ORIGINAL REPORT

A Real-world Meta-analysis Study of Disease Severity in European Patients Who Received Baricitinib for Atopic Dermatitis: Data from BioDay, SCRATCH and TREATgermany

Celeste BOESJES^*, Marjolein de BRUIN-WELLER¹, Nicolaas P. A. ZUITHOFF², Said el BOUHADDANI^2,3,4, Inge HAECK⁵, Marijke KAMSTEEG⁶, Marie-Louise SCHUTTELAAR⁷, Luca LE TREUST⁸, Nick DENHOLM⁸, Ahmet AKKOC⁹, Simon F. THOMSEN¹⁰, Ida VITTRUP¹⁰, Thomas BIRKNER¹¹, Luise HEINRICH¹¹, Jochen SCHMITT¹¹, Stephan WEIDINGER¹² and Thomas WERFEL¹³

¹Department of Dermatology, University Medical Center Utrecht, Utrecht, Netherlands, ²Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, Netherlands, ³Population Health Research, King Abdullah International Medical Research Center, Jeddah, Saudi Arabia, ⁴King Saud bin Abdulaziz University for Health Sciences, Jeddah, Saudi Arabia, ⁵Department of Dermatology, Reinier de Graaf Gasthuis, Delft, Netherlands, ⁶Department of Dermatology, Radboud University Medical Center, Nijmegen, Netherlands, ⁷Department of Dermatology, University Medical Center Groningen, Groningen, Netherlands, ⁸Real-World Evidence, OPEN Health, London, United Kingdom, ⁹Zitelab, Frederiksberg, Denmark, ¹⁰Department of Dermatology, Bispebjerg Hospital, University of Copenhagen, Copenhagen, Denmark, ¹¹Center for Evidence-Based Healthcare, University Hospital Carl Gustav Carus and Carl Gustav Carus Faculty of Medicine, Technische Universität Dresden, Dresden, Germany, ¹²Center for Inflammatory Skin Diseases, Department of Dermatology and Allergy, University Hospital Schleswig-Holstein Campus Kiel, Kiel, Germany, and ¹³Department of Dermatology and Allergy, Hannover Medical School, Hannover, Germany

Corr: Celeste Boesjes, Department of Dermatology, University Medical Center Utrecht, Utrecht, Netherlands. ^*Email: c.m.boesjes-2@umcutrecht.nl

Key words: Atopic Dermatitis; Baricitinib; EASI; Registry.

 

 

Clinical trials have demonstrated the effectiveness of baricitinib, an oral selective Janus kinase 1/2 inhibitor, for patients with moderate-severe atopic dermatitis, but data from real-world practice are limited. This non-interventional cohort study evaluated clinician- and patient-reported disease severity over 16 weeks using 3 national atopic dermatitis registries: BioDay (Netherlands), SCRATCH (Denmark) and TREATgermany (Germany). Absolute scores for Eczema Area and Severity Index (EASI) and itch Numerical Rating Scale (Itch-NRS) were assessed at baseline and follow-up (Week 13/16) for each registry and for the overall pooled cohort. Descriptive analyses, generalized linear mixed modelling and meta-analysis approaches were applied to complete-case and multiply-imputed datasets. In total, 264 patients (89 BioDay, 117 SCRATCH, 58 TREATgermany) were included. Results showed improvements in EASI, with 58%, 68% and 62% achieving EASI≤7 at Week 13/16 in BioDay, SCRATCH and TREATgermany, respectively. In the pooled cohort, the proportion was 62% (95% CI: 50–73%). The pooled mean EASI change was -7.8 [14.4,–1.1]. Effectiveness was generally observed in both biologic-naïve and -experienced patients. Notable heterogeneity was observed, particularly for Itch-NRS, with the proportion achieving NRS≤4 ranging from 20% (BioDay) to 71% (TREATgermany). These real-world findings support the effectiveness of baricitinib while also highlighting variability in patient outcomes.

Atopic dermatitis (AD) is a chronic inflammatory skin disease characterized by severe itch and typically presents with erythematous, infiltrated and excoriated lesions, imposing a substantial burden on patients' quality of life (1).

Historically, broad-acting immunosuppressants were the only systemic options for AD not controlled by topical therapies, but recent developments include the approval of more targeted therapies including Janus kinase (JAK) inhibitors such as baricitinib, upadacitinib and abrocitinib, and the interleukin-4/13/31 inhibitors such as dupilumab, tralokinumab, lebrikizumab and nemolizumab (2).

Baricitinib was the first European Medicine Agency (EMA)-approved oral JAK inhibitor for treatment of adult patients with moderate-to-severe AD who are candidates for systemic therapy (3). Phase III monotherapy (BREEZE-AD1 and BREEZE-AD2) and combined therapy trials (baricitinib with topical corticosteroids [TCS]; BREEZE-AD7) have demonstrated baricitinib’s efficacy in moderate-to-severe AD patients who had inadequate response to topical therapies, improving clinical signs and symptoms in patients within 16 weeks of treatment and inducing rapid reduction of itch (4, 5). Post-hoc analyses also revealed rapid onset of itch improvement, often within a day of the first dose (6). Furthermore, recent data indicate that patients with an Itch Numerical Rating Scale (Itch-NRS)≥7 and<40% affected Body Surface Area (BSA), “itch-dominant” patients, are most likely to see Eczema Area and Severity Index (EASI) improvement or benefit most from combination (baricitinib and TCS) therapy (7, 8).

Since 2016, several national prospective treatment registries for AD have been established to assess the long-term effectiveness and safety of systemic treatments for AD (9, 10). These include the Dutch BioDay (established in 2018 with>2000 patients from 23 centres) (11), Danish SCRATCH (established in 2017 with ~1200 patients from 6 centres) (12) and German TREATgermany (established in 2016, with>2,500 adult AD patients across 66 centres (13)) registries; all include patients who received systemic treatments for moderate-to-severe AD and represent a rich data source of real-world baricitinib-treated patients.

Existing real-world research in systemic treatment of AD patients is fragmented, with separate observational studies from BioDay, SCRATCH and TREATgermany corroborating the effectiveness of baricitinib (9, 14, 15). Initial data from BioDay (N=51) and SCRATCH (N=44) registries showed significant improvements in clinician-assessed and patient-reported outcomes within the first 16 weeks of treatment, with a higher discontinuation rate due to ineffectiveness in BioDay (33.3%) compared to SCRATCH (7%) (14, 16).

This study aimed to demonstrate the real-world clinical effectiveness of baricitinib in AD patients by combining data from 3 European registries and to describe the characteristics of specific subgroups of interest including: patients with and without exposure to biologic treatment prior to baricitinib initiation and patients with a baseline BSA≤40% AND Itch-NRS≥7 (Itch-dominant phenotype).

MATERIALS AND METHODS

Study design and registry populations

This study combined data from BioDay, SCRATCH and TREATgermany; each registry’s patient inclusion criteria is described in Table SI. For this study, each registry aligned patient selection and outcome measurements at different time-points to ensure consistency.

Patient selection

The study population included adults (≥18 years) participating in BioDay, SCRATCH and TREATgermany who were treated with baricitinib for AD within routine clinical practice prior to August 2023 and were baricitinib-naïve prior to registry entry. Patients initiated on another non-baricitinib systemic therapy on the same day as initiation of baricitinib (index date, Day 0) were excluded from the study.

Time periods, variables and outcomes

Patient characteristics, treatment patterns, EASI scores and Itch-NRS scores were extracted from each registry for the period prior to the index date (baseline) until 125 days (16 weeks) post-index date (unless the patient was lost to follow-up, or otherwise censored).

Primary outcomes were EASI (to evaluate clinician-assessed severity) and Itch-NRS (patient-reported severity) and were described for Week 0 (i.e. baseline; includes data documented 14 days prior to and up to 2 days post initiation of baricitinib [Day: -14 to Day 2]) and Week 13/16 (Day 70 to Day 125]. In TREATgermany, patient visits typically occurred closer to Week 13 than Week 16, and therefore, our analysis window was expanded (Table SI). Where a patient had>1 baricitinib visit within a time window, data from the most recent visit were analysed.

The proportion of patients achieving specified clinical outcome thresholds (EASI≤7, Itch-NRS≤4 and EASI≤7 and/or Itch NRS≤4) were assessed at Week 13/16.

Key analyses were conducted for the overall cohort and for cohorts of interest. Biologic-experienced and biologic-naïve patients were defined as patients who did, and did not, receive a prescription of dupilumab and/or tralokinumab prior to initiation of baricitinib, respectively and “Itch-dominant” patients had BSA≤40% AND Itch NRS≥7 at baseline.

Statistical methods

Descriptive statistics for complete cases (patients with non-missing EASI / Itch-NRS scores at baseline and Week 13/16 only) were reported for absolute EASI and Itch-NRS scores (as continuous outcomes). Generalised Linear Mixed Models (GLMM) for normal and binomial distributions were fitted for continuous outcomes (absolute EASI & Itch NRS score) and binary outcomes (EASI≤7, Itch NRS≤4, EASI≤7 and/or Itch NRS≤4), respectively. As a sensitivity analysis, multiple imputation with chained equations (MICE) was applied for missing outcomes to avoid bias due to selection and to optimise statistical power. Imputation models included outcomes (EASI, Itch-NRS) at each available time-point, with predictive mean matching or linear regression (in the case of high proportion of missing data) for continuous variables, and (polytomous) logistic regression methods for imputation for dichotomous or categorical outcomes.

Meta-analytic approaches were used to pool the per-registry results for primary outcomes, with random effects (RE) modelling with inverse variance weighting for meta-analyses of means and regression coefficients. GLMM approaches for meta-analysis were used to pool proportions (17). Results were reported as pooled effects (means, proportions and regression coefficients), 95% confidence interval (CI) with Hartung-Knapp adjustment (where applicable) of the pooled effect size and associated p-values, as well as appropriate heterogeneity measures (e.g. I-Square, Tau², prediction intervals [PI]). Pooled results were presented with and without data imputation for missing data, except for pooled proportions, where only non-imputed results are described due nonconvergence of models for imputed data.

Descriptive analyses were conducted using R Statistical Software® version 4.2.0 or later (18). The meta-analyses were performed by BioDay using the Statistical Software v4.3.0 (or later) adapted package (metafor) (19).

RESULTS

In total, 264 patients (89 patients from BioDay, 117 from SCRATCH and 58 from TREATgermany) were included (Fig. 1). Overall, 60% (BioDay), 28% (SCRATCH) and 31% (TREATgermany) of patients were biologic-experienced at baseline (Table I).

Fig. 1. Total number of patients included in key study cohorts per registry. Red boxes indicate cohort for key effectiveness analyses (primary outcomes); additional sub-analyses were conducted for the cohorts in blue. *Ineffectiveness was a documented reason for all 10 discontinuations in SCRATCH, 22 of 34 in BioDay [11/34 due to side-effects and 3/34 due to patient choice), and 7 of 12 in TREATgermany [4/12 side-effects, 3/12 patient choice or other]).

Table I. Patient demographics, clinical characteristics and treatment history per registry (as observed at baseline/Week 0)

	BioDay N=89	SCRATCH N=117	TREATgermany N=58
Age, index date, years, mean (SD)	44.0 (18.5)	39.0 (14.3)	41.5 (14.6)
Gender, N (%)
Male	54 (61%)	41 (35%)	42 (72%)
Female	35 (39%)	76 (65%)	16 (28%)
Baseline EASIa, N (%)	81 (91%)	75 (68%)	57 (98%)
Mean (SD)	16.8 (13.6)	15.1 (10.8)	19.5 (12.8)
Median (IQR)	13.0 (14.6)	13.2 (14.2)	18 (16.0)
Baseline itch NRSb, N (%)	66 (74%)	72 (62%)	55 (95%)
Mean (SD)	6.3 (2.5)	6.7 (2.7)	6.6 (2.3)
Median (IQR)	7.0 (3.0)	7.0 (3.5)	7 (3.0)
Itch-dominant phenotype, N (%)	20 (22%)	20 (17%)	12 (21%)
Prior treatment history
Biologic-experienced, N (%)	53 (60%)	33 (28%)	18 (31%)
Previous systemic therapy, N (%)*	85 (96%)	107 (91%)	22 (38%)
Note: Baseline EASI/Itch NRS scores are ‘as observed’ for all patients in overall cohort who had a score recorded at Week 0. *The most frequent treatments in BIODAY were cyclosporin (79%), dupilumab (61%) and methotrexate (51%); in SCRATCH were methotrexate (75%), azathioprine (50%) and prednisolone (35%) and in TREATgermany was dupilumab (68%). aEASI scores range from 0 to 72, with higher scores indicating worse AD severity. bItch NRS scores range from 0 (no itch) to 10 (worst itch imaginable). cIn patients with systemic treatment recorded post or on registry entry date and prior to index date. EASI: Eczema Area and Severity Index; IQR: Interquartile range; Itch-NRS: itch Numerical Rating Scale; SD: standard deviation.

Patient demographics, clinical characteristics and treatment history were broadly similar across registries (Table I). Baseline EASI [mean (SD)] was 16.8 (13.6) for BioDay, 15.1 (10.8) for SCRATCH and 19.5 (12.8) for TREATgermany. Mean [SD] Itch-NRS at baseline was broadly similar across registries (6.3 [2.5] for BioDay, 6.7 [2.7] for SCRATCH and 6.6 [2.3] for TREATgermany).

The proportions of patients with the Itch-dominant phenotype (BSA≤40% AND Itch-NRS≥7) at baseline per registry were 22% (BioDay), 17% (SCRATCH) and 21% (TREATgermany); patient characteristics were generally similar to the overall cohort (Table SII).

Effectiveness

The proportions of patients who achieved EASI≤7 at Week 13/16 were 58% (25/43) in BioDay, 68% (25/37) in SCRATCH and 62% (24/39) in TREATgermany, respectively. Of biologic-naïve/experienced patients, 65%/54% (BioDay), 65%/83% (SCRATCH) and 59%/70% (TREATgermany) achieved EASI≤7 at follow-up (Fig. 2).

Fig. 2. Percentage (number) of patients achieving specified outcome thresholds at Week 13/16 per registry, overall and by prior biologic exposure. Total sample size excluding missing data, (i.e.) total patients with complete Week 0 and Week 13/16 information for the specified variable; Abbreviations: EASI: Eczema Area and Severity Index, Itch-NRS: itch Numerical Rating Scale.

Overall, 20% (6/30), 49% (19/39) and 71% (25/35) of the patients reached Itch-NRS≤4 at Week 13/16 in BioDay, SCRATCH and TREATgermany, respectively; when considering prior biologic exposure, 20%/20% (BioDay), 50%/43% (SCRATCH) and 78%/50% (TREATgermany) of biologic naïve/experienced patients reached Itch-NRS≤4 (Fig. 2).

Of patients with the Itch-dominant phenotype at baseline, the proportions who achieved EASI≤7 were 70% (7/10), 67% (6/9) and 67% (8/12) for BioDay, SCRATCH and TREATgermany, respectively (Fig. S1).

Absolute and change scores at baseline and follow-up are shown in Table SIII; general improvements were observed on average for descriptive (non-imputed) EASI and Itch-NRS at follow-up.

Results from the adjusted analysis (GLMM) for continuous outcomes are shown in Table SIV.

Pooled meta-analysis

Meta analysis of EASI in all patients showed a pooled mean reduction of -7.8 (95% CI: -14.4; -1.1; p=0.04) points at Week 13/16 and a reduction in Itch-NRS of -2.2 (95% CI: -4.6; 0.2; p=0.06). (Fig. 3). Heterogeneity was observed, especially in the results of the overall pooled cohort, with I² values of up to 76.55%: patients in the TREATgermany registry generally showed better improvement in EASI and Itch-NRS compared to the patients in BioDay and SCRATCH (Fig. 3). Similar results were observed when using the descriptive (non-imputed) approach (Fig. S2). The pooled mean (95% CI) reduction from baseline at follow-up for EASI (with data-imputation) was -6.6 points (-14.7; 1.6; p=0.07) in biologic-experienced and 7.9 points (-12.5; -3.4; p=0.02) in biologic-naive patients, respectively (Fig. 3; see Fig. S1 for non-imputed results).

Fig. 3. Meta analyses depicting pooled change in EASI and Itch NRS scores at follow-up for overall cohort and by prior biologic exposure (multiple imputation approach). Results shown include mean change (95% CI) for each registry and the pooled cohort (as per random effects modelling), and associated p-values, prediction intervals (PI), Tau² and I² percentage with 95% CI.

For the overall cohort, the pooled proportions of patients meeting the thresholds at follow-up were 0.62 (95% CI: 0.50; 0.73) for EASI≤7, 0.46 (95% CI: 0.05; 0.94) for Itch NRS≤4, and 0.78 (95% CI: 0.63; 0.88) for EASI≤7 and/or Itch-NRS≤4, respectively; broadly similar results were observed when stratifying by prior biologic exposure. Heterogeneity was mostly observed for NRS≤4, with I² values up to 88.71%, whereas heterogeneity for other proportions was generally low (Fig. 4).

Fig. 4. Results from meta-analysis of the proportion of patients meeting specified thresholds at Week 13/16 for overall cohort and by prior biologic exposure (no imputation).

There were 52 (mean age 42 years; 56% male) patients included in the pooled Itch-Dominant cohorts. The pooled mean change [95% CI] estimates (imputed) for EASI and Itch-NRS were -5.03 [-10.70; 0.65]; p=0.06 and -3.94 [-6.78; -1.09]; p=0.03, respectively (Fig. S3).

Treatment discontinuation

Overall, 38%, 9% and 21% of patients in BioDay, SCRATCH and TREATgermany, respectively, discontinued baricitinib treatment by Week 13/16. Ineffectiveness was the most frequently reported reason for discontinuation in all registries (Fig. 1).

DISCUSSION

Main findings

This study provides an updated assessment of the real-world effectiveness of baricitinib in AD for patients treated in the Netherlands, Denmark and Germany.

General improvements based on meta-analyses in absolute EASI scores were observed at follow-up in all patients, with a mean reduction of almost 8 points for the overall pooled cohort. Clinical response was somewhat better in biologic-naïve patients and slightly lower in those with prior biologic exposure. Improvements were also observed for Itch-NRS, with a mean reduction of more than 2 points at Week 13/16.

The descriptive (non-imputed) results showed that 62% of patients in the overall pooled cohort achieved EASI≤7 at follow-up. There was broad consistency across registries, while prediction intervals (indicative of the range into which we can expect the effects of future studies to fall (20)) were moderately supportive of a beneficial effect for EASI.

Interpretation

Moderate-to-high levels of heterogeneity were observed for both EASI and Itch-NRS outcomes. This led to the observation that the improvement over time was statistically significant in each registry separately, but the corresponding p-value of, for example, the pooled Itch-NRS was slightly higher than the traditional threshold of 0.05. Heterogeneity was observed in pooled results of both EASI and Itch-NRS for all patients and in some subgroups of biologic-naïve and biologic-experienced patients. In the meta-analyses of proportions, heterogeneity was mainly observed for NRS≤4. In all results, the heterogeneity estimates (i.e. CIs of I² and tau² values) were uncertain with a wide confidence interval, likely related to the low number of studies in this meta analyses (21). Heterogeneity appears to be caused by a more pronounced improvement in patients in the TREATgermany registry. Notably, in Germany, baricitinib is used as a first-line systemic treatment, while in the Netherlands and Denmark, it is prescribed after the failure of conventional treatments (12, 22). Patients in BioDay may also be more difficult to treat overall (due to higher proportion of biologic experienced patients). Differences in descriptive results between registries may also be influenced by the number of patients in washout of other systemic medications and the need for alignment of definitions and timepoints, which also contributes to variation when comparing with other publications.

When considering data from each registry separately, mean EASI ranged from 15.1 (SCRATCH) to 19.5 (TREATgermany) at baseline, aligning with baseline scores in prior real-world studies for patients with moderate-to-severe AD who had previously had an inadequate response to biologic therapy (14, 15, 23). At follow-up, mean EASI ranged from 7.2 (SCRATCH) to 10.1 (BioDay) in the current study while the proportions of patients reaching EASI≤7 ranged from 58% (25/43) to 68% (25/37). Real-world data on the 16 week effectiveness of baricitinib are somewhat limited, with existing publications primarily based on single-country registries. In a previous BioDay study of moderate-severe patients undergoing baricitinib treatment, the mean EASI was 11.1 at Week 16, with a probability of achieving EASI≤7 of 29.4% (95% CI 13.1–53.5) (14), likely reflective of the more complex, elderly or refractory patient characteristics under study. Similarly, a TREATgermany study found a substantial decrease in EASI at 3 months (mean±SD 21.5 ± 13.2 vs. 9.3 ± 9.0), with a slightly higher proportion of patients (44%) achieving EASI-75 at follow up (15) as baseline EASI were also higher. These results may have been influenced by differences in baseline characteristics, including the distribution of prior systemic therapies, age and EASI baseline. It is however notable that improvements in EASI were observed in this study despite a lower baseline EASI than observed in the clinical trials for baricitinib (4, 5) and for other real-world studies involving other treatments (24, 25, 26): this may indicate that there is a place for oral and more targeted therapies in populations with more moderate disease (e.g. pre- biologic or following biologic-failure).

In this study, results regarding the effectiveness of baricitinib were however particularly mixed when considering the impact on Itch-NRS relative to other RCTs and single-country RWE studies (4, 5, 14). In general, the most favourable results were observed for TREATgermany, although results were less consistent for Itch-NRS compared with EASI. Overall mean reductions in Itch-NRS from baseline to follow-up ranged from 1.2 to 3.2 points while the percentage of patients achieving Itch-NRS≤4 ranged from 20% (BioDay) to 71% (TREATgermany), indicating wide variability across settings; this notion was also reflected in some of the pooled data and wider heterogeneity measures (e.g. relatively high I²/ Tau² and wide CIs). Variability in patient-reported measures such as Itch-NRS has been reported in other AD studies and may be driven by wider factors such as age, gender and race (27, 28). Further research regarding what constitutes a meaningful change in AD severity for patients within real-world vs trial settings may be beneficial. Ultimately, these findings highlight the inherent challenges of treating this population and indicate that treatment effectiveness, particularly in terms of reducing itch severity, can be highly variable. This factor may also have contributed to the relatively high discontinuation rate observed in the current study, and in previous work (14).

In addition, our study provides insights into the effectiveness of baricitinib for specific patient subgroups. In general, EASI results for patients who were biologic-naïve at baseline were consistent with those observed in the overall cohort, supporting the clinical effectiveness of baricitinib for these patients. Although heterogeneity was observed in the results, there was some indication that baricitinib can be effective for patients who had switched from another advanced systemic treatment, both within individual registries and in the pooled cohort. These findings generally align with previous research which showed no significant differences in effectiveness for those with a prior lack of response to biologic treatment (“dupilumab non-responders”) vs those who switched to baricitinib for other reasons (14).

Our study also supports earlier findings that patients with the Itch-dominant phenotype may show greater regional improvements in disease severity (8). Despite small sample sizes, a slightly higher proportion of these patients met study thresholds at follow-up vs the overall cohort, despite the benchmark for improvement arguably being more challenging to reach in these patients (due to lower or more moderate baseline scores vs overall cohort). While real-world data are limited, patients with a more moderate BSA score (<40) yet higher perceived itch may represent a sizeable proportion of the population receiving baricitinib in routine clinical practice (29). These findings may indicate that this is a particular group that could benefit from baricitinib in the real-world, warranting further investigation.

Limitations

Our findings differed somewhat from the BREEZE-AD clinical trials (4, 5), likely due to baseline population differences and the smaller magnitude of change observed in the real-world settings, where patients had more moderate EASI scores (4, 5). Comparisons with trial data are limited by stricter trial inclusion criteria (e.g. worse baseline severity and requirement for treatment wash-out periods) unlike the broader, more flexible registry criteria. This reduces the potential for improvement in the real-world registries and also limits the usability of outcome measures based on relative or percentage change (including EASI-75) typically used in trial settings, prompting our focus on absolute outcomes. Additional factors that may explain differences in results include country-level discrepancies in treatment algorithms, the timing of approvals for relevant treatments (including dupilumab and JAK inhibitors), inter-rater variability, discontinuation rates and broader factors related to finances and access to treatment (12, 22), reflecting real-world complexity. Missing data, partially addressed via multiple imputation, and wider observational study limitations (e.g. unmeasured confounding, model assumptions), may have also introduced bias. Differences in registry definitions of “biologic-experienced” may also affect comparability.

Finally, assessing heterogeneity in meta-analyses is often debated, with no single gold-standard measure (20, 30, 31). We reported several measures (I², Tau², prediction intervals) to address limitations of individual metrics (20). While prediction intervals were broad and sometimes overlapped zero, they generally supported baricitinib’s effectiveness across real-world contexts (20).

Conclusions

Based on data to week 16 from 3 European registries, baricitinib appears effective in reducing clinically-assessed AD severity in a real-world population with relatively moderate - but still impactful - baseline severity, likely partially influenced by bridging therapies. Most patients were refractory to other systemic or biologic treatments. Improvements in EASI appeared to be broadly reflected not only in biologic-naïve patients, but also in those with prior biologic exposure and those with BSA≤40 but NRS≥7 (Itch-Dominant). Nevertheless, the real-world effectiveness of baricitinib appears to be heterogenous, particularly regarding its impact on patient-reported itch severity. Although these data reinforce previously published data, further research into the longer-term effectiveness of baricitinib would be valuable, alongside further evidence derived from moderate patients with predominant itch, to improve understanding of the patient populations likely to derive the greatest benefit from AD treatments.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the valuable contributions of the following research groups and treatment centres that contributed to each nationwide registry:

BioDay: The BioDay registry consists of 20 recruiting centres, for this manuscript patients were included from the University Medical Center Utrecht (prof. dr. de Bruin-Weller), University Medical Center Groningen (dr. M.L. Schuttelaar), Radboud Medical Center (dr. M. Kamphuis), Reinier de Graaf Hospital (dr. I. Haeck), Haga Hospital (dr. F. Garritsen), Medical Center Leeuwarden (dr. K. Politiek), Meander Medical Center (dr. AMT van Lynden-Van Nes), Spaarne Gasthuis (dr. AJ Oosting).

SCRATCH: The SCRATCH Study Group consists of the recruiting centres named in the author’s list and the following recruiting centres: KS Ibler, A Hærskjold, MO Christensen, EH Pinborg, Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark/ J Elberling, Department of Dermatology and Allergy, Copenhagen University Hospital – Herlev and Gentofte, Copenhagen, Denmark/ L Themstrup, Department of Dermatology, Zealand University Hospital, Roskilde, Denmark/ CG Mørtz, Department of Dermatology and Allergy, Odense University Hospital, Odense, Denmark/ C Vestergaard, Department of Dermatology, Aarhus University Hospital, Aarhus, Denmark/ MG Dalager, Department of Dermatology, Aalborg University Hospital, Aalborg, Denmark/ NS Krogh, Zitelab, Frederiksberg, Denmark.

TREATgermany Study Group: The TREATgermany Study Group consists of the following recruiting centres: Abraham, S., Department of Dermatology, University Allergy Center, Carl Gustav Carus Faculty of Medicine, TU Dresden, Dresden/ Adler, N., Hautärzte am Fastnachtsbrunnen, Mainz/ Anders, S., Dermatologic Practice, Munich/ Asefi, M., Dermatology study centre Hunsrueck, Simmern/ Asmussen, A., Dermatology at Lesum, Bremen/ Augustin, M., Institute for Health Services Research in Dermatology, University Medical Center Hamburg Eppendorf, Hamburg/ Bell, M., Dermatologic Practice, Andernach/ Biedermann, T., Department of Dermatology and Allergy, School of Medicine, TU Munich, Munich/ Boashie, U., Dermatologic Practice, Dresden/ Bong, A., Dermatologic Practice, Emmerich/ Brücher, JJ., Dermatologic Practice, Hautambulatorium, Magdeburg/ Buck, P., Dermatologic Practice - Goldbek medical, Hamburg/ Bürkle, CP., Skin and laser medicine Kinzigtal, Haslach/ Ertner, K., Dermatologic Practice, Nuernberg/ Fell, I., Dermatology Bad Soden Study Center, Bad Soden/ Gerlach, B., Dermatologic Practice, Dresden/ Gorriahn-Maiterth, H., Practice Dermasana, Karlsruhe/ Großmann, B., Dermatologic Practice, Koblenz/ Handrick, C., Dermatologic Practice, Berlin/ Herrmann, A., Dermatologic Practice, Greifswald/ Hilgers, M., University Hospital Aachen, Department of Dermatology and Allergology, Aachen/ Hoffmann, M., Dermatologic Practice, Witten/ Homey, B., Department of Dermatology and Allergology, University Hospital Duesseldorf, Düsseldorf/ Hong-Weldemann, SH Dermatologic Practice, Freiburg/ Hünermund, A., Dermatologic Practice, Heiligenstadt/ Jacobs, F., Practice DERMAKULM, Kulmbach/ John, L., Dermatologic Practice, Freital/ Kage, P., Dermatologic Practice, Freiberg/ Kleinheinz, A., Clinics for Dermatology, Elbe Klinikum Buxtehude, Buxtehude/ Krähn-Senftleben, G., Donau Alb Skin Center, Neu-Ulm/ Kreutzer, K., Department of Dermatology, OWL University Hospital of Bielefeld University, Campus Clinic Bielefeld, Bielefeld/ Meinhardt, K., Dermatologic Practice, Beckum/ Merl, V., Dermatology Nordwaldeck, Joint practice Dr. Merl and Dr. Richter, Bad Arolsen/ Neubert, K., Dermatologic Practice, Burgstaedt/ Novak, N., University Hospital Bonn, Clinic and Polyclinic for Dermatology and Allergology, Bonn/ Pinter, A., Department of Dermatology, Venereology and Allergology, Clinical Research, University Hospital, Frankfurt am Main/ Quist, S., Dermatology Clinic, Helix Medical Excellence Center Mainz, Mainz/ Ramaker-Brunke, J., Practice 'Die Hautärzte' Braunschweig, Braunschweig/ Reitenbach-Blindt, I., Practice Dermasana, Eggenstein-Leopoldshafen/ Rossbacher, J., Dermatologic Practice - Hautzentrum, Friedrichshain/ Schaarschmidt, ML., University Medical Center Mannheim Dermatology Clinic, Mannheim/ Schaefer, T., Dermatologic Practice, Derma Koeln, Cologne/ Schäkel, K., Department of Dermatology, University Hospital, Heidelberg/ Schenck, F., Dermatology Center, Hannover/ Schirmer, T., Dermatologic Practice, Berlin/ Schulz-Kiesow, M., Dermatologic Practice, Lübeck/ Schwarz, B., Dermatologic Practice, Langenau/ Schwichtenberg, U., Practices Derma-Nord, Bremen/ Stahl, M., Dermatologic Practice, Osterode/ Staubach-Renz, P., Clinic for Dermatology, University Hospital, Mainz/ Steinke, S., Hohenzollernring Dermatologic Practice, Münster/ Sticherling, M., Department of Dermatology, University, German Center for Immunotherapy, Erlangen/ Stuhlert, A., Dermatologic Practice, Pforzheim/ Suckow, M., Bayreuth Skin and Laser Center, Bayreuth/ Tchitcherina, E., Dermatologic Practice, Friedberg/ Völkel, E., Dermatologic Practice, Nuernberg/ von Kiedrowski, R., Focus Practice for chronic inflammatory dermatoses, skin cancer and allergology and also Study Center CMS3, Selters (Westerwald)/ Weidinger, S., Center for Inflammatory Skin Diseases, Department of Dermatology and Allergy, University Hospital Schleswig-Holstein, Campus Kiel, Kiel/ Weisshaar, E., Division of Occupational Dermatology, Department of Dermatology, Ruprecht-Karls University Heidelberg, Heidelberg/ Werfel, T., Department of Dermatology and Allergy, Hannover Medical School, Hannover/ Wiemers, F., Dermatologic Practice, Leipzig/ Wildfeuer, T., Dermatologic Practice, Berlin/ Wollenberg, A., Dept. of Dermatology and Allergy, Augsburg University Hospital, Augsburg; Dept. of Dermatology and Allergy, University Hospital Munich, LMU Munich/ Worm, M., Department of Dermatology, Allergy and Venereology, Charité Berlin, Berlin/ Yenigün, N., Dermatologic Practice, Kiel.

We are grateful for their support and collaboration in providing access to data, which was critical for the analysis. The efforts of the registry teams in collecting and maintaining high-quality data are sincerely appreciated.

Data access and analysis were performed by the registries and coordinated by OPEN Health. Additional support with data analysis and interpretation was provided by Inmaculada De La Torre, Silvia Sabatino and Alan James Michael Brnabic of Eli Lilly.

Medical writing assistance in the drafting and revision of this manuscript was provided by Fiona Glen, Fatemeh Saberi Hosnijeh and Edward Ottley of OPEN Health (funded by Eli Lilly).

Additionally, we thank all the researchers, clinicians, and staff at these institutions for their ongoing contributions to advancing knowledge in this area of research.

REFERENCES

1. Eichenfield LF, Tom WL, Chamlin SL, Feldman SR, Hanifin JM, Simpson EL, et al. Guidelines of care for the management of atopic dermatitis: section 1. Diagnosis and assessment of atopic dermatitis. J Am Acad Dermatol 2014; 70: 338–351. https://doi.org/10.1016/j.jaad.2013.10.010
2. Wollenberg A, Kinberger M, Arents B, Aszodi N, Avila Valle G, Barbarot S, et al. European guideline (EuroGuiDerm) on atopic eczema: part I - systemic therapy. J Eur Acad Dermatol Venereol 2022; 36: 1409–1431. https://doi.org/10.1111/jdv.18345
3. EMA. European medicines agency. Olumiant; 2018. [cited 2023 March 23]. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/olumiant
4. Simpson EL, Lacour JP, Spelman L, Galimberti R, Eichenfield LF, Bissonnette R, et al. Baricitinib in patients with moderate-to-severe atopic dermatitis and inadequate response to topical corticosteroids: results from two randomized monotherapy phase III trials. Br J Dermatol 2020; 183: 242–255. https://doi.org/10.1111/bjd.18898
5. Reich K, Kabashima K, Peris K, Silverberg JI, Eichenfield LF, Bieber T, et al. Efficacy and safety of baricitinib combined with topical corticosteroids for treatment of moderate to severe atopic dermatitis: a randomized clinical trial. JAMA Dermatol 2020; 156: 1333–1343. https://doi.org/10.1001/jamadermatol.2020.3260
6. Buhl T, Rosmarin D, Serra-Baldrich E, Fernandez-Peñas P, Igarashi A, Konstantinou MP, et al. Itch and sleep improvements with baricitinib in patients with atopic dermatitis: a post hoc analysis of 3 phase 3 studies. Dermatol Ther 2021; 11: 971–982. https://doi.org/10.1007/s13555-021-00534-8
7. Thyssen JP, de Bruin-Weller M, Costanzo A, Grond S, Schuster C, Liu C, et al. Baseline body surface area and itch severity define response to baricitinib in patients with moderate-to-severe atopic dermatitis at week 16. Adv Ther 2023; 40: 3574–3587. https://doi.org/10.1007/s12325-023-02528-8
8. Carrascosa JM, Narcisi A, Nomura T, Ständer S, Vestergaard C, Sabatino S, et al. Baricitinib improvement across regions in atopic dermatitis patients with baseline body surface area up to 40% and severe itch. Dermatol Ther 2024; 14: 1561–1573. https://doi.org/10.1007/s13555-024-01171-7
9. Gerbens LAA, Apfelbacher CJ, Irvine AD, Barbarot S, de Booij RJ, Boyce AE, et al. TREatment of ATopic eczema (TREAT) Registry Taskforce: an international Delphi exercise to identify a core set of domains and domain items for national atopic eczema photo- and systemic therapy registries. Br J Dermatol 2019; 180: 790–801. https://doi.org/10.1111/bjd.16714
10. Vermeulen FM, Gerbens LAA, Bosma AL, Apfelbacher CJ, Irvine AD, Arents BWM, et al. TREatment of ATopic eczema (TREAT) Registry Taskforce: consensus on how and when to measure the core dataset for atopic eczema treatment research registries. Br J Dermatol 2019; 181: 492–504. https://doi.org/10.1111/bjd.17715
11. BioDay eczema and atopic diseases registry. 2023. [cited 2023 March 24]. Available from: https://bioday.nl/
12. Musters AH, Gerbens LAA, van der Gang L, Middelkamp-Hup MA, Ouwerkerk W, Hijnen DJ, et al. Baseline characteristics of atopic eczema patients enrolled in seven European registries united in the TREatment of ATopic eczema (TREAT) registry taskforce. J Eur Acad Dermatol Venereol 2025; 39: 2096–2112. https://doi.org/10.1111/jdv.20876
13. TREATgermany. Deutsches Neurodermitis-Register; . [cited 2025 July 21]. Available from: https://treatgermany.org/
14. Boesjes CM, Kamphuis E, Zuithoff NPA, Bakker DS, Loman L, Spekhorst LS, et al. Daily practice experience of baricitinib treatment for patients with difficult-to-treat atopic dermatitis: results from the BioDay registry. Acta Derm Venereol 2022; 102: adv00820. https://doi.org/10.2340/actadv.v102.3978
15. Traidl S, Heinrich L, Siegels D, Heratizadeh A, Kind B, Haufe E, et al. Treatment of moderate-to-severe atopic dermatitis with baricitinib: results from an interim analysis of the TREATgermany registry. J Eur Acad Dermatol Venereol 2024; 38: e887–e891. https://doi.org/10.1111/jdv.19979
16. Vittrup I, Elberling J, Skov L, Ibler KS, Jemec GBE, Mortz CG, et al. Short-term real-world experience with baricitinib treatment in Danish adults with moderate-severe atopic dermatitis. J Eur Acad Dermatol Venereol 2023; 37: e543–e546. https://doi.org/10.1111/jdv.18804
17. Schwarzer G, Rücker G. Meta-analysis of proportions. Methods Mol Biol 2022; 2345: 159–172. https://doi.org/10.1007/978-1-0716-1566-9_10
18. R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2021. Available from: https://www.R-project.org/
19. Viechtbauer W. Conducting meta-analyses in R with the metafor package. J Stat Softw 2010; 36: 1–48. https://doi.org/10.18637/jss.v036.i03
20. Harrer M, Cuijpers P, Furukawa TA, Ebert DD. Chapter 5 between-study heterogeneity. In: Doing Meta-Analysis in R. Available from: https://bookdown.org/MathiasHarrer/Doing_Meta_Analysis_in_R/heterogeneity.html#het-measure-which
21. Langan D, Higgins JPT, Jackson D, Bowden J, Veroniki AA, Kontopantelis E, et al. A comparison of heterogeneity variance estimators in simulated random-effects meta-analyses. Res Synth Methods 2019; 10: 83–98. https://doi.org/10.1002/jrsm.1316
22. Nederland Z. Zorginstituut Nederland. [cited 2023 May 16]. Available from: https://www.medicijnkosten.nl/
23. Augustin M, Serra-Baldrich E, Seneschal J, Grond S, Lampropoulou A, Elrayes M, et al. Baseline characteristics and treatment patterns of patients with atopic dermatitis treated with oral systemic therapies: an interim analysis of the AD-REAL study. Adv Ther 2025; 42: 2728–2738. https://doi.org/10.1007/s12325-025-03137-3
24. Kamata M, Tada Y. A literature review of real-world effectiveness and safety of dupilumab for atopic dermatitis. JID Innov 2021; 1: 100042. https://doi.org/10.1016/j.xjidi.2021.100042
25. Tolino E, Ambrosio L, Bernardini N, Proietti I, Skroza N, Potenza C. Effectiveness of tralokinumab in different phenotypes of atopic dermatitis: a real-world study. Dermatol Ther 2025; 15: 337–350. https://doi.org/10.1007/s13555-025-01341-1
26. Ibba L, Falcidia C, Di Giulio S, Bianco M, Valenti M, Facheris P, et al. Real-world effectiveness and safety of upadacitinib and abrocitinib in moderate-to-severe atopic dermatitis: a 52-week retrospective study. J Clin Med 2025; 14: 2953. https://doi.org/10.3390/jcm14092953
27. Li Y, Swerlick RA. Capture of patient itch scores in practice reveals disparate itch impact on the basis of age, gender, and race: a cross-sectional survey analysis. JID Innov 2025; 5: 100338. https://doi.org/10.1016/j.xjidi.2024.100338
28. Wei W, Anderson P, Gadkari A, Blackburn S, Moon R, Piercy J, et al. Discordance between physician- and patient-reported disease severity in adults with atopic dermatitis: a US cross-sectional survey. Am J Clin Dermatol 2017; 18: 825–835. https://doi.org/10.1007/s40257-017-0284-y
29. de Wijs L, Schreurs C, Schlösser A, Nijsten T, Hijnen DJ. Baricitinib for atopic dermatitis patients who responded inadequately to dupilumab treatment: first daily practice results. JEADV Clin Pract 2022; 1: 364–371. https://doi.org/10.1002/jvc2.64
30. Lorenc T, Felix L, Petticrew M, Melendez-Torres GJ, Thomas J, Thomas S, et al. Meta-analysis, complexity, and heterogeneity: a qualitative interview study of researchers’ methodological values and practices. Syst Rev 2016; 5: 192. https://doi.org/10.1186/s13643-016-0366-6
31. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21: 1539–1558. https://doi.org/10.1002/sim.1186