Comparison of MRI features of first?episode optic neuritis with positive AQP4 and MOG antibodies

Xilan Liu,Ping Lu,Yang Li,Ting Yuan,Naier Lin,Guohong Tian,Yan Sha
DOI: https://doi.org/10.3760/cma.j.issn.1005-1201.2019.06.009
2019-01-01
Abstract:Objective To explore the MRI features of acute optic neuritis with positive aquaporin?4 (AQP4) antibodies and myelin oligodendrocyte glycoprotein (MOG) antibodies during the first attack. Methods Eighty five patients (105 affected nerves) with first?episode AQP4?seropositive optic neuritis (AQP4?ON; n=58; 64 affected nerves) and MOG?seropositive optic neuritis (MOG?ON; n=27; 41 affected nerves) diagnosed by our hospital Neuro?ophthalmology Department between April 2014 and December 2017 were retrospectively analyzed. Double seronegative patients or double seropositive patients were not included. All patients underwent orbital conventional MRI; 29 patients also underwent RESOLVE?DWI and 55 underwent brain T2 fluid?attenuated inversion recovery (T2FLAIR). Theclinical features (age, gender, disease duration) and MRI features (T2WI, bilateral involvement, the degree and extent of enhancement, the extent of involvement, the affected segment and ADC values of nerve; demyelinating lesions of brain) were analyzed by 2 doctors. The qualitative parameters were compared with χ2 test or Fisher exact test and the quantitative parameters were compared with two independent sample t test (normal distribution) or the Mann?Whitney U test (skewed distribution). The Kappa test was used to test the consistency of the qualitative characteristics of the images evaluated by the 2 doctors. The interclass correlation coefficient (ICC) was used to test the consistency of the ADC obtained by the 2 doctors. Results There were no significant differences in the distributions of age and disease duration between 2 groups (t=1.911, Z=-1.054, P>0.05). AQP4?ON had a higher female proportion (χ2=13.911, P<0.05) and showed mainly unilateral involvement. MOG?ON more commonly showed bilateral involvement (χ2=17.640, P<0.05), but there were no significant differences in the demyelinating lesions of brain (χ2=0.000)and the presence/absence of optic tract involvement (both P>0.05). There was a higher probability of optic nerve swelling, long segmental lesions and optic nerve head involvement in MOG?ON(χ2=16.845, 13.525, 13.264, P<0.05). Moreover, the rate of obvious enhancement and the involvement of optic chiasm was higher in AQP4?ON(χ2=7.152, 5.088, P<0.05). RESOLVE?DWI showed the ADC values of the AQP4?ON were significantly lower than those of MOG?ON (t=-3.300, P<0.05). The qualitative features of image evaluated by the 2 doctors were consistent well (Kappa=0.712-0.932, P<0.001).The ADC values obtained by the 2 doctors were consistent well (ICC=0.942, P<0.001).Conclusions For the first?episode acute optic neuritis, AQP4?ON hasobvious female predilection and the trend of unilateral involvement and obvious enhancement, while MOG?ON isoften involved bilaterally with longitudinally extensive lesions. The ADC values of the AQP4?ON are significantly lower than those of MOG?ON and AQP4?ON are more vulnerable to involve the optic chiasm. Conventional MRI and RESOLVE?DWI are helpful for differentiating AQP4?ON from MOG?ON during the early stage of the disease course.
What problem does this paper attempt to address?