An intensified dosing schedule of subcutaneous methotrexate in patients with moderate to severe plaque-type psoriasis (METOP): a 52 week, multicentre, randomised, double-blind, placebo-controlled, phase 3 trial

Articles

An intensified dosing schedule of subcutaneous methotrexate in patients with moderate to severe plaque-type psoriasis (METOP): a 52 week, multicentre, randomised, double-blind, placebo-controlled, phase 3 trial Richard B Warren, Ulrich Mrowietz, Ralph von Kiedrowski, Johannes Niesmann, Dagmar Wilsmann-Theis, Kamran Ghoreschi, Ina Zschocke, Thomas M Falk, Norbert Blödorn-Schlicht, Kristian Reich

Summary Background Methotrexate is one of the most commonly used systemic drugs for the treatment of moderate to severe psoriasis; however, high-quality evidence for its use is sparse and limited to use of oral dosing. We aimed to assess the effect of an intensified dosing schedule of subcutaneous methotrexate in patients with moderate to severe plaque-type psoriasis. Methods We did this prospective, multicentre, randomised, double-blind, placebo-controlled, phase 3 trial (METOP) at 16 sites in Germany, France, the Netherlands, and the UK. Eligible patients were aged 18 years or older, had a diagnosis of chronic plaque psoriasis for at least 6 months before baseline, had currently moderate to severe disease, and were methotrexate treatment-naive. Participants were randomly assigned (3:1), via a computer-generated random number sequence integrated into an electronic data capture system, to receive either methotrexate at a starting dose of 17·5 mg/week or placebo for the first 16 weeks, followed by methotrexate treatment of all patients up to 52 weeks (methotrexate–methotrexate vs placebo–methotrexate groups). Dose escalation to 22·5 mg/week was allowed after 8 weeks of methotrexate treatment if patients had not achieved at least a 50% reduction in baseline Psoriasis Area and Severity Index score (PASI), with corresponding volume increases in placebo injections. Treatment was combined with folic acid 5 mg/week. Group allocation was concealed from participants and investigators from the time of randomisation until an interim database lock at week 16, and was open label from week 16 onwards, with no masking of participants or investigators. The primary efficacy endpoint was a 75% reduction in PASI score (PASI 75) from baseline to week 16. We did analysis by modified intention to treat, with non-responder imputation. This study is registered with EudraCT, number 2012-002716-10. Findings Between Feb 22, 2013, and May 13, 2015, we randomly assigned 120 patients to receive methotrexate (n=91) or placebo (n=29). At week 16, a PASI 75 response was achieved in 37 (41%) patients in the methotrexate group compared with three (10%) patients in the placebo group (relative risk 3·93, 95% CI 1·31–11·81; p=0·0026). Subcutaneous methotrexate was generally well tolerated; no patients died or had serious infections, malignancies, or major adverse cardiovascular events. Serious adverse events were recorded in three (3%) patients who received methotrexate for the full 52 week treatment period. Interpretation Our findings show a favourable 52 week risk–benefit profile of subcutaneous methotrexate in patients with psoriasis. The route of administration and the intensified dosing schedule should be considered when methotrexate is used in this patient group. Funding Medac.

Introduction Psoriasis is a complex inflammatory skin disease characterised by erythematous and scaly plaques. The disease requires early and continuous control similar to other chronic inflammatory disorders, such as rheumatoid arthritis and inflammatory bowel disease. In 2014, WHO recognised psoriasis as a serious non-communicable disease and, in a report from 2016, emphasised the lifelong nature of the disease and the need to urgently research and make accessible cost-effective medicines worldwide.1 Development of monoclonal antibodies that target cytokines central to the psoriatic disease cascade,

including tumour necrosis factor (TNF)α and interleukins 23 and 17A, has greatly broadened the therapeutic armamentarium;2–4 however, survey findings indicate restricted use of these expensive treatments in fewer than 10% of patients with moderate to severe disease, even in high-income countries.5 Methotrexate has been used in the treatment of psoriasis and psoriatic arthritis for more than 50 years, and is recommended as a first-line systemic drug for the treatment of moderate to severe psoriasis in US and European guidelines,6 mainly on the basis of experience and health-economic considerations. However, in the absence of modern, well

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

Published Online December 21, 2016 http://dx.doi.org/10.1016/ S0140-6736(16)32127-4 See Online/Comment http://dx.doi.org/10.1016/ S0140-6736(16)32574-0 Dermatology Centre, Salford Royal NHS Foundation Trust, University of Manchester and Manchester Academic Health Science Centre, Manchester, UK (R B Warren MD); Psoriasis-Center, Department of Dermatology, University Medical Center SchleswigHolstein, Campus Kiel, Kiel, Germany (Prof U Mrowietz MD); Company for Medical Study & Service Selters (CMS3) GmbH, Selters, Germany (R von Kiedrowski MD); Hautarzt Gemeinschaftspraxis, Bochum, Germany (J Niesmann MD); Department of Dermatology and Allergy, University of Bonn, Bonn, Germany (D Wilsmann-Theis MD); Department of Dermatology, University Medical Center, Eberhard Karls University Tübingen, Tübingen, Germany (K Ghoreschi MD); SCIderm GmbH Hamburg, Hamburg, Germany (I Zschocke PhD, Prof K Reich MD); and Dermatologikum Hamburg, Hamburg, Germany (T M Falk PhD, N Blödorn-Schlicht MD, Prof K Reich) Correspondence to: Prof Kristian Reich, Dermatologikum Hamburg, Hamburg 20354, Germany [email protected]

1

Articles

Research in context Evidence before this study A meta-analysis of methotrexate reported a 75% reduction in Psoriasis Area and Severity Index score (PASI 75) of 45·2% (95% CI 34·1–60·0) from primary endpoints at either 12 weeks or 16 weeks, mainly based on three larger studies of oral methotrexate in patients with psoriasis, of which two were not placebo controlled. The safety meta-analysis showed that roughly 7% of patients had treatment-limiting adverse events, but this analysis included several studies of methotrexate for indications other than psoriasis that differ in their susceptibility to safety events of interest. We searched PubMed, MEDLINE, Embase, and the Cochrane Library for meta-analyses, randomised and non-randomised controlled clinical trials, case series, case reports, and open studies of methotrexate in psoriasis from inception to April 1, 2016. One retrospective study (n=85) investigated the success of subcutaneous methotrexate after failure of oral treatment. A larger retrospective analysis published after April, 2016, included a subgroup of 27 patients who switched from oral to subcutaneous methotrexate. Findings from both studies suggested that patients with failure of oral treatment might benefit from a switch to subcutaneous treatment, but neither included hard efficacy endpoints and both have the inherent bias that comes with the retrospective nature of the studies. Added value of this study Methotrexate is one of the most commonly used systemic drugs for the treatment of moderate to severe psoriasis, yet

designed clinical trials,7 uncertainties remain regarding the optimum dosing scheme, route of administration, and the safety profile, with appropriate monitoring in place. Only recently have high-quality data for methotrexate in psoriasis become available from three studies that tested the drug in comparison to the biological drugs adalimumab,8 briakinumab,9 and infliximab.10 Although these studies differed in their dosing schemes, with methotrexate starting doses of 5 mg/week in conjunction with either a slower8 or more rapid uptitration scheme,9 and a starting dose of 15 mg/week in one study,10 they all showed similar response rates to induction therapy according to a 75% reduction in Psoriasis Area and Severity Index score (PASI 75) in 36–42% of patients after 16–24 weeks of treatment. Long-term data showed a low stability of methotrexate response, with high dropout rates due to loss of efficacy in up to 70% of patients over 1 year.9 The safety profile of methotrexate in these studies was similar with different dosing schemes in the first months of treatment, and generally similar to the profile of the biological drugs during longer-term treatment with regard to events perceived as of special concern with methotrexate, such as hepatopathy and leukopenia. Methotrexate was given orally in all these studies, which might be a suboptimum route of administration in view of a trial in patients with rheumatoid arthritis that showed 2

high-quality evidence for its use is sparse and limited to use of the oral formulation. Dosing strategies include low starting doses in combination with different and sometimes complicated uptitration schemes. To the best of our knowledge, this is the first double-blind, randomised, placebo-controlled study of an intensified two-step dosing schedule of subcutaneous methotrexate in patients with moderate to severe plaque-type psoriasis. Implications of all the available evidence Our findings show a favourable 52 week risk–benefit profile of subcutaneous methotrexate in patients with psoriasis, with significantly more patients in the methotrexate group than the placebo group achieving the primary endpoint of a PASI 75 response at week 16, and a generally acceptable safety profile. Our study suggests that methotrexate can be started at a higher dose with no detrimental effect on risk, that a subcutaneous formulation and dosing scheme could lead to a more rapid and sustained response than that typically seen with oral regimens, and that the clinical response to methotrexate is associated with prominent effects on cutaneous T-helper-cell type 1 and type 17 pathways. These findings have potential implications and applicability for dermatologists and other users of methotrexate, such as rheumatologists and gastroenterologists, when treating inflammatory disorders.

superiority of subcutaneous over oral methotrexate after 24 weeks.11 We did the METOP trial to assess the effect of an intensified subcutaneous dosing scheme of methotrexate in patients with moderate to severe plaque-type psoriasis. Because the exact mechanism of action of methotrexate in psoriasis is unknown, we also investigated pairs of skin biopsies obtained before and during treatment to obtain insight into the effects of the drug on selected inflammatory disease pathways.

Methods Study design and participants We did this investigator-initiated, multicentre, randomised, double-blind, placebo-controlled, phase 3 trial at 16 sites in Germany (n=13), France, the Netherlands, and the UK (n=1 each). Eligible participants were aged 18 years or older, were methotrexate treatment-naive, and had a diagnosis of plaque-type psoriasis for at least 6 months before baseline, with currently moderate to severe disease based on the definition by Finlay.12 Patients were required to have a normal chest radiograph within 6 months before study entry and were excluded if hepatic enzymes (alanine aminotransferase [ALT], aspartate aminotransferase [AST], or γ glutamyl transferase [GGT]) were elevated to more than twice the upper limit of

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

Articles

normal or total leukocyte counts were less than 3·0 × 109 cells per L in screening laboratory tests. Previous treatment with biological drugs had to be discontinued for at least five times their half-life; other systemic therapies and phototherapies used for the treatment of psoriasis were discontinued at least 4 weeks before study entry and topical treatments were discontinued at least 2 weeks before entry. Bland emollients were allowed during the study. Patients with a previous diagnosis of psoriatic arthritis could be enrolled; however, we excluded patients with currently active psoriatic arthritis, as defined by five or more tender or swollen joints and peripheral C-reactive protein concentrations of more than twice the upper limit of normal. The appendix (pp 2, 3) provides a complete list of inclusion and exclusion criteria. The study protocol was approved by the ethics committee of the University of Kiel, Kiel, Germany, and by investigational review boards at each site. The study was done in conformance with good clinical practice guidelines. All patients provided written informed consent.

Randomisation and masking Eligible patients were randomly assigned (3:1), via computer-generated random numbers (RandList 1.2) in an ascending order, to receive either methotrexate or placebo injections for the first 16 weeks of the study (phase 1). Between weeks 16 and 52 (phase 2), patients from both treatment groups were entitled to receive methotrexate injections—ie, were either continued on methotrexate (methotrexate–methotrexate group) or crossed-over from placebo (placebo–methotrexate group). Study phase 1 was done in a double-blind manner, with group allocation concealed from participants and investigators from the time of randomisation until an interim database lock at week 16. From week 16 onwards, methotrexate injections were administered on an open-label basis, with no masking of participants or investigators. All study drugs were supplied in identical packages. The syringes for placebo and active drug were not distinguishable and were fully coated to prevent identification of colour differences between injections.

Procedures From week 0 to week 16, patients received once-weekly subcutaneous injections of either 17·5 mg methotrexate (Metoject or Metex [Medac, Wedel, Germany] concentration 50 mg/mL) in a volume of 0·35 mL, or 0·35 mL placebo injections. The first injection at day 0 and the injection at day 112 were given at the study sites; all other injections were self-administered by the patient. We chose subcutaneous rather than oral administration because of evidence of an improved adherence, a lower risk of accidental overdosing, and a higher concentration of methotrexate polyglutamates—the active methotrexate metabolite.13,14 The starting dosing was based on results of the three previous trials in which starting

doses between 5 mg and 15 mg with different subsequent uptitration schemes were not associated with differences in the safety profile.8–10 If a 50% reduction in PASI score (PASI 50) was not reached after 8 weeks, patients received injections with either 22·5 mg methotrexate per week in a volume of 0·45 mL, or 0·45 mL placebo injections. From week 16 to week 52, patients originally started on methotrexate remained on the same dose unless they were receiving 17·5 mg methotrexate per week and at week 24 had not achieved PASI 75, in which case they could be dose escalated to 22·5 mg methotrexate per week. If patients were already receiving the 22·5 mg dose at week 24 and PASI 50 was not reached, they were excluded from continuing with study treatment. Patients originally started on placebo and switched to methotrexate at week 16 received 17·5 mg of methotrexate weekly, with possible escalation to the 22·5 mg dose at week 24 if PASI 50 was not achieved. In accordance with the study protocol, patients originally given placebo who achieved a PASI 75 response at week 16 received no further injections until the disease relapsed, when they were treated with 17·5 mg methotrexate per week as described. Patients from all treatment groups received 5 mg of oral folic acid 24 h after each injection. Two punch biopsies (3 mm in diameter) were obtained from patients from the edge of a representative psoriatic plaque, each at baseline and from the same plaque at week 16. One biopsy taken at each timepoint was immediately transferred to 4% buffered formalin, the other biopsy was fixed in RNAlater solution (Qiagen, Hilden, Germany) for subsequent RNA extraction. All biopsies were handled and analysed by personnel masked to treatment and timepoints. RNA was isolated by means of the RNeasy Fibrous Tissue Mini Kit (Qiagen). Complementary DNA was synthesised with the Applied Biosystems High-Capacity cDNA Reverse Transcription Kit (Life Technologies, Grand Island, NY, USA). Quantitative real-time PCR assays were done in duplicate with a Bio-Rad iQ5 Cycler (Bio-Rad Laboratories, Hercules, CA, USA) as previously described.15 Results for IL17A, IFNG, and TNFA were assessed with the iQ5 Optical System Software, version 2.0 (Bio-Rad Laboratories). Quantification was based on the comparative CT method. Samples were normalised based on two reference genes: B2M and UBC. Psoriatic microanatomical features (epidermal thickness, parakeratosis, presence of epidermal microabscesses) and immunohistochemical markers (CD3-positive T cells, CD1a-positive and CD11c-positive dendritic cells, Ki-67-positive keratinocytes) were investigated and semiquantitatively assessed on paraffin-embedded sections with the AxioVision SE64 release 4.8 system (Zeiss, Oberkochen, Germany) on digitally scanned images as previously described.15

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

See Online for appendix For the protocol see http://www. dermatologikum.de/forschunglehre/neue-therapien/

3

Articles

142 patients screened for eligibility

120 randomly assigned

91 to the methotrexate–methotrexate group

22 excluded 10 because inclusion criterion not fulfilled or exclusion criterion fulfilled 6 at the patient’s request 6 for other reasons*

29 to the placebo–methotrexate group

14 discontinued during phase 1 10 had adverse events 2 lost to follow-up 1 because of poor efficacy 1 other

7 discontinued during phase 1 4 had adverse events 2 at patients’ choice 1 because of poor efficacy

22 completed phase 1

77 completed phase 1

21 discontinued during phase 2 9 had adverse events 7 because of poor efficacy 2 at patients’ choice 2 lost to follow-up 1 other

7 discontinued during phase 2 4 had adverse events 2 lost to follow-up 1 at patients’ choice

15 completed phase 2

56 completed phase 2

Figure 1: Trial profile *Time window for baseline exceeded, life-threatening cardiomyopathy, loss to follow-up, patient did not attend baseline visit, patient withdrew consent, clarification of hepatitis laboratory results was not possible within 28 days because further tests were necessary.

Outcomes The primary endpoint was achievement of PASI 75 from baseline to week 16. Secondary endpoints included the PASI 75 response rate at week 52 and, at weeks 16 and 52, the PASI 50 and PASI 90 response rates; a static physicians’ global assessment (sPGA), measured on a 7 point scale ranging from 0 (clear) to 6 (severe) inclusive; the Nail Psoriasis Severity Index (NAPSI) of the finger nails; the Dermatology Life Quality Index (DLQI); and the EuroQol five dimensions questionnaire, ranging from level 1 (no problems) to level 5 (extreme problems). Safety was assessed based on reported adverse events, laboratory values, vital signs, physical examinations, and assessments of local drug tolerability. Adverse events were classified as drug related if the association was classified as related or possible by the investigator. Safety data were reviewed at regular intervals by an independent data monitoring committee. Concentrations of aminoterminal propeptide of type III procollagen (PIIINP) were assessed at baseline, week 16, week 32, and week 52.

Statistical analysis We intended to show that an intensified subcutaneous methotrexate dosing scheme would be superior to placebo in inducing a PASI 75 response at week 16 on the basis of a two-sided Pearson’s χ² test with an α level 4

of 0·05. The sample size estimation of 120 treated patients was based on this statistical approach, an assumed PASI 75 response rate of 35% in patients given methotrexate and 10% in those given placebo, a power of 80%, and 3:1 allocation of patients in the intervention versus control groups. All outcomes were analysed in the modified intentionto-treat population of patients who had received at least one injection of study drug, with missing data treated as indicating no response (non-responder imputation). Post-hoc comparisons of secondary and non-prespecified endpoints at week 16 were likewise based on the modified intention-to-treat population with non-responder imputation, and were done with a two-sided Pearson’s χ² with an α level of 0·05. These analyses were deemed exploratory and no adjustment for multiple testing was made. Selected clinical outcomes at week 52 are also presented based on an as-observed analysis. Two-sided 95% CIs for PASI response rates over time were calculated according to the Clopper–Pearson method. A two-sided asymptotic 95% CI was calculated for the relative risk (RR) of a PASI 75 response at week 16. Safety analyses were based on data for all patients who received at least one dose of study drug. Safety data are reported both before and after week 16. Data for the biopsy substudy were analysed with descriptive statistics. Cutaneous cells positive for CD3, CD11c, or Ki67 in biopsies from week 16 were compared with values from paired biopsies taken at baseline by use of the Stuart–Maxwell test. We used the Wilcoxon test to analyse ratios of post-treatment versus pretreatment cutaneous mRNA concentrations of IL17A, IFNG, and TNFA. All statistical analyses were done with SAS (version 9.3), except for analyses of biopsy data, which were done with R (version 3.3.0). This study is registered with EudraCT, number 2012-002716-10.

Role of the funding source The funder of the study had no role in study design, data collection, data analysis, data interpretation or writing of the report. The corresponding author had full access to all the data in the study and all authors had final responsibility for the decision to submit for publication.

Results Between Feb 22, 2013, and May 13, 2015, 120 participants from 13 sites (n=11 in Germany, n=1 each in The Netherlands, and the UK; three sites were initiated, but did not enroll patients) were randomly assigned to receive weekly injections of either methotrexate (n=91) or placebo (n=29; figure 1). 99 (83%) patients completed the study up to week 16, of whom 71 (72%) completed the study up to week 52 (figure 1). Most patients were middle-aged white men with longstanding psoriasis, a mean weight of greater than 90 kg,

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

Articles

and a mean body-mass index of roughly 30 kg/m² (table 1). Main baseline disease activity characteristics were balanced between groups, with a mean PASI score of roughly 15 and a mean DLQI score of roughly 12, indicating moderate to severe disease with a major effect on health-related quality of life according to severity definitions used in clinical trials and guidelines.16 Disease duration tended to be longer, and a previous diagnosis of psoriatic arthritis documented more frequently, in patients assigned to receive methotrexate than in those assigned to receive placebo (table 1). At week 16, a PASI 75 response had been achieved in 37 (41%) patients in the methotrexate group compared with three (10%) patients in the placebo group (RR 3·93, 95% CI 1·31–11·81; p=0·0026; figure 2, table 2). Dose escalation to 22·5 mg/week at week 8 was documented in 28 (31%) patients in the methotrexate group. 25 (27%) patients given methotrexate had an sPGA score of clear (0) or almost clear (1) at 16 weeks compared with two (7%) patients given placebo; 16 (18%) versus no patients, respectively, had a 90% reduction in PASI response (PASI 90) at this timepoint (figure 2, table 2). On the basis of non-responder imputation analysis of the modified intention-to-treat population, PASI 75 response rates at week 52 were 45% in the methotrexate–methotrexate group and 34% in the placebo–methotrexate group (figure 2, table 2). Among patients crossed over from placebo, dose escalation to 22·5 mg/week was documented in five (23%) of 22 patients at week 24 (ie, 8 weeks after initiation of active treatment). Five (55%) patients in the methotrexate–methotrexate group were dose escalated to 22·5 mg/week at week 24. Response rates increased with continuous methotrexate treatment; at week 52, PASI 90 responses were recorded in almost 28% of patients in both the methotrexate–methotrexate and methotrexate–placebo groups, and an sPGA score of 0 or 1 was recorded in almost 40% of patients in both groups (figure 2, table 2). 29 (78%) of 37 patients achieving a PASI 75 response at week 16 with methotrexate treatment still had this response at week 52. In as-observed analyses of patients who received methotrexate for the full 52 week treatment period, PASI 75, 90, and 100 responses were achieved in 73%, 45%, and 18% of patients, respectively; 64% of patients achieved an sPGA 0 or 1 response (table 2). Methotrexate reduced the activity of nail psoriasis within 16 weeks, as measured by the NAPSI score of the worst fingernail as a target nail, whereas placebo had no effect (table 2). After 52 weeks of methotrexate treatment, eight (14%) of 59 patients with an active target nail at baseline showed complete clearance of that nail (table 2). Improvements in the signs of psoriasis were paralleled by improvements in the health-related quality of life as assessed by the DLQI. At week 16, we recorded an absolute DLQI score of 5 or less (mild effect of psoriasis

Placebo– Methotrexate– methotrexate group methotrexate group (n=29) (n=91) Age (years) Mean (SD)

45·9 (12·9)

44·4 (10·8)

Median (range)

48 (18–73)

46 (23–65)

Men

65 (71%)

25 (86%)

Female

26 (29%)

4 (14%)

White

89 (98%)

29 (100%)

Asian

2 (2%)

Sex

Ethnic origin

Weight (kg)

92·4 (18·6)

0 95·9 (20·9)

≥100 kg

26 (29%)

11 (38%)

Body-mass index (kg/m2)*

30·1 (6·3); n=87

30·1 (6·1)

Psoriasis duration (years)

20·7 (13·8)

14·3 (11·3)

sPGA ≥4†

74 (81%)

22 (76%)

PASI

15·4 (5·9)

15·4 (5·3)

Body surface area

20·0 (11·7)

19·6 (12·5)

59 (65%)

20 (69%)

Fingernail psoriasis At least one affected nail NAPSI of target (worst) nail

4·0 (1–8)

4·0 (1–8)

DLQI

12·9 (7·7)

11·6 (6·7)

Confirmed psoriatic arthritis‡

11 (12%)

2 (7%)

Previous treatment Phototherapy

7 (8%)

1 (3%)

Conventional systemic§

29 (32%)

7 (24%)

Fumaric acid esters

24 (26%)

7 (24%)

Other

6 (7%)

2 (7%)

Biological

5 (5%)

1 (3%)

Data are mean (SD), median (range), or n (%), unless otherwise specified. sPGA=static Physicians’ Global Assessment. PASI=Psoriasis Area and Severity Index. NAPSI=Nail Psoriasis Severity Index. DLQI=Dermatology Life Quality Index.*Body mass index could not be calculated for four patients. †Measured on a scale ranging from 0 (clear) to 6 (severe). ‡Refers to a previous diagnosis by a rheumatologist, but does not indicate currently active disease. §Excluding methotrexate.

Table 1: Baseline demographics and clinical characteristics

on DLQI domains) in 59% of patients given methotrexate and a score of 0 or 1 (no effect) in 43% of patients in this group, compared with 34% and 10% of patients, respectively, given placebo (table 2). In as-observed analysis, 40 (71%) of 56 patients continuously treated with methotrexate had a DLQI response of 0 or 1 at week 52 (table 2). The difference from baseline to week 16 in the proportion of patients reporting no problems across the five dimensions of the EQ-5D suggested different effects of methotrexate compared with placebo. For example, at week 16, the proportion of patients who had no problems with usual activities increased by 13% in the methotrexate group compared

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

5

Articles

100

50% improvement in PASI Methotrexate–methotrexate group Placebo–methotrexate group

Proportion of patients (%)

80

75% improvement in PASI

69%

66%

67%

58%

58% 51%

60

51%

45%

41% 45% 40

27%

27%

34% 31%

34% 24%

20

5%

10%

17%

3%

0%

0 90% improvement in PASI

24%

17%

0%

sPGA 0 or 1*

100

Proportion of patients (%)

80

60

40% 38% 29%

40

24% 14%

20

19%

28% 5%

1% 3% 0%

0%

0 4

27%

18%

8

38%

34% 38%

27%

21%

14%

0%

0%

0%

16

24

32

52

Weeks

4

8

7% 16

24

32

52

Weeks

Figure 2: Proportion of patients achieving reductions of 50%, 75%, and 90% in PASI score and an sPGA score of 0 or 1 over the 52 week treatment period Data shown are based on modified intention-to-treat non-responder imputation analysis. Error bars represent exact 95% CIs. Patients received methotrexate (n=91) or placebo (n=29) up to week 16, followed by methotrexate treatment of all patients up to week 52. PASI=Psoriasis Area and Severity Index. sPGA=static Physicians’ Global Assessment. *An sPGA of 0 indicates clear of disease and an sPGA of 1 indicates almost clear of disease.

with a 14% reduction of patients in the placebo group (appendix pp 4, 5). Similarly, the proportion of individuals reporting no problems with anxiety and depression had increased by 7% at week 16 in the methotrexate group, but had decreased by 17% in the placebo group (appendix pp 4, 5). All adverse events appeared during treatment and fulfilled the criterion of treatment-emergent adverse events. No patients died or had malignancies or major adverse cardiovascular events during the study, and no serious adverse events were related to treatment with methotrexate (table 3). During the placebo-controlled study phase, nasopharyngitis, headache, gastrointestinal disorders, and increases in hepatic enzymes were the most frequently reported adverse events (table 3 and appendix p 6), of which gastrointestinal disorders and increases in hepatic enzymes were more common in patients in the methotrexate group than in those in the placebo group (table 3). Among patients given methotrexate, nausea or vomiting accounted for most of the gastrointestinal events in the first 16 weeks and 6

during the overall study period (table 3). Gastrointestinal adverse events were usually mild or moderate and led to permanent discontinuations of study drug in three (3%) patients who received methotrexate for the full 52 week treatment period (table 3). Over 52 weeks, elevation of hepatic enzymes was reported in 23% of patients started on methotrexate, of whom roughly half permanently discontinued study drug (table 3). Increases in hepatic enzymes were not associated with elevated concentrations of PIIINP, which were detected in five (4%) of 120 patients at baseline and in five (7%) of 71 patients at week 52 (n=1 same patient, n=4 different patients). Leukopenia was reported in five (5%) patients started on methotrexate, but no patients had leukopenia of grade 3 or 4 and only one patient had lymphopenia grade 3 (table 3). Rates of infections were similar between groups, with nasopharyngitis accounting for more than 50% of all infectious events (table 3, appendix p 6). A case of severe nasopharyngitis was reported in study phase 1 in one (3%) of 29 patients receiving placebo. No serious or severe infectious

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

Articles

adverse events were recorded during methotrexate treatment. Pairs of baseline and week 16 biopsies were available for 27 patients, of whom six (23%) received placebo (none with a PASI 75 response), seven (26%) were methotrexate PASI 75 non-responders, and 14 (52%) were methotrexate PASI 75 responders. In the patients with a methotrexate response, clinical effects were associated with prominent reductions in numbers of skin-infiltrating CD3-positive T cells (appendix p 7) and CD11c-positive dendritic cells (data not shown), with week 16 numbers almost returning to values in normal skin. These changes were associated with effects on the expression of specific T helper cell types 17 and 1 cytokines. By comparison with baseline values, cutaneous mRNA levels of interleukin 17A at week 16 were largely reduced in methotrexate PASI 75 responders, but were not significantly changed in methotrexate PASI75 nonresponders or patients receiving placebo (appendix p 7). Individual changes in cutaneous mRNA levels of interleukin 17A correlated well with individual PASI responses (Spearman correlation coefficient –0·74; p<0·0001). Similarly, levels of interferon-γ mRNA were reduced in methotrexate responders, with no effects in non-responders or patients receiving placebo (appendix p 7), whereas expression levels of TNFα remained unaffected (data not shown), irrespective of the clinical response. Immunological effects coincided with effects on selected epidermal variables; the most pronounced reductions in methotrexate PASI75 responders were recorded for numbers of Ki-67-positive (proliferating) keratinocytes and numbers of epidermal microabscesses (data not shown).

Week 16

Week 52

Methotrexate Placebo group p value (n=29) group (n=91)

Methotrexate– Placebo– methotrexate methotrexate group group

50% improvement in PASI Non-responder imputation

60 (66%)

As observed

ND

9 (31%) ND

0·0009 ··

53/91 (58%)

10/29 (34%)

53/56 (95%)

10/15 (67%)

41/91 (45%)

10/29 (34%)

41/56 (73%)

10/15 (67%)

25/91 (27%)

8/29 (28%)

25/56 (45%)

8/15 (53%)

10/91 (11%)

5/29 (17%)

10/56 (18%)

5/15 (33%)

42/91 (46%)

10/29 (34%)

42/56 (75%)

10/15 (67%)

75% improvement in PASI Non-responder imputation

37 (41%)

As observed

ND

3 (10%) ND

0·0026 ··

90% improvement in PASI Non-responder imputation

16 (18%)

As observed

ND

0 ND

0·0153 ··

100% improvement in PASI Non-responder imputation As observed

4 (4%) ND

0 ND

0·2508 ··

PASI ≤3 Non-responder imputation

35 (38%)

As observed

ND

3 (10%) ND

0·0046 ··

sPGA 0 or 1* Non-responder imputation

25 (27%)

36/91 (40%)

11/29 (38%)

As observed

ND

ND

2 (7%)

0·0208 ··

36/56 (64%)

11/15 (73%)

Mean (SD)

–9·4 (6·58)

–2·6 (5·83)

··

–11·6 (7·50)

–5·5 (6·09)

Median (range)

–9·0 (–29 to 1)

–3·0 (–11 to 10)

··

DLQI ≤5 (mild effect)

54 (59%)

10 (34%)

0·0195

50/91 (55%)

13/15 (45%)

Non-responder imputation

39 (43%)

3 (10%)

0·0014

40/91 (44%)

13 (45%)

As observed

ND

40 (71%)

13 (87%)

DLQI Absolute change –11·0 (–30 to 1)

–3·0 (–20 to 0)

DLQI 0 or 1 (no effect)

ND

··

NAPSI†‡

Discussion Our findings show a favourable 52 week risk–benefit profile of subcutaneous methotrexate in patients with psoriasis, with significantly more patients in the methotrexate group than the placebo group achieving the primary endpoint of a PASI 75 response at week 16, and a generally acceptable safety profile. These findings should help to close the gap, identified in a systematic review,7 in availability of high-quality data for a drug that has been in use for more than 50 years. We used an intensified dosing scheme with a higher starting dose than used in previous studies of 17·5 mg/week and dose escalation to 22·5 mg/week after 8 weeks of methotrexate if PASI 50 was not achieved. This dose-escalation scheme followed pharmacokinetic considerations to establish a stable pool of methotrexate polyglutamates and was a result of previous clinical observations in psoriasis. In particular, we did not assess dose escalation beyond 22·5 mg/week because PASI 50 non-responders to 20 mg were not previously shown to benefit from further uptitration to 25 mg/week.8 In rheumatoid arthritis, increases in concentrations of methotrexate polyglutamates and improved outcomes have been

Absolute change Mean (SD)

–0·86 (2·02)

0·47 (1·46)

··

Median (range)

–1·00 (–8·0 to 3·0)

0·00 (–1·0 to 4·0)

··

Total clearance

3/59 (5%)

0/20

–1·83 (2·05) –2·0 (–6·0 to 5·0)

0·4033

8/59 (14%)

–1·40 (2·12) –2·0 (–5·0 to 2·0) 5/20 (25%)

Data are n (%) or n/N (%), unless otherwise specified. PASI=Psoriasis Area and Severity Index. ND=not done. sPGA=static Physicians’ Global Assessment. DLQI=Dermatology Life Quality Index. NAPSI=Nail Psoriasis Severity Index. *0 indicates clear of disease and 1 indicates almost clear of disease. †All nail data are non-responder imputation analyses of worst fingernail (target nail) changes. ‡Patients with an active target nail (NAPSI ≥1) at baseline.

Table 2: Clinical responses at weeks 16 and 52

reported after patients switched from oral to subcutaneous methotrexate,17 and retrospective clinical data for patients with psoriasis likewise suggest advantageous efficacy of subcutaneous dosing.18,19 Limitations of our study mainly relate to the relatively small number of patients enrolled and the absence of an active comparator group with oral methotrexate. Additionally, the study population was mainly white, so further study in non-white participants might be necessary to fully understand the efficacy and safety in a more genetically diverse population.

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

7

Articles

Weeks 0–16 Methotrexate group (n=91)

Weeks 16–52 Placebo group (n=29)

Methotrexate– methotrexate group (n=76)*

Weeks 0–52 Placebo– methotrexate group (n=22)

Methotrexate– methotrexate group (n=91)

Any adverse event

75 (82%)

27 (93%)

59 (78%)

17 (77%)

86 (95%)

Any drug-related adverse event†

55 (60%)

14 (48%)

35 (46%)

12 (55%)

66 (73%)

Any serious adverse event

1 (1%)

4 (14%)

2 (3%)

1 (5%)

3 (3%)

Any drug-related serious adverse event†

0

2 (7%)

0

0

0

Adverse events of special interest Death Any infection

0 40 (44%)

0 13 (45%)

0 31 (41%)

0 11 (50%)

0 58 (64%)

Serious infections

0

0

0

0

Severe infections‡

0

1 (3%)

0

0

0 0

Malignancies

0

0

0

0

0

MACE

0

0

0

0

0

Depression

1 (1%)

1 (3%)

1 (1%)

0

2 (2%)

White blood cell count decreased Any

4 (4%)

1 (3%)

1 (1%)

0

5 (5%)

Leukopenia grade 3§

0

0

0

0

0

Lymphopenia grade 3§

0

0

1 (1%)

0

1 (1%)

Permanent discontinuation¶

2 (2%)

0

0

0

2 (2%)

Hepatic enzyme increased||

12 (13%)

2 (7%)

10 (13%)

5 (23%)

>2 × ULN

10 (11%)

1 (3%)

7 (9%)

5 (23%)

21 (23%) 17 (19%)

>3 × ULN

8 (9%)

1 (3%)

6 (8%)

3 (14%)

14 (15%)

Permanent discontinuation¶

6 (7%)

1 (3%)

5 (7%)

2 (9%)

11 (12%)

Any

22 (24%)

3 (10%)

10 (13%)

7 (32%)

30 (33%)

Nausea or vomiting

13 (14%)

1 (3%)

7 (9%)

3 (14%)

20 (22%)

Diarrhoea

3 (3%)

1 (3%)

4 (5%)

3 (14%)

6 (7%)

Abdominal pain

3 (3%)

1 (3%)

0

1 (5%)

3 (3%)

Mouth ulceration

3 (3%)

0

0

1 (5%)

3 (3%)

Permanent discontinuation¶

1 (1%)

1 (3%)

2 (3%)

0

3 (3%)

Gastrointestinal disorders

Data are n (%), unless otherwise specified. MACE=major cardiovascular adverse event (myocardial infarction, stroke, death due to cardiovascular event). ULN=upper limit of normal. *One patient completed phase 1, but never started phase 2. †Adverse and serious adverse events were classified as drug related if the association was classified as related or possibly related by the investigator. ‡Infectious adverse events requiring systemic treatment. §According to Common Terminology Criteria for Adverse Events version 3.0, with grade 3 lymphopenia less than 500 cells per mm³ to 200 cells per mm³ and grade 3 leukopenia less than 2000 cells per mm³ to 1000 cells per mm³. The one patient with grade 3 lymphopenia had 467 lymphocytes per mm³ as the lowest number. No patients had grade 4 lymphopenia or leukopenia. ¶Permanent study drug discontinuations due to the respective adverse events of interest are assigned to the actual time of discontinuation. ||Hepatic enzymes included aspartate aminotransferase, alanine aminotransferase, and γ glutamyltransferase.

Table 3: Adverse events of special interest

The induction efficacy recorded in the present study is similar to that shown in previous trials of oral methotrexate, in that roughly 40% of patients had achieved a PASI 75 response by week 16 (41% in this study; 36–42% in three previous studies8–10). However, different study populations and study designs restrict the value of this indirect comparison. In particular, two9,10 of the previous trials with oral methotrexate were not placebo controlled, but active comparator-controlled, and one trial10 was open label. Efficacy tended to increase up to week 24 in the present study in patients started on methotrexate, but this delayed response is difficult to distinguish from effects related to the termination of masking and the subsequent open-label treatment with active drug from week 16 onwards. The main difference 8

of subcutaneous to oral dosing, in addition to a more rapid onset of efficacy (a PASI 75 response rate at week 8 of 27% in this placebo-controlled study vs roughly 20% in the previous active comparator-controlled study8), was a superior and more stable long-term response beyond week 16 and up to week 52 of treatment. In particular, on the basis of modified intention-to-treat non-responder imputation analyses, the PASI 75 response rate at week 52 was 24% in the only previous 1 year trial with oral methotrexate at doses of up to 25 mg/week in combination with 5 mg folic acid per week9 compared with 45% in this study. Similarly, higher levels of response at week 52 for measures such as PASI 90 and sPGA 0 or 1 were recorded in roughly 18% and 20% of patients, respectively, in the previous trial with oral methotrexate7 compared with 28%

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

Articles

and 40% of patients, respectively, in the present trial with subcutaneous methotrexate. Although the overall dropout rate was greater than 70% (n=118/163) during the 1 year study period in patients receiving oral methotrexate,9 with non-response being the main reason (n=95), the dropout rate over 52 weeks in patients given subcutaneous methotrexate in the present study was 39% (n=35/91), with eight patients discontinuing because of poor efficacy. Among patients started on subcutaneous methotrexate, permanent study drug discontinuation due to adverse events was reported in roughly a fifth of patients and elevation of liver enzymes was the most common adverse event reported as the reason for permanent study drug discontinuation in 12% of patients. All these patients had elevations of more than three times the upper limit of normal of at least one of the tested variables (ALT, AST, GGT). The distribution of age, weight, and body-mass index in patients who had liver enzyme increases of this magnitude was similar to that in the overall patient population. Furthermore, concentrations of PIIINP, which is regarded as a useful early marker of methotrexate-induced liver damage in psoriasis,20 were not associated with liver enzyme increases. Gastrointestinal adverse events, including nausea and diarrhoea, arose in about 30% of patients, but were usually mild and did not lead to study drug discontinuation (n=4 affected patients terminated prematurely). Whether the subcutaneous administration of methotrexate or the setting of a clinical trial had an effect on the dropout rate due to gastrointestinal problems remains to be further investigated. No deaths, severe or serious infections, malignancies, or major adverse cardiovascular events occurred during treatment with subcutaneous methotrexate. The exact mechanisms of action of methotrexate in psoriasis are not clear at present, but immunomodulatory effects rather than direct antiproliferative effects through inhibition of dihydrofolate reductase are assumed to have a role. Increasing evidence shows that immuno-modulatory effects are at least partly mediated via agonistic effects of methotrexate on the adenosine A2A receptor21 and inhibitory effects on NFκB activation through depletion of tetrahydrobiopterin.22 More recently, a modulation of T-cell motogenic pathways by methotrexate has been described.23 Here, we present preliminary evidence that the clinical response to subcutaneous methotrexate involves inhibitory effects on key cellular components of the inflammatory infiltrate in psoriasis—namely, CD11c-positive dendritic cells and T cells, and the cutaneous expression of cytokines central to T helper cell type 1 and type 17 pathways.24 These immunological effects were paralleled by an at least partial normalisation of epidermal changes characteristic of psoriasis, such as keratinocyte hyperproliferation. Our results support and extend previous findings of a suppression of T helper cytokine expression in patients responding to treatment with

methotrexate.25 Although the observed changes are unlikely to represent a unique pattern induced by successful methotrexate treatment,15,25 they add to the emerging concept of a range of cutaneous immunological effects mediated by the drug that warrant further investigation, including the analysis of early skin responses before the onset of a marked clinical improvement. Methotrexate meets the WHO remit of being accessible worldwide. Our findings encourage the use subcutaneous methotrexate for treatment of psoriasis, and suggest longterm clinical outcomes better than previously reported for oral administration, although final confirmation will be needed in a direct head-to-head trial of subcutaneous versus oral dosing. Our findings might also help to guide future recommendations for the optimum dosing of methotrexate. In addition to results from previous trials with oral methotrexate and from patient registries,26 our findings support an acceptable risk–benefit profile of methotrexate, but appropriate patient selection and monitoring are mandatory, particularly with regard to gastrointestinal and hepatic side-effects. There is growing evidence that genetic profiling of enzyme and transporter variants involved in methotrexate metabolism,27,28 and phenotyping of specific T-cell populations,29 could in the future help to better predict patients responding to and tolerating methotrexate therapy. Contributors RBW, UM, and KR designed the study, analysed the data, and prepared the manuscript. All coauthors participated in the collection and final interpretation of the data, reviewed the final manuscript, and made the decision to submit for publication. Declaration of interests RBW has received personal fees from AbbVie, Almirall, Amgen, Boehringer Ingelheim Pharma, Celgene, Janssen-Cilag, Leo, Lilly, Novartis, Pfizer, and Xenoport outside the submitted work. UM has been an advisor to, received speakers honoraria or grants from, or participated in clinical for Abbott/AbbVie, Almirall Hermal, Amgen, BASF, Biogen Idec, Boehringer Ingelheim, Celgene, Centocor, Eli Lilly, Foamix, Forward Pharma, Galderma, Janssen, Leo Pharma, Medac, MSD, Miltenyi Biotech, Novartis, Pfizer, Teva, VBL, and Xenoport. RvK has been an investigator, consultant, advisor, or speaker for Abbvie, Almirall, Amgen, Biogen Idec, Boehringer Ingelheim, Celgene, Eli Lilly, GSK, Leo, Janssen-Cilag, MSD, Novartis, Pfizer, UCB, and VBL Pharma. JN has received grants from Amgen, Novartis, Janssen-Cilag, LEO, Lilly, Medac, Regeneron, and Dermapharm, outside the submitted work. DW-T has has been an advisor to, received speakers honoraria or grants from, or participated in clinical for Abbvie, Almirall, Amgen, Biogen, Boehringer Ingelheim Pharma, Celgene, Forward Pharma, GlaxoSmithKline, Janssen-Cilag, Leo, Lilly, Medac, Merck Sharp & Dohme, Novartis, Pfizer, UCB Pharma, and VBL. KG has been an advisor to, received speakers honoraria or grants from, or participated in clinical for Abbott/AbbVie, Almirall, Biogen, Boehringer Ingelheim, Celgene, Delenex, Eli Lilly, Galderma, Janssen, Medac, MSD, Novartis, and Pfizer. KR has received personal fees from AbbVie, Amgen, Biogen, Boehringer Ingelheim, Celgene, Centocor, Covagen, Forward Pharma, GlaxoSmithKline, Janssen-Cilag, Leo, Lilly, Medac, Merck Sharp & Dohme, Novartis, Pfizer, Regeneron, Takeda, UCB Pharma, and Xenoport, outside the submitted work. IZ, TMF, and NB-S declare no competing interests. Acknowledgments METOP was an investigator-initiated trial supported by a grant from Medac Germany to KR. Medac also supplied study medication. The study was designed by consultant experts in psoriasis (RBW, UM, and

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4

9

Articles

KR) in conjunction with employees of SCIderm GmbH, Germany, which served as the clinical research organisation for study management, data collection, and statistical analysis. References 1 WHO. Global report on psoriasis. 2016. http://apps.who.int/iris/ bitstream/10665/204417/1/9789241565189_eng.pdf?ua=1 (accessed April 1, 2016). 2 Reich K, Nestle FO, Papp K, et al. Infliximab induction and maintenance therapy for moderate-to-severe psoriasis: a phase III, multicentre, double-blind trial. Lancet 2005; 366: 1367–74. 3 Papp KA, Langley RG, Lebwohl M, et al. Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with psoriasis: 52-week results from a randomised, double-blind, placebo-controlled trial (PHOENIX 2). Lancet 2008; 371: 1675–84. 4 Langley RG, Elewski BE, Lebwohl M, et al. Secukinumab in plaque psoriasis—results of two phase 3 trials. N Engl J Med 2014; 371: 326–38. 5 Lebwohl MG, Kavanaugh A, Armstrong AW, Van Voorhees AS. US perspectives in the management of psoriasis and psoriatic arthritis: patient and physician results from the population-based Multinational Assessment of Psoriasis and Psoriatic Arthritis (MAPP) survey. Am J Clin Dermatol 2016; 17: 87–97. 6 Warren RB, Chalmers RJ, Griffiths CE, Menter A. Methotrexate for psoriasis in the era of biological therapy. Clin Exp Dermatol 2008; 33: 551–54. 7 West J, Ogston S, Foerster J. Safety and efficacy of methotrexate in psoriasis: a meta-analysis of published trials. PLoS One 2016; 11: e0153740. 8 Saurat JH, Langley RG, Reich K, Unnebrink K, Sasso EH, Kampman W. Relationship between methotrexate dosing and clinical response in patients with moderate to severe psoriasis: subanalysis of the CHAMPION study. Br J Dermatol 2011; 165: 399–406. 9 Reich K, Langley RG, Papp KA, et al. A 52-week trial comparing briakinumab with methotrexate in patients with psoriasis. N Engl J Med 2011; 365: 1586–96. 10 Barker J, Hoffmann M, Wozel G, et al. Efficacy and safety of infliximab vs. methotrexate in patients with moderate-to-severe plaque psoriasis: results of an open-label, active-controlled, randomized trial (RESTORE1). Br J Dermatol 2011; 165: 1109–17. 11 Braun J, Kästner P, Flaxenberg P, et al. Comparison of the clinical efficacy and safety of subcutaneous versus oral administration of methotrexate in patients with active rheumatoid arthritis: results of a six-month, multicenter, randomized, double-blind, controlled, phase IV trial. Arthritis Rheum 2008; 58: 73–81. 12 Finlay AY. Current severe psoriasis and the rule of tens. Br J Dermatol 2005; 152: 861–67. 13 Hawwa AF, AlBawab A, Rooney M, Wedderburn LR, Beresford MW, McElnay JC. Methotrexate polyglutamates as a potential marker of adherence to long-term therapy in children with juvenile idiopathic arthritis and juvenile dermatomyositis: an observational, cross-sectional study. Arthritis Res Ther 2015; 17: 295. 14 Dervieux T, Furst D, Lein DO, et al. Pharmacogenetic and metabolite measurements are associated with clinical status in patients with rheumatoid arthritis treated with methotrexate: results of a multicentred cross sectional observational study. Ann Rheum Dis 2005; 64: 1180–85.

10

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Reich K, Papp KA, Matheson RT, et al. Evidence that a neutrophil-keratinocyte crosstalk is an early target of IL-17A inhibition in psoriasis. Exp Dermatol 2015; 24: 529–35. Mrowietz U, Kragballe K, Reich K, et al. Definition of treatment goals for moderate to severe psoriasis: a European consensus. Arch Dermatol Res 2011; 303: 1–10. Stamp LK, Barclay ML, O’Donnell JL, et al. Effects of changing from oral to subcutaneous methotrexate on red blood cell methotrexate polyglutamate concentrations and disease activity in patients with rheumatoid arthritis. J Rheumatol 2011; 38: 2540–47. Yesudian PD, Leman J, Balasubramaniam P, et al. Effectiveness of subcutaneous methotrexate in chronic plaque psoriasis. J Drugs Dermatol 2016; 15: 345–49. West J, Ogston S, Palmer C, et al. Methotrexate in psoriasis under real-world conditions: long-term efficacy and tolerability. Br J Dermatol 2016; 174: 1407–10. Montaudié H, Sbidian E, Paul C, et al. Methotrexate in psoriasis: a systematic review of treatment modalities, incidence, risk factors and monitoring of liver toxicity. J Eur Acad Dermatol Venereol 2011; 25: (suppl 2): 2–11. Vincenzi F, Padovan M, Targa M, et al. A(2A) adenosine receptors are differentially modulated by pharmacological treatments in rheumatoid arthritis patients and their stimulation ameliorates adjuvant-induced arthritis in rats. PLoS One 2013; 8: e54195. Spurlock CF 3rd, Tossberg JT, Matlock BK, Olsen NJ, Aune TM. Methotrexate-mediated inhibition of nuclear factor κB activation by distinct pathways in T cells and fibroblast-like synoviocytes. Rheumatology 2015; 54: 178–87. Talme T, Bergdahl E, Sundqvist KG. Methotrexate and its therapeutic antagonists, caffeine and theophylline, target a motogenic T-cell mechanism driven by thrombospondin-1 (TSP-1). Eur J Immunol 2016; 46: 1279–90. Zaba LC, Fuentes-Duculan J, Eungdamrong NJ, et al. Psoriasis is characterized by accumulation of immunostimulatory and Th1/Th17 cell-polarizing myeloid dendritic cells. J Invest Dermatol 2009; 129: 79–88. Goldminz AM, Suárez-Fariñas M, Wang AC, et al. CCL20 and IL22 messenger RNA expression after adalimumab vs methotrexate treatment of psoriasis: a randomized clinical Trial. JAMA Dermatol 2015; 151: 837–46. Reich K, Mrowietz U, Radtke MA, et al. Drug safety of systemic treatments for psoriasis: results from The German Psoriasis Registry PsoBest. Arch Dermatol Res 2015; 307: 875–83 Warren RB, Smith RL, Campalani E, et al. Genetic variation in efflux transporters influences outcome to methotrexate therapy in patients with psoriasis. J Invest Dermatol 2008; 128: 1925–29. Becker ML, Gaedigk R, van Haandel L, et al. The effect of genotype on methotrexate polyglutamate variability in juvenile idiopathic arthritis and association with drug response. Arthritis Rheum 2011; 63: 276–85. Peres RS, Liew FY, Talbot J, et al. Low expression of CD39 on regulatory T cells as a biomarker for resistance to methotrexate therapy in rheumatoid arthritis. Proc Natl Acad Sci USA 2015; 112: 2509–14.

www.thelancet.com Published online December 21, 2016 http://dx.doi.org/10.1016/S0140-6736(16)32127-4