Elicit: CD80/CD86 Blockade and T-Cell Dynamics

Skip to main content

CD80/CD86 Blockade and T-Cell Dynamics

Research reportView only

Create alertChat

May 6, 2026

How does CD80/CD86 blockade by abatacept affect T-cell activation and tolerance?

Abatacept reduces CD4+ T-cell activation by blocking CD28 costimulation and shifts cells toward naive phenotypes, but achieves reversible immunosuppression rather than permanent tolerance, with paradoxical effects on regulatory T cells varying by disease context.

Abstract

CD80/CD86 blockade by abatacept consistently reduces T-cell activation across multiple disease contexts through selective modulation of CD28-mediated costimulation. Treatment decreases expression of activation markers (CD38, ICOS, Ki-67) and reduces proliferation particularly in CD4+ T cells, with 7-10-fold fewer proliferating and activated effector memory CD4+ T cells observed in GVHD prevention. Abatacept shifts CD4+ compartments toward naive phenotypes while reducing central memory populations, with transcriptional analysis revealing decreased expression of genes regulating cell cycle and chromatin dynamics. CD8+ T-cell subsets remain relatively resistant to these effects, indicating lesser reliance on CD28 costimulation. The effects are reversible after treatment discontinuation, with T-cell populations returning to baseline within 6 months, indicating that abatacept maintains an altered immune state through continuous pharmacological inhibition rather than inducing permanent tolerance.

Effects on regulatory T cells are context-dependent and paradoxical: in rheumatoid arthritis, abatacept increases Treg numbers through reduced apoptosis but diminishes their suppressive function because CD80/CD86 blockade on responder T cells reduces susceptibility to Treg-mediated suppression, while in multiple sclerosis, memory Treg frequencies decrease due to their dependence on ongoing costimulation. Clinical translation of these immunological effects varies substantially by disease, with robust benefits observed in GVHD prevention where transient early suppression of donor T-cell activation is sufficient, sustained benefits in established rheumatoid arthritis with high baseline T-cell activation, but no clinical improvement in mild atopic asthma despite similar shifts in T-cell populations. This heterogeneity suggests that costimulation blockade is most effective when CD28-dependent T-cell responses drive disease pathology and baseline immune activation is high.

Methods

We analyzed 10 sources from an initial pool of 200, using 8 screening criteria. Each paper was reviewed for 8 key aspects that mattered most to the research question. More on methods

Records from Elicit search

n = 200

Papers screened using: Abatacept Intervention, T-cell Outcomes, Control Groups, Study Design, Intervention Isolation, Study Setting, Publication Status, Intervention Specificity

n = 200

Papers screened out

n = 190

Papers included for extraction

n = 10

Press enter or space to select a node.You can then use the arrow keys to move the node around. Press delete to remove it and escape to cancel.

Press enter or space to select an edge. You can then press delete to remove it or escape to cancel.

Paper search

We performed a semantic search across over 138 million academic papers from the Elicit search engine, which includes all of Semantic Scholar and OpenAlex.

We ran this query: “How does CD80/CD86 blockade by abatacept affect T-cell activation and tolerance?”

The search returned 200 total results from Elicit.

We retrieved 200 papers most relevant to the query for screening.

Screening

We screened in sources based on their abstracts that met these criteria:

We considered all screening questions together and made a holistic judgement about whether to screen in each paper.

Data extraction

We asked a large language model to extract each data column below from each paper. We gave the model the extraction instructions shown below for each column.

Extract the disease model or clinical context used to study CD80/CD86 blockade effects on T-cells, including:

Extract specific details about abatacept treatment for CD80/CD86 blockade, including:

Extract how T-cell activation was assessed and the specific findings related to CD80/CD86 blockade by abatacept, including:

Extract how T-cell tolerance or regulatory function was assessed and the specific findings related to CD80/CD86 blockade by abatacept, including:

Extract the specific T-cell populations affected by CD80/CD86 blockade and the direction/magnitude of effects, including:

Extract molecular and pathway-level insights into how CD80/CD86 blockade by abatacept affects T-cell activation and tolerance, including:

Extract the temporal pattern of T-cell activation and tolerance changes following CD80/CD86 blockade, including:

Extract factors that could affect interpretation of T-cell activation and tolerance effects from CD80/CD86 blockade, including:

Results

Characteristics of Included Studies

Study

Full text retrieved?

Disease context

Study design

Abatacept regimen

Primary T-cell measures

J. Kremer et al., 2006

Yes

Rheumatoid arthritis with inadequate methotrexate response

Randomized, double-blind, placebo-controlled trial

~10 mg/kg IV monthly for 1 year

Not specified

M. Weisman et al., 2006

No

Active rheumatoid arthritis with inadequate methotrexate response

Phase II trial

10 mg/kg or 2 mg/kg for 12 months

Inflammatory biomarkers (IL-6, soluble IL-2 receptor)

Divya Koura et al., 2013

Yes

Acute GVHD prevention during unrelated-donor HCT

First-in-disease trial

10 mg/kg IV on days -1, +5, +14, +28 post-HCT

Ki-67 proliferation, CD38/HLA-DR activation, FoxP3+ Tregs

M. Eichmann et al., 2020

No

New-onset type 1 diabetes

Clinical trial

Not specified

CD4+ and CD8+ naive and memory subsets by flow/mass cytometry

R. Alten, 2007

No

Rheumatoid arthritis unresponsive to DMARDs or TNF-blockers

Review of trials

~10 mg/kg at weeks 0, 2, 4, then every 4 weeks for 6-12 months

Not specified

T. Orbán et al., 2014

Yes

Recent-onset type 1 diabetes

Phase 2 trial

10 mg/kg IV every 28 days for 24 months

CM and naive CD4 T cells, CD4+CD25high Tregs, CD69 activation

A. Parulekar et al., 2013

No

Mild atopic asthma

Randomized, placebo-controlled, double-blind trial

3 months of treatment

Naive vs. memory CD4+ T cells

B. Watkins et al., 2021

No

Severe acute GVHD following unrelated-donor HCT

Phase II trial, randomized (8/8 matched) and single-arm (7/8 mismatched)

CNI/MTX plus abatacept, measured at day +100 and +180

T-cell activation (method not specified)

S. Glatigny et al., 2019

Yes

Relapsing-remitting multiple sclerosis

Phase II, double-blind, placebo-controlled trial

Weeks 0, 2, 4, then every 4 weeks through week 24

Tfh and Treg frequency, CD38/ICOS activation, transcriptional profiling

M. Bonelli et al., 2016

No

Rheumatoid arthritis

In vivo clinical analysis and ex vivo experiments

Not specified

Treg frequency/function, CD95 expression, activation markers

toof

Pageof

The included studies examined abatacept’s effects across diverse autoimmune and transplant contexts. Five studies had full text available. Disease models included rheumatoid arthritis (4 studies), type 1 diabetes (2 studies), GVHD prevention (2 studies), multiple sclerosis (1 study), and asthma (1 study). Most studies used approximately 10 mg/kg dosing at 4-week intervals, though treatment durations varied from 3 months to 24 months. Concomitant immunosuppression was common, including methotrexate in rheumatoid arthritis studies and calcineurin inhibitors plus methotrexate in GVHD prevention studies.

Effects on T-cell Activation

Activation Marker Expression

Abatacept treatment reduced expression of multiple T-cell activation markers across disease contexts. In GVHD prevention, CD4+ T cells showed 7-fold fewer proliferating cells (Ki-67+) and 10-fold fewer activated cells (CD38+/HLA-DR+) at day +28 compared to controls, with effects concentrated in effector memory populations. In multiple sclerosis, abatacept decreased the proportion of CD38+ and ICOS+ activated cells within both T follicular helper (Tfh) and regulatory T cell populations. Rheumatoid arthritis patients exhibited downregulation of activation-associated markers and CD95 on CD4+ T cells and Tregs.

The activation changes were selective for certain T-cell populations. CD4+ T cells were consistently more affected than CD8+ T cells across studies. In type 1 diabetes, CD4+ conventional and regulatory subsets showed altered frequencies while CD8+ subsets remained relatively resistant, indicating lesser reliance on CD28-mediated costimulation in CD8+ cells. This differential sensitivity extended to specialized subsets, with Tfh cells showing progressive frequency declines and reduced activation marker expression in multiple sclerosis patients.

Proliferation and Cell Cycle Effects

Transcriptional analysis revealed that abatacept disrupted cell cycle and proliferation pathways in activated T cells. In multiple sclerosis patients, abatacept reduced expression of genes regulating cell cycle and chromatin dynamics during cell proliferation, directly linking reduced costimulatory signaling to impaired activation and proliferation. This was consistent with functional proliferation measures showing significant inhibition of CD4+ T cell proliferation in GVHD prevention, affecting both effector and central memory subsets.

Inflammatory Biomarkers

Beyond cellular activation markers, abatacept reduced soluble inflammatory mediators associated with T-cell activation. In rheumatoid arthritis, 12 months of treatment significantly decreased serum levels of IL-6, soluble IL-2 receptor, C-reactive protein, soluble E-selectin, and soluble ICAM-1 compared to placebo. Smaller reductions in TNF-α and rheumatoid factor were also observed. These systemic changes reflected the anti-inflammatory and immunomodulatory effects of selective CD28 costimulation blockade.

Effects on T-cell Subset Distribution

Study

Context

Naive T cells

Memory T cells

Specialized subsets

Effect magnitude

T. Orbán et al., 2014

Type 1 diabetes

Increased naive CD4+ T cells

Decreased central memory CD4+ T cells

Reduced CD4+CD25high Tregs

Significant reduction in absolute CM numbers

M. Eichmann et al., 2020

Type 1 diabetes

Ag-naive subsets increase

Ag-experienced subsets decrease

Not specified

CD4+ affected, CD8+ not affected

A. Parulekar et al., 2013

Asthma

Increased naive CD4+ T cells

Decreased memory CD4+ T cells

Not specified

Not quantified

S. Glatigny et al., 2019

Multiple sclerosis

No effect on CD45RA+ naive Tregs

Decreased CD45RO+ memory Tregs

Decreased Tfh frequency; reduced CD38+/ICOS+ activation

Progressive Tfh decline

M. Bonelli et al., 2016

Rheumatoid arthritis

Not specified

Not specified

Increased Treg numbers; diminished Treg suppressive function

Not quantified

toof

Pageof

Abatacept consistently shifted CD4+ T-cell compartments toward naive phenotypes while reducing memory populations. In type 1 diabetes, abatacept treatment significantly reduced central memory CD4+ T cells (CD45RO+CD62L+) while increasing naive CD4+ T cells (CD45RO-CD62L+), with the reduction driven by decreased absolute numbers in circulation. The ratio of naive to central memory cells increased significantly throughout 24 months of treatment. Similarly, asthma patients showed increased naive and decreased memory CD4+ T cells after 3 months.

Within memory populations, effects were selective. Multiple sclerosis patients showed specific decreases in CD45RO+ memory Tregs while CD45RA+ naive Tregs remained unaffected. Tfh cells, a specialized memory subset, decreased progressively in frequency with accompanying reductions in activated (CD38+/ICOS+) Tfh cells.

Effects on Regulatory T Cells and Tolerance

Abatacept’s effects on regulatory T cells varied by subset and assay. In rheumatoid arthritis, total Treg numbers increased after treatment initiation, but functional suppression assays revealed diminished capacity to suppress responder T-cell proliferation. This occurred because CD80/CD86 blockade on responder T cells reduced their susceptibility to Treg-mediated suppression.

Multiple sclerosis patients exhibited decreased Treg frequency, particularly affecting memory Tregs expressing CD38 and ICOS activation markers. The demethylation status of the TSDR region in the FOXP3 gene also decreased, suggesting reduction in bona fide stable Tregs. In GVHD prevention, CD4+CD25high/CD127low/FoxP3+ putative Tregs transiently decreased in the abatacept cohort early after transplantation compared to standard treatment.

Type 1 diabetes studies showed trends toward reduced CD4+CD25high cells (enriched for thymus-derived Tregs), though associations with disease progression were not statistically significant. The reduction in central memory CD4+ T cells, which can differentiate into effector cells upon activation, accompanied by increased naive cells, represented a shift toward less activated immune states potentially favoring tolerance.

Temporal Dynamics of Effects

Study

Context

Onset of activation effects

Peak effects

Reversibility

Duration assessed

Divya Koura et al., 2013

GVHD

Day +28

Day +28

Yes, effects diminished after treatment cessation

Through day +100

T. Orbán et al., 2014

Type 1 diabetes

Within first 6 months

6, 12, and 24 months

Yes, returned to baseline 6 months post-treatment

30 months (24 + 6 follow-up)

S. Glatigny et al., 2019

Multiple sclerosis

Treg changes at 4 weeks; Tfh at 16 weeks

Not specified

Yes, reversed upon discontinuation

52 weeks

J. Kremer et al., 2006

Rheumatoid arthritis

By 6 months (clinical effects)

Up to 1 year

Not assessed

1 year

toof

Pageof

The onset and kinetics of T-cell effects varied by subset. In GVHD prevention, significant inhibition of CD4+ T-cell proliferation and activation was evident by day +28, but similar levels of activation were observed in treated and control cohorts after day +100, indicating transient effects limited to the period of drug exposure. The terminal half-life of abatacept (approximately 19.5 days) influenced the duration of immunological effects.

Type 1 diabetes studies demonstrated sustained effects during continuous treatment. Significant changes in central memory and naive CD4+ T cells were observed at 6, 12, and 24 months, persisting throughout the treatment period but returning to baseline values 6 months after treatment discontinuation. This reversibility was consistent across disease contexts, with multiple sclerosis patients showing complete reversal of cellular and molecular changes upon abatacept cessation.

Notably, different T-cell subsets exhibited distinct kinetic patterns. In multiple sclerosis, Treg frequency changes occurred as early as 4 weeks post-treatment while Tfh cell declines were observed later at 16 weeks, suggesting differential susceptibility to costimulation blockade across subsets.

Mechanistic Insights into Activation vs. Tolerance

Abatacept’s mechanism extends beyond simple CD80/CD86 blockade to affect multiple cellular processes. Transcriptional profiling in multiple sclerosis revealed that abatacept decreased expression of genes regulating cell cycle and chromatin dynamics during proliferation, directly linking reduced costimulatory signaling to impaired activation and cell division. The selective decrease in activated (CD38+/ICOS+) Tfh and Treg cells indicated that abatacept preferentially affected cells requiring ongoing costimulation.

The differential impact on CD4+ versus CD8+ T cells reflects their relative dependence on CD28 costimulation. CD8+ T-cell subsets showed resistance to abatacept effects in type 1 diabetes, suggesting lesser reliance on the CD28 pathway for activation and maintenance. This mechanistic distinction has important implications for which immune responses are modulated by costimulation blockade.

Rheumatoid arthritis studies revealed an additional mechanism: abatacept reduced CD95-mediated T-cell apoptosis in a dose-dependent manner, leading to increased T-cell numbers including Tregs. However, blockade of CD80/CD86 on responder T cells simultaneously diminished their susceptibility to Treg-mediated suppression, creating a paradox where Treg numbers increased but functional suppression decreased. This finding illustrates how costimulation signals serve dual roles in both effector T-cell activation and tolerance mechanisms.

The modulation of central memory T-cell trafficking may contribute to therapeutic effects. In type 1 diabetes, abatacept inhibited transmigration of central memory CD4+ T cells across endothelial cells, potentially affecting their migration and retention in lymphoid tissues where they encounter antigen. The reduction in this activated, antigen-experienced population accompanied by expansion of naive cells represented a shift toward less inflammatory immune states.

Study Quality and Limitations

Several factors affect interpretation of these findings across studies. Sample size limitations were notable, particularly for immunological endpoints. The GVHD first-in-disease trial included only 10 patients, and not all type 1 diabetes patients could be tested at each timepoint, potentially limiting detection of subtle subset changes. The multiple sclerosis trial enrolled 65 participants, which, while larger, still constrained generalizability. The smaller 7/8-HLA-mismatched GVHD stratum had only 43 recipients.

Technical considerations included high variance in absolute T-cell subset counts in type 1 diabetes studies and exclusion of samples with fewer than 50 events in multiple sclerosis flow cytometry analyses. These technical limitations could introduce bias or reduce precision. Most studies focused on peripheral blood samples, which might not fully represent T-cell responses in tissues or other compartments.

Confounding medications were common. Rheumatoid arthritis trials used methotrexate in combination with abatacept, GVHD prevention studies combined abatacept with cyclosporine and methotrexate, making it difficult to isolate abatacept-specific effects. The type 1 diabetes studies were validation studies conducted with only half of the eventual total subjects, indicating incomplete data at the time of publication.

Several studies acknowledged important gaps. Rheumatoid arthritis trials examined only patients with inadequate methotrexate response, not those with early disease. Type 1 diabetes investigators noted the need for more specific Treg markers and functional assessments to fully characterize tolerance effects. The asthma study’s patient population was limited to mild atopic asthma, potentially limiting generalizability to severe disease.

Synthesis

The evidence reveals a consistent pattern: abatacept reduces T-cell activation markers and shifts CD4+ compartments toward naive phenotypes while showing limited effects on CD8+ cells. However, the relationship between these immunological changes and clinical outcomes varies substantially across diseases, warranting careful examination of why similar T-cell effects produce different therapeutic results.

Context-Specific Clinical Translation

In GVHD prevention, the transient reduction in CD4+ T-cell activation and proliferation at day +28 aligned with favorable clinical outcomes, including low rates of severe GVHD (2.5% grade 3-4 in the 8/8-matched randomized cohort). The 7-10-fold reduction in proliferating and activated effector memory CD4+ T cells occurred during the critical early post-transplant window when donor T cells encounter recipient antigens, potentially explaining clinical benefit despite the transient nature of effects.

In contrast, the asthma study found increased naive and decreased memory CD4+ T cells but no significant reduction in allergen-induced eosinophilic inflammation (17.71% vs. 46.39% in placebo, p=0.26) or improvements in FEV1, methacholine responsiveness, or symptoms. This divergence between immunological and clinical effects suggests that CD28-dependent T-cell responses may not be the dominant driver of allergen-induced inflammation in mild atopic asthma, or that the 3-month treatment duration was insufficient for clinically meaningful effects in this chronic condition.

Subset-Specific Mechanisms

The differential kinetics and magnitude of effects across T-cell subsets explain some heterogeneity. In multiple sclerosis, Treg frequency changes occurred at 4 weeks while Tfh declines appeared at 16 weeks, suggesting that not all CD28-dependent processes have equivalent sensitivity to blockade. Memory Tregs (CD45RO+) decreased while naive Tregs (CD45RA+) remained constant, indicating that antigen-experienced regulatory populations are more dependent on ongoing costimulation than their naive counterparts. This mechanistic insight predicts that abatacept would have greater effects in settings dominated by memory T-cell responses (such as transplant rejection or established autoimmunity) compared to those requiring de novo T-cell priming.

The Treg Paradox

The apparently contradictory findings regarding regulatory T cells—increased numbers but decreased function in rheumatoid arthritis versus decreased frequencies in multiple sclerosis—can be reconciled by considering that CD80/CD86 serve dual roles. In rheumatoid arthritis, abatacept increased Treg numbers by reducing CD95-mediated apoptosis while simultaneously blocking CD80/CD86 on responder T cells, making them less susceptible to Treg suppression. This mechanism would maintain or increase Treg numbers while reducing their functional impact. In multiple sclerosis, the focus on memory Tregs that require costimulation for maintenance explains their selective depletion. Both findings may be correct within their respective contexts: rheumatoid arthritis involving high baseline inflammation and apoptosis versus multiple sclerosis where memory Treg maintenance is critical.

Disease Activity and Baseline State

Studies examining patients with active, established disease (rheumatoid arthritis with inadequate methotrexate response, active GVHD prevention) showed more robust clinical benefits from abatacept compared to those with mild disease activity (mild atopic asthma). This pattern suggests a dose-response relationship where higher baseline T-cell activation provides more opportunity for therapeutic benefit from costimulation blockade. The significant reductions in inflammatory biomarkers in rheumatoid arthritis patients (IL-6, soluble IL-2 receptor, CRP) reflected high baseline inflammation that responded to suppression of activated T cells.

Reversibility and Treatment Duration

The consistent reversibility of T-cell changes after treatment cessation indicates that abatacept does not induce permanent immune reprogramming or tolerance, but rather maintains an altered immune state through continuous inhibition. This explains why type 1 diabetes studies required 24 months of continuous treatment to demonstrate slowed β-cell decline, and why the transient 4-dose regimen in GVHD prevention was sufficient for that acute condition but might not sustain effects in chronic autoimmunity. The temporal alignment between drug exposure (half-life ~19.5 days) and immunological effects supports a pharmacological rather than tolerogenic mechanism.

Confounder Effects

The universal use of concomitant immunosuppression in clinical trials complicates attribution of effects solely to abatacept. Methotrexate, used in all rheumatoid arthritis studies, has independent effects on T-cell activation and proliferation. The combination of calcineurin inhibitors and methotrexate in GVHD prevention provides baseline immunosuppression that may interact synergistically with costimulation blockade. Studies showed that abatacept effects were most evident when the baseline regimen alone was insufficient (inadequate methotrexate response in RA, high-risk HCT), suggesting that costimulation blockade addresses a distinct mechanism not fully covered by other agents.

Clinical Implications by Context

For acute T-cell-mediated conditions like GVHD, the transient but profound reduction in CD4+ effector memory activation during the critical early post-transplant period appears sufficient for clinical benefit. For chronic autoimmune diseases like type 1 diabetes and rheumatoid arthritis, sustained treatment produces gradual shifts in T-cell compartments toward less activated states, translating to slowed disease progression over months to years. In diseases where CD28-independent pathways dominate (mild asthma, possibly multiple sclerosis where the trial showed no clinical benefit despite T-cell changes), costimulation blockade alone may be insufficient regardless of its immunological effects.

References

S. Glatigny, B. Höllbacher, S. Motley, Cathy Tan, C. Hundhausen, and 11 more\ (2019).Abatacept Targets T Follicular Helper and Regulatory T Cells, Disrupting Molecular Pathways That Regulate Their Proliferation and Maintenance. Journal of Immunology

M. Bonelli, L. Göschl, S. Blüml, T. Karonitsch, K. Hirahara, and 5 more\ (2016).Abatacept (CTLA-4Ig) treatment reduces T cell apoptosis and regulatory T cell suppression in patients with rheumatoid arthritis. Rheumatology

J. Kremer, H. Genant, L. Moreland, A. Russell, P. Emery, and 6 more\ (2006).Effects of Abatacept in Patients with Methotrexate-Resistant Active Rheumatoid Arthritis. Annals of Internal Medicine

M. Weisman, P. Durez, D. Hallegua, R. Aranda, J. Becker, and 4 more\ (2006).Reduction of inflammatory biomarker response by abatacept in treatment of rheumatoid arthritis. Journal of Rheumatology

Divya Koura, J. Horan, A. Langston, M. Qayed, A. Mehta, and 18 more\ (2013).In vivo T cell costimulation blockade with abatacept for acute graft-versus-host disease prevention: a first-in-disease trial. Biology of Blood and Marrow Transplantation

M. Eichmann, Roman Baptista, R. Ellis, S. Heck, M. Peakman, and 1 more\ (2020).Costimulation Blockade Disrupts CD4+ T Cell Memory Pathways and Uncouples Their Link to Decline in β-Cell Function in Type 1 Diabetes. Journal of Immunology

R. Alten\ (2007).Die Kostimulationsmodulation mit Abatacept: Ein neuartiges und erfolgreiches Therapieprinzip in der Therapie der rheumatoiden Arthritis - Teil 1: Zum Wirkmechanismus von Abatacept. Aktuelle Rheumatologie

T. Orbán, C. Beam, Ping Xu, Keith J M Moore, Qi Jiang, and 5 more\ (2014).Reduction in CD4 Central Memory T-Cell Subset in Costimulation Modulator Abatacept-Treated Patients With Recent-Onset Type 1 Diabetes Is Associated With Slower C-Peptide Decline. Diabetes

A. Parulekar, J. Boomer, Brenda M Patterson, Huiqing Yin-Declue, Christine M. Deppong, and 4 more\ (2013).A randomized controlled trial to evaluate inhibition of T-cell costimulation in allergen-induced airway inflammation. American Journal of Respiratory and Critical Care Medicine

B. Watkins, M. Qayed, C. McCracken, B. Bratrude, K. Betz, and 41 more\ (2021).Phase II Trial of Costimulation Blockade With Abatacept for Prevention of Acute GVHD. Journal of Clinical Oncology

Download BIBDownload RISDownload TXT

Report

Status

Gather sources

200 sources found

Details

Screen sources

10 sources included

Details

Extract data

80 data points extracted

Details

Generate report

Save PDF

BIBLaTeX, ZoteroRISZotero, MendeleyTXTAPA BibliographyPDFPDFDOCXMicrosoft Word

Chat

Got some follow-up questions?

Sign up or sign in to chat with this report.

Back

Effects of Abatacept in Patients with Methotrexate-Resistant Active Rheumatoid Arthritis

J. Kremer, H. Genant, L. Moreland, A. Russell, P. Emery, C. Abud-Mendoza, J. Szechiński, Tracy Li, Zhiyu Ge, J. Becker, R. Westhovens

Annals of Internal Medicine·

2006·

760 citations

SourceDOI

Plain textPDF

Searching for PDF

Unable to find PDF from source

Back

Study Model

- Disease type/condition studied: Rheumatoid arthritis - Study population characteristics relevant to T-cell function: Patients with persistent and active rheumatoid arthritis despite methotrexate treatment - Experimental approach: Randomized, double-blind, placebo-controlled clinical trial - Control groups used for comparison: Placebo group

Abatacept Regimen

- Dose: Approximately 10 mg/kg of body weight; specific dosages by weight range: 500 mg (<60 kg), 750 mg (60-100 kg), 1000 mg (>100 kg) - Dosing schedule: Once-monthly infusions; initial infusions on days 1, 15, and 29, then every 28 days up to day 337 - Duration of treatment: 1 year - Route of administration: Intravenous infusion over 30 minutes - Timing relative to outcome measurements: Outcomes measured during the 1-year treatment period - Concomitant immunosuppressive treatments: Methotrexate (stable dose during first 6 months, with potential adjustments or additions of other disease-modifying antirheumatic drugs or corticosteroids between 6 and 12 months)

T-cell Activation Measures

Not mentioned (the paper does not provide specific details on T-cell activation measures or findings related to CD80/CD86 blockade by abatacept)

T-cell Tolerance Measures

Not mentioned (the paper does not provide specific data on T-cell tolerance measures or regulatory function related to CD80/CD86 blockade by abatacept)

T-cell Subset Effects

Not mentioned (the paper does not provide specific information on the effects of CD80/CD86 blockade by abatacept on different T-cell subsets or their functional changes)

Mechanistic Insights

Abatacept modulates T-cell activation by blocking the co-stimulatory signal through CD80/CD86, mimicking CTLA-4's downregulation of CD28-mediated activation. No specific molecular or pathway-level insights into transcriptional changes, cell cycle effects, chromatin dynamics, signal transduction pathways, or metabolic changes are provided.

Effect Kinetics

- Time to onset of effects on T-cell activation: As early as 6 months, with significant improvements in ACR responses. - Time to onset of tolerance-related changes: Not explicitly mentioned, but sustained improvements suggest ongoing tolerance effects. - Peak effect timepoints: Continued improvement up to 1 year suggests a sustained peak effect. - Duration and persistence of effects: Sustained improvement over the 1-year study period. - Reversibility after treatment discontinuation: Not explicitly addressed. - Differences in kinetics between activation and tolerance measures: Not explicitly detailed, but rapid onset with sustained effects.

Study Limitations

- Sample size limitations for immunological endpoints: Not explicitly mentioned. - Confounding medications or treatments: Methotrexate was used in combination with abatacept, which could affect T-cell responses. - Technical limitations in T-cell assays: Not mentioned. - Patient population factors that could affect T-cell responses: The study only examined patients with an inadequate response to methotrexate and not those in early disease stages. - Missing timepoints or incomplete immune monitoring: The study's one-year duration may not capture long-term effects. - Potential off-target effects of abatacept beyond CD80/86 blockade: Not mentioned.

Context Abatacept, an agent that selectively modulates the co-stimulatory signal required for T-cell activation, may benefit some patients with rheumatoid arthritis. Contribution This 1-year, randomized, double-blind trial compared once-monthly infusions of abatacept with placebo in 652 patients with symptomatic rheumatoid arthritis despite ongoing methotrexate treatment. Compared with placebo recipients, patients who received abatacept more often had improved physical function, more frequently met standard response criteria, and less often had radiographic progression of joint damage. They also had serious infections (2.5% vs. 0.9%) and infusion reactions more often. Implications Adding abatacept can reduce disease activity in patients with rheumatoid arthritis and an inadequate response to methotrexate. The Editors Rheumatoid arthritis is characterized by synovial membrane hyperplasia and inflammatory cell infiltrate, including activated T cells (1). T cells contribute to the initiation and perpetuation of rheumatoid arthritis immunopathology, leading to inflammation and, ultimately, joint destruction. Activated T cells proliferate and induce monocytes, macrophages, and synovial fibroblasts to produce proinflammatory cytokines, such as tumor necrosis factor-, interleukin-1, and interleukin-6 (1), and stimulate osteoclastogenesis and matrix metalloproteinase secretion (2), as well as immunoglobulin production by B cells (3). The central role of activated T cells in rheumatoid arthritis immunopathology makes T-cell activation a rational therapeutic target. T cells require 2 signals for full activation: an antigen-specific signal (signal 1) and a co-stimulatory signal (signal 2) (4). One of the best-characterized co-stimulatory pathways is the engagement of CD80 or CD86 on antigen-presenting cells with CD28 on T cells (5). In the normal immune response, endogenous cytotoxic T-lymphocyte antigen-4 (CTLA-4) downregulates CD28-mediated T-cell activation by binding to CD80 or CD86 with higher avidity than CD28 (6). Abatacept is a soluble, recombinant, fully human fusion protein, comprising the extracellular domain of CTLA-4 and the Fc portion of IgG1, modified to prevent complement fixation. Abatacept is the first in a new class of agents for treating rheumatoid arthritis that selectively modulate the co-stimulatory signal required for full T-cell activation. A phase IIa study of patients with rheumatoid arthritis and an inadequate response to disease-modifying antirheumatic drugs showed the efficacy of abatacept as monotherapy (7). In a phase IIb study of abatacept plus methotrexate in patients with rheumatoid arthritis and an inadequate response to methotrexate, signs and symptoms of rheumatoid arthritis, physical function, and health-related quality of life statistically significantly improved over 1 year (8, 9). We present findings from the phase III, 1-year Abatacept in Inadequate Responders to Methotrexate (AIM) trial, which was designed to further evaluate the safety and clinical efficacy of abatacept plus methotrexate and to assess the effects of abatacept on the radiographic progression of structural damage. Methods The institutional review boards or independent ethics committees approved a common clinical protocol for each site, and we performed the study in accordance with the ethical principles of the Declaration of Helsinki. All patients provided written informed consent to the study protocol before randomization. Patients Eligible patients were at least 18 years of age, had had rheumatoid arthritis for at least 1 year, and met the American Rheumatism Association criteria for rheumatoid arthritis (10). Rheumatoid arthritis was persistent and active despite methotrexate treatment. All patients must have been treated with methotrexate (15 mg/wk) for 3 months or longer, with a stable dose for 28 days before enrollment. We required patients to undergo a washout of all other disease-modifying antirheumatic drugs at least 28 days before randomization. We allowed corticosteroid use, with dosages equal to 10 mg of prednisone or less per day, stabilized for 25 days before randomization. At randomization, we required patients to have 10 or more swollen joints, 12 or more tender joints, and C-reactive protein levels of 10.0 mg/L or greater (normal range, 1.0 mg/L to 4.0 mg/L) while receiving methotrexate. We required tuberculin skin testing before randomization. We excluded patients with a positive tuberculin skin test result unless they had completed treatment for latent tuberculosis before enrollment. Study Design Our 1-year, multicenter, multinational, randomized, double-blind, placebo-controlled study aimed to compare the efficacy and safety of abatacept versus placebo in combination with methotrexate in patients with rheumatoid arthritis and an inadequate response to methotrexate treatment. We used a central randomization system, and the Drug Management Group within Bristol-Myers Squibb, Princeton, New Jersey, generated the randomization schedule. Stratification per site was not performed. Patients were randomly assigned in a 2:1 ratio to receive either a fixed dose of abatacept, approximately 10 mg/kg of body weight, or placebo. Patients weighing less than 60 kg, 60 to 100 kg, or more than 100 kg received 500 mg, 750 mg, or 1000 mg of abatacept, respectively. We administered study medication by 30-minute intravenous infusion on days 1, 15, and 29 and then every 28 days up to and including day 337. No premedication was required. The protocol specified that all patients were to receive methotrexate, 15 mg or more per week, although methotrexate at 10 mg per week was acceptable if the patient had a history of toxicity. During the first 6 months, we did not allow adjustments in methotrexate dose, except in cases of toxicity. We permitted use of stable dosages of nonsteroidal anti-inflammatory drugs and corticosteroid dosages equal to 10 mg of prednisone or less per day. Between 6 and 12 months, we allowed the following adjustments, as the investigator deemed necessary: 1) adjustment in methotrexate dose, 2) addition of 1 other disease-modifying antirheumatic drug (hydroxychloroquine, sulfasalazine, gold, or azathioprine), or 3) adjustment in corticosteroid dose equal to 10 mg of prednisone or less per day. However, investigators were blinded to treatment group assignment throughout the 1-year study. Clinical Efficacy Measures Our 3 primary objectives were to evaluate the proportion of patients in each group with a 20% improvement in American College of Rheumatology (ACR) response criteria (ACR 20) at 6 months, the proportion of patients in each group with clinically significant improvement (0.3 unit) in the Health Assessment Questionnaire Disability Index (HAQ-DI) score (11) at 1 year, and the radiographic progression of joint erosions (assessed by comparing changes from baseline in the Genant-modified Sharp score) (12, 13) at 1 year. Table 1 summarizes the outcome measures used to assess the response to treatment. Table 1. Outcome Measures for Assessing Response to Treatment of Rheumatoid Arthritis* Secondary objectives included assessing ACR 50 and ACR 70 responses at 6 months and all ACR responses at 1 year. In addition, we determined the proportions of patients achieving a major clinical response and a protocol-defined extended major clinical response at 1 year. We also assessed changes in disease activity by using the Disease Activity Score 28 (DAS28) (20, 21). We assessed improvements in physical function over 1 year by using the HAQ-DI, which measures physical function during daily activities (22). We evaluated changes in health-related quality of life by using the Medical Outcomes Study Short Form-36 Health Survey (SF-36) (17), which evaluates physical and mental health status (Table 1) (18, 19). Physicians blinded to treatment group assignment performed assessments at enrollment and at every visit before treatment administration on days 1, 15, and 29; every 28 days up to and including day 169 (6 months); and on days 225, 281, and 365 (1 year). Radiographic Evaluation We performed standardized radiography of the hands or wrists and feet at baseline and at 1 year or upon early termination (if applicable). Two independent expert readers who were blinded to treatment group assignment, chronological order of radiography, and patients' clinical response assessed all radiographic images for changes in erosion and joint-space narrowing by using the Genant-modified Sharp scoring system. Safety and Immunogenicity We monitored all patients who received at least 1 dose of the study medication for adverse events, serious adverse events, infusion reactions, clinical laboratory test abnormalities, and clinically significant changes in vital signs. Adverse events were self-reported by the patient and elicited by general questioning and examination at each visit. We attributed an adverse event to the study treatment on the basis of the investigator's opinion, and we deemed an event as serious by standard regulatory definition. An external safety advisory panel, consisting of 5 physicians (3 rheumatologists, 1 oncologist, and 1 infectious disease expert), assessed overall safety in a blinded fashion by using reports of adverse events and laboratory results on a quarterly basis. We obtained serum samples before infusions on days 1, 29, 85, 169, 281, and 365 or 28 days after the last dose of the study medication in patients who discontinued before 1 year. We assessed immunogenicity by immunoassay to measure the antibody response to the entire abatacept molecule and also specifically to the CTLA-4 portion of the molecule (7). Statistical Analysis The protocol estimated that 680 patients would need to be enrolled to randomly assign 540 patients. We based sample sizes on a 5% level of significance (2-tailed). The study had 99% power to detect a difference of 20% in ACR 20 between the 2 groups. On the basis of the hierarchic

R heumatoid arthritis is characterized by synovial mem- brane hyperplasia and inflammatory cell infiltrate, including activated T cells (1). T cells contribute to the initiation and perpetuation of rheumatoid arthritis immunopathology, leading to inflammation and, ultimately, joint destruction. Activated T cells proliferate and induce monocytes, macrophages, and synovial fibroblasts to produce proinflammatory cytokines, such as tumor necrosis factor-␣, interleukin-1, and interleukin-6 (1), and stimulate osteoclastogenesis and matrix metalloproteinase secretion (2), as well as immunoglobulin production by B cells (3). The central role of activated T cells in rheumatoid arthritis immunopathology makes T-cell activation a rational therapeutic target.

T cells require 2 signals for full activation: an antigenspecific signal (signal 1) and a co-stimulatory signal (signal 2) (4). One of the best-characterized co-stimulatory pathways is the engagement of CD80 or CD86 on antigen-presenting cells with CD28 on T cells (5). In the normal immune response, endogenous cytotoxic T-lymphocyte antigen-4 (CTLA-4) downregulates CD28-mediated T-cell activation by binding to CD80 or CD86 with higher avidity than CD28 (6).

Abatacept is a soluble, recombinant, fully human fu-sion protein, comprising the extracellular domain of CTLA-4 and the Fc portion of IgG1, modified to prevent complement fixation. Abatacept is the first in a new class of agents for treating rheumatoid arthritis that selectively modulate the co-stimulatory signal required for full T-cell activation. A phase IIa study of patients with rheumatoid arthritis and an inadequate response to disease-modifying antirheumatic drugs showed the efficacy of abatacept as monotherapy (7). In a phase IIb study of abatacept plus methotrexate in patients with rheumatoid arthritis and an inadequate response to methotrexate, signs and symptoms of rheumatoid arthritis, physical function, and health-related quality of life statistically significantly improved over 1 year (8,9). We present findings from the phase III, 1-year Abatacept in Inadequate Responders to Methotrexate (AIM) trial, which was designed to further evaluate the safety and clinical efficacy of abatacept plus methotrexate and to assess the effects of abatacept on the radiographic progression of structural damage.

METHODS

The institutional review boards or independent ethics committees approved a common clinical protocol for each site, and we performed the study in accordance with the ethical principles of the Declaration of Helsinki. All patients provided written informed consent to the study protocol before randomization.

Patients

Eligible patients were at least 18 years of age, had had rheumatoid arthritis for at least 1 year, and met the American Rheumatism Association criteria for rheumatoid ar-thritis (10). Rheumatoid arthritis was persistent and active despite methotrexate treatment. All patients must have been treated with methotrexate (Ն15 mg/wk) for 3 months or longer, with a stable dose for 28 days before enrollment. We required patients to undergo a washout of all other disease-modifying antirheumatic drugs at least 28 days before randomization. We allowed corticosteroid use, with dosages equal to 10 mg of prednisone or less per day, stabilized for 25 days before randomization.

At randomization, we required patients to have 10 or more swollen joints, 12 or more tender joints, and C-reactive protein levels of 10.0 mg/L or greater (normal range, 1.0 mg/L to 4.0 mg/L) while receiving methotrexate. We required tuberculin skin testing before randomization. We excluded patients with a positive tuberculin skin test result unless they had completed treatment for latent tuberculosis before enrollment.

Study Design

Our 1-year, multicenter, multinational, randomized, double-blind, placebo-controlled study aimed to compare the efficacy and safety of abatacept versus placebo in combination with methotrexate in patients with rheumatoid arthritis and an inadequate response to methotrexate treatment. We used a central randomization system, and the Drug Management Group within Bristol-Myers Squibb, Princeton, New Jersey, generated the randomization schedule. Stratification per site was not performed. Patients were randomly assigned in a 2:1 ratio to receive either a fixed dose of abatacept, approximately 10 mg/kg of body weight, or placebo. Patients weighing less than 60 kg, 60 to 100 kg, or more than 100 kg received 500 mg, 750 mg, or 1000 mg of abatacept, respectively. We administered study medication by 30-minute intravenous infusion on days 1, 15, and 29 and then every 28 days up to and including day 337. No premedication was required.

The protocol specified that all patients were to receive methotrexate, 15 mg or more per week, although methotrexate at 10 mg per week was acceptable if the patient had a history of toxicity. During the first 6 months, we did not allow adjustments in methotrexate dose, except in cases of toxicity. We permitted use of stable dosages of nonsteroidal anti-inflammatory drugs and corticosteroid dosages equal to 10 mg of prednisone or less per day. Between 6 and 12 months, we allowed the following adjustments, as the investigator deemed necessary: 1) adjustment in methotrexate dose, 2) addition of 1 other disease-modifying antirheumatic drug (hydroxychloroquine, sulfasalazine, gold, or azathioprine), or 3) adjustment in corticosteroid dose equal to 10 mg of prednisone or less per day. However, investigators were blinded to treatment group assignment throughout the 1-year study.

Clinical Efficacy Measures

Our 3 primary objectives were to evaluate the proportion of patients in each group with a 20% improvement in American College of Rheumatology (ACR) response crite-

Context

Abatacept, an agent that selectively modulates the costimulatory signal required for T-cell activation, may benefit some patients with rheumatoid arthritis.

Contribution

This 1-year, randomized, double-blind trial compared once-monthly infusions of abatacept with placebo in 652 patients with symptomatic rheumatoid arthritis despite ongoing methotrexate treatment. Compared with placebo recipients, patients who received abatacept more often had improved physical function, more frequently met standard response criteria, and less often had radiographic progression of joint damage. They also had serious infections (2.5% vs. 0.9%) and infusion reactions more often.

Implications

Adding abatacept can reduce disease activity in patients with rheumatoid arthritis and an inadequate response to methotrexate.

-The Editors ria (ACR 20) at 6 months, the proportion of patients in each group with clinically significant improvement (Ն0.3 unit) in the Health Assessment Questionnaire Disability Index (HAQ-DI) score (11) at 1 year, and the radiographic progression of joint erosions (assessed by comparing changes from baseline in the Genant-modified Sharp score) (12,13) at 1 year.

Table 1 summarizes the outcome measures used to assess the response to treatment.

Secondary objectives included assessing ACR 50 and ACR 70 responses at 6 months and all ACR responses at 1 year. In addition, we determined the proportions of patients achieving a major clinical response and a protocoldefined extended major clinical response at 1 year. We also assessed changes in disease activity by using the Disease Activity Score 28 (DAS28) (20,21).

We assessed improvements in physical function over 1 year by using the HAQ-DI, which measures physical function during daily activities (22). We evaluated changes in health-related quality of life by using the Medical Outcomes Study Short Form-36 Health Survey (SF-36) (17), which evaluates physical and mental health status (Table 1 ) (18,19).

Physicians blinded to treatment group assignment performed assessments at enrollment and at every visit before treatment administration on days 1, 15, and 29; every 28 days up to and including day 169 (6 months); and on days 225, 281, and 365 (1 year).

Radiographic Evaluation

We performed standardized radiography of the hands or wrists and feet at baseline and at 1 year or upon early termination (if applicable). Two independent expert readers who were blinded to treatment group assignment, chronological order of radiography, and patients' clinical response assessed all radiographic images for changes in erosion and joint-space narrowing by using the Genantmodified Sharp scoring system.

Safety and Immunogenicity

We monitored all patients who received at least 1 dose of the study medication for adverse events, serious adverse events, infusion reactions, clinical laboratory test abnormalities, and clinically significant changes in vital signs. Adverse events were self-reported by the patient and elicited by general questioning and examination at each visit. We attributed an adverse event to the study treatment on the basis of the investigator's opinion, and we deemed an event as serious by standard regulatory definition. An external safety advisory panel, consisting of 5 physicians (3 rheumatologists, 1 oncologist, and 1 infectious disease expert), assessed overall safety in a blinded fashion by using reports of adverse events and laboratory results on a quarterly basis. We obtained serum samples before infusions on days 1, 29, 85, 169, 281, and 365 or 28 days after the last dose of the study medication in patients who discontinued before 1 year. We assessed immunogenicity by immunoassay to measure the antibody response to the entire abatacept molecule and also specifically to the CTLA-4 portion of the molecule (7).

Statistical Analysis

The protocol estimated that 680 patients would need to be enrolled to randomly assign 540 patients. We based sample sizes on a 5% level of significance (2-tailed). The study had 99% power to detect a difference of 20% in ACR 20 between the 2 groups. On the basis of the hierarchical testing procedure for the co-primary measures, this sample size allowed us to detect an 18% difference in HAQ-DI response rate between the 2 groups, with 98% power, and a treatment effect of 60% reduction from placebo (assuming an increase of 1.27 units in placebo for the change from baseline), with 90% power, for change from baseline in the Genant-modified Sharp erosion score. We based the assumptions on the findings of a phase IIb study of patients with rheumatoid arthritis who were using abatacept (8,9).

We performed all efficacy and safety analyses on a modified intention-to-treat population, defined as all randomly assigned patients who received at least 1 dose of study medication. We based all statistical tests on a 2-sided 5% level of significance and used SAS software, version 8.2 (SAS Institute, Cary, North Carolina), for all analyses.

For the co-primary analyses of ACR 20 at 6 months and HAQ-DI responses at 1 year, we used a 2-sided, continuity-corrected chi-square test to compare the responses of the abatacept group with those of the placebo group. We imputed missing data for patients who discontinued as nonresponders subsequent to the discontinuation; thus, we based these analyses on the full modified intention-to-treat denominator. We performed additional sensitivity analyses to assess the effect of the imputation of missing data. These include a "modified worst-case" analysis, where we imputed missing data for placebo recipients who discontinued for reasons other than lack of efficacy by using their last observed response, and a "worst-case" analysis, where we imputed missing data for placebo recipients who discontinued as responders. In both cases, however, we still imputed missing data for abatacept recipients as nonresponders. We performed additional longitudinal analyses by using the generalized estimating equations to assess the treatment effect over time. We used all available data, and the longitudinal analysis assumes that data were missing completely at random and were not dependent on current or future responses. The models included treatment, visit day, and treatment-by-visit interaction as fixed effects, and we used an unstructured covariance to account for withinpatient correlation over time (23,24).

We used a rank-based analysis of covariance (25) to compare the changes from baseline in Genant-modified Sharp scores between treatment groups at 1 year. The model included the ranks for score changes as the depen-dent variable, with treatment group as a main effect, and the ranks for baseline scores as additional covariates. Midranks were assigned for ties. The primary radiographic analyses included all observed data at baseline and at 12 months. We imputed missing annual radiographic data with linear extrapolation for discontinued patients on the basis of the baseline value and the on-treatment assessment at the time of discontinuation, provided that both assessments were available. Summary statistics and a cumulative probability plot were provided for changes from baseline in the Genant-modified Sharp scores at 1 year by treatment group assignment. We performed additional sensitivity analyses to assess the effect of missing annual radiographic data. These included analysis with imputed 12-month values for patients with missing annual assessments on the basis of the responses predicted by the data observed across both treatment groups, clustering patients with similar baseline radiographic scores. In addition, we also performed a "graded worst-case" imputation, where we imputed missing data for abatacept and placebo recipients with progressively worst outcomes and progressively best outcomes, respectively.

To avoid multiple testing, we used a prespecified sequential testing procedure for co-primary end points. We made comparisons only if all preceding co-primary end points were statistically significant, according to the following hierarchy: ACR 20 response at 6 months; functional performance at 1 year, as measured by the HAQ-DI; and change in erosion, by using the Genant-modified Sharp score, at 1 year.

The analysis of covariance with the last observation carried forward (LOCF) approach was the prespecified method for the comparisons between treatment groups of mean changes from baseline in the HAQ-DI and the 8 subscales and the physical and mental component summaries of the SF-36. However, because the limitations of the LOCF approach could yield substantial bias in treatment effects (26), and also on the basis of editorial advice, we used a longitudinal linear mixed-effects model in the comparisons of these end points. We used all available data, and the longitudinal analysis assumes that data were missing at random and were not dependent on current or future responses. The models included treatment, visit day, and treatment-by-visit interaction as fixed effects, and we used an autoregressive (1) covariance to account for within-patient correlation over time (23).

For DAS28, we used a 2-sided, continuity-corrected chi-square test to compare the responses of the abatacept group with those of the placebo group. We summarized the incidence of adverse events by treatment and used 95% CIs for the comparisons between treatment groups.

Role of the Funding Source

analysis. Interpretation of the data was aided by the funding biostatisticians, with input from the authors. The funding source was not involved in the decision to submit the article for publication.

Patient Characteristics

We enrolled 1250 patients with rheumatoid arthritis, and we randomly assigned 652 of them to treatment with abatacept (n ϭ 433) or placebo (n ϭ 219) plus methotrexate (Figure 1 ). The most frequent reason for exclusion was if a patient no longer met the study entry criteria. Baseline demographic or clinical characteristics did not notably differ between treatment groups (Table 2 ). Because of adherence issues identified during the study, we excluded patients from 1 site from all efficacy analyses before unblinding but included them in the safety analysis.

More patients in the abatacept group (89%) than in the placebo group (74%) completed 1 year of treatment (Figure 1 ). Lack of efficacy was the most common reason for discontinuation in the placebo group (18% vs. 3%). Adverse events were the most common reasons for discon-tinuation in the abatacept group (4% vs. 2%). Fewer patients discontinued the study during months 7 through 12 in the abatacept group than in the placebo group (4% vs. 5%, respectively).

During the study, the background methotrexate dosage was stable and similar in both groups (approximately 15 mg/wk), as were nonsteroidal anti-inflammatory drug and corticosteroid dosages. Between 6 and 12 months, 15 (3.7%) abatacept-treated patients versus 25 (14.4%) placebo recipients received additional disease-modifying antirheumatic drugs (azathioprine, 2 [0.5%] patients vs. 3 [1.7%] patients; sulfasalazine, 8 [2.0%] patients vs. 12 [6.9%] patients; hydroxychloroquine, 5 [1.3%] patients vs. 6 [3.5%] patients; P Ͻ 0.001).

Clinical Efficacy ACR Responses and Major Clinical Response

The ACR 20 scores statistically significantly improved at 6 months with abatacept (67.9% for abatacept vs. 39.7% for placebo; P Ͻ 0.001; difference, 28.2 percentage points [95% CI, 19.8 to 36.7 percentage points]) (Figure 2, A ). At 6 months, ACR 50 responses were 39.9% versus *MTX ϭ methotrexate. Nine abatacept-treated patients and 5 placebo recipients from 1 site were excluded from all efficacy analyses before unblinding due to nonadherence but were included in all safety analyses. 16.8% (difference, 23.0 percentage points [CI, 15.0 to 31.1 percentage points]) and ACR 70 responses were 19.8% versus 6.5% (difference, 13.3 percentage points [CI, 7.0 to 19.5 percentage points]) for abatacept versus placebo, respectively (P Ͻ 0.001 for both) (Figure 2, B and C ).

Between 6 and 12 months, all ACR responses continually improved in patients receiving abatacept, while responses in placebo recipients were largely unchanged from month 6. At 1 year, ACR 20 responses had increased to 73.1% versus 39.7% (difference, 33. 4 2 , A to C). Post hoc analyses of the abatacept group showed that the proportion of patients with ACR 50 and ACR 70 responses statistically significantly increased from 6 months to 12 months (P Ͻ 0.001 for 6 months vs. 12 months). Of the abatacept-treated patients with an ACR 70 response at 1 year, 45% maintained the response for 6 consecutive months (major clinical response overall, 14.2%) and 21% maintained the response for 9 consecutive months (extended major clinical response overall, 6.1%).

In the modified worst-case and worst-case sensitivity analyses, we observed an ACR 20 response at 6 months in more patients in the abatacept group (68% in both cases) than in the placebo group (40% in the modified worst-case and 57% in the worst-case analyses). An additional longitudinal analysis confirmed the significant increase in ACR 20 response for abatacept versus placebo (P Ͻ 0.001) and, more specifically, at day 15 (P ϭ 0.008). This early response was largely driven by rapid improvements in pain and by patients' and physicians' assessments of disease activity, which were statistically significant from day 15 onward compared with placebo (data not shown). Longitudinal analyses also demonstrated significant increases in ACR 50 and ACR 70 responses with abatacept compared with

Physical Function

At the start of the study, patients' physical function was considerably impaired (HAQ-DI score of 1.7 in both groups) (Table 2 ). At 1 year, physical function clinically significantly improved in statistically significantly more abatacept-treated patients (11) than placebo recipients (63.7% vs. 39.3%; P Ͻ 0.001; difference, 24.4 percentage points [CI, 15.9 to 32.9 percentage points]) (Figure 2, D ).

In the modified worst-case sensitivity analysis, more patients in the abatacept group (64%) had an HAQ-DI response at 1 year than those in the placebo group (42%). In the worst-case analysis, similar proportions in both treatment groups (64%) had an HAQ-DI response. How-ever, because of the extreme nature of the response imputation rule, the resulting high response rate in the placebo group does not represent an observable placebo response rate. The longitudinal analysis using the generalized estimating equations confirmed the significant increase in the proportion of patients with an HAQ-DI response for abatacept versus placebo (P Ͻ 0.001).

When we used the longitudinal linear mixed-effects approach, the mean improvement from baseline in the HAQ-DI was statistically significantly better in abatacepttreated patients than in placebo recipients at both 6 months and 1 year (P Ͻ 0.001). Results were consistent with the LOCF approach.

Radiographic Progression

We collected radiographic data for 586 (92%) randomly assigned patients at baseline and at 1 postbaseline time point. Baseline erosion, joint-space narrowing, and total scores were similar between the groups (Table 2 ). At 1 year, abatacept-treated patients demonstrated statistically significant slowing of structural damage progression compared with placebo recipients, with an approximately 50% reduction in change from baseline in Genant-modified Sharp scores compared with that of placebo. The median change from baseline in erosion score was 0.0 (25th and 75th percentiles, 0.0 and 1.0, respectively) for abatacept versus 0.27 (25th and 75th percentiles, 0.0 and 1.3, respectively) for placebo (P ϭ 0.029; Figure 3 ). Median changes in the joint-space narrowing and total scores were similar between the groups. The median change in joint-space narrowing score was 0.0 (25th and 75th percentiles, 0.0 and 0.5, respectively) for abatacept versus 0.0 (25th and 75th percentiles, 0.0 and 1.0, respectively) for placebo (P ϭ 0.009) (Figure 3 ). The median change in total score was 0.25 (25th and 75th percentiles, 0.0 and 1.8, respectively) for abatacept versus 0.53 (25th and 75th percentiles, 0.0 and 2.5, respectively) for placebo (P ϭ 0.012) (Figure 3 ). The mean change from baseline was 0.63 for abatacept versus 1.14 for placebo in erosion score, 0.58 for abatacept versus 1.18 for placebo in joint-space narrowing score, and 1.21 for abatacept versus 2.32 for placebo in total score.

The sensitivity analysis suggested that the few missing data did not statistically significantly affect the robustness of slowing the progression of structural damage. For the sensitivity analysis, in which we imputed missing 1-year radiographic values, the median changes in total score were 0.26 (interquartile range, 0.00 to 1.84) and 0.53 (interquartile range, 0.00 to 3.14) in the abatacept and placebo groups, respectively. These changes for both treatment groups were the same as those in the primary analysis, with differences observed only in the 75th percentile and a more notable increase in the placebo group. In the graded worstcase sensitivity analyses, in which several imputations increasingly favored placebo, the trend for the benefit of abatacept was maintained compared with that of placebo, even in the extreme case (data not shown).

Disease Activity

Patients exhibited high baseline disease activity (DAS28 of 6.4 for both groups [15]). At 6 months and 12 months, 30.1% and 42.5% of the abatacept group, respectively, had a DAS28 of 3.2 or less, compared with 10.0% and 9.9% of the placebo group, respectively (P Ͻ 0.001). Abatacept induced DAS28 less than 2.6 in 14.8% of abatacept recipients versus 2.8% of placebo recipients at 6 months and in 23.8% of abatacept recipients versus 1.9% of placebo recipients at 1 year (P Ͻ 0.001).

Health-Related Quality of Life

When we used the linear mixed-effects approach, both the physical (P Ͻ 0.001) and mental (P ϭ 0.009) component summaries significantly improved from baseline to 6 months (increase of Ն3 units) (18,19) in the abatacept group compared with the placebo group. At 1 year, both summary scores for patients treated with abatacept were still significant (physical component summary, P Ͻ 0.001; difference, 3.8 [CI, 2.4 to 5.2]; mental component summary, P ϭ 0.038; difference, 1.76 [CI, 0.1 to 3.4]). Results were also significant at both 6 months and 1 year with the LOCF approach.

Safety

The overall incidence of adverse events was similar in both the abatacept and placebo groups (87.3% vs. 84.0% [CI, Ϫ2.5 to 9.1 percentage points]) (Table 3 ). The most Interquartile range changes from baseline in Genant-modified Sharp erosion, joint-space narrowing (JSN), and total scores were evaluated at 1 year or at early termination (if applicable). The median (solid circles), interquartile range, and 10th and 90th percentiles (dotted lines) are shown. Data shown are from all randomly assigned and treated patients with baseline and follow-up radiography. MTX ϭ methotrexate. frequently reported adverse events (Ͼ5% in either group) included headache, nasopharyngitis, and nausea. More patients discontinued because of adverse events in the abatacept group than in the placebo group (4.2% vs. 1.8%) (Table 3).

The incidence of serious adverse events increased with abatacept treatment; rates of discontinuation due to serious adverse events were similar between the groups (Table 4 ). The most frequently reported serious adverse events were musculoskeletal, primarily related to hospitalizations for rheumatoid arthritis flares or elective surgery for rheumatoid arthritis.

The incidence of infection reported as a serious adverse event was higher with abatacept than with placebo (Table 4 ). Discontinuations due to serious infections were similar between groups (2 discontinuations [0.5%] for abatacept vs. 1 discontinuation [0.5%] for placebo). More patients in the abatacept group than in the placebo group had prespecified infections that met the criteria for a serious adverse event (a subset of all serious adverse events). We observed an increase in cases of pneumonia with abatacept treatment versus placebo treatment (Table 4 ). One abatacept-treated patient reported an enlarged lymph node, which revealed histologic findings on biopsy compatible with possible tuberculosis; however, the patient did not experience any symptom of tuberculosis and we found no bacterial evidence of tuberculosis. One case of unconfirmed tuberculosis was reported in the placebo group. Two deaths due to infections occurred. One abatacepttreated patient with underlying pulmonary charac-terized by a history of tuberculosis, asbestos exposure, and pulmonary fibrosis, died of bronchopneumonia, pulmonary aspergillosis, and Pseudomonas aeruginosa septicemia. A placebo recipient died after P. aeruginosa pneumonia, sepsis, and multiorgan failure.

The incidence of neoplasms (benign or malignant) and hematologic disorders was similar in both groups. A large B-cell lymphoma of the thyroid on a background of Hashimoto thyroiditis was reported in 1 patient receiving abatacept, and 1 endometrial carcinoma was reported in a patient receiving placebo.

No major autoimmune disorders, such as multiple sclerosis or lupus, were reported (Tables 3 and 4 ).

More infusion reactions (acute and peri-infusional) occurred with abatacept than with placebo (Table 4 ). Two patients discontinued because of severe acute infusion reactions. One patient experienced hypersensitivity (rash and chest pain) after the second infusion; the second patient experienced severe hypotension during the fourth infusion. Both events resolved shortly after cessation of infusions. Severe peri-infusional events were infrequent.

Immunogenicity

Six patients (1.4%) demonstrated antibody reactivity to abatacept. The pattern of ACR 20 responses in these patients was similar before and after the time of antibody response, and no patient had a hypersensitivity reaction.

DISCUSSION

The phase III AIM study confirmed and extended the findings of a phase IIb study in a similar patient population (8,9) by demonstrating that abatacept, with background methotrexate, is effective in reducing the signs and symptoms of rheumatoid arthritis and improving physical function and health-related quality of life. The study also demonstrated that abatacept slows the progression of structural damage in patients with moderate-to-severe disease and an inadequate response to methotrexate treatment.

We observed a rapid and sustained increase in all ACR responses through 1 year with abatacept. Both pain and self-assessment of disease activity statistically significantly improved in abatacept-treated patients, as early as day 15 in some patients. A total of 28.8% of abatacept-treated patients exhibited an ACR 70 response at 1 year, and a statistically significant number of these patients maintained their ACR 70 response for 6 and 9 consecutive months. Statistically significant proportions of patients in the abatacept group had DAS28 less than 2.6 or 3.2 or less at 6 months, and scores continued to increase through 1 year. The increasing efficacy observed during 1 year of abatacept treatment was not due to the allowance of additional medications during the second half of the study. Few abatacept-treated patients received additional disease-modifying antirheumatic drugs (3.7% for abatacept vs. 14.4% for placebo) during months 7 through 12 of treatment.

We believe that our study is the first to demonstrate that abatacept statistically significantly slows the progression of structural damage. One-year radiographic data, from 92% of patients, indicate that the progression of structural damage was reduced by approximately 50% with abatacept compared with placebo. Additional sensitivity analyses to assess the effect of imputation suggest that the few missing data (8%) did not statistically significantly affect these findings. Abatacept demonstrated this protective effect in a wide range of patients, including those with relatively long-standing disease, with irreversible damage and highly progressed lesions. These findings correspond well with the observed improvements in clinical and functional end points. We should note that the actual degree of progression in both groups was relatively low in our study, although the primary end point of reduction in the progression of erosion was met and was statistically significant for abatacept versus placebo. The clinical relevance of the finding may require additional long-term observations in patients with rheumatoid arthritis who receive abatacept for prolonged periods of time.

Our safety findings were consistent with those of a previous phase IIb abatacept study in a similar patient population (8,9). Other than headache and nasopharyngitis, which were more frequent with abatacept, the incidence of the most commonly reported adverse events was similar for both treatment groups. Discontinuations due to adverse events occurred in only 4.2% of patients receiving abatacept compared with 1.8% of those receiving placebo. Serious adverse events occurred in 15% of abatacept-treated patients and 11.9% of placebo recipients.

Prespecified serious infections occurred in 2.5% of patients receiving abatacept and 0.9% of patients receiving † Acute infusional events occurred within 1 hour of the start of the infusion; peri-infusional events occurred within 24 hours of the start of the infusion. ‡ A subset of overall infections classified as serious adverse events, prespecified as those serious infections that may be associated with use of immunomodulatory drugs. § Unconfirmed. Not reported as a serious infection.

placebo. One case of aspergillosis occurred in the abatacept group. One case of tuberculosis was reported in each group; however, neither case of tuberculosis was confirmed bacteriologically. We included patients who were from countries where tuberculosis is endemic; however, as recommended with antitumor necrosis factor agents used for treating rheumatoid arthritis, which have shown an increased incidence of tuberculosis, we screened all patients by tuberculin skin test before study entry. We excluded patients with a positive antituberculin skin test result, and therefore, comparison of the rates of tuberculosis with the early experience of antitumor necrosis factor agents, in which routine tuberculin skin testing was not performed, would not be appropriate. Additional longer-term information is needed to determine whether an increased relative risk for tuberculosis or other opportunistic infection is associated with the use of this agent. Increases in hematologic and hepatic abnormalities and malignant conditions were not associated with abatacept treatment. In addition, no major autoimmune diseases were reported (for example, lupus or demyelination), as observed with anticytokine therapies (27). These findings should be considered within the context of our study's limitations. The trial's 1-year duration precludes the determination of whether longer-term treatment will be associated with the emergence of other possible toxicities. Detection of a range of toxicities will require more widespread study in more patients. Furthermore, we examined the safety and efficacy of abatacept in only 1 subset of the patient population with rheumatoid arthritis-those with an inadequate response to methotrexate. An additional phase III study of abatacept in patients with rheumatoid arthritis and an inadequate response to antitumor necrosis factor-␣ therapy demonstrated statistically significant clinical benefits with a similar safety and tolerability profile (28). Our current trial assessed the effects of abatacept in patients with established rheumatoid arthritis (mean duration of about 9 years) and was not designed to investigate the effects of abatacept in early disease. Further studies of abatacept in the longer-term treatment of patients with rheumatoid arthritis and varying disease and treatment histories are required to substantiate the efficacy and safety findings with abatacept to date.

The data from the phase IIb trial (8,9) and the larger, optimal-dose, phase III investigation indicate that the strategy of selective T-cell inhibition by abatacept provides consistent and statistically significant additional therapeutic value for treating patients with rheumatoid arthritis and an inadequate response to methotrexate. Abatacept treatment is, thus, an alternative strategy to inhibit tumor necrosis factor-␣ in these patients, although the relative merits of each approach may require several years to determine via information derived from large databases or registries.

In our study, the clinical benefits seen with the fixed dosage of abatacept encompassed clinical and radiographic efficacy, statistically significant and clinically meaningful improvements in patients' physical function and healthrelated quality of life, and a consistent safety profile. Observations of the slowing of radiographic progression by abatacept, which we believe that our study is the first to demonstrate, as well as the safety and clinical findings, are expected to be extended with longer-term observations in this and other patient populations. Overall, abatacept seems to be a rational and effective treatment strategy for patients with rheumatoid arthritis who have an inadequate response to weekly methotrexate.

From the Center for Rheumatology, Albany, New York; University of California, San Francisco, and Synarc Inc., San Francisco, California; University of Alabama at Birmingham School of Medicine, Birmingham, Alabama; University of Alberta, Edmonton, Alberta, Canada; Leeds General Infirmary, Leeds, United Kingdom; Hospital Central, San Luis Potosi, Mexico; University of Medical Sciences, Wroclaw, Poland; Bristol-Myers Squibb, Princeton, New Jersey; and University Hospital Leuven, Leuven, Belgium.

Grant Support: By Bristol-Myers Squibb. Requests for Single Reprints: Joel M. Kremer, MD, Center for Rheumatology, 1367 Washington Avenue, Suite 1, Albany, NY 12206; email, jkremer@joint-docs.com.

Potential Financial Conflicts of

Current author addresses and author contributions are available at www .annals.org.

Failed to load PDF:

Paper sources

Abstract screening pilot

Abstract screening results

Extraction pilot

Extraction results

Research report

Modify setup

StripeM-Inner