Adalimumab Reduces Epicardial Adipose Tissue in Psoriasis: A Post Hoc Analysis of a Randomized Placebo-Controlled Clinical Trial

Abstract

Background

Psoriasis is associated with increased epicardial adipose tissue (EAT), a metabolically active fat depot in direct contact with the myocardium, and a mediator of coronary artery disease and atrial fibrillation. The effects of psoriasis treatments on EAT are largely unknown. We conducted a post hoc analysis of the Vascular Inflammation in Psoriasis (VIP) trial (NCT01553058, NCT01866592).

Methods

VIP compared, in a 1:1:1 randomized manner, placebo (n=27), adalimumab (n=31), and phototherapy (n=28) from baseline to week 12, with all subjects then receiving adalimumab for 52 weeks. EAT volume was quantified from CT imaging at baseline, week 12, and week 52, with concurrent assessment of clinical variables and metabolic and inflammatory biomarkers.

Results

At baseline (N=86), EAT volume was positively correlated with age (r = 0.51, p= 2.44 × 10⁻), body mass index (r = 0.48, p= 2.01×10⁻⁶), and weight (r = 0.52, p= 3.17 × 10⁻), and with inflammatory biomarkers GlycA (r = 0.27, p= 9.48×10⁻³), C-reactive protein (r = 0.25, p= 9.94×10⁻³), and IL-6 (r = 0.23, p= 0.02). Over 12 weeks, adalimumab treatment reduced EAT volume (mean change −11.9 cm³, p=0.04; baseline mean EAT 183.1 cm³), with no improvement observed in the placebo or phototherapy groups. When compared with the placebo and phototherapy arms combined, adalimumab was associated with a greater reduction in EAT at week 12 (between-group difference −13.9 cm³, p=0.03). The reduction in EAT was not significantly associated with changes in weight or PASI. No additional reduction in EAT was observed after 52 weeks of adalimumab treatment.

Conclusion

Adalimumab was associated with an early reduction in EAT that was not significantly correlated with changes in weight or psoriasis activity. These findings indicate treatment-specific effects on a high-risk cardiac fat depot and suggest that TNF inhibition may influence cardiac adiposity relevant to cardiovascular disease.

Introduction

Obesity is a major contributor to psoriasis morbidity and is causally implicated in the development of psoriasis.^1,2 It is associated with greater disease severity, an increased risk of psoriatic arthritis, reduced response to treatment, and loss of response to biologic therapies over time.^3–5 Obesity also contributes substantially to adverse cardiometabolic health in people with psoriasis.⁶

Patients with psoriasis have increased visceral adipose tissue (VAT), a metabolically active fat depot that surrounds internal organs and contributes to insulin resistance, dyslipidaemia, systemic inflammation, and atherosclerosis.^7,8 More recently, epicardial adipose tissue (EAT), a distinct subtype of VAT, has been identified as an independent risk factor for major adverse cardiovascular events and is increased in individuals with psoriasis.⁹ EAT lies between the myocardium and the visceral pericardium, in direct contact with the coronary vasculature and adjacent cardiac structures, without a separating fascia. EAT is a highly inflammatory depot with dense macrophage infiltration.¹⁰ Through the secretion of pro-inflammatory cytokines, including interleukin-6 (IL-6) and tumour necrosis factor (TNF), EAT is thought to exert paracrine effects on the adjacent myocardium and coronary arteries, thereby contributing to local vascular inflammation, atherosclerosis, and myocardial fibrosis.¹¹

Despite the strong epidemiological and mechanistic links between psoriasis, adiposity, and cardiometabolic disease, clinical trial evidence to date suggests limited benefit of psoriasis treatment on traditional markers of cardiovascular risk.¹² Several systemic therapies used in psoriasis have been associated with increases in overall adiposity or adverse metabolic effects. Tumour necrosis factor (TNF) inhibitors and Janus kinase (JAK) inhibitors have been linked to weight gain, while JAK inhibitors and acitretin have been associated with dyslipidaemia.^13–15 Treatment with apremilast, a phosphodiesterase-4 inhibitor, has also been associated with reductions in visceral and epicardial adiposity at 12 weeks, with effects sustained through 52 weeks; however, this trial did not have a placebo or active comparator control group.^16,17

The increasing recognition of EAT as a clinically relevant inflammatory cardiometabolic fat depot, combined with the scarce data on the impact of psoriasis treatment on this important imaging biomarker of cardiovascular risk, motivated us to conduct a post hoc analysis of data from the Vascular Inflammation in Psoriasis (VIP). VIP was a randomized study comparing adalimumab, phototherapy, and placebo.¹⁸ We hypothesized that systemic treatment of psoriasis with adalimumab would reduce EAT volume.

Methods

The following methods are reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) 2010 statement.¹⁹

Trial Design and Oversight

This study was a post hoc analysis of the VIP trial, a multicenter, randomized, placebo-controlled clinical trial registered at ClinicalTrials. gov (NCT01553058, NCT01866592). The design, conduct, and primary outcomes of the VIP trial have been reported previously.¹⁸ The trial was initiated in July 2012 and completed in October 2016. Briefly, VIP was designed to evaluate the effects of psoriasis treatments on cardiometabolic and inflammatory pathways and aortic vascular inflammation in patients with moderate-to-severe plaque psoriasis.

All participants provided written informed consent. The study protocol and all amendments were approved by the institutional review board at the University of Pennsylvania (VIP/ VIP-E protocol numbers: 814278 and 817552) and, where applicable, by local institutional review boards at participating centers. The trial was conducted in accordance with the Declaration of Helsinki and principles of good clinical practice.

Participants

Participants were eligible for inclusion if they met all of the following criteria at baseline:

• Age 18 years or older

• Diagnosis of chronic plaque psoriasis for at least six months

• Moderate-to-severe psoriasis, defined by:

  • Psoriasis Area and Severity Index (PASI) score of at least 12, and
  • Body surface area (BSA) involvement of at least 10%

• Considered candidates for systemic therapy or phototherapy based on clinical assessment

• Able to undergo FDG-PET/ computed tomography (CT) imaging

Participants were excluded if any of the following were present:

• Use of prohibited psoriasis therapies within protocol-defined washout periods

• Participants with poorly controlled or unstable cardiovascular disease were excluded, including unstable ischaemic heart disease, congestive heart failure, recent cerebrovascular events, and uncontrolled hypertension (systolic blood pressure >180 mmHg or diastolic blood pressure >90 mmHg), in accordance with the parent VIP trial protocol. Active infection or clinically significant inflammatory or infectious condition

• Poorly controlled medical comorbidities that could interfere with study participation

• Contraindications to FDG-PET/ computed tomography imaging

Randomization and Interventions

Participants were randomized in a 1:1:1 ratio to receive adalimumab, NB-UVB, or placebo during the 12-week double-blind randomized phase. Randomization was performed using a computer-generated allocation sequence with allocation concealment.

Adalimumab was administered as a subcutaneous injection according to the approved dosing regimen for plaque psoriasis. Participants assigned to phototherapy were not blinded and received supervised NB-UVB treatment using a standardized protocol based on Fitzpatrick skin type and estimated minimal erythema dose. Participants assigned to placebo received matched injections.

After completion of the 12-week randomized phase, eligible participants entered an open-label extension phase in which all participants received adalimumab and were followed for a total of 52 weeks from initiation of adalimumab (week 52 for those randomized to adalimumab; week 64 for those randomized to placebo or phototherapy and crossed over to adalimumab).

Epicardial Adipose Tissue Quantification

Computed Tomography Acquisition

Non-contrast CT scans suitable for cardiac fat assessment were obtained at baseline, week 12, and week 52 using standardized acquisition protocols. Imaging parameters, including slice thickness, tube voltage, and reconstruction algorithms, were kept consistent within participants across study visits to minimize measurement variability.

Image Analysis and EAT Measurement

EAT volume was quantified using previously published methods for EAT assessment.²⁰ Analyses at each time point included all participants with available EAT measurements at the relevant study visits. Accordingly, analyses of change from baseline to week 12 included 86 participants with baseline and week 12 CT imaging, whereas analyses through week 52 included the 61 participants with complete CT imaging at all three study visits. EAT was defined as adipose tissue located between the outer surface of the myocardium and the visceral layer of the pericardium. EAT was measured from the level of the bifurcation of the pulmonary trunk to the apex of the heart. Segmentation was performed using a predefined attenuation threshold for adipose tissue on CT imaging, ranging from −190 to −30 Hounsfield units. Manual and semi-automated contouring techniques were applied to delineate the pericardial boundary. EAT volume was calculated by summing voxel volumes within the defined pericardial contour and expressed in cubic centimetres.

Image analyses were performed using dedicated image-processing software (OsiriX MD, Pixmeo SARL, Bernex, Switzerland) by three trained readers who were blinded to treatment assignment, clinical data, and imaging time point. For the reproducibility assessment, repeat image analyses were performed with scans presented in randomized order to minimise recall bias 4 weeks after the initial test.

Reproducibility Assessment

EAT measurements in the VIP trial were performed by three independent readers. Scans from the VIP study were randomly allocated and divided equally among the three readers, such that each reader analyzed one third of the imaging dataset.

Inter-reader and intra-reader reproducibility assessments were performed using a separate subset of CT scans from the VIP-A study (NCT03082729). These scans were used exclusively for reproducibility assessment and were not included in the primary analyses of the VIP trial dataset.

Inter-reader reproducibility was assessed by comparing EAT measurements from two readers against measurements performed by a third reader, an attending radiologist with expertise in cardiac imaging. Intra-reader reproducibility was assessed for each reader using repeat measurements of the same scans, with readers blinded to their prior measurements.

Inter-reader reliability of EAT volume measurements was assessed in a subset of 10 scans independently evaluated by three readers (R.R, F.K, and E.R). Intraclass correlation coefficients (ICCs) were calculated using a two-way mixed-effects model with raters specified as fixed effects and scans as random effects. We report the consistency-of-agreement ICC, which evaluates whether raters provide proportionally consistent measurements across scans, permitting systematic differences in absolute values.

The individual consistency-of-agreement ICC was 0.94 (95% CI 0.83–0.98), and the average-measures ICC across three readers was 0.98 (95% CI 0.94–0.99), indicating excellent reliability. The corresponding absolute-agreement ICC was 0.88 (95% CI 0.56–0.97) for individual measurements and 0.96 (95% CI 0.79–0.99) for average measurements. The F-test for ICC > 0 was statistically significant (F(9,18) = 45.88, p < 0.001).

Clinical and Laboratory Assessments

Psoriasis severity was assessed at each study visit using PASI. Anthropometric measurements, including body weight and height, were obtained using standardized procedures, and body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared.

Blood samples were collected at baseline and follow-up visits for assessment of metabolic and inflammatory biomarkers using validated laboratory assays, as described in the parent VIP trial.¹⁸ Laboratory personnel were blinded to treatment assignment and imaging outcomes.

Outcomes

Primary Outcome

The primary outcome of this post hoc analysis was the absolute change in EAT volume from baseline to week 12.

Secondary outcomes

Secondary outcomes included EAT volume at baseline, week 12, and end of study, absolute changes in EAT volume from baseline to week 12 and from baseline to end of study among participants completing all imaging assessments, and between-group differences in EAT change at week 12.

Exploratory Outcomes

Prespecified exploratory outcomes included associations between baseline EAT volume and selected demographic, anthropometric, and inflammatory measures, correlations between changes in EAT volume and concurrent changes in body weight, BMI, and inflammatory biomarkers at week 12.

Statistical Analysis

Analysis Population

All analyses were conducted using a complete-case approach. Participants with available EAT measurements at the relevant time points were included in each analysis, and missing data were not imputed.

Descriptive Analyses

Continuous variables are presented as mean with standard deviation or median with interquartile range, as appropriate based on distribution. Categorical variables are presented as counts and percentages.

Association Analyses

Associations between baseline EAT volume and demographic, clinical, and inflammatory variables were assessed using Spearman rank correlation coefficients. Associations between changes in EAT volume and concurrent changes in body weight, BMI, and inflammatory biomarkers from baseline to week 12 were also examined using Spearman rank correlation coefficients.

Longitudinal Change Analyses

Within-group changes in EAT volume, body weight, and BMI from baseline to week 12 were evaluated using paired statistical tests. Between-group differences in absolute change in EAT volume at week 12 were assessed using linear regression models. An exploratory analysis assessed change in EAT volume from the start of adalimumab treatment to end of study among participants who completed all three imaging assessments.

As a sensitivity analysis, missing end-of-study CT measurements were imputed using multiple imputation by chained equations with predictive mean matching (50 imputations). The imputation model included treatment group, serial EAT measurements, demographic characteristics, and baseline clinical and inflammatory variables. Estimates were pooled using Rubin’s rules.

Statistical Inference

All statistical tests were two-sided, and a p value less than 0.05 was considered statistically significant. No adjustment was made for multiple comparisons, as analyses were exploratory and post hoc in nature. Statistical analyses were performed using Stata (StataNow, version 19.5).

Results

Study Population and Baseline Characteristics

A total of 179 participants were screened for eligibility, of whom 97 were randomized in a 1:1:1 ratio to placebo, adalimumab, or phototherapy. 92 participants (95%) completed the 12-week randomized phase, and 86 had both baseline and week 12 CT scans and were included in the primary analysis (27 randomized to placebo, 31 to adalimumab, and 28 to phototherapy). 81 participants entered the open-label adalimumab extension phase, of whom 61 (71%) completed 52 weeks of adalimumab treatment and had an additional end-of-study scan. The CONSORT flow diagram outlining participant screening, randomization, follow-up, and attrition is available in the primary report of the VIP trial.¹⁸

Baseline demographic and clinical characteristics were similar between treatment groups (Table 1), with standardized mean differences demonstrating good baseline balance (Supplementary Table 1). Participants had moderate to severe psoriasis at baseline, with median PASI scores of 15.0 to 17.6 and median BSA involvement of approximately 17% to 21%. Mean BMI was >30 across all treatment groups. Baseline demographic and clinical characteristics also did not differ meaningfully between participants who completed all three imaging assessments and those who completed baseline and week 12 scans only.

Table 1.

Baseline Demographic and Clinical Characteristics According to Imaging Completion Status

	Participants Completing Baseline and Week 12 Imaging Assessments				Participants Completing Baseline, Week 12 and Week 52 Imaging Assessments
	Placebo	Adalimumab	Phototherapy	Total	Placebo	Adalimumab	Phototherapy	Total
N	27	31	28	86	21	20	20	61
Age Mean (SD)	44.9 (14.4)	44.3 (14.1)	41.0 (13.4)	43.4 (13.9)	47.6 (13.4)	43.2 (14.9)	40.5 (14.9)	43.9 (14.5)
Sex (%)
Female	10 (37.0)	8 (25.8)	7 (25)	25 (29.1)	9 (45)	4 (20)	7 (35)	20 (30)
Male	17 (63.0)	23 (74.2)	21 (75)	61 (70.9)	12 (32)	16 (36)	13 (32)	41 (70)
Pso Duration (Y) Median (IQR)	20 (7–29)	11.5 (3–22)	11.5 (7–17)	13.5 (6–25)	20.5 (8.5–29-5)	10 (4.5–18.5)	10.5 (6–17)	12.5 (6–25)
PsA (%)	1(3.7)	0 (0.0)	0 (0.0)	1(1.1)	1(4.2)	0 (0.0)	0 (0.0)	1(1.5)
BMI Mean (SD)	32.1 (7.8)	31.1 (7.5)	32.6 (9.0)	31.9 (8.1)	32.2 (8.1)	31.4 (7.5)	32.2 (9.3)	31.9 (8.2)
History of cardiovascular disease	2 (7.4)	2 (6.5)	2 (7.1)	6 (7.0)	3 (12.5)	0 (0.0)	2 (9.1)	5 (7.5)
History of diabetes	1 (3.7)	3 (9.7)	0	4 (4.7)	1 (4.2)	1 (4.8)	0 (0.0)	2 (3.0)
History of hypertension	6 (22.2)	6 (19.4)	2 (7.1)	14 (16.3)	6 (25.0)	4 (19.0)	1 (4.6)	11 (16.4)
History of hyperlipidemia	4 (14.8)	5 (16.1)	4 (14.3)	13 (15.1)	4 (16.7)	2 (9.5)	3 (13.6)	9 (13.4)
10-year Framingham Risk Mean (SD)	8.2 (8.9)	8.4 (7.7)	6.1 (7.6)	7.6 (8.1)	9 (8.5)	7.2 (5.6)	5.7 (8.0)	7.6 (8.0)
Baseline BSA % Median (IQR)	20.5 (16–33)	18.3 (14.5–27.1)	17.3 (15–26)	19 (15–30)	17.7 (14.5–31.5)	18.5 (15–25.3)	20.8 (15–28)	18.5 (15–29)
Baseline PASI Median (IQR)	15.0 (13.3–20.6)	17.6 (14.9–22.3)	16.4 (14.5–20.9)	16.7 (13.9–21.3)	14.7 (13.3–20)	17.7 (14.3–21.6)	16.8 (15–20.9)	16 (13.6–20.8)
Previous phototherapy use	10 (37.0)	5 (16.1)	12 (42.9)	27 (31.4)	9 (62.5)	2 (10.0)	10 (45.4)	21 (31.8)
Previous oral systemic use	9 (33.3)	10 (32.3)	9 (32.1)	28 (32.5)	8 (33.3)	7 (35.0)	6 (27.3)	21 (31.8)
Previous biologic use	10 (37.0)	10 (32.3)	8 (28.6)	28 (32.6)	8 (33.3)	5 (25.0)	5 (22.7)	18 (7.3)

Notes: Baseline demographic, clinical, cardiometabolic, and psoriasis-related characteristics stratified by randomized treatment group among participants who completed baseline and week 12 imaging assessments and among those who completed all three imaging assessments (baseline, week 12, and week 52). Continuous variables are presented as mean (standard deviation) or median (interquartile range), as appropriate. Categorical variables are presented as number (percentage).

Abbreviations: BMI, body mass index; BSA, body surface area; IQR, interquartile range; PASI, Psoriasis Area and Severity Index; PsA, psoriatic arthritis; SD, standard deviation.

Baseline EAT is Associated with Age, Adiposity, and Systemic Inflammation

At baseline, higher EAT volume was positively correlated with age (ρ = 0.588, p = 2.60 × 10=⁻⁹), weight (ρ = 0.552, p = 3.64×10⁻⁸), and BMI (ρ = 0.536, p = 1.03×10⁻⁷) (Table 2). Baseline EAT volume was also positively associated with markers of systemic inflammation, including C-reactive protein (CRP) (ρ = 0.223, p = 0.048), GlycA (ρ = 0.302, p = 0.008), and IL-6 (ρ = 0.308, p = 0.006). No significant association was observed between baseline EAT volume and TNF (ρ = 0.080, p = 0.486).

Table 2.

Spearman Rank Correlations Between Baseline Epicardial Adipose Tissue Volume and Demographic, Anthropometric, and Inflammatory Measures

	Baseline EAT	Age	Weight (kg)	BMI	CRP	TNF	GlycA	Log IL-6
Baseline EAT		0.588 (p=2.60×10−9)*	0.552 (p=3.64×10−8)*	0.536 (p=1.03×10−7)*	0.223 (p=0.04)*	0.080 (p=0.486)	0.302 (p=0.008)*	0.308 (p=0.006)*
Age	0.588 (p=2.60×10−9)*		0.155 (p=0.153)	0.169 (p=0.120)	0.226 (p=0.045)*	0.101 (p=0.379)	0.171 (p=0.140)	0.288 (p=0.010)*
Weight (kg)	0.552 (p=3.64×10−8)*	0.155 (p=0.153)		0.885 (p=2×10−29)*	0.208 (p=0.066)	0.047 (p=0.680)	0.193 (p=0.095)	0.107 (p=0.350)
BMI	0.536 (p=1.03×10−7)*	0.169 (p=0.120)	0.885 (p=2×10−29)*		0.314 (p=0.005)*	0.121 (p=0.291)	0.287 (p=0.012)*	0.240 (p=0.034)*

Notes: Spearman rank correlation coefficients (ρ) between baseline epicardial adipose tissue (EAT) volume and selected demographic, anthropometric, and inflammatory measures among participants with baseline and week 12 CT imaging. Correlations were calculated using pairwise complete observations. Asterisks denote statistically significant correlations (p<0.05).

Abbreviations: BMI, body mass index; CRP, C-reactive protein; EAT, epicardial adipose tissue; GlycA, glycoprotein acetylation; IL-6, interleukin-6; TNF, tumour necrosis factor.

Adalimumab Reduces EAT Volume at Week 12

Among the 86 participants with complete baseline and week 12 imaging assessments, mean EAT volume decreased from baseline to week 12 in the adalimumab group, whereas no significant change was observed in the placebo or phototherapy groups (Table 3). The mean change in EAT volume in the adalimumab group was −11.9 cm³ (95% CI, −23.2 to −0.6 cm³; p = 0.04). In contrast, within-group changes in EAT volume were not statistically significant in the placebo or phototherapy groups. When the placebo and phototherapy arms were combined, adalimumab was associated with a greater reduction in EAT at week 12 (between-group difference, −13.9 cm³; 95% CI, −26.4 to −1.4 cm³; p = 0.03). Exploratory pairwise comparisons of adalimumab versus placebo and versus narrowband ultraviolet B phototherapy are shown in Supplementary Table 2.

Table 3.

Change in Epicardial Adipose Tissue (EAT) Volume from Baseline to Week 12 Among Participants with Complete Baseline and Week 12 Imaging Assessments

Outcome	Placebo (n = 27)	Phototherapy (n = 28)	Adalimumab (n = 31)	Placebo + Phototherapy (n = 55)	p value^
Baseline EAT, cm3	172.3 (82.7)	167.2 (97.8)	183.1 (109.4)	169.7 (89.9)	—
Week 12 EAT, cm3	174.5 (87.8)	169.1 (93.7)	171.2 (100.6)	171.7 (90.1)	—
Absolute change, cm3	+2.1 (4.8)	+1.9 (4.8)	−11.9 (5.6)	+2.0 (3.4)	0.026*
Within-group p value	0.657	0.694	0.043*	—	—

Notes: Values are mean (SD) for baseline and week 12 EAT volume and mean (SE) for absolute change. Within-group p values represent paired comparisons of change from baseline within each randomized treatment arm. The between-group p value represents the comparison of absolute change in EAT volume between adalimumab and the combined placebo and phototherapy groups using linear regression.

Abbreviations: EAT, epicardial adipose tissue; SD, standard deviation; SE, standard error. Significance * at 0.05. ^ Comparison of adalimumab versus combined placebo/phototherapy.

Within the adalimumab group, reductions in EAT volume were not significantly correlated with concurrent changes in weight (r = 0.34, p = 0.07), BMI (r = 0.32, p = 0.08), CRP (r = 0.37, p = 0.054), PASI (r = −0.25, p = 0.16), TNF (r = −0.19, p = 0.33), GlycA (r = 0.24, p = 0.24), or IL-6 (r = 0.24, p = 0.22).

Early Reductions in EAT are Attenuated Over Longer Follow-up

Among the 61 participants who completed all three imaging assessments, the reduction in EAT volume observed at week 12 in the adalimumab group was attenuated by the end of study (Figure 1). Similarly, no significant change in EAT volume was observed from the start of adalimumab treatment to the end of study (n = 61; mean change, −3.0 cm³; 95% CI, −10.2 to 4.2 cm³; p = 0.41).-

Figure 1.

Longitudinal changes in epicardial adipose tissue (EAT) volume by treatment group. Mean EAT volume at baseline, week 12, and end of study by randomized treatment group. Analyses from baseline to week 12 included the 86 participants with baseline and week 12 CT imaging, whereas longitudinal analyses through end of study included the 61 participants with complete CT imaging at all three study visits. During the randomized phase (baseline to week 12), participants received placebo, Adalimumab, or phototherapy; from week 12 onward, all participants received open-label adalimumab. The dashed line represents the overall mean across treatment groups.

Abbreviations: EAT, epicardial adipose tissue.

Baseline characteristics of participants with complete and missing end-of-study CT imaging (Supplementary Table 3). Sensitivity analysis using multiple imputation produced similar results (Supplementary Table 4).

Discussion

In this post hoc analysis of a randomized, placebo-controlled clinical trial, adalimumab was associated with an early reduction in EAT volume over 12 weeks compared with placebo and phototherapy. Changes in EAT volume showed little evidence of an association with concurrent changes in weight or PASI and were not observed in the comparator arms. No further reduction in EAT was seen with longer-term adalimumab treatment during the open-label extension phase. Collectively, these findings suggest a treatment-specific effect of TNF inhibition on a high-risk cardiac fat depot in psoriasis. The findings are especially novel as the impact of adalimumab on EAT has not been previously described and this is the first randomized controlled trial to demonstrate benefit of a psoriasis treatment on an imaging marker of cardiovascular risk.

Several important observations emerge from these data. At baseline, EAT volume demonstrated strong associations with age and measures of adiposity, including weight and BMI, consistent with established cardiometabolic paradigms.²¹ EAT was also positively correlated with systemic inflammatory biomarkers, including CRP, GlycA, and IL-6. These findings reinforce the concept that EAT in psoriasis represents an inflammatory adipose tissue compartment that may contribute to heightened cardiovascular risk, rather than serving solely as a marker of generalized adiposity.^9,11

The key observation is that adalimumab was associated with a statistically significant reduction in EAT volume over 12 weeks, both within-group and compared with placebo and phototherapy during the randomized phase. This finding builds on a previous cross-sectional study of patients with rheumatoid arthritis, in which treatment with TNF inhibitors was associated with reduced EAT thickness compared with non-biologic disease-modifying therapies.²² The present randomized analysis therefore provides prospective, randomized, controlled evidence that targeted TNF blockade is associated with reduced EAT volume over a defined treatment interval, supporting the hypothesis that TNF signalling contributes to regulation of this inflammatory adipose tissue compartment.

Short-term changes in EAT were not consistently correlated with concurrent changes in weight or circulating inflammatory markers within treatment arms. In the parent VIP analyses, adalimumab was associated with significant reductions in CRP, IL-6, and GlycA, reflecting systemic anti-inflammatory effects.¹⁸ However, in the present analysis, the reduction in EAT volume observed with adalimumab was not significantly associated with changes in these biomarkers. These findings indicate that although adalimumab improves systemic inflammatory markers, the observed reduction in EAT does not appear to be directly mediated by short-term changes in weight or circulating inflammation. This contrasts with findings from the VIP-A study, in which reductions in EAT were accompanied by reductions in weight, suggesting distinct mechanistic pathways across therapies.¹⁷ Given the anatomic proximity of EAT to the coronary vasculature and myocardium and its capacity for paracrine cytokine signalling, selective modulation of this depot may represent a local effect of TNF inhibition that is not fully captured by peripheral inflammatory measures.²³ Although a clinically meaningful threshold for EAT reduction has not been established, the magnitude of EAT reduction observed with adalimumab is comparable to that reported with apremilast and with SGLT2 inhibitors.^17,24 While SGLT2 inhibitors are associated with reductions in major cardiovascular events, whether EAT reduction itself mediates these benefits, or whether a similar reduction achieved through other therapies confers comparable cardiovascular benefit, remains uncertain.

The observation that adalimumab reduced EAT despite minimal weight changes is particularly notable in light of prior observations that TNF inhibitor treatment is associated with weight gain.¹⁵ These findings indicate that systemic TNF inhibition may differentially influence regional adipose depots, potentially reducing inflammatory activity within epicardial fat even in the absence of reductions in overall adiposity. This distinction refines current understanding of the cardiometabolic effects of biologic therapies in psoriasis and underscores the importance of compartment-specific phenotyping rather than reliance on global measures such as weight or BMI alone.

With longer follow-up, the early reduction in EAT observed at week 12 was not sustained through end of study. Interpretation of this finding requires consideration of both biological and methodological factors. First, variability in the measurement of EAT, and patient dropout during the extension phase of the study may be impacted our ability to detect EAT changes over time. It is also possible that early anti-inflammatory effects within epicardial fat plateau with continued therapy or that adaptive remodelling processes mitigate initial changes. However, EAT thickness previously showed early and rapid reduction with drugs modulating the adipose tissue.²⁵ These considerations highlight the need for prospectively designed studies with adequate power to evaluate longer-term trajectories of EAT under sustained cytokine inhibition, together with mechanistic studies to elucidate underlying pathways.

In the parent VIP trial, adalimumab did not reduce aortic vascular inflammation,¹⁸ whereas phototherapy was associated with a within-group reduction in vascular inflammation. In contrast, the present analysis demonstrates a reduction in EAT volume with adalimumab but not with phototherapy. This discordance suggests that distinct cardiometabolic compartments in psoriasis may respond differentially to therapeutic interventions. Vascular inflammation and epicardial adiposity likely represent related but biologically non-identical inflammatory processes. Phototherapy may exert favorable effects on vascular inflammation through pathways linked to improvement in cutaneous disease and systemic immune modulation, whereas TNF inhibition may preferentially influence inflammatory adipose tissue depots such as EAT. Taken together, these findings reinforce the concept that cardiometabolic risk in psoriasis is multidimensional and that therapeutic effects may vary across vascular, adipose, and systemic inflammatory domains. Comprehensive phenotyping of these compartments may therefore be necessary to fully characterize the cardiovascular implications of targeted immune therapies.

This study has several strengths. The randomized design of the parent trial, blinded imaging assessment, rigorous quantification of EAT, and reproducibility testing enhance internal validity. The availability of imaging at multiple time points permitted evaluation of both short-term and longer-term changes. Additionally, detailed phenotyping enabled assessment of associations with anthropometric and inflammatory variables.

As with all research, there are limitations to consider. This was a post hoc analysis of a RCT and EAT was not a prespecified primary outcome, hence no formal sample size calculation was performed for the EAT endpoint. As no adjustment was made for multiple comparisons, the possibility of type I (alpha) error should also be considered when interpreting these exploratory findings. Imaging of EAT represents a surrogate biomarker rather than a clinical cardiovascular endpoint. Participants with established cardiovascular disease were excluded, potentially limiting generalizability to higher-risk populations. Additional studies are therefore needed to confirm and extend our findings.

The clinical implications of EAT reduction in psoriasis remain to be established. EAT has been independently associated with coronary atherosclerosis, atrial fibrillation, and adverse cardiovascular events.^26,27 Whether short-term modulation of this depot translates into durable changes in cardiovascular risk is uncertain. However, the present findings provide evidence that targeted cytokine inhibition can modify a high-risk inflammatory adipose tissue compartment that lies in direct anatomic and functional proximity to the coronary vasculature. These observations extend the cardiometabolic framework of psoriasis beyond traditional risk factors and circulating biomarkers, highlighting the potential relevance of tissue-level inflammatory remodelling.

Conclusion

Adalimumab was associated with an early reduction in EAT volume in patients with moderate-to-severe psoriasis, that was not associated with concurrent changes in weight or PASI. These data provide prospective randomized evidence that TNF inhibition may modulate a biologically relevant cardiac fat depot. Incorporation of advanced imaging biomarkers into psoriasis trials could therefore enhance understanding of the systemic effects of immune-targeted therapies and help clarify their broader cardiometabolic impact.

Funding Statement

This study was supported by grants (NHLBI R01-HL111293, K24-AR-064310) and by an unrestricted grant from AbbVie to the Trustees of the University of Pennsylvania). This work was funded by the International Psoriasis Council, National Psoriasis Foundation, and the Center for Clinical Sciences in Dermatology and the University of Pennsylvania Skin Biology and Diseases Resource-based Center. The above funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Abbreviations

BMI, body mass index; BSA, body surface area; CONSORT, Consolidated Standards of Reporting Trials; CRP, C-reactive protein; CT computed tomography; EAT, epicardial adipose tissue; FDG-PET/CT, fluorodeoxyglucose positron emission tomography/computed tomography; GlycA, glycoprotein acetylation; ICC, intraclass correlation coefficient; IL-6, interleukin-6; JAK, Janus kinase; NB-UVB, narrowband ultraviolet B; PASI, Psoriasis Area and Severity Index; PsA, psoriatic arthritis; TNF, tumour necrosis factor; VAT, visceral adipose tissue; VIP, Vascular Inflammation in Psoriasis.

Data Sharing Statement

All study protocols, data, and other materials are available by request made to the corresponding author.

Disclosure

Mrs Faradia Kernizan reports grants or contracts from Ortho Dermatologics, Johnson and Johnson, outside the submitted work. Dr Nehal Mehta reports other interests from Amgen, Lilly, Medscape, WebMD, outside the submitted work. Dr Daniel Shin reports data safety/advisory board participation from DSMB for EVOLUTION, outside the submitted work. Dr Joel Gelfand has served as a consultant for AbbVie, Artax (DSMB), Bristol Myers Squibb, Boehringer Ingelheim, Celldex (DSMB), FIDE (which is sponsored by multiple pharmaceutical companies) GSK, Inmagene (DSMB), Lilly, Leo, Moonlake (DSMB), Janssen Biologics, Novartis Corp, UCB (DSMB), Neuroderm (DSMB), Oruka, Inc, Teva (DSMB; and receives research grants (to the Trustees of the University of Pennsylvania) from Amgen, Bristol Myers Squibb, and Pfizer Inc.; and received payment for continuing medical education work related to psoriasis that was supported indirectly pharmaceutical sponsors. Dr Joel Gelfand is a Deputy Editor for the Journal of Investigative Dermatology receiving honoraria from the Society for Investigative Dermatology, is Chief Medical Editor for Healio Dermatology (receiving honoraria) and is a member of the Board of Directors for the International Psoriasis Council and the Medical Dermatology Society, receiving no honoraria. The abstract of this paper was presented at the SID annual meeting 2026 in Chicago, United States, as a conference talk with interim findings (abstract ID 0497). The presentation’s abstract was published in the Journal of Investigative Dermatology in August 2026 (https://www.jidonline.org/article/S0022-202X(26)01808-7/fulltext).