ACR 50 response was defined as a 50% improvement from baseline to Month 12 in tender and swollen joint counts and 50% improvement in 3 of the remaining 5 core set measures (subject global assessment of pain, subject global assessment of disease activity, physician global assessment of disease activity, subject assessment of physical function), and 1 acute phase reactant value [ie, CRP].

Intention-to-Treat=All randomized and treated; all missing data subsequent to discontinuation are considered non-responders.

MCR was defined as 6 months of consecutive ACR 70 response at Month 12. ACR 70, the American College of Rheumatology (ACR) definition of 70% improvement was based on a 70% improvement (compared to baseline values) in tender and swollen joint counts and 70% improvement in 3 of the remaining 5 core set measures (subject global assessment of pain, subject global assessment of disease activity, physician global assessment of disease activity, subject assessment of physical function, and 1 acute phase reactant value [ie, CRP]).

Intention-to-Treat=All randomized and treated; all missing data subsequent to discontinuation are considered non-responders.

DAS 28-CRP is a continuous variable that is a composite of 4 variables: the number of tender joints out of 28 joints, the number of swollen joints out of 28 joints, CRP in milligrams/Liter (mg/L), and subject assessment of disease activity measure on a Visual Analogue Scale (VAS) of 100 millimeters (mm). The DAS28 scale=0 to 10, indicating the current activity of the rheumatoid arthritis. A DAS28 >5.1=high disease activity; <3.2=low disease activity; <2.6=remission. Change from Baseline=Post-baseline - Baseline value; Adjusted for baseline value.

Last Observation Carried Forward (LOCF) Intent to Treat population = all randomized and treated. As change from baseline analysis, only those with baseline and post-baseline included.

Physical function was evaluated using the HAQ-disability index (HAQ-DI), a questionnaire with 20 questions assessing function in 8 domains: dressing, arising, eating, walking, hygiene, reach, grip, and common activities. The questions are evaluated on a 4-point scale: 0 = without any difficulty; 1 = with some difficulty; 2 = with much difficulty; 3 = unable to do. The 8 category scores are averaged into an overall HAQ-DI score on a scale from 0 (no disability) to 3 (completely disabled). Higher scores indicate greater dysfunction. HAQ response=improvement of at least 0.3 units from baseline.

Intention-to-Treat=All randomized and treated; all missing data subsequent to discontinuation are considered non-responders.

The SF-36 covers 8 health dimensions: 4 physical subscales (physical function, role-physical, bodily pain, and general health) and 4 mental subscales (vitality, social function, role-emotional, and mental health). The scores range from 0 to 100, with a higher score indicating better quality of life. Two summary scores (physical and mental component summaries) were produced taking a weighted linear combination of the 8 individual subscales. Change from Baseline=Post-baseline - Baseline value; adjusted for baseline value.

Last Observation Carried Forward (LOCF) Intent to Treat population = all randomized and treated. As change from baseline analysis, only those with baseline and post-baseline included.

To assess joint damage progression, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and JSN. The erosion score range is 0 (no radiographic damage) to 145 (worst possible radiographic damage). The joint space narrowing score range is 0 (no radiographic damage) to 145 (worst possible radiographic damage). Higher scores indicated more damage. Change from baseline = Post-baseline - Baseline value

Intention-to-Treat - linear extrapolation imputation. Analysis of change from baseline restricts subjects included in to the analysis to those with baseline and post-baseline.

Serum samples were analyzed by ELISA to detect antibodies against the whole molecule (both CTLA4 and Ig [anti-abatacept antibody]) or solely to CTLA4 (anti-CTLA4-T antibody). Reported as titer, the reciprocal of the sample dilution which yielded a signal equivalent to the statistically set cut point for the assay. For the anti-abatacept assay, minimum required dilution is 400-fold, therefore seronegative samples are those < lowest reportable titer (<400). For the anti-CTLA4-T assay, minimum required dilution is 25-fold, therefore seronegative samples are those < lowest reportable titer (<25).

Treated participants in the double-blind period who were evaluated for anti-abatacept or anti-CTLA4-T responses

Number of participants who achieved remission at Month 12 of treatment, as defined by a Disease Activity Score (DAS) 28-CRP score of <2.6. DAS 28-CRP is a continuous measure, a composite of 4 variables: number of tender joints out of 28 joints, number of swollen joints out of 28 joints, CRP (in mg/L), and subject assessment of disease activity measure on a Visual Analogue Scale (VAS) of 100 millimeters (mm). The DAS28 scale=0 (best) to 10 (worst), indicating the current activity of the rheumatoid arthritis. A DAS28 >5.1 = high disease activity; <=3.2 = low disease activity; <2.6 = remission.

Intent to treat = all randomized and treated subjects. Those with missing data post-discontinuation were considered non-responders.

To assess joint damage progression, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and joint space narrowing (JSN). The total Genant-modified Sharp score ranges from 0 (no radiographic damage) to 290 (worst possible radiographic damage) and is the sum of the erosion score (range 0-145) and the joint space narrowing score (range 0-145). Higher scores indicated more damage.

The analysis was intent-to-treat. Because the analysis was change from baseline, only those with baseline and post-baseline were included. Linear extrapolation imputation was applied.

AE=any new untoward medical occurrence or worsening of a pre-existing medical condition in a subject administered an investigational product and that does not necessarily have a causal relationship with this treatment. Related AE/SAE=Certain, Probable, Possible, or Missing. SAE=any untoward medical occurrence that results in death, is life-threatening, requires or prolongs inpatient hospitalization (including elective surgery), results in persistent or significant disability/incapacity, is a congenital anomaly/birth defect, or is an important medical event.

All subjects who received at least 1 dose of ABA in the open-label period were included in the safety analyses.

SAE=any untoward medical occurrence that results in death, is life-threatening, requires or prolongs inpatient hospitalization (including elective surgery), results in persistent or significant disability/incapacity, is a congenital anomaly/birth defect, or is an important medical event.

All subjects who received at least 1 dose of ABA in the open-label period were included in the safety analyses.

Serum samples were analyzed by ELISA to detect antibodies against the whole molecule (both CTLA4 and Ig [anti-abatacept antibody]) or solely to CTLA4 (anti-CTLA4-T antibody). Reported as titer, the reciprocal of the sample dilution which yielded a signal equivalent to the statistically set cut point for the assay. For the anti-abatacept assay, minimum required dilution is 400-fold, therefore seronegative samples are those < lowest reportable titer (<400). For the anti-CTLA4-T assay, minimum required dilution is 25-fold, therefore seronegative samples are those < lowest reportable titer (<25).

Treated participants in the open-label period were evaluated for anti-abatacept or anti-CTLA4-T responses

Physical function was evaluated using the HAQ-disability index (HAQ-DI), a questionnaire with 20 questions assessing function in 8 domains: dressing, arising, eating, walking, hygiene, reach, grip, and common activities. The questions are evaluated on a 4-point scale: 0 = without any difficulty; 1 = with some difficulty; 2 = with much difficulty; 3 = unable to do. The 8 category scores are averaged into an overall HAQ-DI score on a scale from 0 (no disability) to 3 (completely disabled). Higher scores indicate greater dysfunction. HAQ response=improvement of at least 0.3 units from baseline.

All treated participants in the Open-label period. Treatment groups represent treatment received in the Double Blind Period.

To assess joint damage progression, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and JSN. The total Genant-modified Sharp score ranges from 0 (no radiographic damage) to 290 (worst possible radiographic damage) and is the sum of the erosion score (range 0-145) and the joint space narrowing score (range 0-145). Higher scores indicated more damage. Change from baseline = Postbaseline - baseline value.

All treated participants in the open-label period. Because the analysis was change from baseline, only those with baseline and post-baseline were included. Linear extrapolation imputation was applied. Treatment groups represent treatment received in the double-blind period.

Any untoward medical occurrence (SAE) that resulted in death

All subjects who received at least 1 dose of ABA in the open-label period were included in the safety analyses.

The incidence rates of autoimmune disorders are defined as the (number of patients experiencing the event/exposure within the period)*100 and are expressed in 100 person-years. Subjects experiencing the event had their exposure censored at the time of the 1st event.

All subjects who received at least 1 dose of abatacept. Double-blind (DB) period: all treated in DB period; Open-label (OL) Period: all treated in OL period.

The incidence rates of infections and infestations are defined as the (number of patients experiencing the event /exposure within the period)*100 and are expressed in 100 person-years. Subjects experiencing the event had their exposure censored at the time of the 1st event.

All subjects who received at least 1 dose of abatacept. Double-blind (DB) period: all treated in DB period; Open-label (OL) Period: all treated in OL period.

The incidence rates of malignant neoplasms are defined as the (number of patients experiencing the event /exposure within the period)*100 and are expressed in 100 person-years. Subjects experiencing the event had their exposure censored at the time of the 1st event.

All subjects who received at least 1 dose of abatacept. Double-blind (DB) period: all treated in DB period; Open-label (OL) Period: all treated in OL period.

There were 107 Prespecified, acute-infusional SAEs (occurring within 1 hour after the start of study drug infusion) pre-specified in the protocol; anaphylactic shock was the only one occuring in this study.

All participants treated during the Open-Label period.

Number of subjects with high liver function and kinedy tests: alkaline phosphatase (ALP) >2x upper limit of normal (ULN) or if pretreatment (PRE-RX) >ULN then >3x PRE-RX; aspartate aminotransferase (AST) >3x ULN or if PRE-RX >ULN then >4x PRE-RX; alanine aminotransferase (ALT) >3x ULN or if PRE-RX >ULN then >4x PRE-RX; g-glutamyl transferase (GGT)>2x ULN or if PRE-RX >ULN then >3x PRE-RX; total bilirubin >2x ULN or if PRE-RX >ULN then >4x PRE-RX; blood urea nitrogen >2x PRE-RX; creatinine >1.5x PRE-RX.

All Treated participants in the Open-label Period; n=number of participants evaluated for this measure.

Marked abnormalities in hemoglobin >3 g/dL decrease from PRE-RX; hematocrit <0.75x PRE-RX; erythrocytes <0.75x PRE-RX; platelet count <0.67x lower limit of normal (LLN) or >1.5x ULN or if PRE-RX <LLN then <0.5x PRE-RX and <100,000/mm3; leukocytes <0.75x LLN or >1.25x ULN or if PRE-RX <LLN then <0.8x PRE-RX or >ULN if PRE-RX >ULN then >1.2x PRE-RX or <LLN; neutrophils if value <1.00 x10^3 c/uL; lymphocytes if value <.750 x10^3 c/uL or if value >7.50 x10^3 c/uL; monocytes if value >2000/MM3; basophils if value >400/mm3; eosinophils if value >.750 x10^3 c/uL

All participants treated during the open-label period; n= number of participants evaluated for this measure.

Participants with no radiographic progression (defined as change in score <=0 or <=0.5), from baseline to Month 24. To assess joint damage progression, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and JSN. The total Genant-modified Sharp score ranges from 0 (no radiographic damage) to 290 (worst possible radiographic damage) and is the sum of the erosion score (range 0-145) and the joint space narrowing score (range 0-145). Higher scores indicated more damage.

All treated participants in the open-label period. Treatment groups represent treatment received in the double-blind period.

Participants with no radiographic progression ((defined as change in score <=0 or <=0.5), sustained from Month 12 and Month 24. To assess joint damage progression, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and JSN. The total Genant-modified Sharp score ranges from 0 (no radiographic damage) to 290 (worst possible radiographic damage) and is the sum of the erosion score (range 0-145) and the joint space narrowing score (range 0-145). Higher scores indicated more damage.

Number of Participants Analyzed=All treated participants in the open-label period. (Treatment groups represent treatment received in the double-blind period.) n=the number of subjects with observed data included in the analysis.

Mean difference observed in change from baseline to Month 12 and between Month 12 and Month 24 in radiographic scores (Total Score). To assess joint damage, the Genant-modified Sharp scoring method was used to evaluate radiographs of hands/wrists and feet for erosions and JSN. The total Genant-modified Sharp score ranges from 0 (no radiographic damage) to 290 (worst possible radiographic damage) and is the sum of the erosion score (range 0-145) and the joint space narrowing score (range 0-145). Higher scores indicated more damage.

Analysis includes all treated participants in the open-label period originally randomized to abatacept. Analysis includes all participants with observed assessments collected at Baseline (Day 1), Day 365 (Month 12), and Day 729 (Month 24)

AE=any new untoward medical occurrence or worsening of a pre-existing medical condition in a subject administered an investigational product and that does not necessarily have a causal relationship with this treatment. Related AE/SAE=Certain, Probable, Possible, or Missing. SAE=any untoward medical occurrence that results in death, is life-threatening, requires or prolongs inpatient hospitalization (including elective surgery), results in persistent or significant disability/incapacity, is a congenital anomaly/birth defect, or is an important medical event.

All treated participants in the Double-Blind period

Number of participants with laboratory values (hematology, liver and kidney functions, electrolytes, glucose tests, protein tests, metabolite tests, and urine chemistry tests) considered markedly abnormal according to prespecified protocol criteria

All treated in the DB period. n=Number of participants evaluated for this measure.