engl J Med 2012; Fig S2. Risk of bias of the included trials.
366:1181–9. Fig S3. Scoring heat map of the quality of evidence for each
outcome across pooled studies.
Supporting Information Fig S4. Forest and funnel plots: Verum vs placebo.
Fig S5. Forest and funnel plots: Verum vs verum.
Additional Supporting Information may be found in the online Fig S6. Surface Under the Cumulative RAnking curves
version of this article at the publisher’s website: (SUCRA) for all biologic treatments in the psoriasis network
Appendix S1. Supplementary materials and methods. for efficacy outcomes.
Appendix S2. STATA script for NMA analysis. Fig S7. Surface Under the Cumulative RAnking curves
Table S1. PROSPERO register file. (SUCRA) for all biologic treatments in the psoriasis network
Table S2. PRISMA extension for NMA 2015 checklist. for efficacy outcomes.
Table S3. Baseline characteristics and results of each study.
Table S4. Summary of finding tables (SoF).

© 2016 British Association of Dermatologists British Journal of Dermatology (2017) 176, pp594–603