Introduction
Psoriasis is a chronic, immune-mediated inflammatory disease with a global prevalence of approximately 2–3% and affects more than 125 million people worldwide.1 Plaque psoriasis is the predominant clinical subtype, accounting for approximately 80% of cases, of whom 20–30% are moderate-to-severe. In addition to cutaneous erythema and scaling, psoriasis is frequently accompanied by comorbidities such as psoriatic arthritis, cardiovascular disease, metabolic syndrome, and depression, which substantially increase disability and mortality and markedly impair health-related quality of life.2 For mild psoriasis, topical agents, principally corticosteroids and vitamin D analogues, remain the mainstay of treatment. For moderate-to-severe plaque psoriasis, guidelines recommend systemic biologic therapies targeting TNF-α, IL-12/23, IL-17, and IL-23 as first-line options, with oral agents such as apremilast and methotrexate as alternatives.3,4 However, current therapies have important limitations.
Biologics require parenteral administration, are costly, may be contraindicated in patients with active infection, advanced heart failure, or a history of malignancy, and may lose effectiveness over time; conventional oral agents have limited efficacy and may cause adverse effects, including myelosuppression and hepatic or renal toxicity.5
In China, a dermatologist-confirmed population survey in six cities reported a standardized psoriasis prevalence of 0.47%.6 The National Psoriasis Center Registry subsequently collected 55,402 baseline observations from 282 centers during 2020–2022; 37.4% of assessed patients had PASI ≥10 and 53.2% had BSA involvement ≥10%.7 These clinic-based estimates reflect a substantial disease burden among patients seeking care and support the need for treatment evidence from Chinese clinical practice.
Deucravacitinib is the first oral, selective allosteric inhibitor of tyrosine kinase 2 (TYK2) inhibitor approved for moderate-to-severe plaque psoriasis. By binding to the TYK2 pseudokinase (JH2) domain and stabilizing its autoinhibited conformation, deucravacitinib achieves >1000-fold selectivity over JAK1, JAK2, and JAK3, while avoiding the hematologic toxicity and increased the infection risks associated with non-selective JAK inhibition.8–10 TYK2 is a central mediator of the JAK-STAT pathway and plays a pivotal role in transducing IL-12, IL-23 and type I interferon signals implicated in psoriasis pathogenesis.10,11 Large, global Phase 3 trials (POETYK PSO-1 and PSO-2) demonstrated that deucravacitinib 6 mg once daily produced significantly higher PASI-75 response rates at week 16 than placebo and apremilast, with a cumulative clinical benefit more than 50% greater than that of apremilast.12–14 Regional studies, including the POETYK PSO-3 trial and Japanese subgroup analyses, further confirmed efficacy in Asian patients, with a PASI 75 response rate of 62% at week 16, responses sustained through week 52, and clinically meaningful effects at difficult-to-treat sites such as the scalp.15–17 Long-term follow-up and pooled analyses have shown that efficacy is maintained for up to 3 years, with a low incidence of serious adverse events and a favorable safety profile.16,18,19 A multicenter real-world study published in 2026 reported a scalp-specific Physician Global Assessment (PGA) 0/1 clearance rate of 89.8% at week 16, further supporting its effectiveness at recalcitrant sites.16,20 Targeting the TYK2 pathway has therefore emerged as an important therapeutic strategy for psoriasis, with accumulating clinical evidence.21
However, most available evidence for deucravacitinib comes from randomized multicenter trials, and real-world data from Chinese patients remain limited. There is a paucity of studies addressing short-term (12-week) responses, anatomical site-specific outcomes, and the predictive value of baseline biomarkers, particularly predictors that could help identify likely responders.21,22 Network meta-analyses provide indirect comparisons with selected biologic therapies, but do not establish equivalent efficacy in routine Chinese practice.22 To address these gaps, we conducted a prospective, single-center, real-world study to evaluate the 12-week efficacy and safety of deucravacitinib in Chinese patients with moderate-to-severe plaque psoriasis, characterize response patterns across anatomical sites, and explore baseline biomarker predictors, thereby generating real-world evidence for its use in Asian populations.
Our group previously characterized several mechanisms implicated in psoriasis pathogenesis. These include CXCR6-mediated tissue-resident memory T cells (TRM) chemotaxis via CXCL16 that sustains disease relapse;23 fibroblast subpopulations with high matrix metalloproteinase 2 expression regulating CD8+ T-cell residency and inflammation through CD100 signaling;24 and glycoprotein Ib-CD11b+ monocyte-derived macrophages serving as a cellular bridge between atherosclerosis and exacerbated psoriatic inflammation.25 Building on these mechanistic findings, the present study provides a prospective real-world evaluation of deucravacitinib response predictors in Chinese patients with moderate-to-severe psoriasis, thereby bridging molecular pathogenesis with clinical patient stratification for selective TYK2 inhibition.
Materials and Methods
Study Design and Patients
This prospective, single-center, real-world study enrolled 120 patients with moderate-to-severe plaque psoriasis who attended the Department of Dermatology, Huashan Hospital, Fudan University, between November 1, 2024 and December 30, 2025. Disease severity was assessed using the Psoriasis Area and Severity Index (PASI) and body surface area (BSA) at baseline and at weeks 4, 8, and 12. Eligible adults met at least one of the following disease-severity criteria: PASI ≥10, BSA involvement ≥10%, or Dermatology Life Quality Index (DLQI) ≥10.
Regional PASI subscores were recorded for four anatomical regions (head/neck, trunk, upper limbs, and lower limbs). The study was registered with the Chinese Clinical Trial Registry (ChiCTR2400091511). The study was conducted in accordance with the Declaration of Helsinki, and all participants provided written informed consent before enrollment.
Treatment
Prior to enrolment, all participants underwent routine laboratory screening, including complete blood count (CBC), serum biochemistry, lipid profile, and screening for hepatitis and tuberculosis. No washout period was required for previous therapies. All patients received oral deucravacitinib 6 mg once daily for 12 weeks.
During treatment, patients were advised to continue supportive topical care (emollients, topical corticosteroids, and topical calcineurin inhibitors).
Outcome Measures
At the baseline visit, demographic and clinical data were collected, including sex, age, age at psoriasis onset, body mass index (BMI), clinical manifestations, personal and family medical history, and previous treatment regimens. BMI was classified into two categories: normal weight (<24 kg/m2) and overweight (≥24 kg/m2).
The primary efficacy endpoint was the proportion of patients achieving a PASI 75 response at week 12. Secondary endpoints included the proportions of patients who achieved PASI 90 and PASI 100 responses at weeks 4, 8 and 12; changes from baseline in absolute PASI and BSA scores at each time point; and regional efficacy outcomes, including PASI response rates and score changes for each anatomical site.
Safety
Throughout the 12-week treatment period, all adverse events (AEs) and laboratory parameters were monitored prospectively. The type, incidence, and severity of AEs occurring during treatment were recorded, and changes from baseline in clinical laboratory values and vital signs were systematically assessed.
Biomarker Analysis
To explore potential predictive biomarkers of treatment response, peripheral blood samples were collected prior to therapy from a subset of patients (n=43). Cytokine and chemokine concentrations were quantified using RayPlex Human Chemokine Bead Array and RayPlex Human Inflammation Bead Array. The measured analytes included, but were not limited to, TNF-α, IL-1β, IL-23p19, IFN-γ, G-CSF, MIG, MIP-1α, IL-8, GRO-α, IL-17A, and IL-4.
Statistical Analysis
Efficacy analyses were performed using the full analysis set (FAS). Categorical variables are presented as counts and percentages, and continuous variables as mean ± standard deviation (SD) unless otherwise indicated. Potential predictors of PASI 75 and PASI 90 responses were evaluated using multivariable logistic regression models, with results expressed as odds ratios (ORs) and 95% confidence intervals (CIs).
Within-patient comparisons of PASI and BSA scores before and after treatment were performed using paired t tests or corresponding nonparametric tests. Cytokine levels were compared between PASI 75 achievers and non-achievers using unpaired t tests or nonparametric equivalents. All hypothesis tests were two-sided, and P < 0.05 was considered statistically significant.
Data were analyzed using GraphPad Prism 10.0 and SPSS 26.0.
Results
Baseline Patient Demographic and Clinical Characteristics
A total of 120 patients with moderate-to-severe plaque psoriasis were enrolled. Of these, 101 (84.2%) were naïve to systemic treatment (initial-treatment group), whereas 19 (15.8%) had received previous systemic therapy (prior-systemic group). The mean age of the overall cohort was 37.2 years (SD 13.3); 70 patients (58.3%) were male and 50 (41.7%) were female.
The mean baseline PASI score was 7.4 (SD 2.7) and mean BSA involvement was 15.3% (SD 10.5%). The mean disease duration was 9.2 years (SD 9.3). Fifty-four patients (45.0%) had a BMI <24 kg/m2, 48 (40.0%) had a BMI ≥24 kg/m2, and BMI data were missing for 18 (15.0%).
Comorbidities included psoriatic arthritis in 9 patients (7.5%), nail psoriasis in 31 (25.8%), hypertension in 31 (25.8%), and diabetes mellitus in 21 (17.5%).
Among the 19 patients in the prior-systemic group, 11 (57.9%) had previously received IL-17A inhibitors, 3 (15.8%) had received IL-23 inhibitors, 3 (15.8%) had received apremilast, and 2 (10.5%) had received upadacitinib. Detailed baseline characteristics by group are presented in Table 1.Table 1Demographic and Clinical Characteristics of Psoriasis Patients at BaselineVariableTotal Group (n=120)Naïve Group (n=101)Prior Systemic Therapy Group (n=19)Gender Male, n (%)70 (58.3)58 (57.4)12 (63.2) Female, n (%)50 (41.7)43 (42.6)7 (36.8)Age, mean (SD), years37.2 (13.3)36.6 (12.9)40.4 (14.9)Body mass index, kg/m2 <24, n (%)54 (45.0)49 (48.5)5 (26.3) ≥24, n (%)48 (40.0)38 (37.6)10 (52.6) Missing, n (%)18 (15.0)14 (13.9)4 (21.1)Itching, n (%)66 (55.0)56 (55.4)10 (52.6)Family history, n (%)31 (25.8)26 (25.7)5 (26.3)Smoking, n (%)23 (19.2)18 (17.8)5 (26.3)Drinking, n (%)26 (21.7)21 (20.8)5 (26.3)PASI, mean (SD)7.4 (2.7)7.5 (2.8)6.8 (1.9)BSA, mean (SD), %15.3 (10.5)15.8 (10.9)12.8 (8.0)Duration of plaque psoriasis, mean (SD), years9.2 (9.3)8.6 (8.9)12.4 (1.8)Comorbidity, n (%) Psoriatic arthritis9 (7.5)4 (4.0)5 (26.3) Nail psoriasis31 (25.8)26 (25.7)5 (26.3) Hypertension31 (25.8)25 (24.8)6 (31.6) Diabetes21 (17.5)16 (15.8)5 (26.3)Previous treatment, n (%) Apremilast3 (2.5)0 (0.0)3 (15.8) Upadacitinib2 (1.7)0 (0.0)2 (10.5) IL-17A monoclonal antibody11 (9.2)0 (0.0)11 (57.9) IL-23 monoclonal antibody3 (2.5)0 (0.0)3 (15.8)Abbreviations: BMI, body mass index; BSA, body surface area; IL, interleukin; PASI, Psoriasis Area and Severity Index; SD, standard deviation.
Efficacy
Treatment with deucravacitinib was associated with significant reductions in both PASI and BSA scores at weeks 4, 8, and 12 relative to baseline (Figure 1). At week 4, 5.8% of patients achieved a PASI 75 response. By week 8, the PASI 75 and PASI 90 response rates had increased to 25.8% and 6.5%, respectively.
At week 12, 59.6% of patients achieved PASI 75, 24.0% achieved PASI 90, and 4.8% achieved complete clearance (PASI 100).Figure 1Changes in Efficacy Outcomes During Deucravacitinib Treatment. (A) Changes in PASI scores and BSA from baseline to weeks 4, 8, and 12. (B) Proportions of patients achieving PASI 75, PASI 90, and PASI 100 at weeks 4, 8, and 12.
(C) PASI response rates over time by anatomical region (head and neck, trunk, upper limbs and lower limbs). ***P < 0.001 (paired t test, compared with baseline).A mixed chart showing PASI and BSA changes and PASI response rates over time and by body region.Image A: Two box plots. Left: PASI Score (0-25), x-axis 0w, 4w, 8w, 12w. Boxes decrease over time; asterisks at 4w, 8w, 12w.
Right: BSA Score (0-60), same x-axis. Boxes decrease; asterisks at 4w, 8w, 12w. Image B: Line graph. x-axis time (w) 0, 4, 8, 12; y-axis PASI Response rate (percent) 0-80.
Legend: PASI 75, 90, 100. PASI 75: 5.75 at 4w, 25.81 at 8w, 59.62 at 12w. PASI 90: 6.45 at 8w, 24.04 at 12w.
PASI 100: 4.81 at 12w. Image C: Four line graphs with error bars. x-axis 0w, 4w, 8w, 12w; y-axis PASI Response rate (0.0-1.0). Head/neck: 0.39 at 4w, 0.64 at 8w, 0.79 at 12w.
Trunk: 0.39 at 4w, 0.63 at 8w, 0.76 at 12w. Upper limbs: 0.38 at 4w, 0.62 at 8w, 0.72 at 12w. Lower limbs: 0.42 at 4w, 0.58 at 8w, 0.74 at 12w.Abbreviations: BSA, body surface area; PASI, Psoriasis Area and Severity Index.
Regional PASI 75 response rates also increased over time. The rates for the head and neck were 39% at week 4, 64% at week 8, and 79% at week 12; the corresponding rates were 39%, 63%, and 76% for the trunk; 38%, 62%, and 72% for the upper limbs; and 42%, 58%, and 74% for the lower limbs (Supplementary Tables 1 and 2).
Efficacy Prediction
In the multivariable logistic regression analysis (Figure 2), baseline BMI <24 kg/m2 was the only variable independently associated with achieving a PASI 90 response (OR, 5.34; 95% CI, 1.28–22.21). No other evaluated baseline characteristic, including disease duration, previous systemic therapy, smoking, alcohol use, family history, nail involvement, hypertension, or diabetes, was independently associated with either PASI 75 or PASI 90 (all P > 0.05).Figure 2Adjusted Odds Ratios and 95% CIs for Factors Associated with PASI 75 and PASI 90 at Week 12. *P < 0.05.A forest plot of adjusted odds ratios for PASI 75 and PASI 90 across baseline factors.Two forest plots show adjusted odds ratios (OR) with 95 percent confidence intervals (CI) for PASI 75 and PASI 90. For PASI 75, the x-axis ranges from e-2 to e3, with a reference line at e0.
Factors include: Duration greater than or equal to 84 months (OR 0.786, CI 0.312-1.980), Systemic Treatment (OR 1.985, CI 0.563-7.006), Smoking (OR 1.295, CI 0.344-4.868), Alcohol (OR 0.978, CI 0.295-3.244), Family history (OR 0.839, CI 0.315-2.236), Nail involvement (OR 0.558, CI 0.202-1.545), BMI less than 24 (OR 1.046, CI 0.369-2.962), Hypertension (OR 1.952, CI 0.674-5.652), Diabetes (OR 2.315, CI 0.658-8.137). For PASI 90, the x-axis spans e-3 to e4, with factors: Duration greater than or equal to 84 months (OR 1.186, CI 0.383-3.674), Systemic Treatment (OR 0.666, CI 0.131-3.374), Smoking (OR 1.608, CI 0.257-10.055), Alcohol (OR 1.625, CI 0.318-8.301), Family history (OR 2.156, CI 0.601-7.735), Nail involvement (OR 0.385, CI 0.117-1.27), BMI less than 24 (OR 5.336, CI 1.282-22.213*), Hypertension (OR 0.559, CI 0.128-2.443), Diabetes (OR 3.386, CI 0.33-34.789).Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; PASI, Psoriasis Area and Severity Index.
Pretreatment peripheral blood from 43 patients was analyzed using RayPlex Human Chemokine Bead Array and RayPlex Human Inflammation Bead Array. Of these patients, 18 were classified as PASI 75 responders and 19 as non-responders; the remaining six had unknown response status because they were lost to follow-up. Cytokine profiles are summarized in Figure 3A.
Exploratory comparisons (Figure 3B and Supplementary Table 3) showed higher baseline TNF-α, IL-1β, IL-23p19, IFN-γ, and G-CSF levels in non-responders. No statistically significant between-group differences were observed for the other measured cytokines and inflammatory mediators.Figure 3Pretreatment Serum Biomarker Profiles and Their Association with Treatment Response. (A) Heatmap of 23 pretreatment biomarkers for the overall sample and the two classified response groups.
(B) Pretreatment biomarker distributions for all sampled patients, PASI 75 non-responders, and PASI 75 responders. Samples were obtained from 43 patients; 18 were classified as responders and 19 as non-responders, whereas six had unknown response status because of loss to follow-up. The responder/non-responder comparison included the 37 classified patients. **P < 0.01, *P < 0.05 (t test comparing PASI 75 responders with non-responders only).A heatmap and box plots showing pretreatment biomarker levels by PASI 75 response groups.The composite image includes a heatmap and five box plots.
The heatmap displays standardized levels of 23 pretreatment biomarkers across all patients, PASI 75 non-responders and responders. The x-axis lists biomarkers like IFNg, IL-12p70 and others, while the y-axis categorizes groups as Inflammation, Treg, Th17, Th2 and Th1. The color scale ranges from negative 2 to positive 2, indicating biomarker level deviations.
The box plots focus on TNFa, IL-1b, IL-23p19, IFNg and G-CSF. The x-axis represents biomarker levels and the y-axis shows response groups: All, less than PASI75 and greater than or equal to PASI75. Significant differences are marked with asterisks.
The heatmap provides an overview, while the box plots highlight specific biomarkers with group comparisons. The legend differentiates groups by color. PASI 75 non-responders generally show higher levels in TNFa, IL-1b and IL-23p19.
The image illustrates the relationship between biomarker levels and treatment response.
Safety
Among the 120 treated patients, 41 (34.2%) reported at least one adverse event during treatment (Table 2). The most common events were acne (24/120, 20.0%), folliculitis (13/120, 10.8%), upper respiratory tract infection (7/120, 5.8%), eczema (3/120, 2.5%), urticaria (2/120, 1.7%), and tinea pedis (2/120, 1.7%). Less common events, each reported in one patient (1/120, 0.8%), included headache, diarrhea, nausea, oral ulcer, periodontitis, and herpes simplex virus infection.
No serious adverse events were reported during the 12-week observation period.Table 2Adverse Events During 12 Weeks of Deucravacitinib Treatment in the Total Treated Population (N=120)Adverse EventPatients, nIncidence, %Any adverse event4134.2Acne2420.0Folliculitis1310.8Upper respiratory tract infection75.8Urticaria21.7Eczema32.5Headache10.8Diarrhea10.8Nausea10.8Oral ulcer10.8Tinea pedis21.7Periodontitis10.8Herpes simplex virus infection10.8
Discussion
This prospective real-world study evaluated deucravacitinib 6 mg once daily for 12 weeks in Chinese patients with moderate-to-severe plaque psoriasis. The week 12 PASI 75 response rate in our cohort was broadly consistent with the week 16 rates reported in the global POETYK PSO-1 trial (58.4%)12 and the Asian POETYK PSO-3 trial (62%).15 Improvements were observed across all anatomical regions (head and neck, trunk, and upper and lower limbs) and increased over time. In POETYK PSO-1, the PASI 75 response rate at week 16 was higher with deucravacitinib than with apremilast (58.4% vs 35.1%).12 Network meta-analyses have also provided indirect comparisons with biologic therapies.20,22
Previous studies have shown that deucravacitinib remains effective in patients previously treated with biologics, although response rates in biologic-experienced patients tend to be modestly lower than in biologic-naïve populations.26 Mechanistically, TYK2 inhibition may promote regulatory T cell (Treg) differentiation and function while suppressing Th1 and Th17 cell differentiation, which may contribute to preserved efficacy in treatment-experienced patients.27
Notably, the highest regional response in our study was observed in the head and neck, exceeding the responses in the trunk, upper, and lower limbs. This finding aligns with results from the POETYK SCALP phase 3b/4 program, in which deucravacitinib produced a significantly higher scalp-specific PGA 0/1 response rather than placebo at week 16.28 A 2026 multicenter real-world study similarly reported a high scalp PGA 0/1 clearance rate of 89.8% at week 16 among patients with scalp involvement at baseline.16 These findings supported suggest that the scalp and head and neck may respond favorably to TYK2 inhibition. Because depression is a common comorbidity of psoriasis, previous reports of associations between TYK2-pathway treatment, tryptophan metabolism, and depressive symptoms29 suggest that patients with mood comorbidities may represent a clinically relevant subgroup for further study.
In the analysis of efficacy predictors, baseline BMI < 24 kg/m2 was independently associated with achieving a PASI 90 response (OR, 5.34; 95% CI, 1.28–22.21). Hagino et al retrospectively evaluated 74 Japanese patients treated with deucravacitinib 6 mg/day and found that lower BMI was associated with a week 16 sPGA 0/1 response with a ≥2-point improvement (OR 0.825; 95% CI 0.713–0.955).30 Takahashi et al also reported a lower baseline BMI among patients attaining complete skin clearance at week 52 in a prospective cohort of 77 patients.31 This association may reflect the higher baseline inflammatory burden commonly observed in patients with obesity, which could attenuate the clinical response.
Among the 37 patients with classified responses, baseline TNF-α, IL-1β, IL-23p19, IFN-γ, and G-CSF levels were higher in PASI 75 non-responders. These findings suggest that the baseline concentrations of specific proinflammatory cytokines may be related to therapeutic outcomes. Catlett et al demonstrated treatment-related modulation of IL-23/Th17 and type I interferon pathway biomarkers by deucravacitinib.32 This pharmacodynamic evidence provides biological context, but longitudinal validation using prespecified biomarker analyses is still needed.
Adverse events occurred in 34.2% of the total treated cohort. Acne, folliculitis, and upper respiratory tract infection were the most common events. These event types are consistent with those reported in pooled trial safety analyses.33 No serious adverse events were reported; however, a cohort of 120 patients followed for 12 weeks is insufficient to characterize uncommon or delayed risks.
This study has several limitations. First, it was a single-center investigation restricted to Chinese patients and had a modest sample size, which may have introduced selection bias and limited generalizability. Second, the 12-week follow-up precluded assessment of long-term efficacy and safety.
Third, biomarker analyses were performed in only 43 patients without post-treatment cytokine measurements, limiting causal inference between baseline biomarkers and clinical outcomes. Fourth, concomitant topical therapies were not systematically recorded or standardized, and no washout period was required, which could confound efficacy assessments. Finally, the absence of a comparator group limits causal inference and prevents the assessment of treatment-specific predictors.
Larger multicenter cohorts with longer follow-up are needed in the future.
Conclusions
In this prospective single-center Chinese cohort, deucravacitinib 6 mg once daily was associated with improvement in psoriasis across anatomical sites over 12 weeks. Adverse events were predominantly mild to moderate, and longer follow-up is needed to characterize safety. Baseline BMI <24 kg/m2 and lower pretreatment levels of TNF-α, IL-1β, IL-23p19, IFN-γ, and G-CSF were associated with better treatment responses.