Table 2

Summary of efficacy outcomes

	Total population			Seropositive (RF +ve and/or ACPA +ve) subgroup		Seronegative (RF -ve and ACPA -ve) subgroup
	Placebo + MTX	Rituximab (2×500 mg) + MTX	Rituximab (2×1000 mg) + MTX	Placebo + MTX	Rituximab (2×1000 mg) + MTX	Placebo + MTX	Rituximab (2×1000 mg) + MTX
Joint damage outcomes (mITT population)	n=232	n=239	n=244^†	n=211	n=218^†	n=21	n=24^†
Week 52
Change in mTSS (mean)	1.079	0.646	0.359^**	1.148	0.354^**	0.387	0.352
Change in erosion score (mean)	0.738	0.453	0.233^**	nc	nc	nc	nc
% patients with no progression	53	58	64^*	54	62	52	79
OR (unadjusted)		1.190	1.517		1.438		3.455
95% CI		0.827 to 1.712	1.051 to 2.189		0.979 to 2.114		0.936 to 12.743
Week 24
Change in mTSS (mean)	0.701	0.580	0.328^*	nc	nc	nc	nc
% patients with no progression	59	63	70^*	nc	nc	nc	nc
ITT	n=249	n=249	n=250	n=227	n=224	n=22	n=24
Disease activity outcomes (ITT population) week 52
ACR20	64%	77%^*	80%^***	64%	81%	64%	71%
ACR50	42%	59%^***	65%^***	41%	67%	55%	54%
OR (unadjusted)		2.043	2.567		2.915		0.985
95% CI		1.430 to 2.919	1.788 to 3.685		1.986 to 4.278		0.308 to 3.146
ACR70	25%	42%^***	47%^***	25%	50%	23%	25%
ACR90	9%	17%^*	16%^*	8.8%	18.3%	13.6%	0%
Major clinical response	8%	18%^*	21%^***	nc	nc	nc	nc
Mean ACRn	19.5	42.9^***	46.0^***	29.8	58.4	33.6	40.8
EULAR good response	18%	39%^***	42%^***	18%	43%	18%	33%
DAS28-ESR LDA	20%	40%^***	43%^***	20%	44%	23%	33%
DAS28-ESR remission	13%	25%^**	31%^***	12%	31%	18%	25%
Change in DAS28-ESR (mean)^‡	−2.06	−3.05^***	−3.21^***	−2.60^§	−3.64^§	−2.95^§	−3.08^‡
Physical function outcomes week 52
Change in HAQ-DI (mean)^‡	−0.628	−0.905^***	−0.916^***	nc	nc	nc	nc
% With HAQ-DI decrease ≥0.22	77	87^*	88^*	nc	nc	nc	nc

↵* p<0.05,
↵** p<0.001,
↵*** p<0.0001 for differences vs placebo+MTX.
↵† Total number of patients included in rituximab 2×1000 mg autoantibody subgroups differed by two versus total number of patients included in rituximab 2×1000 mg group throughout because two patients could not be classified (RF-negative but no ACPA data, so unable to be assigned to either group).
↵‡ Adjusted mean presented for total population; standard mean presented for subgroups (where available). Analyses included baseline scores as additional covariates.
↵§ No statistical analysis was performed on the subgroup data.
Van Elteren test for difference in distribution of changes in radiographic variables; ANOVA model adjusted for stratification factors (RF, region) (adjusted mean changes shown in table) for all other continuous variables; Cochran–Mantel–Haenszel test for categorical variables; non-responder imputation used for ACR major clinical response and EULAR response variables, last observation carried forward.
ACPA, anti citrullinated peptide antibodies; ACR, American College of Rheumatology; ACRn, American College of Rheumatology Index of Improvement in Rheumatoid Arthritis; ANOVA, analysis of variance; DAS28, Disease Activity Score in 28 joints; ESR, erythrocyte sedimentation rate; EULAR, European League Against Rheumatism; HAQ-DI, Health Assessment Questionnaire-Disability Index; ITT, intent-to-treat; LDA, low disease activity; mITT, modified intent-to-treat; mTSS, Genant-modified total Sharp score; MTX, methotrexate; nc, not calculated; RF, rheumatoid factor.