A Study to Evaluate the Safety and Efficacy of Rituximab in Combination With Methotrexate Compared to Methotrexate Alone in Patients With Active Rheumatoid Arthritis

To achieve an ACR20 required at least a 20% improvement compared with Baseline in both tender joint counts (68 joints assessed for tenderness) and swollen joint counts (66 joints assessed for swelling), as well as a 20% improvement in three of the following five additional measurements: * Physician's global assessment of disease activity (assessed using a 100 mm Visual Analog Scale \[VAS\]); * Patient's global assessment of disease activity (assessed using a 100 mm VAS); * Patient's assessment of pain (assessed using a 100 mm VAS); * Health Assessment Questionnaire (HAQ; a patient completed questionnaire consisting of 20 questions, scored from 0-3); * Acute phase reactant: C-reactive protein (CRP) or, if CRP was missing, erythrocyte sedimentation rate (ESR). Participants who withdrew prematurely from the study prior to Week 24, who received rescue therapy or had insufficient data in order to calculate a clinical response were considered to be non-responders.

Time frame: Baseline and Week 24

Population: Intent to treat population included all randomized participants who received at least 1 or part of an infusion. ACR was calculated using the last observation carried forward (LOCF) values for each component. Participants who withdrew prior to week 24, received rescue therapy or had insufficient data to calculate ACR were considered non-responders.

The DAS28 is a composite score to measure disease activity in patients with rheumatoid arthritis, derived from the following variables: * The number of swollen and tender joints assessed using the 28-joint count; * Erythrocyte sedimentation rate (ESR); * Patient's global assessment of disease activity measured on a 100 mm visual analog scale. The DAS28 score ranges from zero to ten. A DAS28 score above 5.1 means high disease activity whereas a DAS28 less than or equal to 3.2 indicates low disease activity. Remission is achieved by a DAS28 lower than 2.6.

Time frame: Baseline and Week 24

Population: Intent to treat population with available data. DAS28 was calculated using last observation carried forward values for each of the component variables. If any of the components were missing then the DAS28 value will be missing.

The FACIT-Fatigue questionnaire is a self-administered patient questionnaire that consists of 13 questions designed to measure the degree of fatigue experienced by participants in the previous 7 days. Participants respond to the questions using a value in the range of 0 (not at all) to 4 (very much). The scale score is computed by summing the item scores, after reversing those items that are worded in the negative direction. The FACIT-Fatigue subscale score ranges from 0 to 52, where higher scores represent less fatigue. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions in eight domains (physical function, pain, general and mental health, vitality, social function, physical and emotional health). The individual domain scores are calculated and transformed to range from 0 to 100, with higher scores indicating a better level of functioning. A positive change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population. N indicates the number of participants with non-missing data at each time point.

To achieve an ACR50 required at least a 50% improvement compared with Baseline in both tender joint counts (68 joints assessed for tenderness) and swollen joint counts (66 joints assessed for swelling), as well as a 50% improvement in three of the following five additional measurements: * Physician's global assessment of disease activity (assessed using a 100 mm Visual Analog Scale \[VAS\]); * Patient's global assessment of disease activity (assessed using a 100 mm VAS); * Patient's assessment of pain (assessed using a 100 mm VAS); * Health Assessment Questionnaire (HAQ; a patient completed questionnaire consisting of 20 questions, scored from 0-3); * Acute phase reactant: C-reactive protein (CRP) or, if CRP was missing, erythrocyte sedimentation rate (ESR). Participants who withdrew prematurely from the study prior to Week 24, who received rescue therapy or had insufficient data in order to calculate a clinical response were considered to be non-responders.

Time frame: Baseline and Week 24

Population: Intent to treat population. ACR was calculated using LOCF values for each component. Participants who withdrew prior to Week 24, received rescue therapy or had insufficient data to calculate ACR were considered non-responders.

To achieve an ACR50 required at least a 50% improvement compared with Baseline in both tender joint counts (68 joints assessed for tenderness) and swollen joint counts (66 joints assessed for swelling), as well as a 50% improvement in three of the following five additional measurements: * Physician's global assessment of disease activity (assessed using a 100 mm Visual Analog Scale \[VAS\]); * Patient's global assessment of disease activity (assessed using a 100 mm VAS); * Patient's assessment of pain (assessed using a 100 mm VAS); * Health Assessment Questionnaire (HAQ; a patient completed questionnaire consisting of 20 questions, scored from 0-3); * Acute phase reactant: C-reactive protein (CRP) or, if CRP was missing, erythrocyte sedimentation rate (ESR). Participants who withdrew prematurely from the study prior to week 48, who received rescue therapy or had insufficient data in order to calculate a clinical response were considered to be non-responders.

Time frame: Baseline and Week 48

Population: Intent to treat population. ACR was calculated using LOCF values for each component. Participants who withdrew prior to Week 48, received rescue therapy or had insufficient data to calculate ACR were considered non-responders.

To achieve an ACR70 required at least a 70% improvement compared with Baseline in both tender joint counts (68 joints assessed for tenderness) and swollen joint counts (66 joints assessed for swelling), as well as a 70% improvement in three of the following five additional measurements: * Physician's global assessment of disease activity (assessed using a 100 mm Visual Analog Scale \[VAS\]); * Patient's global assessment of disease activity (assessed using a 100 mm VAS); * Patient's assessment of pain (assessed using a 100 mm VAS); * Health Assessment Questionnaire (HAQ; a patient completed questionnaire consisting of 20 questions, scored from 0-3); * Acute phase reactant: C-reactive protein (CRP) or, if CRP was missing, erythrocyte sedimentation rate (ESR). Participants who withdrew prematurely from the study prior to Week 24, who received rescue therapy or had insufficient data in order to calculate a clinical response were considered to be non-responders.

Time frame: Baseline and Week 24

Population: Intent to treat population. ACR was calculated using LOCF values for each component. Participants who withdrew prior to Week 24, received rescue therapy or had insufficient data to calculate ACR were considered non-responders.

To achieve an ACR70 required at least a 70% improvement compared with baseline in both tender joint counts (68 joints assessed for tenderness) and swollen joint counts (66 joints assessed for swelling), as well as a 70% improvement in three of the following five additional measurements: * Physician's global assessment of disease activity (assessed using a 100 mm Visual Analog Scale \[VAS\]); * Patient's global assessment of disease activity (assessed using a 100 mm VAS); * Patient's assessment of pain (assessed using a 100 mm VAS); * Health Assessment Questionnaire (HAQ; a patient completed questionnaire consisting of 20 questions, scored from 0-3); * Acute phase reactant: C-reactive protein (CRP) or, if CRP was missing, erythrocyte sedimentation rate (ESR). Participants who withdrew prematurely from the study prior to Week 48, who received rescue therapy or had insufficient data in order to calculate a clinical response were considered to be non-responders.

Time frame: Baseline and Week 48

Population: Intent to treat population. ACR was calculated using LOCF values for each component. Participants who withdrew prior to Week 48, received rescue therapy or had insufficient data to calculate ACR were considered non-responders.

The DAS28 is a composite score to measure disease activity in patients with rheumatoid arthritis, derived from the following variables: * The number of swollen and tender joints assessed using the 28-joint count; * Erythrocyte sedimentation rate (ESR); * Patient's global assessment of disease activity measured on a 100 mm visual analog scale. The DAS28 score ranges from zero to ten. DAS28 above 5.1 indicates high disease activity. Low disease activity is defined by a DAS28 score less than or equal to 3.2. Remission is defined by a DAS28 score less than 2.6.

Time frame: Week 24

Population: Intent to treat population with available data. DAS28 was calculated using last observation carried forward values for each of the component variables. If any of the components were missing then the DAS28 value will be missing. Number of participants analyzed = participants who were evaluable for this outcome.

The DAS28 is a composite score to measure disease activity in patients with rheumatoid arthritis, derived from the following variables: * The number of swollen and tender joints assessed using the 28-joint count; * Erythrocyte sedimentation rate (ESR); * Patient's global assessment of disease activity measured on a 100 mm visual analog scale. The DAS28 score ranges from zero to ten. DAS28 above 5.1 indicates high disease activity. Low disease activity is defined by a DAS28 score less than or equal to 3.2. Remission is defined by a DAS28 score less than 2.6.

Time frame: Week 48

Population: Intent to treat population including participants with available data. DAS28 was calculated using last observation carried forward values for each of the component variables. If any of the components were missing then the DAS28 value will be missing.

A EULAR response reflects an improvement in disease activity and an attainment of a lower degree of disease activity based on the DAS28 score. The DAS28 score ranges from 0-10, with higher scores indicating more disease activity. A Good Response is defined as an improvement (decrease) in the DAS28 of more than 1.2 compared with Baseline and attainment of a DAS28 score of less than or equal to 3.2. A Moderate Response is defined as either: * an improvement (decrease) in the DAS28 of greater than 0.6 and less than or equal to 1.2 from Baseline and attainment of a DAS28 score of less than or equal to 5.1 or, * an improvement (decrease) in the DAS28 of more than 1.2 from Baseline and attainment of a DAS28 score of greater than 3.2. No Response is defined as either an improvement (decrease) in the DAS28 of less than or equal to 0.6, or an improvement (decrease) in the DAS28 of greater than 0.6 and less than or equal to 1.2 and attainment of a DAS28 of more than 5.1.

Time frame: Baseline and Week 24

Population: Intent-to-treat population. LOCF was used for the individual components of the DAS-28. Non-responder imputation was used. Participants who withdrew prior to Week 24, who received rescue therapy or had insufficient data in order to calculate a EULAR response were considered non-responders.

A EULAR response reflects an improvement in disease activity and an attainment of a lower degree of disease activity based on the DAS-28 score. A Good Response is defined as an improvement (decrease) in the DAS28 of more than 1.2 compared with Baseline and attainment of a DAS28 score less than or equal to 3.2. A Moderate Response is defined as either: * an improvement (decrease) in the DAS28 of greater than 0.6 and less than or equal to 1.2 and attainment of a DAS28 score of less than or equal to 5.1 or, * an improvement (decrease) in the DAS28 of more than 1.2 and attainment of a DAS28 score of greater than 3.2. No Response is defined as either an improvement (decrease) in the DAS28 of less than or equal to 0.6, or an improvement (decrease) in the DAS28 of greater than 0.6 and less than or equal to 1.2 and attainment of a DAS28 score of 5.1 or higher.

Time frame: Baseline and Week 48

Population: Intent-to-treat population. LOCF was used for the individual components of the DAS-28. Non-responder imputation was used. Patients who withdrew prior to week 48, who received rescue therapy or had insufficient data in order to calculate a EULAR response were considered non-responders.

The Stanford Health Assessment Questionnaire disability index is a patient-reported questionnaire specific for rheumatoid arthritis. It consists of 20 questions referring to eight domains: dressing/grooming, arising, eating, walking, hygiene, reach, grip, and daily activities. Participants choose from four response categories, ranging from 'without any difficulty' (Score=0) to 'unable to do' (Score=3). The overall score is the average of each of the 8 category scores and ranges from 0 to 3, where zero represents no disability and three very severe, high-dependency disability. A negative change from baseline score indicates an improvement. Improved HAQ-DI is defined as a change from Baseline score less than or equal to -0.22. An Unchanged HAQ-DI is defined as a change from Baseline score greater than -0.22 and less than 0.22. A worsened HAQ-DI score is defined as a change from Baseline score of greater than or equal to 0.22.

Time frame: Baseline and Week 24

Population: Intent-to-treat population including participants with available data. LOCF was used.

The Stanford Health Assessment Questionnaire disability index is a patient-reported questionnaire specific for rheumatoid arthritis. It consists of 20 questions referring to eight domains: dressing/grooming, arising, eating, walking, hygiene, reach, grip, and daily activities. Participants choose from four response categories, ranging from 'without any difficulty' (Score=0) to 'unable to do' (Score=3). The overall score is the average of each of the 8 category scores and ranges from 0 to 3, where zero represents no disability and three very severe, high-dependency disability. A negative change from baseline score indicates an improvement. Improved HAQ-DI is defined as a change from Baseline score less than or equal to -0.22. An Unchanged HAQ-DI is defined as a change from Baseline score greater than -0.22 and less than 0.22. A worsened HAQ-DI score is defined as a change from Baseline score of greater than or equal to 0.22.

Time frame: Baseline and Week 48

Population: Intent-to-treat population including participants with available data. LOCF was used.

C-Reactive Protein (CRP) was measured from blood samples by a central laboratory as a marker for inflammation. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

Erythrocyte sedimentation rate (ESR) indirectly measures how much inflammation is in the body. A higher ESR is indicative of increased inflammation. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The Stanford Health Assessment Questionnaire disability index is a patient-reported questionnaire specific for rheumatoid arthritis. It consists of 20 questions referring to eight domains: dressing/grooming, arising, eating, walking, hygiene, reach, grip, and daily activities. Participants choose from four response categories, ranging from 'without any difficulty' (Score=0) to 'unable to do' (Score=3). The overall score is the average of each of the 8 category scores and ranges from 0 to 3, where zero represents no disability and three very severe, high-dependency disability. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The participant's overall assessment of their current disease activity measured on a 100 mm horizontal visual analog scale (VAS). The left-hand extreme of the line (0 mm) was described as no disease activity (symptom-free and no arthritis symptoms) and the right-hand extreme (100 mm) as maximum disease activity (maximum arthritis disease activity). The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The participant's assessment of their current level of pain on a 100 mm horizontal visual analog scale (VAS), where the left-hand extreme of the line (0 mm) was described as no pain and the right-hand extreme (100 mm) as unbearable pain. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The physician's assessment of the participant's current disease activity on a 100 mm horizontal VAS, where the left-hand extreme of the line (0 mm) was described as no disease activity (symptom-free and no arthritis symptoms) and the right-hand extreme (100 mm) as maximum disease activity. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

The SF-36 measures the impact of disease on overall quality of life and consists of 36 questions split into two major components: physical health and mental health. Under physical health are the following four domains: physical health, bodily pain, physical functioning and physical role limitations. Under the mental health domain there are four domains; mental health, vitality, social functioning, and emotional role limitation. The individual domain scores are aggregated to derive a physical-component summary score and a mental-component summary score which range from 0 to 100, with higher scores indicating a better level of functioning. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A positive percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

Sixty-six joints were assessed and classified as swollen/not swollen by pressure and joint manipulation on physical examination. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

Sixty-eight joints were assessed and classified as tender/not tender by pressure and joint manipulation on physical examination. The percentage change from baseline at each post-baseline visit was calculated as: \[(post-baseline value minus baseline value) divided by Baseline value\]\*100. A negative percentage change from baseline score indicates an improvement.

Time frame: Baseline, Week 24 and Week 48

Population: Intent-to-treat population, LOCF was used. N indicates the number of participants with non-missing data at each time point.

A low immunoglobulin concentration was defined as a concentration below the lower level of normal.

Time frame: Baseline (pre-rituximab), Beginning of the safety follow-up period to the end of the study (approximately 6 years) (post-rituximab)

Population: Safety follow-up population: All participants who were randomized and received any part of a rituximab infusion. Number of participants analyzed = participants with available data. N indicates the number of participants with non-missing data at each time point.

Peripheral CD19+ B-cell repletion was defined as a CD19+ B-cell count that returned to the Baseline value or returned to ≥ the lower limit of normal, whichever was lower.

Time frame: Beginning of the first infusion (Day 1) in the last treatment cycle until repletion or the end of the study (approximately 6.5 years)

Population: Extended safety follow-up population: All participants who were randomized, received any part of a rituximab infusion, and entered the extended safety follow-up period.