B lymphocyte-typing for prediction of clinical response to rituximab

Table 2 Baseline clinical variables and B-cell subsets in responders and non-responders, as determined by univariate logistic regression

Clinical variable or B-cell subset	Responders	Non-responders	OR (95% CI)	Pvalue
Age in years, median (interquartile range)	62.9 (58.2-67.7)	60.6 (49.7-71.4)	0.99 (0.92-1.05)	0.647
sDMARD use, median (interquartile range)	2.5 (2.1-2.8)	2.4 (1.5-3.3)	0.84 (0.33-2.14)	0.956
Prior anti-TNF use, median (interquartile range)	1.1 (0.8-1.4)	1.1 (0.3-2.0)	1.02 (0.37-2.80)	0.978
ESR^a, median (interquartile range)	6.1 (5.4-6.8)	5.0 (2.9-7.2)	0.74 (0.47-1.17)	0.202
Rheumatoid factor, number (percentage)
Positive	27 (90.0)	7 (100.0)	1.91 (0.43-8.32)	0.389^b
Negative	3 (10.0)	0 (0.0)
B-cell subset, median (interquartile range)
Naïve cells^c	8.9 (8.6-9.3)	9.2 (8.6-9.8)	1.51 (0.56-4.11)	0.416
Post-switch memory^c	7.4 (7.1-7.6)	7.9 (7.4-8.5)	3.91 (0.93-16.45)	0.063
Plasmablasts^d	3.8 (3.3-4.4)	5.2 (3.8-6.6)	2.22 (1.04-4.74)	0.040
B-cell depletion, number (percentage)
Complete	2 (6.7)	1 (14.3)	2.33 (0.18-30.10)	0.516
Incomplete	28 (93.3)	6 (85.7)

^aSquare root values; ^bcalculated from contingency table, Pearson-Mantel-Haenszel P value; ^cnatural log (absolute number of cells); ^dnatural log (absolute number of cells + 1). anti-TNF, anti-tumor necrosis factor; CI, confidence interval; ESR, erythrocyte sedimentation rate; OR, odds ratio; sDMARD, synthetic disease-modifying anti-rheumatic drug.

ISSN: 1478-6362