A Study of Retreatment With MabThera (Rituximab) in Combination With Methotrexate in Patients With Rheumatoid Arthritis (RA)

ACR20 defined as overall score of ≥20 in ACR number (ACRn) calculation. Overall score defined as lowest percent improvement from baseline (BL) of following 3 measures: tender joint count (TJC; 68 joints), swollen joint count (SJC: 66 joints), and the 3rd lowest improvement achieved by at least 3 of 5 remaining ACR core parameters: physician's global assessment of disease activity, participant's global assessment of disease activity, participant's assessment of pain (visual analog assessment [VAS]), Health Assessment Questionnaire (HAQ), and C-Reactive Protein (CRP). If CRP missing, erythrocyte sedimentation rate (ESR) was used. In order for improvements in the ACRn score to be expressed as a positive result, rather than the negative changes that improvements represent, the final ACRn results were multiplied by negative 1. Last observation carried forward (LOCF) for TJC/SJC, HAQ, CRP/ESR, VAS. If change in CRP incalculable, change in ESR used. ACR20 set to Non-Responder if ACRn missing

ACR50 was defined as an overall score of 50 in the ACRn calculation. Overall score defined as lowest percent improvement from BL of following 3 measures: TJC (68 joints), SJC (66 joints), and the 3rd lowest improvement achieved by at least 3 of 5 remaining ACR core parameters: physician's global assessment of disease activity, participant's global assessment of disease activity, participant's assessment of pain (VAS), HAQ, and CRP. If CRP missing, ESR was used. In order for improvements in the ACRn score to be expressed as a positive result, rather than the negative changes that improvements represent, the final ACRn results were multiplied by negative 1. LOCF for TJC/SJC, HAQ, CRP/ESR, VAS. If change in CRP incalculable, change in ESR used. ACR50 set to Non-Responder if ACRn missing.

ACR70 was defined as an overall score of 70 in the ACRn calculation. The Overall score defined as lowest percent improvement from BL of following 3 measures: TJC (68 joints), SJC (66 joints), and the 3rd lowest improvement achieved by at least 3 of 5 remaining ACR core parameters: physician's global assessment of disease activity, participant's global assessment of disease activity, participant's assessment of pain (VAS), HAQ, and CRP. If CRP missing, ESR was used. In order for improvements in the ACRn score to be expressed as a positive result, rather than the negative changes that improvements represent, the final ACRn results were multiplied by negative 1. LOCF for TJC/SJC, HAQ, CRP/ESR, VAS. If change in CRP incalculable, change in ESR used. ACR70 set to Non-Responder if ACRn missing.

DAS28 was calculated according to the following formula: DAS28 equals (=) [0.56 multiplied by (*) the square root (√) of TJC] plus (+) [0.28 * √ of SJC] + (0.70 * the natural logarithm (ln) ESR in millimeters per hour (mm/h)] + [0.014 * participant's global assessment of disease activity (GH)]. DAS28-ESR ≥ 5.1 = high disease activity, DAS28-ESR less than or equal to (≤) 3.2 = low disease activity, DAS28-ESR less than (<) 2.6 = remission.

EULAR responses were categorized according to DAS28-ESR score. DAS28-ESR ≤ 3.2 at Week 48 and a change from BL to Week 48 < -1.2 = good response, DAS28-ESR ≤ 3.2 or greater than (>) 3.2 and ≤ 5.1 at Week 48 and a change from BL to Week 48 < -0.6 and ≥ -1.2 = moderate response, DAS28-ESR > 3.2 and ≤ 5.1 at Week 48 and a change from BL to Week 48 < -1.2 = moderate response, DAS28-ESR > 5.1 at Week 48 and a change from BL to Week 48 < -1.2 = moderate response, DAS28-ESR ≤ 3.2 or > 3.2 and ≤ 5.1 at Week 48 and a change from BL to Week 48 ≥ -0.6 = no response, DAS28-ESR > 5.1 at Week 48 and a change from BL to Week 48 < -0.6 and ≥ -1.2 or ≥ -0.6 = no response.

FACIT-F scores were obtained from a 13 question self-administered participant questionnaire designed to measure the degree of fatigue experienced by participants in the previous 7 days. Participants responded to the questions using a value between 0 and 4, where 0 indicated "not at all" and 4 indicated "very much." 11 of the 13 questions were negatively stated; indicating the higher the score of the participant's response, the greater their fatigue. These questions were calculated as 4 minus the participants' response, so that a higher score indicated an improvement in health. The scores for the 2 positively stated questions were not changed. The participants' responses were summed to result in an overall score, which are scored 0 to 52 (52 = highest level of functioning). A positive change from BL indicated improvement.

SF-36 scores were obtained by scoring participants' responses to a 36 item questionnaire. SF-36 evaluated 8 aspects of functional health and well being: physical and social functioning, physical and emotional role limitations, bodily pain, general health, vitality, mental health from a range of 1 (better) to 5 (worst). The score for each section was an average of the individual question scores, which were scaled 0-100 (100=highest level of functioning). These 8 aspects were summarized as physical and mental component scores.

SF-36 scores were obtained by scoring participants' responses to a 36 item questionnaire. SF-36 evaluated 8 aspects of functional health and well being: physical and social functioning, physical and emotional role limitations, bodily pain, general health, vitality, mental health from a range of 1 (better) to 5 (worst). The score for each section was an average of the individual question scores, which were scaled 0-100 (100=highest level of functioning). These 8 aspects were summarized as physical and mental component scores.

Cfirst values were estimated from rituximab serum concentrations by non-compartmental methods using the software WinNonlin Enterprise Version 5.2.

Csecond values were estimated from rituximab serum concentrations by non-compartmental methods using the software WinNonlin Enterprise Version 5.2.

t1/2 values were estimated from rituximab serum concentrations by non-compartmental methods using the software WinNonlin Enterprise Version 5.2.

Surface expression of CD19 was assessed by fluorescence-activated cell sorting (FACS) analysis as a marker of absolute B lymphocyte count.

Surface expression of CD19 was assessed by FACS analysis as a marker of absolute B lymphocyte count. The LLN was defined as < 80 cells/µL.

Surface expression of CD20 was assessed by FACS analysis as a marker of mature and memory B lymphocyte count.

Surface expression of CD22 was assessed by FACS analysis as a marker of mature lymphocyte count.

Simultaneous surface expression of CD19 and CD27 was assessed by FACS analysis as a marker of memory B lymphocyte count.

Surface expression of CD19 in the absence of CD27 expression was assessed by FACS analysis as a marker of naive B lymphocyte count.

Surface expression of CD3 was assessed by FACS analysis as a marker of absolute T lymphocyte count. The normal range of CD3+ T cells was defined as 723-2737 cells/µL.

Surface expression of CD3 was assessed by FACS analysis as a marker of absolute T lymphocyte count. The normal range of CD3+ T cells was defined as 723-2737 cells/µL.

Surface expression of CD4 was assessed by FACS analysis as a marker of T helper cell count. The normal range of CD4+ T cells was defined as 404-1612 cells/µL.

Surface expression of CD4 was assessed by FACS analysis as a marker of T helper cell count. The normal range of CD4+ T cells was defined as 404-1612 cells/µL.

Surface expression of CD8 was assessed by FACS analysis as a marker of cytotoxic T lymphocyte count. The normal range of CD8+ T cells was defined as 220-1129 cells/µL.

Surface expression of CD8 was assessed by FACS analysis as a marker of cytotoxic T lymphocyte count. The normal range of CD8+ T cells was defined as 220-1129 cells/µL.

Simultaneous surface expression of CD16 and CD56 was assessed by FACS analysis as a marker of NK cell count.

Simultaneous surface expression of CD16 and CD56 was assessed by FACS analysis as a marker of NK cell count.

The LLNs for total Ig, IgA, IgG, and IgM were defined as 6.75 grams per liter (g/L), 0.70 g/L, 65 g/L, and 0.40 g/L, respectively.

Percentage of participants who were RF seropositive at BL who became RF seronegative over the course of the study. RF seropositive status was defined as RF ≥ 20 international units (IU) per mL. RF seronegative status was defined as RF < 20 IU/mL.

The LLN for C3 protein was defined as <0.9 grams per liter (g/L).

The LLN of C3 protein was defined as <0.9 g/L.

The LLN of C4 protein was defined as < 0.1 g/L.

A participant was defined as being HACA positive if the HACA serum level was ≥ 5 relative units (RU) per mL and the physician comment read that participant was "immunodepletable with rituximab".

ANA titers were obtained by the following serum dilution schema: negative = negative, borderline = 1 diluted to (:) 40 or 1:80, and positive ≥ 1:160. The change categories were defined for the change from BL to Weeks 24 and 48 according to this schema. Negative to borderline was defined as any change from negative to borderline as no dilution is given for negative results. Negative to positive was defined as at least a two-fold positive change in dilution from BL. Borderline to negative was defined as any change from borderline to negative as no dilution is given for negative results. Borderline to positive was defined as at least a two-fold positive change in dilution from BL. Positive to borderline was defined as at least a two-fold negative change in dilution from BL. Positive to negative was defined as at least a two-fold negative change in dilution from BL. Unchanged was defined as any difference in dilution less than two-fold.

A positive titer result to recall antigens was defined as a serum antibody level equal to or above the following protective levels: tetanus toxoid ≥ 0.1 IU/mL, influenza A > 12 U/mL, influenza B > 12 U/mL, and streptococcus (S.) pneumococcus ≥ 1.0 mg/L.