Abatacept in the treatment of patients with psoriatic arthritis: Results of a six-month, multicenter, randomized, double-blind, placebo-controlled, phase II trial

Authors

  • Philip Mease,

    Corresponding author
    1. Swedish Medical Center and University of Washington, Seattle, Washington
    • Seattle Rheumatology Associates, 1101 Madison Street, Suite 1000, Seattle, WA 98104
    Search for more papers by this author
    • Drs. Mease, Genovese, Gladstein, and Kivitz have received consulting fees, speaking fees, and/or honoraria from Bristol-Myers Squibb (less than $10,000), and Dr. Mease has also received a research grant from Bristol-Myers Squibb.

  • Mark C. Genovese,

    1. Stanford University, Palo Alto, California
    Search for more papers by this author
    • Drs. Mease, Genovese, Gladstein, and Kivitz have received consulting fees, speaking fees, and/or honoraria from Bristol-Myers Squibb (less than $10,000), and Dr. Mease has also received a research grant from Bristol-Myers Squibb.

  • Geoffrey Gladstein,

    1. New England Research Associates, Trumbull, Connecticut
    Search for more papers by this author
    • Drs. Mease, Genovese, Gladstein, and Kivitz have received consulting fees, speaking fees, and/or honoraria from Bristol-Myers Squibb (less than $10,000), and Dr. Mease has also received a research grant from Bristol-Myers Squibb.

  • Alan J. Kivitz,

    1. Altoona Center for Clinical Research, Duncansville, Pennsylvania
    Search for more papers by this author
    • Drs. Mease, Genovese, Gladstein, and Kivitz have received consulting fees, speaking fees, and/or honoraria from Bristol-Myers Squibb (less than $10,000), and Dr. Mease has also received a research grant from Bristol-Myers Squibb.

  • Christopher Ritchlin,

    1. University of Rochester, Rochester, New York
    Search for more papers by this author
  • Paul P. Tak,

    1. Academic Medical Center and University of Amsterdam, Amsterdam, The Netherlands
    Search for more papers by this author
    • Dr. Tak has received consulting fees from Bristol-Myers Squibb (less than $10,000).

  • Jürgen Wollenhaupt,

    1. Schoen Klinik, Hamburg, Germany
    Search for more papers by this author
    • Dr. Wollenhaupt has received honoraria from Bristol-Myers Squibb (less than $10,000).

  • Orna Bahary,

    1. Bristol-Myers Squibb, Princeton, New Jersey
    Search for more papers by this author
    • Ms Bahary and Drs. Becker, Kelly, Sigal, and Teng own stock or stock options in Bristol-Myers Squibb, and Dr. Becker holds a patent for abatacept, which is used in the clinical treatment of autoimmune diseases.

  • Jean-Claude Becker,

    1. Bristol-Myers Squibb, Princeton, New Jersey
    Search for more papers by this author
    • Ms Bahary and Drs. Becker, Kelly, Sigal, and Teng own stock or stock options in Bristol-Myers Squibb, and Dr. Becker holds a patent for abatacept, which is used in the clinical treatment of autoimmune diseases.

  • Sheila Kelly,

    1. Bristol-Myers Squibb, Princeton, New Jersey
    Search for more papers by this author
    • Ms Bahary and Drs. Becker, Kelly, Sigal, and Teng own stock or stock options in Bristol-Myers Squibb, and Dr. Becker holds a patent for abatacept, which is used in the clinical treatment of autoimmune diseases.

  • Leonard Sigal,

    1. Bristol-Myers Squibb, Princeton, New Jersey
    Current affiliation:
    1. University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School, New Brunswick
    Search for more papers by this author
    • Ms Bahary and Drs. Becker, Kelly, Sigal, and Teng own stock or stock options in Bristol-Myers Squibb, and Dr. Becker holds a patent for abatacept, which is used in the clinical treatment of autoimmune diseases.

  • Julie Teng,

    1. Bristol-Myers Squibb, Princeton, New Jersey
    Search for more papers by this author
    • Ms Bahary and Drs. Becker, Kelly, Sigal, and Teng own stock or stock options in Bristol-Myers Squibb, and Dr. Becker holds a patent for abatacept, which is used in the clinical treatment of autoimmune diseases.

  • Dafna Gladman

    1. University of Toronto and Toronto Western Hospital, Toronto, Ontario, Canada
    Search for more papers by this author
    • Dr. Gladman has received consulting fees, speaking fees, and/or honoraria from Abbott, Amgen, Pfizer, and Bristol-Myers Squibb (less than $10,000 each) and has received grants from Bristol-Myers Squibb.


Abstract

Objective

To assess the safety and efficacy of abatacept, a selective T cell costimulation modulator, in patients with psoriatic arthritis (PsA).

Methods

In this 6-month, multicenter, randomized, double-blind, placebo-controlled, phase II study, 170 PsA patients with a psoriasis target lesion (TL) ≥2 cm who had previously taken disease-modifying antirheumatic drugs (DMARDs), including anti–tumor necrosis factor (anti-TNF) agents, were randomized (1:1:1:1) to receive placebo or abatacept at doses of 3 mg/kg, 10 mg/kg, or 30/10 mg/kg (2 initial doses of 30 mg/kg, followed by 10 mg/kg) on days 1, 15, and 29 and then once every 28 days thereafter. The primary end point was the American College of Rheumatology 20% criteria for improvement (ACR20 response) on day 169. Other key end points were magnetic resonance imaging (MRI) scores for joint erosion, osteitis, and synovitis, scores on the Health Assessment Questionnaire (HAQ) and the Short Form-36 (SF-36) health survey, the investigator's global assessment of psoriasis, the TL score, and the Psoriasis Area and Severity Index (PASI) score.

Results

Proportions of patients achieving an ACR20 response were 19%, 33%, 48%, and 42% in the placebo, the abatacept 3 mg/kg, the abatacept 10 mg/kg, and the abatacept 30/10 mg/kg groups, respectively. Compared with placebo, improvements were significantly higher for the abatacept 10 mg/kg (P = 0.006) and 30/10 mg/kg (P = 0.022) groups, but not for 3 mg/kg group (P = 0.121). All abatacept regimens resulted in improved MRI, HAQ, and SF-36 scores, with 10 mg/kg showing the greatest improvements. Improvements in the TL and PASI scores were observed in all abatacept arms; a response according to the investigator's global assessment was seen only with 3 mg/kg of abatacept. The safety profiles were similar among the treatment arms.

Conclusion

The results of this study suggest that 10 mg/kg of abatacept, the approved dosage for rheumatoid arthritis and juvenile idiopathic arthritis, may be an effective treatment option for PsA.

Psoriatic arthritis (PsA) is a chronic inflammatory disease characterized by both arthritis and psoriasis that can cause significant impairment of patients' physical function and quality of life (1, 2). Up to 30% of patients with psoriasis develop PsA, with the skin lesions typically preceding the joint inflammation by several years (3–6). Nonbiologic disease-modifying antirheumatic drugs (DMARDs), methotrexate (MTX) in particular, are the first-choice treatment of PsA, despite disappointing efficacy in clinical trials (7). Recent studies have demonstrated that tumor necrosis factor (TNF) inhibitors are efficacious against both the psoriatic and the arthritic components of PsA (8–16).

T cells have been shown to play a central role in the pathogenesis of PsA (17). Activated T cells bearing markers of antigen activation are abundant in the inflamed joint and skin (18). Alefacept, a fusion protein that inhibits T cell activation by blocking the interaction of lymphocyte function-associated antigen with CD2, has been shown to improve the psoriatic skin lesions and the signs and symptoms of arthritis in patients with active PsA (19–21). The beneficial effect of TNF blockade in PsA is associated with a marked decrease in synovial T cell infiltration (22). Thus, inhibition of T cell activation may lead to clinical benefits in patients with PsA. Full antigen-induced activation of naive T cells requires 2 discrete signals from the antigen-presenting cell (23). Antigen is presented to the T cell receptor in the context of a major histocompatibility complex molecule, but full activation occurs only when the binding of CD80 or CD86 to the CD28 molecule on the T cell produces a second costimulatory signal. After activation, the T cell expresses CTLA-4, which competes with CD28 binding to CD80 or CD86, leading to homeostatic down-modulation of activated T cells.

Abatacept is a soluble, fully human fusion protein consisting of the extracellular domain of CTLA-4 linked to a modified Fc portion of human IgG1. Abatacept selectively inhibits T cell activation via competitive binding to CD80 or CD86 and decreases serum levels of cytokines and inflammatory proteins implicated in the pathogenesis of PsA (24). Abatacept is an approved treatment for chronic inflammatory conditions such as rheumatoid arthritis (RA) and juvenile idiopathic arthritis (JIA), in which T cells are involved in the pathophysiology of the disease (25–30). In a phase I study of abatacept in patients with psoriasis vulgaris, clinical improvement was associated with a reduction of the intralesional T cell population (31, 32). Taken together, these observations suggest that abatacept could be effective in the treatment of PsA. In the present study, we examined the efficacy and safety of 3 different dosing regimens of abatacept compared to placebo in patients with PsA.

PATIENTS AND METHODS

Patient population.

Adult patients who met the criteria of the Classification of Psoriatic Arthritis (CASPAR) Study Group (33) and had active arthritis (defined as the presence of ≥3 swollen joints and ≥3 tender joints), active plaque psoriasis (with at least 1 qualifying target lesion [TL] ≥2 cm in diameter), and a disease duration of ≥3 months were eligible for enrollment in the study. Patients were required to have had an inadequate response to DMARDs, including, but not limited to, MTX or anti-TNF agents. Response to MTX was considered inadequate if it had been taken at a dosage of ≥15 mg/week for ≥2 months prior to randomization. Patients with intolerance of, or an inadequate response to, infliximab, adalimumab, or etanercept discontinued these anti-TNF therapies at screening, and following a washout period of ≥28 days, these patients were assessed for arthritis and psoriasis before randomization.

Key exclusion criteria were as follows: use of any investigational drug within 28 days before randomization, any prior treatment with abatacept, evidence of latent or active tuberculosis, or evidence of chronic or clinically significant infection or malignancy. Women who were pregnant or lactating were also excluded.

Purified protein derivative skin tests were performed on all patients. Those with positive results were treated prophylactically prior to receiving the study drug.

Aside from MTX, no DMARD was continued during the 6-month double-blind treatment period. MTX was continued at a stable dosage only if it had been taken at a stable dosage for ≥3 months prior to screening. A decrease in the MTX dosage was allowed in cases of toxicity. The dosage of nonsteroidal antiinflammatory drug (NSAID) remained unchanged throughout the study unless a decrease in dosage was required because of toxicity. Concomitant corticosteroid treatment was allowed if the dosage (no more than 10 mg of prednisone or its equivalent) had been stable for ≥28 days.

Study protocol and treatment.

This was a 6-month, multicenter, randomized, double-blind, multiple dose–level, placebo-controlled, phase II study. Patients were stratified according to the percentage of body surface area affected by psoriasis at baseline (>3% or <3%). The study compared 3 abatacept dosing regimens to placebo, with a 1:1:1:1 randomization ratio. Treatments were administered as 30-minute intravenous infusions on days 1, 15, and 29, and every 28 days thereafter. The 3 abatacept dosing regimens were as follows: 3 mg/kg (calculated based on body weight at screening); 10 mg/kg (currently approved weight-tiered dosing for RA is 500 mg for patients weighing <60 kg, 750 mg for those weighing 60–100 kg, and 1,000 mg for those weighing >100 kg); and 30/10 mg/kg dose (30 mg/kg [calculated based on body weight at screening] given on days 1 and 15, followed by the weight-tiered dose of 10 mg/kg). Patients across all treatment arms who completed the 6-month double-blind period were given the weight-tiered dose of 10 mg/kg, administered monthly starting on day 169, for the duration of the 18-month open-label period.

The protocol was approved by the Institutional Review Board or the Ethics Committee at each participating center. All patients gave written informed consent before enrollment.

Efficacy assessments.

The primary end point was the American College of Rheumatology 20 (ACR20) response on day 169. The ACR20 response was defined as an improvement of 20% or more in the ACR core criteria (34).

Key secondary end points included the investigator's global assessment of psoriasis and the scores for TLs, the disability index (DI) of the Health Assessment Questionnaire (HAQ) (35), and the Short Form 36 (SF-36) health survey (36). The investigator's global assessment was based on a static evaluation by the investigator of the degree of psoriasis, with ratings ranging from 1 (clear) to 4 (severe) (37). The investigator's global assessment response was defined as investigator's rating of psoriatic lesions as “clear or almost clear.” The TL score was based on the rating of skin erythema, induration, and scaling (scale of 0–4 for each) and was assessed in all patients irrespective of the body surface area affected by psoriasis at baseline. Improvement in physical function was based on a decrease in the HAQ DI scores from baseline; a decrease of at least 0.3 was considered the minimum clinically important difference (MCID) (38). Increases in the SF-36 scores from baseline were used to assess improvement in health-related quality of life. Scores were aggregated to calculate the physical component summary (PCS) score (includes 4 physical subscales: physical function, physical role, bodily pain, and general health) and the mental component summary (MCS) score (includes 4 mental subscales: vitality, social function, emotional role, and mental health).

Exploratory end points were the ACR50 or ACR70 response (a 50% or 70% improvement in the ACR criteria), the Psoriasis Area and Severity Index (PASI), and the assessment of joint damage by magnetic resonance imaging (MRI). The PASI, a composite score ranging from 0 to 72, is based on the grading of the area of skin involved, erythema, induration, and scaling (39). The PASI was assessed in patients with ≥3% of the body surface area affected by psoriasis at baseline. MRI images of arthritic joints in a single hand or single foot obtained at baseline and on days 85 and 169 were digitized. Digitized images were read and graded by an Independent Review Committee consisting of 2 certified radiologists who were blinded with regard to the treatment group and the chronological sequence of the images. Images were graded for bone erosion (0–10 scale), bone marrow osteitis (0–3 scale), and synovitis (0–3 scale), using a modified version (40) of the Rheumatoid Arthritis Magnetic Resonance Imaging Scoring (RAMRIS) system (41).

Safety assessments.

Safety evaluations included monitoring of adverse events and routine laboratory tests. These evaluations were performed on patients who received at least 1 treatment infusion.

Statistical analysis.

All randomized patients who received at least 1 dose of study drug were included in the data analysis according to the intention-to-treat principle. Assuming an ACR20 response rate of 20% with placebo treatment, each arm was estimated to require 41 patients, yielding 92% power to detect a 35% difference in the ACR20 response rate between the 30/10 mg/kg abatacept arm and the placebo arm, and at least 84% power to detect the same difference in ACR20 response rate between the 10 mg/kg abatacept arm and the placebo arm, using the sequential test procedure. The proportions of patients achieving a response according to the ACR20, ACR50, ACR70, investigator's global assessment, TL50 (≥50% improvement in the TL score), TL75 (≥75% improvement in the TL score), PASI50 (≥50% improvement in the PASI score), or PASI75 (≥75% improvement in the PASI score) criteria, along with their 95% confidence intervals (95% CIs), were summarized according to the treatment group. A Cochran-Mantel-Haenszel test (adjusted for a baseline body surface area of ≥3% or <3%) was used to compare the ACR20 response in each abatacept arm with that in the placebo arm. No statistical comparisons were performed for the other response assessments. For the determination of responses according to the ACR, PASI, and investigator's global assessment criteria, all patients who discontinued prematurely or had missing data were considered nonresponders and the data were imputed. A last observation carried forward analysis was performed for the TL, HAQ DI, and SF-36 on day 169.

RESULTS

Characteristics of the study patients.

A total of 170 patients were randomized. Baseline patient demographic and clinical characteristics were comparable across the 4 treatment arms (Table 1), except for the proportion of patients previously exposed to anti-TNF agents, which was greater in the 30/10 mg/kg arm than in the other arms (51% versus 29–36%). Patients had active PsA at baseline, as evidenced by the tender and swollen joint counts, the extent of body surface area affected by psoriasis, the PASI, TL, and HAQ DI scores, and the serum C-reactive protein levels. Nearly three-fourths of the patients had previously taken MTX. MTX was continued in ∼60% of patients. Approximately 20% of patients were receiving concomitant corticosteroids.

Table 1. Baseline demographic and clinical characteristics of the PsA patients, by treatment group*
 Patients taking abataceptPatients taking placebo (n = 42)
30/10 mg/kg (n = 43)10 mg/kg (n = 40)3 mg/kg (n = 45)
  • *

    PsA = psoriatic arthritis; VAS = visual analog scale (0–100 mm); HAQ = Health Assessment Questionnaire; DI = disability index; BSA = body surface area; PASI = Psoriasis Area and Severity Index; NSAIDs = nonsteroidal antiinflammatory drugs; anti-TNF = anti–tumor necrosis factor; DMARDs = disease-modifying antirheumatic drugs.

  • Two initial doses of 30 mg/kg followed by 10 mg/kg.

Demographic characteristics    
 Age, mean ± SD years51.5 ± 9.850.8 ± 10.550.3 ± 9.952.6 ± 12.0
 % female54355145
 % Caucasian100959898
 PsA duration, mean ± SD years7.8 ± 7.710.6 ± 9.47.2 ± 7.47.4 ± 8.0
Arthritis characteristics, mean ± SD    
 No. of tender joints19.6 ± 11.425.2 ± 15.622.7 ± 14.621.3 ± 15.3
 No. of swollen joints10.3 ± 7.112.5 ± 8.710.3 ± 6.910.5 ± 7.9
 C-reactive protein, mg/dl2.2 ± 5.21.5 ± 1.81.8 ± 2.41.3 ± 2.7
 Physician's global assessment, by VAS, mm56.7 ± 19.457.3 ± 19.854.7 ± 17.752.6 ± 18.6
 Patient's global assessment, by VAS, mm55.6 ± 21.560.8 ± 22.959.5 ± 22.158.6 ± 26.5
 Patient's assessment of pain, by VAS, mm58.0 ± 20.867.5 ± 20.658.2 ± 27.462.1 ± 25.8
 HAQ DI (range 0–3)1.2 ± 0.81.3 ± 0.71.1 ± 0.71.2 ± 0.7
Psoriasis characteristics    
 No. with psoriasis covering ≥3% of BSA20212121
 Investigator's global assessment of psoriasis, mean ± SD score (range 0-4)2.6 ± 0.82.6 ± 0.72.4 ± 0.82.5 ± 0.8
 Target lesions, mean ± SD score (range 0–12)6.6 ± 2.86.0 ± 2.45.5 ± 2.76.4 ± 2.7
 PASI score, mean ± SD (range 0–72)16.3 ± 17.89.4 ± 9.211.7 ± 8.413.1 ± 10.5
Previous medications, % of patients    
 Any1001009893
 Methotrexate72858269
 Corticosteroids23333121
 NSAIDs58687355
 Anti-TNF therapy51333629
 Adalimumab26102010
 Etanercept30152224
 Infliximab161547
Concomitant medications, % of patients    
 Any93909179
 Methotrexate58606055
 Corticosteroids21282719
 Other DMARDs2375

Of the 170 patients who were randomized, 147 patients (86%) completed the 6-month double-blind period (Table 2). Fourteen percent of patients discontinued treatment during this period, with adverse event (4%) and lack of efficacy (6%) being the major causes.

Table 2. Disposition of the study patients after 6 months of treatment, by treatment group
 No. (%) taking abataceptNo. (%) taking placebo (n = 42)
30/10 mg/kg (n = 43)*10 mg/kg (n = 40)3 mg/kg (n = 45)
  • *

    Two initial doses of 30 mg/kg followed by 10 mg/kg.

Discontinued treatment6 (14)6 (15)2 (4)9 (21)
 Adverse event1 (2)2 (5)1 (2)3 (7)
 Lack of efficacy3 (7)4 (10)03 (7)
 Withdrew consent2 (5)002 (5)
 Became pregnant001 (2)0
 Other0001 (2)
Completed treatment37 (86)34 (85)43 (96)33 (79)

Arthritis responses.

All 3 abatacept dosing regimens resulted in improved ACR responses as compared to placebo, with the 10 mg/kg regimen showing the greatest improvement (Figure 1). On day 169, the proportions of patients achieving an ACR20 response were 42% of those taking 30/10 mg/kg of abatacept, 48% taking 10 mg/kg, 33% taking 3 mg/kg, and 19% taking placebo (Figures 1A and B). Compared to placebo, the ACR20 responses were statistically significant for the 30/10 mg/kg (P = 0.022) and the 10 mg/kg (P = 0.006), but not the 3 mg/kg (P = 0.121), doses of abatacept.

Figure 1.

Effects of abatacept treatment on the joints in patients with psoriatic arthritis (PsA). A, Percentages of patients achieving a response according to the American College of Rheumatology criteria for 20% improvement (ACR20), 50% improvement (ACR50), and 70% improvement (ACR70) on day 169, by treatment group. ∗ = P = 0.022; † = P = 0.006; ‡ = P = 0.121 versus placebo. B, Percentages of patients achieving an ACR20 response over time, by treatment group. ∗ = P = 0.022; † = P = 0.006; ‡ = P = 0.121 versus placebo on day 169. C, Percentages of patients achieving an ACR20 response on day 169 among those who had and those who had not previously been treated with anti–tumor necrosis factor (anti-TNF) agents. D, Changes in magnetic resonance imaging (MRI) scores for bone erosion, bone marrow osteitis, and synovitis on day 169 compared with baseline, by treatment group. Images were evaluated using the Rheumatoid Arthritis Magnetic Resonance Imaging Scoring system, as modified for use in PsA (see Patients and Methods for details). Values in A and C are the mean; error bars are the 95% confidence interval.

On day 169, the proportions of patients achieving an ACR50 or ACR70 response were higher among those receiving abatacept than among those receiving placebo: the proportions were highest in the 10 mg/kg group, with 25% achieving an ACR50 response and 13% achieving an ACR70 response (Figure 1A). Improvements in the ACR20 response were observed as early as day 29 (the time of first measurements), had leveled off by day 85, and were maintained through the last evaluation on day 169 (Figure 1B). Abatacept induced an ACR20 response regardless of whether the patients had previously received anti-TNF therapy (Figure 1C). At the 10 mg/kg dose of abatacept (but not at the other doses), the proportion achieving an ACR20 response was higher in patients who had never received anti-TNF agents than in those who had previously received anti-TNF agents (56% versus 31%).

Effects on joint damage.

On day 169, the MRI scores for erosion, osteitis, and synovitis appeared to be higher relative to baseline in patients receiving placebo, indicating a worsening of joint inflammation (Figure 1D). On day 169, MRI scores for osteitis and synovitis appeared to be lower relative to baseline in patients receiving abatacept, indicating a trend toward improvement of joint inflammation (Figure 1D). The mean ± SD changes from baseline in the MRI scores for erosion, osteitis, and synovitis, respectively, were as follows: 0.3 ± 3.5, –0.5 ± 1.9, and –0.8 ± 3.1 in the 30/10 mg/kg arm; –0.6 ± 4.2, –1.1 ± 2.6, and –1.4 ± 3.0 in the 10 mg/kg arm; 0.5 ± 2.4, –0.3 ± 1.7, and –0.2 ± 2.9 in the 3 mg/kg arm; and 1.5 ± 7.4, 0.4 ± 3.3, and 0.8 ± 4.3 in the placebo arm (Figure 1D).

Values for all 3 measures in each abatacept arm differed from those in the placebo arm, with the 10 mg/kg regimen showing the largest differences. The median (25th percentile, 75th percentile) changes from baseline in the MRI scores for erosion, osteitis, and synovitis, respectively, were 0.0 (–1.0, 1.0), 0.0 (–1.0, 0.0), and –0.5 (–2.0, 1.0) in the 30/10 mg/kg arm; –0.5 (–2.0, 2.0), –1.0 (–3.5, 0.5), and –1.3 (–3.5, 0.5) in the 10 mg/kg arm; 0.0 (–0.5, 1.3), 0.0 (–1.0, 0.5), and 0.0 (–1.3, 1.5) in the 3 mg/kg arm, and 0.0 (–0.5, 1.0), 0.0 (–0.5, 1.0), and 1.0 (–3.0, 3.0) in the placebo arm. Similar results were seen on day 85 (data not shown).

Psoriatic skin responses.

On day 169, the proportions of patients achieving an investigator's global assessment response (clear or almost clear) were as follows: 21% of those taking 30/10 mg/kg of abatacept, 25% taking 10 mg/kg, 38% taking 3 mg/kg, and 26% taking placebo, with only the 3 mg/kg group showing a modest improvement compared to placebo (Figure 2A). The TL score was assessed in all patients, and the PASI was determined in patients with at least 3% of the body surface areas affected by psoriasis at baseline (∼20 patients per treatment group). Compared to placebo, all 3 abatacept regimens resulted in improved TL50 and PASI50 responses in a time-dependent manner (Figures 2B and C). Improvements in the TL50 and the PASI50 were discernible as early as day 29. Maximum improvements were achieved by day 85 in the 10 and 3 mg/kg arms and by day 169 in the 30/10 mg/kg arm (Figures 2B and C).

Figure 2.

Effects of abatacept treatment on psoriatic skin lesions in patients with psoriatic arthritis (PsA). A, Percentages of patients achieving an investigator's global assessment (IGA) response (lesions judged to be “clear or almost clear”) on day 169, by treatment group. Values are the mean, error bars are the 95% confidence interval (95% CI). B, Percentages of patients achieving an improvement in the target lesion score of ≥50% versus baseline (TL50 response) over time, by treatment group. Values shown at the top are the percentages and 95% CIs of patients with a TL50 response or a TL75 response on day 169. C, Percentages of patients achieving an improvement in the Psoriasis Area and Severity Index of ≥50% versus baseline (PASI50 response) over time, by treatment group. Values shown at the top are the percentages and 95% CIs of patients with a PASI50 response or a PASI75 response on day 169. The PASI was determined only in patients who had psoriasis on ≥3% of their body surface area. There were 21 evaluable patients in each treatment group, except for the 30/10 mg/kg abatacept group, which had 20 patients.

On day 169, the proportions of patients achieving a TL50 response were 36%, 33%, and 30% of those taking 3 mg/kg, 10 mg/kg, and 30/10 mg/kg of abatacept, respectively, compared to 17% for placebo (Figure 2B). The corresponding proportions of patients achieving a TL75 response on day 169 were 29%, 10%, 16%, and 10%, respectively, with only the 3 mg/kg dose of abatacept showing a positive trend. On day 169, the proportions of patients achieving a PASI50 response were 43%, 29%, and 35% at 3 mg/kg, 10 mg/kg, and 30/10 mg/kg, respectively, versus 14% for placebo (Figure 2C). The corresponding proportions achieving a PASI75 response on day 169 were 38%, 14%, 10%, and 5%, respectively, with only those taking the 3 mg/kg dose of abatacept showing improvement compared to placebo.

Physical function and quality of life responses.

On day 169, the proportions of patients achieving an MCID (a decrease of at least 0.3 in the HAQ DI score from baseline [37]) were higher with the abatacept regimens than with placebo (35–45% versus 19%) (Table 3). The proportion achieving an MCID response was highest in the group receiving 10 mg/kg of abatacept (45%). Improvements in the MCID response in the 10 mg/kg and 3 mg/kg arms were evident by the time of the first measurements on day 15 and were maintained throughout the 6-month treatment period. Mean changes in PCS scores on the SF-36 health survey on day 169 as compared with baseline were greater in the abatacept treatment arms than in the placebo arm (6.3–9.3 versus 0.2); the greatest change (9.3) was seen in the 10 mg/kg arm (Table 3). These changes in PCS scores exceeded the threshold (a score of 3) used previously to define clinically meaningful improvement in patients with RA (42). Abatacept treatment (all regimens) also led to increases in the MCS scores on the SF-36 healthy survey in excess of 3 points (3.2–4.5) (Table 3), indicating clinically meaningful improvements.

Table 3. Improvements in disability and quality of life assessed on day 169, by treatment group
Treatment responseNo. (%) taking abataceptNo. (%) taking placebo (n = 42)
30/10 mg/kg (n = 43)*10 mg/kg (n = 40)3 mg/kg (n = 45)
  • *

    Two initial doses of 30 mg/kg followed by 10 mg/kg.

  • Patients achieving a minimum clinically important difference (MCID), which was defined as a ≥0.3-point decrease in scores on the disability index (DI) of the Health Assessment Questionnaire (HAQ).

  • Adjusted mean change from baseline in scores on the physical component summary (PCS) and the mental component summary (MCS) of the Short Form 36 (SF-36) health survey.

MCID in HAQ DI versus baseline    
 No. of patients43404542
 % (95% CI) of patients35 (21, 49)45 (30, 60)36 (22, 50)19 (7, 31)
SF-36 change from baseline    
 No. of patients42404341
 PCS score, mean ± SEM7.3 ± 1.99.3 ± 1.96.3 ± 1.80.2 ± 1.9
 MCS score, mean ± SEM4.5 ± 2.54.4 ± 2.53.2 ± 2.42.4 ± 2.5

Safety.

Adverse events were reported in ∼70% of patients in each treatment arm (Table 4). Seven patients experienced serious adverse events, 4 of whom were in the 30/10 mg/kg arm, 2 in the 10 mg/kg arm, and 1 in the placebo arm. One serious adverse event each in the 30/10 mg/kg (osteomyelitis) and 10 mg/kg (gastroenteritis) arms was considered to be drug-related. Osteomyelitis led to treatment discontinuation. One patient in the 30/10 mg/kg arm developed basal cell carcinoma of the skin. Infusion reactions were reported for 4 patients: 2 in the 30/10 mg/kg arm and 2 in the 10 mg/kg arm. There were no severe or very severe cases of acute infusion reaction. Seven patients prematurely discontinued treatment because of adverse events: 3 in the placebo arm and 4 across all abatacept arms (Table 2).

Table 4. Adverse events, by treatment group
 No. (%) taking abataceptNo. (%) taking placebo (n = 42)
30/10 mg/kg (n = 43)*10 mg/kg (n = 40)3 mg/kg (n = 45)
  • *

    Two initial doses of 30 mg/kg followed by 10 mg/kg.

  • Event reported as being related to treatment.

  • Event led to discontinuation.

Any adverse event29 (67)31 (78)31 (69)30 (71)
Serious adverse event4 (9)2 (5)01 (2)
 Cholecystitis1 (2)000
 Osteomyelitis1 (2)000
 Gastroenteritis01 (3)00
 Basal cell carcinoma1 (2)000
 Dizziness01 (3)00
Personality disorder0001 (2)
 Psychiatric decompensation0001 (2)
 Overdose1 (2)000
Infusion reaction2 (5)2 (5)00

DISCUSSION

This report provides the results after 6 months of treatment from a randomized, placebo-controlled trial evaluating 3 dosing regimens of abatacept in patients with PsA previously exposed to biologic and nonbiologic DMARDs. During the 6-month double-blind period, abatacept treatment given concomitantly with DMARDs was associated with improvement in both joint and skin symptoms. At the end of the 6-month treatment period (day 169), abatacept therapy given at the currently approved dose for patients with RA resulted in a significant difference in the proportion of patients achieving an ACR20 response, the primary efficacy measure, as compared to placebo (48% versus 19%; P = 0.006). This regimen was also associated with reduced joint damage, as measured by the MRI score, and with clinically meaningful improvements in physical function and quality of life. Notable improvements in psoriatic lesions were also observed. These responses were evident within 29 days of treatment initiation (i.e., after 2 injections), leveled off by day 85, and were maintained through the last evaluations on day 169, suggesting durable benefits. Abatacept was well-tolerated during the 6 months of treatment. These results are consistent with those previously reported for abatacept therapy in RA (25), JIA (30), and psoriasis (31), suggesting that abatacept may provide clinicians another alternative therapy for PsA.

In this study, the proportion of patients achieving an ACR20 response at an abatacept dosage of 10 mg/kg was higher in those who had not previously been treated with an anti-TNF agent than in those who had (56% versus 31%). However, such a separation was less pronounced with the 30/10 mg/kg or 3 mg/kg regimen, precluding a firm conclusion.

Across all direct and indirect measures of arthritis we evaluated, including the ACR, joint damage, and HAQ DI scores, the 10 mg/kg dose showed more therapeutic activity than the 3 mg/kg regimen. The TL50 and PASI50 responses with the 10 mg/kg regimen were also generally comparable to those with the 3 mg/kg regimen. Notably, the PASI75, the investigator's global assessment, and the TL75 responses on day 169 were seen only with the 3 mg/kg dose. This dose-response relationship is at variance with the findings of a dose-ranging phase I study of abatacept in psoriasis patients, where the antipsoriatic activity was most pronounced at the highest dose (50 mg/kg) tested (31). An initial loading dose of 30 mg/kg, followed by a dose of 10 mg/kg, did not enhance efficacy, which may be explained in part by the fact that the 30/10 mg/kg arm had a higher proportion of patients that had previously received anti-TNF agents than did the 10 mg/kg arm (51% versus 30%). Small numbers of patients (particularly for the PASI responses, where only 20 patients per arm were evaluated) and inconsistent results among the different measures of skin lesions warrant caution in interpreting these data. Nevertheless, the observed trend toward differential effects of abatacept on the joints and skin may indicate that inflammation in the skin differs from that in the joints.

Although inter-trial comparisons may be misleading, the results of multiple studies indicate that anti-TNF agents are more effective than abatacept in the treatment of psoriasis in patients with PsA (9, 10, 12). Up to 68% of patients receiving infliximab or adalimumab achieved a PASI75 response (9, 10, 12); the highest PASI75 response rate in the present study was 38%, which occurred in the 3 mg/kg group. Our inclusion of patients in the present study who had had an inadequate response to anti-TNF therapy may have resulted in a more treatment-resistant population and may thus have lessened the abatacept response.

As assessed by MRI, 6 months of abatacept treatment showed a trend toward reduced erosion, osteitis, and synovitis scores in this study. Although the MRI methodology has not been validated as an outcome measure for PsA, these results are consistent with the results of radiographic measurements of joint damage previously reported in RA patients treated with abatacept (26). Suppression of joint damage, as assessed by various imaging modalities, has also been reported in PsA patients treated with anti-TNF agents (43–45). A link between joint damage and disability is widely accepted in RA (46), and a similar relationship in PsA is highly likely (43).

Abatacept was generally safe and well-tolerated during this 6-month, double-blind period. The incidence of adverse events was similar between treatment groups. Two serious cases of infection, 1 each in the 30/10 mg/kg group (osteomyelitis) and 10 mg/kg group (gastroenteritis), were reported to be drug-related. No safety events were unique to this patient population. Overall, the safety profile of abatacept in this study was generally similar to that reported in studies of abatacept in RA (25–29).

Limitations of the present analysis include the small population size, the inclusion of patients who had previously been treated with anti-TNF agents, and the relatively short treatment duration. The lack of a dose response of skin lesions may be partly due to the small numbers of patients in each treatment arm and the imbalance in prior use of anti-TNF agents among the treatment arms. The higher proportion of patients who had previously undergone anti-TNF therapy in the 30/10 mg/kg arm confounded the interpretation of the impact of the initial loading dose on efficacy. Only two 30 mg/kg infusions were used; a larger number of loading doses might have been more efficacious. Also of note, the study was powered to detect only the differences in the ACR20 response rate on day 169 between the abatacept arm and the placebo arm and not differences in the rates of the other responses we measured. Despite these limitations, the results presented here support a role of T cell activation in the pathogenesis of PsA and suggest that mechanisms that are insensitive to abatacept are also involved.

In conclusion, the present study comparing 3 abatacept dosing regimens with placebo in patients with PsA showed that the 10 mg/kg regimen, which is currently approved for the treatment of RA and JIA, resulted in the greatest improvement over placebo in the signs and symptoms of arthritis, joint damage, physical function, and quality of life, relative to the 3 mg/kg regimen or the 30/10 mg/kg regimen. The 10 mg/kg regimen also showed activity against skin lesions, which was generally comparable to that of the 3 mg/kg dose. Initial loading with 2 doses of abatacept at 30 mg/kg followed by 10 mg/kg did not enhance efficacy. Abatacept was efficacious, regardless of whether patients had previously been exposed to anti-TNF agents or not. Data on the long-term effectiveness and safety of abatacept in PsA will be reported from the ongoing open-label, extension phase. The results presented here demonstrate that abatacept 10 mg/kg may be a treatment option for patients with active PsA who were previously exposed to DMARDs, including anti-TNF agents.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Mease had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Mease, Genovese, Ritchlin, Bahary, Becker, Kelly, Sigal, Teng, Gladman.

Acquisition of data. Mease, Genovese, Gladstein, Kivitz, Tak, Wollenhaupt, Bahary, Becker, Sigal, Gladman.

Analysis and interpretation of data. Mease, Genovese, Ritchlin, Becker, Kelly, Sigal, Teng, Gladman.

Protocol management. Bahary.

ROLE OF THE STUDY SPONSOR

Bristol-Myers Squibb facilitated the study design, provided writing assistance for the manuscript, and reviewed and approved the manuscript prior to submission. The authors independently collected the data, interpreted the results, and had the final decision to submit the manuscript for publication. Writing and editorial assistance was provided by Motasim Billah, an employee of Bristol-Myers Squibb. Publication of this article was not contingent upon approval by Bristol-Myers Squibb.

Acknowledgements

We thank all of the patients and investigators who participated in this trial.

Ancillary