Quantitative analysis of aquaporin-4 (AQP4) and myelin oligodendrocyte glycoprotein (MOG) antibodies titres: correlation with relapses

Detection of antibodies against aquaporin-4 (AQP4-IgG) and myelin oligodendrocyte glycoprotein (MOG-IgG) is essential for diagnosis of AQP4-IgG+ neuromyelitis optica spectrum disorder (NMOSD) and MOG antibody-associated disease (MOGAD). Uncertainties persist regarding the clinical significance of serum antibody titres to predict relapses. We aimed to analyse the trend of serum AQP4-IgG and MOG-IgG titres during relapses and periods of clinical stability. In this retrospective study we analysed serum AQP4-IgG and MOG-IgG titres from live cell-based assays from the Oxford Autoimmune Neurology Diagnostic Laboratory for the UK specialist NMO Service between February 2010 and June 2024. We recruited seropositive AQP4-NMOSD and MOGAD patients with serum samples available within 30 days of an attack (‘attack’) and at least one another sample for comparison within 1 year (‘pre-attack’ and ‘post-attack’). Up to 3 further serum samples were selected within 2, 3 or 4 years after the attack. 117 attacks in 92 AQP4-IgG+ patients (81.5% female) and 127 attacks in 111 MOG-IgG+ patients (62.2% female) had appropriate samples. Antibody titres significantly increased from ‘pre-attack’ to ‘attack’, decreased from ‘attack’ to ‘post-attack’, and remained stable from 2 to 4 years after attacks. Long-term immunosuppressant treatments (ISTs) induced a further decrease in titres during remission in both cohorts. In 40% of samples in both groups there was an increase in titre at relapse. A ≥2-fold increase in AQP4-IgG titres had an Odds Ratio (OR) of 6.59 and 91.5% specificity of being associated with a relapse, in MOG-IgG+ an OR of 4.87 and 88.6% specificity. 29% (26/92) AQP4-IgG+ patients became ‘seronegative’ during follow: most patients (64%) had low attack titres (≤100), and the time to sero-reversion related the attack titre. In contrast, 41% (46/111) MOG-IgG+ patients became ‘seronegative’, mostly within the first year after the attack regardless of attack titre. In conclusion, in 60% of longitudinal serum samples from patients with AQP4-IgG or MOG-IgG there is no increase in antibody titre. When there is an increase, it is most often at relapse. A ≥2-fold increase in titres should be a risk particularly in case of treatment de-escalation, non-adherence to treatment or if a pseudo-relapse is suspected.