Key Summary Points
Why carry out this study?Moderate-to-severe plaque psoriasis is a chronic, immune-mediated disease associated with substantial clinical, psychosocial, and economic burden; although multiple biologic treatments are available, long-term real-world evidence of their effectiveness, quality of life impact, and safety remains limited.Tildrakizumab, an interleukin-23 p19 inhibitor approved for the treatment of adults with moderate-to-severe plaque psoriasis, has demonstrated efficacy and safety in randomized controlled trials, but the extent to which the results are generalizable to routine clinical practice is unclear.This systematic review and meta-analysis aimed to characterize and synthesize the available real-world evidence on the clinical effectiveness and safety of tildrakizumab in patients with moderate-to-severe plaque psoriasis, alone and relative to guselkumab and risankizumab.What was learned from the study?This systematic literature review and meta-analysis of real-world data provides evidence that patients with moderate-to-severe plaque psoriasis using tildrakizumab had improvements in disease severity and quality of life as early as 3 months and up to 1 year after starting treatment, with safety findings similar to those for guselkumab and risankizumab.These results are consistent with randomized clinical trial data, confirming tildrakizumab as an effective treatment for moderate-to-severe plaque psoriasis in real-world practice.
Introduction
Psoriasis (PsO) is a chronic inflammatory skin condition affecting approximately 2% of US adults [1]; plaque PsO, the most common form, encompasses 80–90% of cases [2]. PsO is characterized by red, scaly plaques on the skin, commonly visible on the elbows, knees, and scalp, and may become generalized on the body over time [3]. Patients with moderate-to-severe PsO benefit greatly from biologic therapies, which provide improved symptom control vs topical treatments or phototherapy [4].
Biologic classes used to treat moderate-to-severe plaque PsO include inhibitors of tumor necrosis factor alpha, interleukin (IL)-17, IL-12/23, and IL-23 p19 [4]. Tildrakizumab is an IL-23 antagonist indicated for the treatment of moderate-to-severe plaque PsO in adults. It is a humanized IgG1κ monoclonal antibody that selectively binds to the IL-23 p19 subunit, blocking IL-23 signaling and downstream activation of the T helper 17 cell pathway [5].
The clinical efficacy and safety of tildrakizumab for the treatment of moderate-to-severe plaque psoriasis were demonstrated in two phase 3 randomized clinical trials, reSURFACE 1 and reSURFACE 2. In both studies, significantly higher proportions of patients treated with tildrakizumab 100 mg or 200 mg vs placebo achieved ≥ 75% improvement from baseline in Psoriasis Area and Severity Index (PASI 75) and Physician Global Assessment (PGA) score of 0 (clear) or 1 (minimal) by week 12, with responses sustained through week 28. Efficacy and safety results were similar between the two doses except for a potential efficacy advantage of tildrakizumab 200 mg in patients weighing ≥ 120 kg [6, 7].
After week 28, patients in both trials achieving at least a partial response (≥ 50% improvement from baseline in PASI) were rerandomized to continue the same treatment, a different dose of tildrakizumab, or placebo, while nonresponders discontinued the study [6].
On the basis of these results, tildrakizumab was approved in the USA and EU at a recommended dose of 100 mg; in the EU, tildrakizumab 200 mg may also be used at the prescriber’s discretion for patients with high disease burden or body weight > 90 kg [5, 8]. The efficacy and safety of tildrakizumab were subsequently confirmed through further analyses of randomized trials [9–12].
A key question for clinicians is to what extent are the efficacy and safety data on tildrakizumab obtained within the tightly controlled conditions of randomized controlled trials (RCTs) reflected in routine clinical practice? Rerandomization in the clinical trials provided important evidence of the durability of tildrakizumab responses but limited outcomes of continuous tildrakizumab treatment at the same dose for more than 28 weeks to selected subgroups of patients, which may not fully reflect results in the more heterogeneous patient populations encountered in routine clinical practice. Real-world data after week 28 could simplify the translation of pivotal trial data into appropriate expectations for patients and providers.
To date, the efficacy and safety of tildrakizumab in real-world clinical practice have only been evaluated in isolated small studies of varying methodology and scientific rigor, without systematic analysis. We postulated that a meta-analysis of real-world evidence (RWE) would provide a coherent view of the impact of tildrakizumab on patients with moderate-to-severe PsO being treated in the real world, including time points and measures unavailable in the randomized clinical trials. This systematic review and meta-analysis characterized and synthesized currently available RWE regarding the clinical effectiveness and safety of tildrakizumab.
Methods
The review was performed in consideration of methods guidance from the Cochrane Handbook of Systematic Reviews of Interventions (version 6.4; updated August 2023) [13]. Findings are reported on the basis of the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement recommendations [14]. A completed PRISMA checklist is provided in Supplementary Table 1.
A protocol was developed a priori to guide conduct of the review and is available upon request from the study authors.
Ethical Approval
This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.
Study Selection Criteria
Study eligibility criteria were defined using the PICOS (Population, Intervention, Comparator, Outcomes, Study design) framework as follows:
Population
Studies involving adults (aged > 18 years) with chronic moderate-to-severe plaque PsO were included. Subgroups of interest (data permitting) included geographic region (US vs ex-US), prior biologic exposure (bio-naïve vs bio-experienced), patients with comorbidities (e.g., cardiovascular disease, obesity, diabetes mellitus, metabolic disorders, depression, Crohn disease, ulcerative colitis, malignancy, myocardial infarction), and tildrakizumab dose (excluding studies that permitted use of 200 mg in select patients or did not report dose).
Intervention and Comparators
The focal treatment of interest was tildrakizumab-asmn. Both uncontrolled studies and studies involving comparisons with any of the following treatments were of interest: certolizumab pegol, secukinumab, etanercept, adalimumab, infliximab, brodalumab, risankizumab, ustekinumab, ixekizumab, guselkumab, standard care, and no treatment.
Outcomes
Studies that provided data related to specific commonly used measures of clinical effectiveness, quality of life (QoL), and safety were of interest: PASI including absolute PASI score (range 0–72, where values 0–4 indicate no-to-mild PsO, 5–10 is moderate PsO, and greater than 10 is severe PsO); PASI 75/90/100 response (i.e., the proportion of patients who achieved ≥ 75%/≥ 90%/100% improvement in absolute PASI scores from baseline); percentage of body surface area (%BSA) impacted by PsO (range 0–100%, where values < 3% represent mild disease, 3–10% is moderate, and > 10% is severe); Physician Global Assessment (PGA; range 0–4 or 0–5, where a value of 0 indicates no signs of PsO, 1 is almost clear, 2 is mild, 3 is moderate, 4 is severe, and 5 is very severe); Dermatology Life Quality Index (DLQI) total score (range 0–30, where values 0–1 indicate no effect on the patient’s life, 2–5 is a small effect, 6–10 is a moderate effect, 11–20 is a very large effect, and 21–30 is an extremely large effect); any AEs; serious AEs (SAEs); treatment-related AEs (TRAEs); and withdrawals due to AEs (WDAEs).
Study Design
Observational studies including retrospective cohort and case–control studies, as well as prospective cohort studies, single-arm studies, phase 4 trials, and pragmatic trials, were sought. Clinical trials (phase 1–3) and real-world studies and publication types providing a lower level of evidence and/or limited statistical power (books, letters, commentaries, case reports, case series with < 5 patients, narrative reviews, cross-sectional studies, and cost studies), and duplicative reports of the same study were excluded.
Study Identification Process
Systematic literature searches were designed and carried out using the OvidSP platform to search MEDLINE® and Embase; the search was most recently updated in November 2023. Details of the literature searches are provided in Supplementary Tables 2–3. Gray literature searches were conducted to locate unpublished and difficult-to-find literature.
This included searching the abstracts from key rheumatology and dermatology conferences from 2021 to 2023. Bibliographies of relevant review articles were also hand searched. Only studies published in English were sought for this review.
Study Selection Process
A senior reviewer (OL) reviewed abstracts, conference proceedings, and gray literature sources according to the pre-defined selection criteria. All citations deemed potentially relevant during stage 1 screening were acquired in full text. These texts were screened by two independent reviewers at stage 2 screening to establish the final set of included studies.
A third reviewer (KH) was consulted to reach consensus on any unresolved discrepancies. The screening process was summarized in a flow diagram as recommended by the PRISMA statement [14].
Data Collection and Risk of Bias Appraisal
Extraction of data from each included study was performed using Microsoft Excel (Microsoft Corporation, Seattle, WA, USA) and involved two reviewers, with a third reviewer consulted to resolve any discrepancies. Detailed data were gathered from all included studies including information related to study characteristics (authors and title, publication year, design, countries and setting [single center vs multicenter] of performance, population data source [e.g., name of database or hospitals from which patients were enrolled], sample size, dates defining the study period, length of follow-up, and patient eligibility criteria), patient characteristics (age, sex, race/ethnicity, comorbidities present), intervention characteristics (interventions and comparators studied, administration information [dose, frequency, route], concomitant/background therapies, duration of treatment, setting for administration [e.g., home, office]), and the aforementioned outcomes. For all outcomes, the corresponding definitions and timing of measurement were captured, and results among subgroups of interest (i.e., age group, history of exposure to biologics, geographic region [US vs ex-US data], patients with comorbidities, and tildrakizumab dose) were gathered.
The risk of bias of each study was appraised using the Newcastle–Ottawa Scale for non-randomized studies and the Joanna Briggs Institute Scale for case series [15, 16].
Approach to Evidence Synthesis
Prior to performance of meta-analyses, a feasibility assessment was conducted using methods adapted from Cope et al. to assess the potential for meta-analyses. Briefly, we evaluated whether sufficient data were available for each outcome and time point to conduct an analysis. Studies were then compared with respect to study characteristics, intervention characteristics, patient characteristics, and distributions of treatment effects to determine whether quantitative synthesis was appropriate [17].
Data on tildrakizumab from RWE studies (either single group or comparative) reporting on the a priori outcomes of interest were meta-analyzed to establish estimates of effectiveness and safety and observe trends in measures over time. Data from studies of tildrakizumab involving a comparison group were combined separately using meta-analysis to assess tildrakizumab’s comparative effectiveness and safety. Meta-analyses were performed for outcomes at grouped time points of interest reflecting short-term (12–16 weeks), mid-term (24–28 weeks), and longer-term (36–52 weeks) outcomes when data were available from two or more studies.
Random effects meta-analyses were performed using the DerSimonian and Laird inverse variance approach when the I2 measure of variance was more than 0% [18]. Continuous outcomes were meta-analyzed in terms of mean values, and dichotomous outcomes were meta-analyzed for the risk estimate or the percentage of patients experiencing the specific outcome. To account for 0 events for dichotomous outcomes, Freeman–Tukey’s double arcsine transformation was applied [19].
For a head-to-head data comparison of tildrakizumab with other available treatments, we used the National Institute for Health and Care Excellence (NICE)-endorsed method in the Cochrane Handbook to conduct pairwise meta-analysis at each time point [18]. Meta-analyses were performed where two or more studies reported the same outcome per comparator. Between-study heterogeneity was assessed using the I2 statistic, which quantifies the proportion of variability in effect estimates due to heterogeneity rather than chance.
I2 values were interpreted as low (≤ 40%), moderate (30–60%), substantial (50–90%), and considerable (≥ 75%) heterogeneity, per Cochrane guidance. When I2 > 0, a random effects meta-analysis model using the DerSimonian and Laird inverse variance method was used to account for between-study variability. When I2 = 0, a fixed-effects model was applied [18].
We estimated the mean difference between tildrakizumab vs other comparators for each continuous outcome and the risk ratio for tildrakizumab vs other comparators for dichotomous outcomes. To increase the robustness within the meta-analysis, we grouped closely spaced time points (i.e., weeks 12–16, 24–28, 36–52).
Findings from all meta-analyses are reported as point estimates with corresponding 95% confidence intervals (CIs). Subgroup analyses were conducted where possible for data stratified by age group, geographic region (US vs ex-US data), history of exposure to biologics, patient comorbidities, and tildrakizumab dose.
Results
Study Selection Process
Of 9865 records identified, 43 unique studies from 62 publications were included in the review (Fig. 1); 37 of these unique studies (described in 45 publications) were eligible for meta-analyses. All included studies are listed in Supplementary Table 4 and all excluded studies in Supplementary Table 5.
Six unique studies from 17 publications were excluded from meta-analyses because they did not report outcomes reported by at least one other study [20–23] or did not report the number of patients per intervention [24, 25].Fig. 1Process of study selection
Study Characteristics
Characteristics of the 43 included studies are presented in Table 1. Studies were published between 2021 and 2023, and sample size ranged from 8 to 4178 participants (median 85). Designs included 34 retrospective cohorts, 5 prospective cohorts, 3 phase 4 observational studies, and 1 case series.
Most were single-arm studies that measured outcomes for tildrakizumab only; 6 studies compared tildrakizumab directly with guselkumab and 6 compared with risankizumab. Studies were generally conducted in Europe (Italy [n = 17], Spain [n = 8], Germany [n = 2], and single studies in Austria, Denmark, Poland, Romania, Switzerland, and the UK) and the USA (n = 3). Four multinational studies were also identified; these largely involved European countries, although 1 involved 240 centers across 23 countries (including Europe, North and South America, Australia, the Middle East, and Eastern Asia).
The median length of follow-up was 12 months (range 3–48 months), and outcomes were most often reported at weeks 24–28 and 36–52. Data sources encompassed 31 studies from hospitals or clinics and 4 from databases or registries; 8 studies did not report a data source. Fourteen studies were published only as abstracts, limiting the available information regarding study methods and findings.Table 1Characteristics of included studies (n = 43)ReferencesNStudy designCountryStudy datesStudy lengthReported outcomes of interestIncluded in meta-analysisTildrakizumab dose Alarcon 2022 [41]91Retrospective cohort; multicenterSpainNR52 weeksDLQI, any SAEsYesNR Bagel 2023 [42]25Prospective cohort; single centerUSANR24 weeksPASI 100, PASI 90, PASI 75, DLQI, %BSA, PGA, any AEs, any treatment-related AEs, discontinuation due to AEs, any SAEsYes100 mg Baratta 2023 [43]26Retrospective cohort; single centerItalyNR156 weeksAbsolute PASI, DLQIYesNR Becher 2022, 2023 [26, 44]122Retrospective cohort; multicenterUKApr 2019–Jan 202212 monthsAbsolute PASI, DLQI, any treatment-related AEs, discontinuation due to AEsYes100 mg or 200 mga Berenguer-Ruiz 2023 [27]190Retrospective cohort; multicenterSpainNR24 weeksAbsolute PASI, PASI 90, DLQI, %BSA, any AEs, discontinuation due to AEsYes100 mg or 200 mga,b Bhatia 2022, 2023 (NCT03718299) [45–51]55Phase 4 observational; multicenterUSAJul 2019–Nov 202164 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, DLQI, %BSA, PGA, discontinuation due to AEs, any SAEsYes100 mg Bris 2023 [22]81Retrospective cohort; multicenterSpainNR12 monthsAbsolute PASINoNR Brunasso 2022 [52]8Retrospective cohort; NRItalyNR20 weeksAbsolute PASIYesNR Burlando 2021 [53]26Retrospective cohort; single centerItalyMar 2020–Sep 202024 weeksAbsolute PASI, PASI 100, PASI 90, DLQI, PGA, any AEsYes100 mg Burlando 2022 [54]89Retrospective cohort; multicenterItaly, SwitzerlandMar 2020–Nov 202112 monthsAbsolute PASI, PASI 100, PASI 90, PASI 75Yes100 mg Burzi 2023 [25]1053Retrospective cohort; NRItalyNR52 weeksPGANoNR Caldarola 2022 [55]59Retrospective cohort; multicenterItalyFeb 2020–Mar 202128 weeksAbsolute PASI, PASI 90, PASI 75, any AEs, any SAEsYes100 mg Campione 2023 [28, 56]106Retrospective cohort; NRItalyNR52 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, DLQI, any AEs, discontinuation due to AEsYes100 mg or 200 mga Costanzo 2022, 2023; TRIBUTE [57–61]177Phase 4 observational; multicenterSpain, ItalyNR24 weeksAbsolute PASI, DLQI, PGAYes100 mg Di Brizzi 2023 [62]51Retrospective cohort; multicenterItalyNR48 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, DLQI, %BSA, any AEs, discontinuation due to AEsYes100 mg Downs 2023 [23]1078Prospective cohort; multicenterUKMay 2019–Mar 202352 weeksDiscontinuation due to any reasonNo100 mg or 200 mga Drerup 2022, 2023 [63, 64]150Prospective cohort; single centerGermanyDec 2018–Aug 20, 202176 weeksAbsolute PASI, DLQI, %BSAYes100 mg Elgaard 2023 [31]80Retrospective cohort; single centerDenmarkJun 11, 2018–Jul 14, 2021Mean 61.4 (SD 43.7) weeksAbsolute PASI, DLQIYesNR Escalas 2023 [65]51Retrospective cohort; single centerSpainNR12 weeksAbsolute PASIYes100 mg Fornons-Servent 2023 [66]104Retrospective cohort; multicenterSpainNov 2019–Jan 202324 monthsAbsolute PASI, DLQI, %BSAYesNR Galluzzo 2022 [67]18Retrospective cohort; single centerItalyUp to Apr 202128 weeksAbsolute PASI, any AEs, discontinuation due to AEs, any SAEsYes100 mg Gambardella 2023 [68]30Retrospective cohort; NRItalyNR36 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, DLQI, PGAYes100 mg Grabowski 2023 [32]97Retrospective cohort; single centerPolandNR30 monthsAbsolute PASI, PASI 90, PASI 75, DLQI, %BSA, discontinuation due to AEs, any SAEsYesNR Graier 2022 [33]197Retrospective cohort; multicenterAustriaMar 2018–Feb 202112 monthsAbsolute PASI, PASI 100, PASI 90, PASI 75, any AEsYes; head-to-head comparisonNR Kahen 2022 [24]NRRetrospective cohort; multicenterUSANRNRCerebrovascular accidents safety signalNoNR Mastorino 2022 [69]129Retrospective cohort; single centerItalyNR52 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, DLQIYesNRMaul 2021 [70]23Retrospective cohort; multicenterSwitzerlandNR2 yearsAbsolute PASI, DLQIYesNR Megna 2022 [36]150Retrospective cohort; single centerItalyNov 2019–Nov 202128 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, %BSA, any AEs, discontinuation due to AEs, any SAEsYes; head-to-head comparison100 mg Merchán 2023 [34]58Retrospective cohort; single centerSpainAug 2020–Aug 202216 weeksAbsolute PASIYes; head-to-head comparisonNR Narcisi 2022 [20]127Retrospective cohort; single centerItalyJun 2019–Jun 2021144 weeksPGA, discontinuation due to AEs, any SAEsNoNR Narcisi 2023 [71]237Retrospective cohort; multicenterItalyJan 2020–Jul 202152 weeksAbsolute PASI, PASI 100, PASI 90, PASI 75, any AEs, discontinuation due to AEs, any SAEsYes100 mg Pinter 2022 and Khattri 2023; PSoHO [29, 30]1981Prospective cohort; multicenterArgentina, Australia, Austria, Brazil, Canada, Colombia, France, Germany, Hungary, Israel, Italy, Mexico, Netherlands, Poland, Portugal, Romania, Saudi Arabia, South Korea, Spain, Switzerland, Taiwan, UAE, UKNR (ongoing 3-year study)12 weeksPASI 75, PASI 90, PASI 100Yes; head-to-head comparison100 mg or 200 mga Ricceri 2022 [35]113Retrospective cohort; multicenterItalyJan 2021–Dec 202152 weeksAny AEs, discontinuation due to AEs, any SAEsYes; head-to-head comparisonNR Rodica 2023 [40]11Case series; single centerRomaniaNR3 monthsPASI 75YesNR Rodriguez Fernandez-Freire 2023 [72]32Retrospective cohort; NRSpainNR52 weeksAbsolute PASI, DLQI, %BSA, PGAYesNR Ruggiero 2022a [73]34Retrospective cohort; single centerItalyNov 2019–Apr 202228 weeksAbsolute PASI, PASI 100, PASI 90, %BSA, discontinuation due to AEs, any SAEsYes100 mg Ruggiero 2022b [37]34Retrospective cohort; single centerItalyOct 2018–Mar 2021Mean 43.7 (SD 8.2) weeksAbsolute PASI, PASI 100, PASI 90, %BSA, discontinuation due to AEs, any SAEsYes; head-to-head comparisonNR Ruggiero 2023 [74]42Retrospective cohort; single centerItalyNov 2019–Nov 202252 weeksAbsolute PASI, PASI 100, PASI 90, DLQI, %BSA, discontinuation due to AEs, any SAEsYes100 mg Ruiz-Villaverde 2022 [75]61Retrospective cohort; multicenterSpainFeb 2019–Feb 202252 weeksAbsolute PASI, DLQI, %BSA, PGA, any AEs, discontinuation due to AEs, any SAEsYes100 mg Sommer 2023; POSITIVE [76–78]263Phase 4 observational; multicenterAustria, Belgium, France, Germany, Italy, Spain, Switzerland, Netherlands, UKNR16, 28 weeksAny SAEs, discontinuation due to AEs, any treatment-related AEs, any AEs, PASIYesNR Torres 2022 [21]4178Retrospective cohort; multicenterCanada, Czech Republic, Italy, Greece, Portugal, Spain, SwitzerlandFeb 2015–Oct 202148 monthsAny AEs, discontinuation due to AEs, discontinuation due to any reasonNoNR Tsianakas 2022, 2023; TILOT [79–82]412Prospective cohort; multicenterGermanyNR52 weeksPGA, absolute PASI, DLQI, any treatment-related AEs, PASI 90, PASI 75, %BSAYes100 mg Wei 2022 [83]30Retrospective cohort; single centerUSANov 2018–Nov 202012 monthsAbsolute PASI, PASI 100, PASI 90, PASI 75, any SAEsYes100 mgAE adverse event, %BSA percentage of body surface area, DLQI Dermatology Life Quality Index, NR not reported, PASI Psoriasis Area and Severity Index, PASI 75/90/100 the proportion of patients who achieved ≥ 75%/≥ 90%/100% improvement in absolute PASI scores from baseline, PGA Physician’s Global Assessment, SAE serious adverse event, UAE United Arab Emiratesa200 mg of tildrakizumab was permitted in select patients (e.g., those with obesity)bOne patient received tildrakizumab 200 mg from week 4 due to insufficient response
Intervention Characteristics
Tildrakizumab
Overall, 23 studies reported use of tildrakizumab 100 mg consistent with the recommended dose in the USA (Table 1). Of these, five mentioned potential use of tildrakizumab 200 mg [23, 26–30] consistent with the European Medicines Agency drug label. Twenty studies provided no dosage information.
Among studies reporting on administration setting, all but one mentioned administration within the clinic setting [23].
Comparator Therapies
Overall, 12 included studies reported on the use of other biologics, including guselkumab (n = 11) [20, 21, 24, 30–37], risankizumab (n = 11) [20, 21, 24, 30–37], brodalumab (n = 4) [20, 21, 25, 30], ixekizumab (n = 4) [20, 21, 25, 30], secukinumab (n = 4) [20, 21, 25, 30], adalimumab (n = 1) [30], and ustekinumab (n = 1) [30]. Six comparative studies included in meta-analyses involved either guselkumab or risankizumab as a comparator. Among these, two studies reported treatment dosage and frequency, which were consistent with the approved labeling for plaque PsO for both guselkumab and risankizumab [30, 36, 38, 39].
All six studies administered treatment in a clinical setting. None of the studies included patients undergoing background therapy.
Patient Characteristics
Studies were generally considered homogeneous in terms of age (median 49.9; range 41.9–70.5 years) and sex (median 61.2%; range 36.4–76.5% male); one study (Ruggiero et al., 2022b [37]) was considered an outlier on the basis of its objective to assess the safety and effectiveness of tildrakizumab, guselkumab, and risankizumab in elderly patients (mean age 70.5 years; 32.4% with diabetes, 70.6% with hypertension, 52.9% with psoriatic arthritis). Among studies providing information regarding comorbidities, prevalent conditions were diabetes mellitus (median 15.7%; range 5.49–32.35%; n = 23 studies), hypertension (median 33.5%; range 17.95–70.59%; n = 21), and psoriatic arthritis (median 19.2%; range 6.73–52.94%; n = 18). Only one study (Torres et al., 2022) reported the use of previous therapies [21].
Study Quality Assessment
Details of judgments from quality assessment of included studies are provided in Supplementary Tables 6–7. Cohort studies were judged to be of moderate-to-high quality, with most scoring at least 5 out of a possible 9 points (n = 37), reflecting the use of representative samples (e.g., multicenter data sources with sufficient sample sizes) and use of secure records to gather outcomes data rather than self-reporting. The case series by Rodica et al., published as a conference abstract, was judged to be of low quality due to limited methodological details [40].
However, it was retained for meta-analyses, as the available data were not considered so biased as to warrant exclusion.
Findings from Meta-Analyses of Tildrakizumab Data
Clinical Effectiveness (PASI, %BSA, PGA Score, and QoL [DLQI])
Findings from meta-analysis of studies reporting data for tildrakizumab (single group or comparative) related to measures of clinical effectiveness and QoL demonstrated growing and prolonged estimates of benefit with tildrakizumab over time. Meta-analyses of absolute PASI scores for tildrakizumab demonstrated a reduction in mean score from 12.81 at baseline (95% CI 11.69, 13.92) to 1.62 (95% CI 1.03, 2.20) after 36–52 weeks (Supplementary Fig. 1A). %BSA improved from 16.21% at baseline (95% CI 13.72%, 18.70%) to 3.27% (95% CI 1.26%, 5.28%) at 36–52 weeks (Supplementary Fig.
1B). Clinically important reductions were also identified at 36–52 weeks for PGA score (Supplementary Fig. 1C) and DLQI score (Supplementary Fig.
1D). Overall, tildrakizumab-treated patients experienced between 78% and 87% improvement across all four clinical outcomes at 36–52 weeks (87% improvement in mean absolute PASI scores, 80% improvement in %BSA, 78% improvement in PGA scores, and 87% improvement in DLQI). Forest plots of tildrakizumab-only data are provided in Supplementary Figs.
2–17.
Subgroup analyses of tildrakizumab data are presented in Supplementary Figs. 18–21. Where data were available from the included studies, analyses related to geographic setting, history of biologic exposure, presence of comorbidities, and tildrakizumab dose yielded similar results.
The percentages of tildrakizumab-treated patients reaching the measures of PASI 75, PASI 90, and PASI 100 responses are provided in Supplementary Fig. 22. Overall, high proportions of patients tended to reach these landmarks by weeks 24–28 and maintained at least the same level of response at weeks 36–52.
Findings from Comparative Studies
Six studies reported findings from head-to-head comparisons of tildrakizumab against other biologics. Most studies included in comparative meta-analyses were conducted in Europe (Table 2). As a result of data availability, only comparisons of tildrakizumab and guselkumab or risankizumab could be conducted.Table 2Studies included for comparative meta-analysis of the overall populationReferences; (sample size)Study design; country/regionComparatorsStudy periodNumber of patients per treatment armKey patient characteristics Graier 2022 [33]; (N = 197)Retrospective; AustriaGuselkumab, risankizumabMar 2018–Feb 2021Tildrakizumab, n = 15; guselkumab, n = 127; risankizumab, n = 55NA Megna 2022 [36]; (N = 150)Retrospective; ItalyGuselkumab, risankizumabNov 2019–Nov 2021Tildrakizumab, n = 21; guselkumab, n = 63; risankizumab, n = 66NAMerchan 2023 [34]; (N = 58)Retrospective; SpainGuselkumab, risankizumabAug 2020–Aug 2022Tildrakizumab, n = 11; guselkumab, n = 20; risankizumab, n = 27NA Pinter 2022 [30]; (N = 1981)Prospective; Argentina, Australia, Austria, Brazil, Canada, Colombia, France, Germany, Hungary, Israel, Italy, Mexico, Netherlands, Poland, Portugal, Romania, Saudi Arabia, South Korea, Spain, Switzerland, Taiwan, UAE, UKGuselkumab, risankizumabOngoing 3-year studyTildrakizumab, n = 95; guselkumab, n = 303 risankizumab, n = 259NA Ricceri 2022 [35]; (N = 113)Retrospective; ItalyGuselkumab, risankizumabJan 2021–Dec 2021Tildrakizumab, n = 24; guselkumab, n = 56; risankizumab, n = 33Patients with overweight (BMI 25–30 kg/m2) or obesity (BMI ≥ 30 kg/m2) Ruggiero 2022b [37]; (N = 34)Retrospective; ItalyGuselkumab, risankizumabOct 2018–Mar 2021Tildrakizumab, n = 6; guselkumab, n = 20; risankizumab, n = 8Elderly patients (≥ 65 years)BMI body mass index, NA not applicable, UAE United Arab Emirates
Clinical Effectiveness (PASI and %BSA)
Based on available data, meta-analyses of comparative clinical effectiveness data were performed for differences in absolute PASI score and %BSA observed over time in comparisons of tildrakizumab with guselkumab and risankizumab.
Differences in Absolute PASI Score
Four studies reported absolute PASI score for tildrakizumab vs guselkumab and risankizumab at a minimum of one time point among baseline, weeks 12–16, and weeks 24–28. Neither the pooled baseline difference between tildrakizumab and guselkumab nor study-level baseline differences showed important differences between groups; there were also no statistically significant differences between tildrakizumab and guselkumab at weeks 12–16 or weeks 24–28 (Table 3). Comparisons between tildrakizumab and risankizumab displayed similar patterns and findings between baseline and weeks 24–28.Table 3Meta-analyses of tildrakizumab comparative effectiveness dataTime of measurementTildrakizumab vs guselkumab, mean difference and 95% CITildrakizumab vs risankizumab, mean difference and 95% CIPASI score%BSAPASI score%BSABaseline− 1.21(− 4.52, 2.10)n = 4, N = 283I2 = 30.63%a− 10.51(− 18.44, − 2.59)n = 2, N = 110− 1.13(− 3.87, 1.61)n = 4, N = 209I2 = 9.49%a− 6.19(− 14.23, 1.85)n = 2, N = 101Weeks 12–161.22(− 1.09, 3.53)n = 3, N = 257–0.85(− 1.71, 3.41)n = 3, N = 195–Weeks 24–281.64(− 0.87, 4.14)n = 3, N = 252− 1.48(− 4.97, 2.02)n = 2, N = 1102.05(− 0.49, 4.59)n = 3, N = 171− 1.99(− 6.00, 2.01)n = 2, N = 101Mean differences from meta-analysis along with 95% CIs are presented for comparative effectiveness outcomes associated with tildrakizumab compared to guselkumab and risankizumab.
For both PASI score and %BSA, values < 0 favor tildrakizumab. Absolute PASI scores range from 0 to 72, where values 0 to 5 indicate no-to-mild psoriasis, 6 to 10 is moderate, and 11 or higher is severe. %BSA values range from 0 to 100%, where values < 3% represent mild disease, 3% to 10% is moderate, and > 10% is severeBSA body surface area, CI confidence interval, PASI Psoriasis Area and Severity IndexaDenotes that a random effects model was used. In all other cases, a fixed-effect model was used for meta-analysis, and I2 values were not applicable
Differences in %BSA
Two studies reported on %BSA for tildrakizumab, guselkumab, and risankizumab. In comparisons with guselkumab, the pooled difference from two studies at baseline favored tildrakizumab (Table 3) as a result of a significant difference in one of the two underlying studies (Megna et al. (2022) [36]; mean difference, − 11.80; 95% CI − 20.94, − 2.66).
At weeks 24–28, no significant difference between tildrakizumab and guselkumab was present. Comparisons with risankizumab followed a similar pattern. The pooled difference between tildrakizumab and risankizumab did not reach statistical significance, and no significant difference between treatments was observed at weeks 24–28.
Safety of Tildrakizumab
Findings from meta-analyses of studies reporting tildrakizumab safety and meta-analyses of comparative studies involving guselkumab and risankizumab are presented in Table 4. The numbers of studies reporting tildrakizumab data varied across outcomes. At both weeks 24–28 and weeks 36–52, the percentages of patients experiencing AEs, SAEs, TRAEs, and WDAEs were small; the highest estimate was observed for TRAEs at 36–52 weeks.
Forest plots of tildrakizumab-only data are provided in Supplementary Figs. 23–30.Table 4Tildrakizumab safety: findings from meta-analysisTildrakizumab-only analyses: pooled percentage (95% CI) of patients with adverse eventsFollow-up time% with any AE (n = 13, N = 1043)% with SAE (n = 15, N = 941)% with TRAE (n = 4, N = 843)% with WDAE (n = 16, N = 1227)Weeks 24–284.79 (0.31, 12.56); n = 8, N = 626, I2 = 87.84%a0 (0.00, 0.12); n = 9, N = 4753.30 (0.00, 10.95); n = 2, N = 288, I2 = 56.02%a0 (0.00, 0.12); n = 9, N = 606Weeks 36–521.43 (0.00, 4.39); n = 6, N = 447, I2 = 51.66%a0.18 (0.00, 1.83); n = 7, N = 496, I2 = 49.82%a4.85 (0.00, 29.73); n = 2, N = 555, I2 = 97.83%a0.24 (0.00, 1.35); n = 8, N = 651, I2 = 35.10%aComparative studies: relative risk (95% CI) of adverse eventsFollow-up time and comparatorAny AESAETRAEWDAEWeeks 24–28 (vs guselkumab)0.80 (0.31, 2.07); n = 2, N = 1642.58 (0.16, 40.34); n = 2, N = 164–0.77 (0.07, 7.95); n = 2, N = 164Weeks 24–28 (vs risankizumab)0.92 (0.35, 2.42); n = 2, N = 1442.03 (0.13, 31.80); n = 2, N = 144–2.03 (0.13, 31.80); n = 2, N = 144Findings from meta-analyses of RWE studies of tildrakizumab are presented. For relative risk measures from comparative RWE studies, values < 1 suggest fewer events with tildrakizumabAE adverse event, CI confidence interval, RWE real-world evidence, SAE serious adverse event, TRAE treatment-related adverse event, WDAE withdrawal due to adverse eventaDenotes a random effects model was used.
In all other cases, a fixed effect model was used for meta-analysis, and I2 values were not applicable
Meta-analyses of comparative RWE studies involving guselkumab and risankizumab as comparators each involved two studies, and no data for TRAEs were reported. For AEs, SAEs, and WDAEs, no statistically significant differences between tildrakizumab and either comparator were observed from meta-analyses.
Discussion
In this real-world meta-analysis, we confirm the findings of the primary clinical studies and network meta-analyses thereof [9–12] and extend the results beyond 28 weeks and PASI 75 without the potential confounding of rerandomization. In the present analysis, tildrakizumab treatment resulted in sustained improvement in patients’ symptoms over time. Meta-analyses across follow-up times of up to 1 year showed that patients could improve from severe disease at baseline (i.e., PASI score > 12) to mild disease at 36–52 weeks (i.e., PASI score < 2), with improvement to mild disease (i.e., PASI score < 5) as early as 12 weeks following treatment initiation.
Clinically important improvements in %BSA (5%, moderate disease), PGA (score of 1, almost clear), and DLQI (score of 3, small effect on patient’s life) were also observed within the first 3 months of treatment and sustained or even further improved for up to 1 year.
Comparative RWE analyses demonstrated that tildrakizumab provided similar benefits relative to other IL-23 inhibitors in clinical practice, such as guselkumab and risankizumab, with no significant differences for PASI or %BSA at any posttreatment time point. In analyses of safety outcomes (AEs, SAEs, and WDAEs), tildrakizumab was associated with few tolerability issues, with no increased risk identified among comparative RWE studies. Comparative findings are consistent with a published network meta-analysis of RCTs that showed no significant differences between tildrakizumab and both guselkumab and risankizumab for PASI 75, PASI 100, PGA, AEs, SAEs, and WDAEs at weeks 12–16 [10].
Our analysis contributes additional data for up to 1 year, a time point that is rarely the focus of randomized trials. Nevertheless, the findings from the present analysis should also be confirmed in future RCTs.
To our knowledge, this is the first comprehensive systematic review and meta-analysis to focus solely on real-world studies of patients with plaque PsO who received tildrakizumab. These findings reinforce the efficacy and safety of tildrakizumab in real-world practice and align with randomized trial results. The close correspondence of real-world and clinical trial results may be related to the every-12-week dosing regimen of tildrakizumab or the requirement in the USA for in-office administration, which mitigate challenges related to frequent dosing or to potentially inexpert administration in a home setting.
The findings from our analysis benefit the dermatology community by reflecting treatment outcomes in broader, more diverse patient populations, providing insights directly relevant to everyday clinical practice. Synthesizing only RWE gives clinicians, patients, and decision-makers a clearer picture of how tildrakizumab performs under clinical practice conditions as opposed to carefully selected clinical trial populations in tightly controlled settings, which can help to guide treatment expectations, optimal therapy selection, and informed decision-making. Subgroup analyses by geography, prior biologic exposure, presence of any comorbidities, and tildrakizumab dose yielded generally similar results, which is reassuring and suggests that heterogeneity of study populations did not materially affect the overall findings.
Our findings have implications for health technology assessment and formulary decision-making. Given the absence of significant differences across IL-23 inhibitors in comparative RWE, decisions around therapy selection may be better guided by physician judgment and patient preferences than by small and potentially confounded differences in real-world outcomes. This highlights the complementary role of RWE to that of randomized controlled studies in informing clinical practice and payer decisions.
Limitations
Although this review has many strengths, several limitations inherent to RWE studies should be noted. While well-accepted statistical methods were used to synthesize published data, selection bias is unavoidable given the observational nature of included studies, heterogeneity in populations, and variability in data collection methods. Our meta-analyses consistently demonstrated improvements in clinical outcomes with tildrakizumab, but the lack of randomization and smaller sample sizes in comparative studies require the findings be interpreted with caution, particularly for the exploratory comparative meta-analysis of tildrakizumab vs other treatments.
Furthermore, because of limited reporting from included studies, meta-analyses were conducted on the basis of raw study data instead of adjusted estimates of effect; the latter are preferable in syntheses of RWE studies given the reduced effects of confounding. As a result of the small numbers of available comparative studies, meta-regression analyses to account for study-level confounders were not feasible. Future studies using adjusted effect estimates and meta-regression are needed to address residual confounding more rigorously.
Inclusion of only English-language publications may have introduced language bias, as potentially relevant studies in other languages could not be assessed and might have influenced the findings. Since the majority of the tildrakizumab RWE studies were conducted in European populations, the generalizability of results to other regions may be limited. Although our review was conducted according to standard guidelines for systematic literature reviews and included comprehensive searches of major electronic databases and gray literature, it remains possible that relevant articles were missed.
Nevertheless, our rigorous search methods likely captured all large and influential eligible studies, and therefore the analyzed estimates likely reflect the true effectiveness, safety, and QoL outcomes associated with tildrakizumab use in real-world practice.
Conclusions
This systematic review and meta-analysis of RWE studies found that tildrakizumab was consistently associated with effectiveness maintained for up to 12 months, improvements in skin-related QoL, and strong tolerability in patients with moderate-to-severe plaque PsO. Pooled estimates indicated that patients with severe disease at baseline experienced improvements in PASI, %BSA, PGA, and DLQI within the first 3 months of tildrakizumab treatment, which were maintained for up to 1 year, suggesting a shift toward mild disease severity status over this period. Findings from this review are consistent with those from RCTs, supporting tildrakizumab as an impactful treatment option for the management of moderate-to-severe plaque PsO.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (PDF 3752 KB)