Table 2.

Association between baseline autoantibody status and radiological destruction

	RF+ (n = 162) vs RF- (n = 94)	Anti-CCP + (n = 148) vs anti-CCP- (n = 108)	Anti-MCV+ (n = 180) vs anti-MCV- (n = 74)	Anti-CII + (> 200 AU/mL; n = 8) versus anti-CII- (n = 248)
Larsen score baseline	4.000 versus 5.000, P = NS (0.13)	4.000 versus 5.000, P = NS (0.16)	4.375 versus 5.000, P = NS (0.21)	13.750 versus 4.375, P = 0.0486*
Larsen score 1 year	10,500 versus10.500, P = NS (0.65)	10.500 versus 10,750, P = NS (0.66)	11.000 versus 10.000, P = NS (0.99)	21.500 versus 10.500, P = NS (0.0799)
Larsen score 2 years	14.000 versus 13.250, P = NS (0.69)	14.000 versus 13.000, P = NS (0.57)	14.375 versus 10,250, P = NS (0.24)	24.000 versus 13.500, P = NS (0.19)
ΔLarsen score 1 yr -baseline	4.500 versus 4.250, P = NS (0.32)	4.500 versus 4.250, P = NS (0.29)	4.500 versus 3.750, P = NS (0.19)	7.375 versus 4.250, P = NS (0.27)
ΔLarsen score 2 yrs -baseline	7.250 versus 6.250, P = 0.0354*	7.500 versus 6.250 P = 0.0102*	7.750 versus 5.250, P = 0.0028*	9.250 versus 6.500, P = NS (0.50)
ΔLarsen score 2 yrs -1 yr	2.500 versus 1.250, P = 0.0009*	2.750 versus 1.000, p < 0.0001*	2.750 versus 1.000, p < 0.0001*	3.000 versus 2,.250, P = NS (0.48)

Differences in median Larsen score between patients with and without different autoantibodies and changes in median Larsen score are given for 256 patients with early rheumatoid arthritis. Anti-MCV data were missing for two patients. ΔLarsen score, difference in Larsen score; RF, rheumatoid factor; Anti-CCP, anti-cyclic citrullinated peptide antibodies; Anti-MCV, antibodies against modified citrullinated vimentin; Anti-CII, anti-native human collagen type II. *P < 0.05.