REVIEW

Dynamic Visual Representation of Clinical Efficacyof Ixekizumab in Psoriasis

Jason E. Hawkes . Kyoungah See . Russel Burge . Stephanie Strakbein .

Missy McKean-Matthews . Daniel Saure . Melinda Gooderham .

Craig Leonardi

Received: February 12, 2021 / Published online: May 29, 2021� The Author(s) 2021

ABSTRACT

Introduction: Ixekizumab, a high-affinitymonoclonal antibody that selectively targetsinterleukin-17A, is an approved treatment forplaque psoriasis. This study aimed to use ani-mated visualizations as a tool to simplify com-plex data from ixekizumab clinical trials.Methods: Animated visualizations were devel-oped to show outcomes from ixekizumab clin-ical trials and a Bayesian network meta-analysisof 11 approved biologics. The visualizationssimultaneously highlighted both aggregate

scores and the individual progression ofpatients over the course of treatment.Results: The animations provided key messagesand information from the complex data inefficient and scientific ways that were alsovisually pleasing and simple to understand. Theanimations highlighted (1) rapid reduction indisease severity from baseline; (2) sustainedefficacy of ixekizumab in the treatment of skinand nail psoriasis; (3) side-by-side comparisonsof treatment efficacy and clinical improvementacross trials; (4) simultaneous visual presenta-tion of individual results with summaryresponse over time; and (5) indirect comparisonof relative treatment effects with other biologicsbased on Bayesian network meta-analysis.Conclusion: The rapid and sustained efficacy ofixekizumab in the treatment of psoriasis wasdemonstrated using multiple dynamic visual-izations with different clinical endpoints. Ani-mated visualizations provided a simpler andmore comprehensive understanding of complexdata than conventional static figures.

Keywords: Animated visualization; Biologics;Ixekizumab; Meta-analysis

Supplementary Information The online versioncontains supplementary material available at https://doi.org/10.1007/s13555-021-00548-2.

J. E. Hawkes (&)Department of Dermatology, University ofCalifornia-Davis, Sacramento, CA, USAe-mail: jehawkes@ucdavis.edu

K. See � R. Burge � S. Strakbein � D. SaureEli Lilly and Company, Indianapolis, IN, USA

R. BurgeUniversity of Cincinnati, Cincinnati, OH, USA

M. McKean-MatthewsSyneos Health, Raleigh, NC, USA

M. GooderhamSKiN Centre for Dermatology, Peterborough, ON,Canada

C. LeonardiCentral Dermatology, St. Louis, MO, USA

Dermatol Ther (Heidelb) (2021) 11:1107–1118

https://doi.org/10.1007/s13555-021-00548-2

Key Summary Points

Why carry out this study?

Ixekizumab, a high-affinity monoclonalantibody that selectively targetsinterleukin-17A, is an approved treatmentfor plaque psoriasis.

This study aimed to use animatedvisualizations as a tool to simplifycomplex data from ixekizumab clinicaltrials.

What was learned from this study?

Dynamic animated visualizations weredeveloped by applying a combination ofstatistical methods, design principles, anddata science.

The animations highlighted both theaggregate and individual response ratesfrom treatment with ixekizumab andother biologics and showed the indirectcomparison of relative treatment effectsbased on Bayesian network meta-analysis.

Animated visualizations can be used toidentify meaningful aspects of clinicaltrial data, such as individual patientvariability and trends during clinical trialtesting.

DIGITAL FEATURES

This article is published with digital features,including a summary slide and video anima-tions to facilitate understanding of the article.To view digital features for this article go tohttps://doi.org/10.6084/m9.figshare.14518881.

INTRODUCTION

Plaque psoriasis is a chronic inflammatory dis-ease that causes characteristic scaly plaques onthe scalp, buttocks, trunk, and extremities [1]

and may increase the risk of other psoriasis-as-sociated conditions such as cardiovascular dis-ease, obesity, diabetes, psoriatic arthritis, andpsychological comorbidities [2, 3]. Patients withplaque psoriasis may face significant quality-of-life challenges [4], highlighting the need forhighly effective long-term treatments [1].

Biologics, such as antibodies that inhibittumor necrosis factor alpha (TNF-a) and inter-leukins (IL)-17 and IL-23, have transformed thetreatment landscape of psoriasis compared totraditional topical therapies or phototherapy,which may not adequately control moderate-to-severe disease [5, 6]. The comparative effective-ness of biologics used to treat psoriasis has beenevaluated in head-to-head studies as well asindirect comparisons using network meta-anal-ysis (NMA) [7–11]. Ixekizumab, a high-affinitymonoclonal antibody that selectively targets IL-17A, is an example of one such biologic; it hasbeen studied in head-to-head trials versusetanercept, ustekinumab, adalimumab, andguselkumab [10, 12–14] and demonstrated rapidefficacy, followed by long-term durability oftreatment response [7–11]. The efficacy ofixekizumab and other phase III-approved bio-logics was highlighted in a Bayesian NMA,showing the competitive landscape of psoriasistreatments.

Comparative studies and NMAs generatelarge amounts of complex data, and staticgraphs and figures are limited in how theydepict changes in data over time, especially asthey relate to individual variability. Animatedvisualizations can be used to present key mes-sages and information from complex data in anefficient and scientific manner that is alsovisually pleasing and simple to understand.These dynamic animations are more engagingand may appeal to a broader audience withheterogeneous learning styles. Here we presentseveral types of animated visualizations thathighlight data on the efficacy of ixekizumab inthe treatment of moderate-to-severe psoriasis.

1108 Dermatol Ther (Heidelb) (2021) 11:1107–1118

METHODS

Studies on Ixekizumab Efficacy

Clinical efficacy data on ixekizumab and com-parator drugs were taken from published stud-ies, including the UNCOVER-2 (NCT01597245),UNCOVER-3 (NCT01646177), IXORA-R(NCT03573323), IXORA-S (NCT02561806), andSPIRIT-H2H (NCT03151551) trials. Detailsregarding the study design, patient selection,treatment arms, and outcome measures werepreviously described for all studies [10, 12–14]and are outlined in Electronic SupplementaryMaterial (ESM) Table 1.

AssessmentsThe following parameters were evaluated tocompare onset, durability of response, andrapidity of skin clearance with biologics.

Mean percentage improvement in PsoriasisArea and Severity Index (PASI) score was asses-sed for:

• Ixekizumab versus etanercept (integratedUNCOVER-2 and UNCOVER-3) from week0 to week 12.

• Ixekizumab versus adalimumab from week 0to week 12 (SPIRIT-H2H is a psoriatic arthri-tis study with the primary endpoint at week24; however, only data up to week 12 ispresented in this paper for consistency withthe other studies).

• Ixekizumab versus guselkumab from week 0to week 12 (IXORA-R).

• Ixekizumab versus ustekinumab from week 0to week 12 (IXORA-S).

Static Physician’s Global Assessment (sPGA)0 and sPGA (0, 1) response rates were comparedbetween ixekizumab and ustekinumab fromweek 0 to week 52 (IXORA-S). In UNCOVER-3,percentage improvement in Nail PsoriasisSeverity Index (NAPSI) scores from baseline toweek 264 was evaluated. After 12 weeks oftreatment, patients who received ixekizumabonce every 2 weeks entered the long-termextension period; ixekizumab was then admin-istered as open-label treatment once every4 weeks.

Studies and Procedure for NMA

Bayesian and frequentist NMAs were conductedto compare the rapidity of response of 11 bio-logics in the treatment of moderate-to-severepsoriasis in adults with IL-17, IL-23p19, IL-12/23p40, and TNF-a (all inhibitors are approvedby the U.S. Food and Drug Administration) [8].A systematic literature review was conducted toidentify phase 3 randomized studies of biologicswith PASI response rates at weeks 2, 4, 8, and 12and Dermatology Life Quality Index (DLQI; 0,1) response rates at week 12. Relative treatmenteffects with the biologics studied compared toplacebo were evaluated. Details of the 33 studiesincluded for this analysis have been previouslydescribed [8]. Data related to DLQI were notincluded in this paper.

Data were utilized from an NMA on theaverage treatment response with biologics,based on posterior distribution of the relativetreatment effects compared to placebo over12 weeks [8].

Statistical Methods and Analysis

The PASI total score and mean percentageimprovement were provided by least squaremeans from mixed models using repeatedmeasures (MMRM) for ixekizumab and placeboin the induction period of UNCOVER-3 (Fig. 1).For categorical PASI response rates, the numberof patients and percentages were summarizedwith non-responder imputation to account formissing data. The model included the fixedeffects of treatment, study, visit, and the treat-ment-by-visit interaction, and the fixed covari-ate of the baseline value. The individual patientrecords of percentage improvement in PASIscore from baseline through week 12 by baselinePASI score are displayed in Video 1 (onlineversion only). PASI baseline score[40 is pre-sented in a gray area to capture outliers.

Mean percentage improvements in NAPSIscore was shown from baseline to 5 years in theUNCOVER-3 study (induction and long-termextension) using a modified baseline observa-tion carried forward imputation (Fig. 2). MeanNAPSI percentage improvement (see Video 2

Dermatol Ther (Heidelb) (2021) 11:1107–1118 1109

[online version only]) at each planned visit wascalculated using MMRM following the modelsfor each study and then plotted as a smoothedline. The smoothed line was generated usingBezier spline to determine mean percentageimprovement in PASI/NAPSI score betweenvisits to create the effect of line-filling overtime.

For the UNCOVER-2 and UNCOVER-3 trials,the model included factors related to individualstudies, such as the fixed covariates of baselinevalue, treatment, study, visit, and the treat-ment-by-visit interaction. In the IXORA-Ranalysis, the model included treatment, pooledcenter, baseline value, visit, treatment-by-visitinteraction, and the baseline value by visit. Forthe IXORA-S analysis, the model includedbaseline value, treatment, visit, weight group(B 100 kg,[ 100 kg), region group (WesternEuropean Union [EU], Eastern EU, and NorthAmerica), and the treatment-by-weight andtreatment-by-region interaction terms. Themodel for the SPIRIT-H2H study includedtreatment group, concomitant conventionalsynthetic disease-modifying anti-rheumaticdrug use at baseline, moderate-to-severe plaquepsoriasis involvement, visit, baseline value, andbaseline-by-visit and treatment-by-visit interac-tions terms.

For the head-to-head comparison animation(see Video 3 [online version only]), mean per-centage improvement in PASI score at eachplanned visit was calculated using MMRM fol-lowing the models for each study and thenplotted as a smoothed line. Similar to Fig. 2, thesmoothed line for Fig. 3 was also generatedusing Bezier spline.

An animation with two separate compo-nents—patient-level data at each planned visiton the left and summary sPGA (0) and sPGA (0,1) response rates for observed records on theright—is shown in Video 4 (online versiononly). The lines connecting response rates weresmoothed using Bezier curves. Solid lines con-nected sPGA (0, 1) response rates and dashedlines connected sPGA (0) response rates. The leftside of the figure shows individual patientrecords of sPGA at each pre-planned study visitin the 52-week treatment period. Patientrecords were arranged into columns by sPGA

value (0–5) and by treatment arm, which wasalso designated by color. Observed records weredenoted by a solid circle in the color of therespective treatment arm. Patients who discon-tinued treatment were denoted by a solid blackcircle in the column of their respective priortreatment and their sPGA position was deter-mined using mBOCF methodology. Patientswho missed a planned visit, but did not dis-continue from the study, appeared as a hollowblack circle at the missing visit and were posi-tioned according to their last observed sPGAvalue until the next observed visit occurred. Theradius of the patients’ circles were scaled pro-portionately to the inverse of the sample size bytreatment arm to make the relative heights ofthe dot-columns representative of the propor-tion of patients in the treatment arm achievingthe respective endpoint. Without adjustment,unequal sample sizes in treatment arms bias theimpact of the stacked circles when presentedside by side. Movement of the patients’ dotsbetween sPGA columns was controlled by aBezier curve drawn between the coordinates ofthe proceeding and subsequent positions of thedots at planned visits.

A representative sample of posterior treat-ment effects was generated at each time pointand treatment effects from this sample wereinterpolated using a Bezier spline function tocreate smooth transitions between posteriordistribution curves at each time point (seeVideo 5 [online version only]; ESM Fig. 2).Himalayan and ridgeline curves that were far-ther along on the x-axis (right-hand side) rep-resented a greater estimated treatment effect.Taller and narrow curves represented morereliable estimates of response rates compared toshorter and wider curves.

Figures 1, 2, 3, 5 and ESM Fig. 2 were createdusing R versions 3.6.0, grid 3.6.0 (R DevelopmentCore Team 98-2013; R Foundation for StatisticalComputing, Vienna, Austria ), gridExtra 2.3 [15],plyr 1.8.4 [16]. Figure 4 was created using D3.js(Bostock 2019).

The Bayesian NMA results were obtainedbased on Bayesian Tool for Meta-Analysis ofNetworks (rjags 4-9 via R) with fixed-treatmenteffect and random-baseline effect (ESM Fig. 1).Deviance is a goodness-of-fit statistic and was

1110 Dermatol Ther (Heidelb) (2021) 11:1107–1118

used as a model for hypothesis testing. Normalindependent models and model fit were asses-sed using the deviance information criterionand residual deviance.

This article is based on previously conductedstudies and does not contain any new studieswith human participants or animals performedby any of the authors.

RESULTS

Rapid Onset of Response Comparedto Placebo (UNCOVER-3 Trial)

An animation of patients who received eitherplacebo or ixekizumab was presented in twopanels, where each dot is an individual patient(Fig. 1); it showed how PASI response changesfrom baseline PASI to week 12 PASI based on thepatients’ baseline scores. At week 0, the dots

reflected the baseline PASI distribution on thehorizontal axis. When running the animation,the dots moved to PASI improvement levels(vertical axis) at each time point. PASIimprovement endpoints, including PASI 50, 75,90, and 100 (improvement of 50, 75, 90, and100% [complete resolution of disease]) weredepicted in a gradation of green with the dark-est being 100%. When the animation waspaused at week 1, the mean percentage PASIimprovements and mean total score PASI wereseen at the top; patients who received ixek-izumab had a rapid improvement in PASI scoresversus those on placebo, including those whowere at the lower end of disease severity in thesetrials (7.3 and 19.6% for placebo; 35.1 and13.6% for ixekizumab). As the animation playedout to week 12, the distribution of dots (pa-tients) reflected the differences in efficacyobserved in the underlying clinical trial.

Fig. 1 Skyward plot at week 12. PASI response ratesamong ixekizumab-treated and placebo-treated patients inthe UNCOVER-3 trial. In the skyward scatter plotanimation, percentage improvement in PASI scores frombaseline through week 12 is highlighted for patients who

received either placebo or ixekizumab once every 2 weeksin the UNCOVER-3 trial. MMRM Mixed models usingrepeated measures, NRI non-responder imputation, PASIPsoriasis Area and Severity Index, PASI 75 C 75%reduction in PASI scores from baseline

Dermatol Ther (Heidelb) (2021) 11:1107–1118 1111

Improvement in Nail Psoriasis Comparedto Placebo (UNCOVER-3 Trial)

Improvement in nail psoriasis was highlightedin the time-course tornado animation, whichincluded the percentage improvement in NAPSIscore from baseline to 5 years as the outcomemeasure (Fig. 2; also see Video 2 [online versiononly]). The mean percentage improvements inNAPSI values were presented next to the line forthe means and 95% confidence interval, whilethe individual patient values were simultane-ously displayed by the moving blue dots up toweek 264. Based on the animated visualization,patients who received ixekizumab generallymaintained high response rates, although theyfluctuated briefly for some patients. Mostpatients consistently achieved complete clear-ance in the nails, as noted by the cluster ofpatients in the NAPSI 100 band (darkest greenregion).

Onset of Action Compared to OtherBiologics (Head-to-Head Studies)

Side-by-side time-course tornado animations(Fig. 3) were used to present data side by sidecomparing percentage improvement in PASIscore from baseline to week 12 between ixek-izumab and etanercept (UNCOVER-2 andUNCOVER-3 trials), ixekizumab and ustek-inumab (IXORA-S trial), ixekizumab and adali-mumab (SPIRIT-H2H trial), and ixekizumab andguselkumab (IXORA-R trial) (see Video 3 [onlineversion only]). Patients who received ixek-izumab had higher average PASI percentageimprovement than patients who received com-parator drugs at all time points within 12 weeksin head-to-head studies. The PASI percentageimprovement for individual patients was rep-resented by the blue and orange dots that moveup or down as they progress from left to rightover time across each figure panel. The

Fig. 2 Time-course tornado plot at week 264. Percentageimprovement in NAPSI score in patients treated with IXEin the UNCOVER-3 trial. Percentage improvement inNAPSI score is highlighted in this time-course tornadoplot visual animation from the UNCOVER-3 trial.

Patients received IXE once every 2 weeks for 12 weeks(Q2W) then IXE once every 4 weeks (Q4W) from week 16to week 264 (excluding titrated visits). IXE Ixekizumab,mBOCF modified baseline observation carried forward,NAPSI Nail Psoriasis Severity Index

1112 Dermatol Ther (Heidelb) (2021) 11:1107–1118

ixekizumab patients (blue dots) were clusteredat generally higher PASI percentage improve-ment values, which were represented by thegray gradation bands versus the competitorpatients (orange dots) who showed more vari-ability in response.

sPGA Response Rates (0, 0/1) for Ixek-izumab and Ustekinumab (IXORA-S trial)

Stacked circle animations were used to representindividual sPGA results, alongside the time-course animation with overall sPGA (0) andsPGA (0, 1) response rates for ixekizumab andustekinumab over 52 weeks (Fig. 4; also seeVideo 4 [online version only]). On the left side,this animation showed individual patient’ssPGA results across time for each time point,with the overall response rate for each sPGAvalue displayed directly under the stacked cir-cles; the dots moved between the six bars basedon their sPGA scores (0–5). The time course on

the right side showed the summary responserates for two key endpoints: those achievingsPGA (0) and those achieving sPGA (0, 1). Thistype of animation allowed individual andsummary data to be displayed simultaneously.For instance, the figure showed that at week 24,56.6% of ixekizumab-treated patients achievedsPGA 0 versus 25.0% of ustekinumab-treatedpatients. The response for individual patientscan be viewed simultaneously with the overallresponse rate for each treatment group. Scalingthe circles to adjust for unequal sample sizesallowed for the stacked circles to be interpretedlike a conventional bar chart while still showinghow individual patients move to different sPGAclassifications over time.

Treatment Effects Compared to OtherBiologics (Multiple Studies from the NMA)

Ridgeline plot animations presented relativetreatment effects with biologics compared to

Fig. 3 Side-by-side time-course tornado plot at week 52.Percentage improvement in PASI score in patients treatedwith IXE and other biologics. In the side-by-side time-courseanimation, percentage improvement in PASI score is

highlighted in this comparison of treatment response in fourpsoriasis trials with IXE vs. comparator drugs from baselinethrough week 52. ADAAdalimumab, ETN etanercept,GUSguselkumab, IXE ixekizumab, UST ustekinumab

Dermatol Ther (Heidelb) (2021) 11:1107–1118 1113

placebo over time (see Video 5 [online versiononly]). At week 12, ixekizumab, brodalumab,guselkumab, and risankizumab demonstratedthe highest treatment effects for PASI 90 withsome overlap as displayed in the animation(Fig. 5). The Himalayan animation simultane-ously showed how quickly on average changesoccur for treatments in relation to one another,as well as the treatment effect on response rates.This type of animation was useful for makingdirect and indirect comparisons across multiplecompounds and studies, which allowed forcomparison without head-to-head data. Ixek-izumab and brodalumab treatment groupsshowed more rapid response rates on PASI 75,and ixekizumab overlapped with risankizumabat week 12 (ESM Fig. 2).

DISCUSSION

Static graphs and figures are limited in depictinghow changes in data occur over time. Animatedvisualization of complex data from head-to-head and NMA psoriasis studies can presentmore detailed results and may be more engag-ing for end-users [17]; it also provides an avenuefor healthcare providers to explore data trendsand to gain clinical insights with increasedefficiency.

In our study, we developed dynamic ani-mated visualizations by applying a combinationof statistical methods, design principles, anddata science. In these novel animations, clinicalimprovement was observed from week to weekfor each treatment group—the rate of responsewas illustrated by the movement of dots

Fig. 4 Simultaneous stacked circle with time-course plotat week 52. sPGA response rates in patients treated witheither ixekizumab or ustekinumab. In the stacked circlewith time-course plot animation, response rates per the

sPGA are shown for patients treated with either ixek-izumab or ustekinumab from baseline through week 52.sPGA Static Physician’s Global Assessment, Obs Observedand mBOCF Modified Baseline Carried Forward

1114 Dermatol Ther (Heidelb) (2021) 11:1107–1118

representing patients. The scatterplot anima-tion showed the progress of individual patients,including gains and losses in improvement thatoccurred from 1 week to the next, providing amore real-world representation of what indi-viduals experience during their psoriasis treat-ment course. The dots represented percentageimprovement in PASI scores among individuals,and the thick clustering of dots indicated thatmost patients who received ixekizumab hadcomplete clearance. However, in the case oflong-term trend analysis, tornado plot anima-tions would be preferable to using a scatterplot.

A tornado time-course plot was used to showsustained efficacy and highlight percentageimprovement in NAPSI scores from baseline inpatients treated with ixekizumab. After264 weeks of treatment, both the aggregate andindividual response rates were evident. The

cluster of patients in the dark-green band alongthe top of the animation showed the proportionof patients who achieved NAPSI 100, whichvisually highlighted treatment effectivenessweek by week and over time.

The psoriasis treatment pipeline is robust,and animations that compare head-to-headstudies are useful to demonstrate the meanresponse to treatment and the overall trendsassociated with each comparator. In side-by-side time-course animations, treatment efficacyat multiple time points can be visually com-pared across multiple studies. Individual patientdata are synced by study time point so thatvariability trends can be compared both withina study and across studies. The patterns ofvariability within ixekizumab treatment wasconsistent in all four studies, while competitorsin the same studies had more variable treatment

Fig. 5 Ridgeline ‘‘Racetrack’’ plot at week 12. In theridgeline ‘‘racetrack’’ animation, the competitive landscapeof phase 3-approved biologics is shown based on a Bayesiannetwork meta-analysis, which depicts a ridgeline plot withPASI 90 scores through weeks 48–52. BRO Brodalumab,CZP certolizumab pegol GUS guselkumab, IFX infliximab,

IXE ixekizumab, PBO placebo, PASI 90 C 90%improvement in PASI from baseline, RIS risankizumab,SEC secukinumab, TIL tildrakizumab, UST ustekinumab,ETN etanercept, ADA adalimumab

Dermatol Ther (Heidelb) (2021) 11:1107–1118 1115

responses. The relative efficacy of ixekizumabversus each competitor differed across studiesbut was shown to be superior in all, as shown bythe mean lines.

When ordinal data are available, stackedcircle animations can be used to depict overalldata trends and individual progress at each timepoint. However, tracking individual patients’trajectories may be difficult. Therefore, thisanimation is best used to highlight grossimpressions of data variability. Himalayan andridgeline plots can be used to show the results ofBayesian meta-analysis. The average response toeach treatment was calculated relative to pla-cebo and represented as a Himalayan plot orRidgeline plot. Ridgeline plots are also usefulwhen several treatments are compared, andsubstantial overlap between treatments is easilyvisualized.

The natural disease course in psoriasis ismore clearly represented using animations thanstatic figures given that dynamic animationshighlight disease variations and treatmentresponses over time. This can be illustrated forspecific body sites, such intertriginous skin,scalp, or nails, to provide important clinicalinsights into differential treatment responses totargeted psoriatic therapies. Many clinical trialsand real-world studies have previously shownthat special sites of skin (e.g., scalp, nails, joints,palms/soles) do not always respond in line withthe observed skin response [18, 19]. Variedindividual response could be related to treat-ment compliance, natural fluctuations in apatient’s immune response, and/or externaltriggers of disease, such as trauma, infections,scratching and stress. Animations can be used asvisual aids to help investigators identify inter-esting clinical observations and formulatehypotheses for further investigation.

The participation of healthcare providers inend-user testing of animated visualizations canconfirm their usability and provide informationon whether healthcare providers benefit fromdetailed data displays compared to conven-tional static presentations [20]. Carefullydesigned studies or surveys evaluating the users’preferences for these types of novel animationsare needed and could directly impact the uptakeof clinical information by healthcare providers.

The comparison of clinical trial data shouldbe done carefully due to differences in studydesign, length of treatment, severity of disease,and patient characteristics. Since IL-17 inhibi-tors have a faster onset than IL-23 or TNFblockade, the rate of treatment response mayappear to be more advantageous at week 12 astreatment difference may not yet be apparent;looking at multiple time points would provide amore comprehensive view of the data. Addi-tionally, while animated visualizations allow forimproved understanding of complex data, theyare primarily a descriptive instrument of dataand, therefore, provide only nominal insightsinto existing information. Animated visualiza-tions should be used in combination withresults that are provided in tables with respec-tive p-values to directly compare psoriasistreatments.

Ixekizumab has demonstrated high efficacyin the treatment of plaque psoriasis based onhead-to-head clinical trials with comparatortreatments. Data visualization via novel ani-mations is an engaging way to represent find-ings from clinical trials; this approach mayincrease efficiency and reduce time spent ondata mining. Animated visualizations can beused to identify meaningful aspects of clinicaltrial data, such as individual patient variabilityand trends during clinical trial testing. Furtherinvestigation of methods to represent clinicaltrial data is an important endeavor to improvethe dissemination and understanding of com-plex clinical information by healthcareproviders.

ACKNOWLEDGEMENTS

Medical Writing and Editorial Assis-tance. Writing assistance was provided by KatieCrosslin, PhD, of Evidera/PPD, and ManjuJanardhanan, MD, full-time employee of EliLilly and Company and was funded by Eli Lillyand Company. Animations support was pro-vided by Nikhil Das Nomula (Fig. 5), a formercontracted employee of Eli Lilly and Company.

1116 Dermatol Ther (Heidelb) (2021) 11:1107–1118

Funding. Sponsorship for this study andRapid Service Fee were funded by Eli Lilly andCompany (Indianapolis, IN, USA).

Authorship. All named authors meet theInternational Committee of Medical JournalEditors criteria for authorship for this article,take responsibility for the integrity of the workas a whole, and have given their approval forthis version to be published.

Disclosures. Craig Leonardi has been on anadvisory board for, and/or been a consultantfor, and/or received speaker’s bureau from, and/or is/was an investigator at the following com-panies: AbbVie, Actavis, Allergan, Amgen,Boehringer Ingelheim, Celgene, Coherus, Cell-ceutix, Corrona, Dermira, Eli Lilly and Com-pany, Galderma, Glenmark, Janssen, LeoPharma, Merck, Novartis, Novella, Ortho Der-matologics, Pfizer, Sandoz, Sienna, Stiefel, SunPharmaceuticals, UCB, Vitae, and Wyeth.Melinda Gooderham has been an investigator,speaker and/or advisor for: AbbVie, Amgen,Akros, Arcutis, Bausch Health, Bausch, BMS,Boehringer Ingelheim, Celgene, Dermira, Der-mavant, Eli Lilly and Company, Galderma, GSK,Incyte, Janssen, Kyowa Kirin, LEO Pharma,MedImmune, Merck, Novartis, Pfizer, Regen-eron, Roche, Sanofi Genzyme, Sun Pharma, andUCB. Jason E. Hawkes currently serves on themedical board of the National Psoriasis Foun-dation and has been a consultant for AbbVie, EliLilly and Company, Janssen, LearnSkin, LEO,Novartis, Pfizer, Regeneron-Sanofi, UCB, Visu-alDx, and UpToDate. Kyoungah See, RusselBurge, Stephanie Strakbein, Missy McKean-Matthews, and Daniel Saure are employees ofEli Lilly and Company and may bestockholders.

Compliance with Ethics Guidelines. Thisarticle is based on previously conducted studiesand does not contain any new studies withhuman participants or animals performed byany of the authors.

Data Availability. All data generated oranalyzed during this study are included in this

published article/as supplementary informationfiles.

Open Access. This article is licensed under aCreative Commons Attribution-NonCommer-cial 4.0 International License, which permitsany non-commercial use, sharing, adaptation,distribution and reproduction in any mediumor format, as long as you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons licence, andindicate if changes were made. The images orother third party material in this article areincluded in the article’s Creative Commonslicence, unless indicated otherwise in a creditline to the material. If material is not includedin the article’s Creative Commons licence andyour intended use is not permitted by statutoryregulation or exceeds the permitted use, youwill need to obtain permission directly from thecopyright holder. To view a copy of this licence,visit http://creativecommons.org/licenses/by-nc/4.0/.

REFERENCES

1. Kim WB, Jerome D, Yeung J. Diagnosis and man-agement of psoriasis. Can Fam Phys. 2017;63(4):278–85.

2. Pariser D, Schenkel B, Carter C, et al. A multicenter,non-interventional study to evaluate patient-re-ported experiences of living with psoriasis. J Der-matol Treat. 2016;27(1):19–26.

3. Strober B, Karki C, Mason M, et al. Characterizationof disease burden, comorbidities, and treatment usein a large, US-based cohort: results from the Cor-rona Psoriasis Registry. J Am Acad Dermatol.2018;78(2):323–32.

4. Kowalewska B, Cybulski M, Jankowiak B, Krajewska-Kułak E. Acceptance of illness, satisfaction with life,sense of stigmatization, and quality of life amongpeople with psoriasis: a cross-sectional study. Der-matol Ther. 2020;10(3):413–30.

5. Cather JC, Young M, Bergman MJ. Psoriasis andpsoriatic arthritis. J Clin Aesthet Dermatol.2017;10(3):S16-s25.

6. Menter A, Strober BE, Kaplan DH, et al. Joint AAD-NPF guidelines of care for the management and

Dermatol Ther (Heidelb) (2021) 11:1107–1118 1117

treatment of psoriasis with biologics. J Am AcadDermatol. 2019;80(4):1029–72.

7. Kemeny L, Berggren L, Dossenbach M, Dutronc Y,Paul C. Efficacy and safety of ixekizumab in patientswith plaque psoriasis across different degrees ofdisease severity: results from UNCOVER-2 andUNCOVER-3. J Dermatol Treat. 2019;30(1):19–26.

8. Warren RB, See K, Burge R, et al. Rapid response ofbiologic treatments of moderate-to-severe plaquepsoriasis: a comprehensive investigation usingbayesian and frequentist network meta-analyses.Dermatol Ther. 2020;10(1):73–86.

9. Paul C, Griffiths CEM, van de Kerkhof PCM, et al.Ixekizumab provides superior efficacy comparedwith ustekinumab over 52 weeks of treatment:results from IXORA-S, a phase 3 study. J Am AcadDermatol. 2019;80(1):70-9.e3.

10. Blauvelt A, Papp K, Gottlieb A, et al. A head-to-headcomparison of ixekizumab vs. guselkumab inpatients with moderate-to-severe plaque psoriasis:12-week efficacy, safety and speed of response froma randomized, double-blinded trial. Br J Dermatol.2020;182(6):1348–58.

11. van der Heijde D, Gladman DD, Kishimoto M, et al.Efficacy and safety of ixekizumab in patients withactive psoriatic arthritis: 52-week results from aphase III study (SPIRIT-P1). J Rheumatol.2018;45(3):367–77.

12. Gordon KB, Blauvelt A, Papp KA, et al. Phase 3 trialsof ixekizumab in moderate-to-severe plaque psori-asis. N Engl J Med. 2016;375(4):345–56.

13. Reich K, Pinter A, Lacour JP, et al. Comparison ofixekizumab with ustekinumab in moderate-to-

severe psoriasis: 24-week results from IXORA-S, aphase III study. Br J Dermatol. 2017;177(4):1014–23.

14. Mease PJ, Smolen JS, Behrens F, et al. A head-to-head comparison of the efficacy and safety ofixekizumab and adalimumab in biological-naıvepatients with active psoriatic arthritis: 24-weekresults of a randomised, open-label, blinded-asses-sor trial. Ann Rheum Dis. 2020;79(1):123–31.

15. Auguie B. Package ‘gridExtra.’ https://cran.r-project.org/web/packages/gridExtra/gridExtra.pdf. Acces-sed 10 May 2021.

16. Wickham H. The Split-Apply-Combine Strategy forData Analysis. 2011. https://www.jstatsoft.org/article/view/v040i01. Accessed 10 May 2021.

17. Nakakoji K TA, Yamamoto Y. Cognitive effects ofanimated visualization in exploratory visual dataanalysis. In: 5th International Conference onInformation Visualisation, 25–26 July 2001; Lon-don. https://doi.org/10.1109/IV.2001.942042

18. Crowley JJ. Challenges in psoriasis treatment: nail,scalp, and palmoplantar involvement. In: WeinbergJM, Lebwohl M, editors. Advances in psoriasis: amultisystemic guide. Cham: Springer InternationalPublishing; 2021. p. 343–62.

19. Egeberg A, See K, Garrelts A, Burge R. Epidemiology ofpsoriasis in hard-to-treat body locations: data fromthe Danish skin cohort. BMC Dermatol. 2020;20(1):3.

20. Pohl M, Endl H, Fels U. Animated scatterplot—analysis of time-oriented data of diabetes patients.Stud Health Technol Inform. 2016;223:191–8.

1118 Dermatol Ther (Heidelb) (2021) 11:1107–1118