A, Mean difference in the Psoriasis Area and Severity Index (PASI) scores between the usual care (UC) and dose reduction (DR) groups after 12 months. B, Mean difference in the Dermatology Life Quality Index (DLQI) scores between the DR and UC groups after 12 months. C, Comparison of the PASI scores between the UC and DR groups. The PASI scores are depicted at months 0, 3, 6, 9, and 12. D, Comparison of the DLQI scores between the UC and DR groups at months 0, 3, 6, 9, and 12. Errors bars indicate 95% CIs. The dotted vertical line depicts the noninferiority margin in context of the PASI and DLQI difference.
aP < .05.
DLQI indicates Dermatology Life Quality Index; PASI, Psoriasis Area and Severity Index.
eMethods. Detailed Methods
eFigure 1. Study Population and Decision Steps
eFigure 2. Linear Mixed Model Analyses PASI, DLQI and SF-36
eFigure 3. Kaplan-Meier Curve Presenting Time Until Persistent Flare in the Usual Care and Dose Reduction Group
eTable 1. Subgroup Analyses: Outcome Measures at Baseline and End of Study Month
eTable 2. Proportions of Patients With Dosing Steps After 12 Months
eTable 3. Factors Associated With Failure of Dose Tapering
eTable 4. Cumulative Dose Per Patients in 12 Months
eResults. Adverse Events
Data Sharing Statement
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Atalay S, van den Reek JMPA, den Broeder AA, et al. Comparison of Tightly Controlled Dose Reduction of Biologics With Usual Care for Patients With Psoriasis: A Randomized Clinical Trial. JAMA Dermatol. 2020;156(4):393–400. doi:10.1001/jamadermatol.2019.4897
Is a dose-reduction strategy of biologics noninferior to usual care in patients with stable psoriasis?
In this noninferiority randomized clinical trial of patients with plaque psoriasis, noninferiority could not be demonstrated for the Psoriasis Area and Severity Index at 12 months (dose reduction strategy vs usual care), although the mean Psoriasis Area and Severity Index difference was small. For the Dermatology Life Quality Index, noninferiority could be demonstrated, and dose tapering did not lead to more persistent flares.
Dose reduction in real-life situations is possible, but a tight control scheme that monitors the Psoriasis Area and Severity Index and Dermatology Life Quality Index is warranted.
Biologics revolutionized the treatment of psoriasis. Biologics are given in a fixed dose, but lower doses might be possible.
To investigate whether dose reduction (DR) of biologics in patients with stable psoriasis is noninferior to usual care (UC).
Design, Setting, and Participants
This pragmatic, open-label, prospective, controlled, noninferiority randomized clinical trial was conducted from March 1, 2016, to July 22, 2018, at 6 dermatology departments in the Netherlands. A total of 120 patients with plaque psoriasis and stable low disease activity who were receiving treatment with adalimumab, etanercept, or ustekinumab were studied.
Patients were randomized 1:1 to DR (n = 60) or UC (n = 60). In the DR group, injection intervals were prolonged stepwise, leading to 67% and 50% of the original dose.
Main Outcomes and Measures
The primary outcome was between-group difference in disease activity corrected for baseline at 12 months compared with the predefined noninferiority margin of 0.5. Secondary outcomes were Psoriasis Area and Severity Index (PASI) score and health-related quality of life (including Dermatology Life Quality Index [DLQI] and Medical Outcomes Study 36-Item Short Form Health Survey scores), proportion of patients with short and persistent flares (defined as PASI and/or DLQI scores >5 for ≥3 months), and proportion of patients with successful dose tapering.
Of 120 patients (mean [SD] age, 54.0 [13.2] years; 82 [68%] male), 2 patients were lost to follow-up, 2 patients had a protocol violation, and 5 patients had a protocol deviation, leaving 111 patients for the per-protocol analysis (53 in the DR group and 58 in the UC group). The median PASI scores at month 12 were 3.4 (interquartile range [IQR], 2.2-4.5) in the DR group and 2.1 (IQR, 0.6-3.6) in the UC group (mean difference, 1.2; 95% CI, 0.7-1.8). This indicates that noninferiority was not demonstrated for DR compared to UC. The median DLQI score at month 12 was 1.0 (IQR, 0.0-2.0) in the DR group and 0.0 (IQR, 0.0-2.0) in the UC group (mean difference, 0.8; 95% CI, 0.3-1.3), indicating noninferiority for DR compared with UC. No significant difference was found regarding persistent flares between groups (n = 5 in both groups). Twenty-eight patients (53%; 95% CI, 39%-67%) in the DR group tapered their dose successfully at 12 months. No severe adverse events related to the intervention occurred.
Conclusions and Relevance
In this trial, noninferiority was not demonstrated for DR of adalimumab, etanercept, and ustekinumab based on the PASI in patients with psoriasis compared with UC with the chosen noninferiority margin. However, the strategy was noninferior based on the DLQI. Dose tapering did not lead to persistent flares or safety issues.
ClinicalTrials.gov Identifier: NCT02602925
Psoriasis is an immune-mediated inflammatory skin disease for which several biologics are available. These drugs play an important role in the management of psoriasis, with adalimumab, etanercept, and ustekinumab being frequently used.1 However, long-term immunosuppressive effects of biologics may not be without risk, and biologics are expensive, placing a high financial burden on health care systems.2 For both reasons, aiming for the lowest effective dose is important.
Several studies3-5 have indicated that withdrawal of biologics frequently leads to quick relapse of psoriasis. However, dose reduction (DR) might lead to lower cumulative exposure without losing clinical efficacy in patients with psoriasis.6-12 These studies were not randomized and lacked a control group.
We hypothesized that when disease activity is monitored intensively while the biologic interval is prolonged stepwise, timely action can be taken to prevent persistent flares. Such a treatment strategy would allow physicians to determine safely the lowest effective dose for the individual patient. The aim of this trial was to investigate whether a disease activity–guided DR strategy is noninferior compared with usual care (UC) in patients with psoriasis.
This pragmatic, open-label, noninferiority randomized clinical trial of adalimumab, etanercept, and ustekinumab was conducted from March 1, 2016, to July 22, 2018, at 1 academic and 5 regional Dutch hospitals. The study protocol is available in Supplement 1. Patients 18 years or older with plaque psoriasis and stable low disease activity using standard doses of adalimumab, etanercept, or ustekinumab for at least 6 months were included. We defined stable low disease activity as a Psoriasis Area and Severity Index (PASI) score of 5 or lower at 2 subsequent visits in the past 6 months and a Dermatology Life Quality Index (DLQI) score of 5 or lower at study inclusion.13 Exclusion criteria were a biologic indication other than psoriasis, concomitant use of systemic immunosuppressants, or severe comorbidities with short life expectancy. The study protocol and amendments were approved by the ethics committee of Radboud University Medical Center. All patients provided written informed consent. Data were deidentified, but a coding list was kept in a locked file separately; therefore, data were pseudonymized. The trial was performed in accordance with the Declaration of Helsinki14 and Good Clinical Practice guidelines. This trial followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. The rationale and design are described elsewhere.13
In total, 120 patients were randomly assigned (1:1) to DR (n = 60) or UC (n = 60), and web-based randomization was performed with stratification by agent using variable permuted blocks with concealment allocation.15 All patients were seen once every 3 months. Dose reduction steps are described in the eMethods and eFigure 1 in Supplement 2. The PASI scores were assessed by treating physicians and research nurses with specific PASI training (S.A., J.M.P.A.v.d.R., L.J.v.V., M.D.N., J.M.M., P.M.O., M.I.K., M.A.B., and E.M.G.J.d.J.). No strict treatment rules were applied in the UC group, and in both groups, patients were allowed to continue using methotrexate and acitretin or start or stop topical therapy. To monitor disease activity and flares, the PASI and DLQI scores were assessed at study visits. We defined short disease flares as a single PASI score greater than 5 and/or a single DLQI score greater than 5. When short disease flares occurred, patients returned to their previous effective dose or original dose (eFigure 1 in Supplement 2). No further attempts for DR were made. Patients were instructed to contact their physician between visits if needed; this was followed by an additional visit or telephone contact within a week.
The primary outcome was between-group difference in disease activity (PASI) corrected for baseline PASI at 12 months compared with the noninferiority margin. Secondary outcomes were PASI score and health-related quality of life (including DLQI and Medical Outcomes Study 36-Item Short Form Health Survey [SF-36] scores), proportion of patients with short and persistent flares (we defined persistent flares as PASI and/or DLQI scores >5 for ≥3 months), and proportion of patients with successful dose tapering. Successful dose tapering was defined as patients with a lower dose than normal while maintaining PASI and DLQI scores of 5 or lower. Other secondary outcomes were indicators of successful dose tapering, treatment characteristics, level of C-reactive protein, severe adverse events (AEs), severe AEs related to the intervention, and AEs of special interest.
Statistical analyses were performed according to the previously published analysis plan.13 Per-protocol analyses were performed because this type of analysis is the preferred and most conservative for noninferiority studies.16 Because no generally accepted noninferiority margin was available for absolute PASI, the noninferiority margin was set as one-third of the SD of PASI based on a rule of thumb for calculating standardized effect sizes (difference between groups = standardized effect size × pooled population SD).17 The standardized effect size was set at 0.3.18 The SD of PASI in the BioCapture registry19,20 was 1.4, leading to a margin of 0.5. Because we performed our analysis with correction for the baseline PASI score, a formula that increases power significantly was used to account for the correlation between baseline and follow-up PASI scores.21 This correlation was 0.67 in the BioCapture registry. When performing an unpaired, 2-tailed t test for sample size calculation, the sample size could be multiplied by 1 − ρ2 after adding 1 extra person per group (ρ is the correlation between the baseline measure and the follow-up measure). Thus, when performing an analysis of covariance corrected for baseline PASI score with the noninferiority margin set at 0.5, α = .05, and ρ = 0.67, we achieved 80% power when enrolling 54 patients per arm. To account for a 10% rate of missing individuals, 60 patients were included per arm.
For the primary outcome, an additional intention-to-treat analysis was performed. Patients who deviated from the protocol were included for the intention-to-treat analysis. The primary outcome of PASI score at 12 months was corrected for the baseline PASI score using a general linear model (analysis of covariance). Noninferiority was assumed if the upper limit of the 95% CI of this difference did not exceed the noninferiority margin.
For all outcome measures (PASI, DLQI, and SF-36), descriptive statistics were computed at all time points, and scores were, per time point, compared between the treatment groups. To gain insight into the differences in the course of these outcomes between groups, we decided ad hoc to perform additional linear mixed-model analyses (eMethods in Supplement 2). For the DLQI, we performed a prespecified noninferiority analysis at 12 months with a predefined noninferiority margin that was based on the rule of thumb (one-third SD) and clinical relevance. An SD of 6 on the DLQI for patients from the BioCapture registry led to a noninferiority margin of 2.13 Proportions of patients with short and persistent flares were compared between groups using the Fisher exact test.
The proportion of patients with successful dose tapering was calculated. Indicators of successful dose tapering, expressed as relative risks, were analyzed with a log binomial regression model. Possible indicators of successful dose tapering were selected a priori based on clinical relevance and the literature. When more than 1 variable in univariable log binomial regression analysis produced a P < .20, a multivariable log binomial regression analysis was planned. A backward selection procedure was used based on P < .05 to arrive at a final model. Because the log binomial model was unable to fit in a multivariable setting, PASI score as the continuous variable was recoded into 2 categories: higher than 3.0 and 3.0 or lower.
Baseline characteristics and treatment characteristics were analyzed using descriptive statistics. For nominal variables, numbers (percentages) are presented, and for continuous variables, means (SDs) or medians (interquartile ranges [IQRs]) are presented depending on skewness of the data. The mean cumulative dose per patient for each biologic throughout the study was calculated. The percentage of the dose used in DR compared with UC was calculated.
The severe AEs were described per group with corresponding monthly event rates and rate ratios to compare the groups. All AEs and their association with dose tapering were assessed by 2 reviewers (S.A., J.M.P.A.v.d.R.). We decided ad hoc to perform additional analyses on AEs of special interest, including musculoskeletal concerns, given that a substantial number of patients also had psoriatic arthritis. All musculoskeletal AEs were checked by a rheumatologist (A.A.d.B.) as a third reviewer. Two classifications were made: evident psoriatic arthritis (exacerbations) and all musculoskeletal concerns.
Statistical analyses were performed with SPSS software, version 23.0 (SPSS Inc) and SAS software, version 9.4 (SAS Institute Inc).
In total, 120 patients were included in the study (mean [SD] age, 54.0 [13.2] years; 82 [68%] male), with 60 assigned to DR and 60 to UC (Figure 1). Two patients were lost to follow-up, 2 patients had a protocol violation, and 5 patients had a protocol deviation, leaving 111 patients for the per-protocol analysis (53 in the DR group and 58 in the UC group). Of the 2 patients lost to follow-up, one was lost to follow-up because of psychiatric illness and the other patient continued treatment in a nonparticipating hospital. Demographic and baseline characteristics were well balanced in the DR and UC groups (Table 1). These characteristics mimic a typical cohort of psoriasis patients taking biologics except for low disease activity at baseline. The number of missing visits was small (1% of all planned visits [1% for PASI, 2% for DLQI, and 6% for SF-36]), and multiple imputation was deemed to be unnecessary.
The per-protocol analysis found a median PASI score at month 12 of 3.4 (IQR, 2.2-4.5) in the DR group and 2.1 (IQR, 0.6-3.6) in the UC group. The mean difference in the PASI score at month 12, corrected for baseline PASI score, was 1.2 (95% CI, 0.7-1.8). The lower limit of the 95% CI exceeded the noninferiority margin by 0.5 (Figure 2A). The intention-to-treat analysis found similar results; the mean difference in PASI score at month 12, corrected for baseline PASI score, was 1.1 (95% CI, 0.6-1.7).
Disease activity (PASI) was compared at all time points between the UC and DR groups (Figure 2B). A significant difference was seen between the UC and DR scores at 9 months (mean UC score, 2.3; 95% CI, 1.8-2.9; mean DR score, 3.4; 95% CI, 2.8-3.9; P = .003) and 12 months (mean UC score, 2.2; 95% CI, 1.8-2.6; mean DR score, 3.4; 95% CI, 3.0-3.9; P = .001). The PASI linear mixed model is presented in eFigure 2A and eTable 1 in Supplement 2.
The median DLQI score at month 12 was 1.0 (IQR, 0.0-2.0) for the DR group and 0.0 (IQR, 0.0-2.0) for the UC group. The mean difference for the DLQI score at month 12, corrected for baseline DLQI score, was 0.8 (95% CI, 0.3-1.3). The upper limit of the 95% CI for the mean difference did not cross the noninferiority margin of 2.0 (Figure 2C), indicating noninferiority for DR for DLQI. The median DLQI scores were compared at all time points (Figure 2D). A significant difference was observed only at 9 months (mean DLQI score in UC group, 1.4; 95% CI, 0.7-2.1; mean DLQI score in DR group, 1.6; 95% CI, 1.1- 2.1; P = .01). Linear mixed models for the DLQI and SF-36 are presented in eFigure 2B-D and eTable 1 in Supplement 2. A significant difference regarding short flares was found between the DR and UC groups during the study (19 of 53 patients [36%; 95% CI, 24%-50%] in the DR group and 8 of 58 patients [14%; 95% CI, 7%-26%] in the UC group; P = .04). No significant difference regarding persistent flares was found between the DR and UC groups during the study period (5 of 53 patients [9%; 95% CI, 4%-21%] in the DR group and 5 of 58 patients [9%; 95% CI, 3%-20%] in the UC group; P > .99). The relative risk was 1.1 (95% CI, 0.3-3.6) in the DR vs UC group (P > .99). Time until persistent flares is shown in eFigure 3 in Supplement 2. Four patients in the DR group with persistent flares did not have PASI and DLQI scores that returned to 5 or lower after 12 months. In the UC group, 1 patient did not have PASI and DLQI scores that returned to 5 or lower after 12 months. In both groups, no switch to another biologic therapy was required. Eighteen additional visits were performed in the DR group vs 1 in the UC group.
Twenty-eight patients (53%; 95% CI, 39%-67%) tapered their dose successfully after 12 months. Ten patients (19%; 95% CI, 10%-32%) were taking two-thirds of their original dose, and 18 patients (34%; 95% CI, 22%-48%) were taking half of their original dose (Figure 3). The proportion of patients with successful DR split per biologic is given in eTable 2 in Supplement 2. Thirty-six patients (68%; 95% CI, 54%-80%) used a lower dose until their 12-month visit, but 8 of the DRs (15%; 95% CI, 7%-28%) were not deemed to be successful because of a PASI or DLQI score higher than 5 at that time. No variables were associated with successful DR in the final model (eTable 3 in Supplement 2).
The mean cumulative doses during 12 months per patient for each biologic are given in eTable 4 in Supplement 2. The percentage DR varied between 23% and 34%. A significant difference was seen in topical corticosteroid use between both groups. In the DR group, 44 patients (73%; 95% CI, 60%-84%) used topical steroids for a mean of 87 days (95% CI, 65-109 days), and in the UC group, 21 patients (35%; 95% CI, 23%-48%) used topical steroids for a mean of 35 days (95% CI, 20-50 days) in 12 months (P < .001). Rates of additional methotrexate or acitretin use were 13% in the DR group and 12% in the UC group at baseline. During the study, the number of patients using methotrexate or acitretin (including doses) remained stable.
Six patients (11%; 95% CI, 5%-24%) in the DR group and 6 patients (10%; 95% CI, 4%-22%) in the UC group reported at least 1 severe AE (P > .99). None of these severe AEs were assumed to be causally related to DR. One patient had 5 severe AEs because of preexistent peripheral arterial disease. Relevant AEs and severe AEs and corresponding rates and rate ratios are given in Table 2. No significant differences in AEs were seen in either group except for general musculoskeletal concerns, which occurred more often in the DR group (rate ratio, 4.92; 95% CI, 2.04-11.87; P < .001). In each group, 1 patient presented with evident psoriatic arthritis concerns during the study, and both patients had been diagnosed as having psoriatic arthritis before study inclusion. Adverse events are described in the eResults in Supplement 2.
On the basis of the examined DR strategy, noninferiority of the PASI score at 12 months was not demonstrated for DR of adalimumab, etanercept, and ustekinumab in patients with psoriasis compared with UC. The mean difference in the PASI score between groups was 1.2 (95% CI, 0.7-1.8) and was larger than the noninferiority margin of 0.5. However, noninferiority of the DLQI was demonstrated. The mean difference in the DLQI score between both groups at 12 months was 0.8 (95% CI, 0.3-1.3), which did not exceed the predefined noninferiority margin of 2.0. No significant differences were found regarding persistent flares. No severe AEs related to intervention and no differences in quality of life between groups were seen. Successful DR was possible in 53% (95% CI, 39%-67%) of patients. The mean cumulative dose for DR was 34% lower compared with the UC dose in patients taking adalimumab, 26% lower in patients taking etanercept, 23% lower in patients taking 45 mg of ustekinumab, and 34% lower in patients taking 90 mg of ustekinumab.
As stated, noninferiority could not be demonstrated for the primary end point. We found a mean difference in PASI score of 1.2, which exceeded the noninferiority margin. This finding means that physicians and patients need to be aware of the chance of an increase in PASI score when trying to reduce the dose. It is important to weigh the risk of a PASI score increase against the benefits of using a lower dose of the biologic. A significant increase in nonspecific musculoskeletal concerns in the DR group compared with the UC group was reported. These concerns were mainly mild, and evident (exacerbation of) psoriatic arthritis was seen once in both the DR and UC groups. Because of the open-label design of this study, reporting bias could be present for these concerns. Patients in the DR group used more topical treatment than those in the UC group, which was inherent to the DR strategy.
For psoriasis, noncomparative, observational studies6,10-12,22,23 have been conducted with interval prolongation, showing success rates that range from 22% to 90%. Our study found successful DR in 53% of patients. Differences that were observed with other studies6,10-12,22,23 can be explained by study design, population, and the tight control character of our strategy. The present strategy was based on prevailing treatment goals, which defined low disease activity as PASI and DLQI scores of 5 or lower.24 However, current treatment goals are shifting toward a PASI score of 3 or lower.25 In our multivariate regression model, we analyzed the threshold PASI score of 3 or lower vs higher than 3 as well, which was not identified as an indicator of successful DR.
A strength of our study was the high internal validity by means of the randomized design, use of validated outcome measures, and good data integrity. The pragmatic design led to a high external validity of our results. The number of patients needed to achieve enough statistical power was met, and missing data were scarce. The risk of selection bias or confounding was deemed to be low because both groups seemed to be balanced in baseline characteristics. Incorporation of the DLQI in the tapering strategy gave important insight from the perspective of the patient.
The open-label design might be regarded as a limitation. Patients and physicians could return to using a higher dose, assuming that patients generally fear a flare. However, few patients returned at their own initiative to higher (n = 4) or original (n = 2) doses. Moreover, the goal of this pragmatic study was to provide high external validity, and blinding would have lowered external validity. An absolute PASI score was used instead of a relative PASI or Psoriasis Global Assessment score. By design, patients had low disease activity, and a PASI score of 75 of 90 is almost unachievable with a starting PASI score lower than 5. Dose reduction was accompanied by higher topical treatment use and more visits. Tumor necrosis factor and interleukin 12 and 23 blockers were analyzed as a group because the focus was on the DR strategy and not on the working mechanism itself. Thus, for per-drug conclusions, this study was underpowered but still provides relevant pilot data. Another limitation is the absence of a validated flare criterion for psoriasis. Our flare criterion is therefore a definition based on expert opinion and combined with a patient-reported outcome (DLQI). Both DLQI and PASI weighed equally on the algorithm of clinical decision-making in case a flare occurred.
In this first, to our knowledge, randomized, tightly controlled noninferiority trial, a DR strategy for adalimumab, etanercept, and ustekinumab in patients with psoriasis, noninferiority of DR compared with standard dose was not demonstrated for PASI at 12 months. Nevertheless, many patients were able to prolong the interval of their biologic use while maintaining low PASI and DLQI scores. The cumulative DR varied from 23% to 34% depending on the biologic. In 53% of patients, dose was successfully tapered, and no severe AEs related to the intervention were observed. No difference regarding persistent flares was found in either group. This disease activity–guided DR strategy is expected to have an effect on health care expenditures.
Accepted for Publication: December 16, 2019.
Corresponding Author: Elke M. G. J. de Jong, MD, PhD, Department of Dermatology, Radboud University Medical Center, Rene Descartesdreef 1, 6525 GL Nijmegen, the Netherlands (firstname.lastname@example.org).
Published Online: February 12, 2020. doi:10.1001/jamadermatol.2019.4897
Author Contributions: Drs Atalay and van den Reek share first authorship. Drs de Jong and Atalay had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: van den Reek, den Broeder, van de Kerkhof, Kievit, De Jong.
Acquisition, analysis, or interpretation of data: Atalay, van den Reek, den Broeder, van Vugt, Otero, Njoo, Mommers, Ossenkoppele, Koetsier, Berends, Groenewoud, Kievit, De Jong.
Drafting of the manuscript: Atalay, van den Reek, Groenewoud, Kievit, De Jong.
Critical revision of the manuscript for important intellectual content: Atalay, van den Reek, den Broeder, van Vugt, Otero, Mommers, Ossenkoppele, Koetsier, Berends, van de Kerkhof, Groenewoud, Kievit, De Jong.
Statistical analysis: Atalay, van den Reek, den Broeder, Groenewoud, De Jong.
Obtained funding: van den Reek, den Broeder, Kievit, De Jong.
Administrative, technical, or material support: Atalay, van Vugt, Otero, Ossenkoppele, Koetsier, van de Kerkhof, Groenewoud, De Jong.
Supervision: van den Reek, den Broeder, Njoo, Ossenkoppele, De Jong.
Conflict of Interest Disclosures: Dr Atalay reported receiving grants from the Netherlands Organisation for Health Research and Development (ZonMw) during the conduct of the study and participating in trials sponsored by Janssen Pharmaceuticals, AbbVie, Celgene, and Novartis outside the submitted work. Dr van den Reek reported receiving grants from ZonMw during the conduct of the study; participating in trials sponsored by Janssen Pharmaceuticals, AbbVie, and Celgene outside the submitted work; receiving speaking fees from Janssen Pharmaceuticals and AbbVie; and receiving reimbursement for attending a symposium from Janssen Pharmaceuticals, Pfizer, Celgene, and AbbVie. Dr van Vugt reported receiving grants from ZonMw during the conduct of the study and participating in trials sponsored by AbbVie, Janssen, Celgene, and Novartis outside the submitted work. Dr Mommers reported receiving personal fees from Novartis during the conduct of the study. Dr Ossenkoppele reported performing consulting services for AbbVie outside the submitted work. Dr van de Kerkhof reported receiving fees for consultancy service or lectureships from Celgene, Almirall, AbbVie, Amgen, Eli Lilly and Company, Novartis, Janssen Pharmaceutica, Leo Pharma, Bristol-Myers Squibb, and Dermavant. Dr de Jong reported receiving grants from ZonMw during the conduct of the study and grants from AbbVie, Janssen, Novartis, and Leo Pharma outside the submitted work. Dr de Jong reported receiving research grants from the independent research fund of the Department of Dermatology, Radboud University Medical Center, Nijmegen, the Netherlands from AbbVie, Pfizer, Novartis, Leo Pharma, and Janssen Pharmaceutica and acting as a consultant and/or paid speaker for and/or participating in research sponsored by AbbVie, Janssen Pharmaceutica, Novartis, Eli Lilly and Company, Celgene, and Leo Pharma. No other disclosures were reported.
Funding/Support: This study was funded by ZonMw, The Hague, the Netherlands (principal investigator/main applicant: Dr de Jong; coapplicants: Drs van den Reek, Kievit, den Broeder, and van de Kerkhof).
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Data Sharing Statement: See Supplement 3.
Additional Contributions: Lia Schalkwijk, BSc, Department of Dermatology, Radboud University Medical Center, Nijmegen, the Netherlands, provided data entry. She was compensated for her work. We thank all the patients who participated in this study.