Article Text
PDF
Abstract
Background IgG antibodies against myelin oligodendrocyte glycoprotein (MOG-IgG) define a subset of associated disorders (myelin oligodendrocyte glycoprotein associated disorders (MOGAD)) that can have a relapsing course. However, information on relapse predictors is scarce. The utility of retesting MOG-IgG over time and measuring their titres is uncertain. We aimed to evaluate the clinical relevance of longitudinal MOG-IgG titre measurement to predict relapses in patients with MOGAD.
Methods In this retrospective multicentre Italian cohort study, we recruited patients with MOGAD and available longitudinal samples (at least one >3 months after disease onset) and tested them with a live cell-based assay with endpoint titration (1:160 cut-off). Samples were classified as ‘attack’ (within 30 days since a disease attack (n=59, 17%)) and ‘remission’ (≥31 days after attack (n=295, 83%)).
Results We included 102 patients with MOGAD (57% adult and 43% paediatric) with a total of 354 samples (83% from remission and 17% from attack). Median titres were higher during attacks (1:1280 vs 1:640, p=0.001). Median onset titres did not correlate with attack-related disability, age or relapses. Remission titres were higher in relapsing patients (p=0.02). When considering the first remission sample available for each patient, titres >1:2560 were predictors of relapsing course in survival (log rank, p<0.001) and multivariate analysis (p<0.001, HR: 10.9, 95% CI 3.4 to 35.2). MOG-IgG seroconversion to negative was associated with a 95% relapse incidence rate reduction (incidence rate ratio: 0.05, p<0.001).
Conclusions Persistent MOG-IgG positivity and high remission titres are associated with an increased relapse risk. Longitudinal MOG-IgG titres could be useful to stratify patients to be treated with long term immunosuppression.
- MULTIPLE SCLEROSIS
- MYELOPATHY
- MYELIN
Data availability statement
Data are available in a public, open access repository. Raw data are available at the Zenodo repository (DOI: 10.5281/zenodo.6631485)
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
IgG antibodies against myelin oligodendrocyte glycoprotein (MOG-IgG) are useful biomarkers to identify a distinct demyelinating disorder (myelin oligodendrocyte glycoprotein associated disorder). The utility of retesting MOG-IgG over time and measuring their titres is still uncertain.
WHAT THIS STUDY ADDS
Quantitative and longitudinal MOG-IgG testing provides useful information in clinical practice as high remission titres and persistent positivity are associated with a higher risk of a relapsing course
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Measuring and quantifying remission MOG-IgG titres could be used as a tool to stratify patients at high relapse risk, improving the treatment choice.
Introduction
IgG antibodies against the myelin oligodendrocyte glycoprotein (MOG-IgG) detected using a cell-based assay (CBA) employing full-length human myelin oligodendrocyte glycoprotein (MOG) protein as antigenic substrate identify a spectrum of demyelinating disorders (MOG-IgG associated disorders, MOGAD) distinct from aquaporin-4 IgG-associated neuromyelitis optica spectrum disorders (AQP4-IgG NMOSD1 2) and multiple sclerosis (MS).3–9 MOGAD involves all age groups and is associated with a relapsing course in around 40%–50% of patients.4 5 10–13
Accrual of disability in MOGAD is thought to be attack-related, and patients with a relapsing course are often treated with long-term immunosuppression. Early identification of patients with relapsing disease is of the utmost importance, but, so far, no validated predictors of relapse risk have been identified.14
Despite the clear diagnostic value of MOG-IgG, the relevance and the routine use of MOG-IgG quantitative determination and longitudinal testing in clinical practice are still open issues. MOG-IgG titres correlate with the disease phase, being higher during attacks compared with remission.5 6 Initial studies suggest that high titres at onset11 and persistent positivity over time15–18 may be associated with a relapsing course, but this has not been confirmed in other investigations involving both children and adults.6 19
In this study, we aimed to assess the clinical relevance of quantitative and longitudinal MOG-IgG testing in a large population of paediatric and adult patients with MOGAD.
Methods
Patients
This is a multicentre retrospective cohort study involving 34 paediatric and adult neurology hospitals in Italy. We included patients observed at the participating hospitals that performed MOG-IgG testing between 2018 and 2021 fulfilling the following criteria: (1) diagnosis of MOGAD according to published criteria,17 20 (2) serum MOG-IgG detected with a CBA at disease onset in any of the 34 participating centres and (3) at least one available follow-up serum sample collected more than 3 months after disease onset titrated with a live CBA in one of the two referral laboratories participating in this study (Pavia and Verona). The titration in one of the two referral laboratories of onset samples was not mandatory for the inclusion in the study.
Information regarding clinical presentation at disease onset and follow-up, outcome data as well as laboratory data was collected from patient records. Patients were grouped as having acute optic neuritis (ON), transverse myelitis (TM) or longitudinally extensive transverse myelitis (LETM) when spinal cord lesions extended over ≥3 vertebral segments, acute disseminated encephalomyelitis (ADEM) according to published criteria,21 ADEM-like when they fulfilled ADEM criteria but without encephalopathy, and NMOSD.1 Patients not fulfilling any of the previous groups were classified as ‘other inflammatory disorders’. No patient with clear cortical encephalitis was included in our cohort. MRI scans were performed with various timing, with different scanners, and as part of the diagnostic workup. Data were collected from MRI reports, and MRI scans were not specifically revised for this study. MRI was performed within 30 days after disease onset in 69 of 86 patients (80.2%), for which information on timing was available.
Patients ≤16 years old were considered as paediatric. Attacks were divided as ‘onset’ (the first clinical manifestation of the disease) and ‘relapses’ (clinical events that occurred after onset). Relapses were defined as clinical and radiological (or alternatively visual evoked potential alterations with prolonged P100 latency in case of ON) evidence of new disease activity at least 30 days after the index event (90 days in ADEM patients17).
Disability was assessed with the Expanded Disability Status Scale (EDSS)22 and the modified Rankin Scale.23
In the two referral laboratories, serum samples were all tested on arrival using live CBAs for total MOG-IgG and for MOG-IgG1, as previously described.24 25 Both laboratories used a full-length MOG construct (kind gift of Professor Markus Reindl) and shared similar protocols (see online supplemental methods). We classified samples with titres of ≥1:160 as positive. Samples positive at the starting dilution of 1:20 were titrated at doubling dilutions. The last dilution producing a positive staining of the transfected cells detected by use of a fluorescence microscope was considered as the endpoint titre. Only samples titrated in one of the two referral laboratories were considered for statistical analysis.
Supplemental material
Considering the observational nature of the study, the timing for retesting was not defined ‘per protocol’ but depended on the choice of the treating physician.
We defined ‘attack samples’ those collected within 30 days before or 30 days after the first day of an attack, either at onset or at relapse, and ‘remission samples’ those collected more than 31 days after the first day of an attack.
Statistical analysis
Qualitative variables were given as percentages and quantitative variables as median with IQRs. Differences in qualitative variables were measured through χ2 or Fisher’s exact test. The differences in quantitative variables were tested using t-test or Mann-Whitney non-parametric test, depending on whether data were normally distributed.
Survival analyses were performed using Kaplan-Meier curves and analysed with the log-rank test. Univariate and multivariate survival analyses were performed using a Cox model. We considered as baseline variables those collected at disease onset. Results are presented in terms of HRs with 95% CIs. We included in the multivariate analysis all the variables that were significant from univariate analysis.
In order to assess the prognostic relevance of high versus low titres in terms of relapse risk, we identified a discriminatory titration through the method proposed in Contal and O'Quigley.26 This method selects the value which enhances the log-rank statistic to maximise the difference between the two groups.
Predictive value of remission titres and interaction with immunosuppressant therapy were investigated by a time-dependent Cox model, where the introduction of a therapy represents the time-dependent covariate. We also included other variables that could represent confounding factors.
The longitudinal impact of MOG-IgG seroconversion from positive to negative on the relapse rate was investigated using a random effects Poisson regression model. Here, MOG-IgG negativity was included as a fixed effect, and the duration of the observed time was included as an adjustment. The patient identifier was included as a random effect.
P values of ≤0.05 were considered significant (two-sided). All analyses were performed with Stata/SE for Windows V.14 and R V.4.0.2 (R Core Team, USA).
Results
Patients and samples
We analysed a total of 354 samples from 102 patients with MOGAD, including 59 attack samples (16.7%), 72.9% of which were collected at onset and 27.1% during relapses, and 295 remission samples (83.3%). All attack samples were obtained before patients had any acute treatment such as intravenous immunoglobulins, intravenous steroids and plasma exchange (PLEX). The median time from disease onset to remission sample collection was 18 months (IQR 9–32). Considering all samples, the median number of samples available for each patient was 2 (IQR 2–4).
Relevant clinical information is reported in table 1, while detailed laboratory, neuroradiological and outcome information is available in online supplemental table 1. In the total cohort of 102 patients, 59 (42.2%) were men (female to male ratio=1.0:1.4) and 58 (56.9%) were ≥16 years old. Ethnicities included white European (95.1%), black (3.9%) and Asian (1%). In adult patients, the most frequent clinical presentations were ON (50.0%), TM (32.8%, 26.1% of which were LETM) or a combination of the two (5.2%), while ADEM-like was the most common presentation in paediatric patients (50%, p<0.001). Forty-four of 102 patients (43.1%) had relapses, without significant differences between adult and paediatric populations, with a median annualised relapse rate of 0.54 (IQR 0.23–1.55) (table 1). Median time to first relapse was 7 months (IQR 2–23), and the median number of relapses experienced by each patient was 1 (IQR 1–3). Relapsing patients showed partial/complete recovery after the first attack more frequently than monophasic patients (p=0.002) and more frequently periventricular lesions on brain MRI (online supplemental table 1, p=0.04).
Supplemental material
Clinical characteristics of patients with MOGAD included in the study
We performed a univariate analysis to investigate baseline variables. Periventricular lesions, complete recovery after the attack and intravenous steroid therapy were significantly associated with an increased risk of relapse (table 2). However, among these variables, only a complete recovery after the onset attack remained significant in a multivariate Cox model (table 2; HR 0.037, 95% CI 0.016 to 0.883).
Univariate and multivariate analyses for relapse predictors in patients with MOGAD
Attack and remission samples
Considering all samples, the median MOG-IgG titre was 1:640 (IQR 1:160–1:2560). Attack samples had higher median titres (1:1280, IQR 1:320–1:4480) than remission samples (1:640, IQR 160–2560; p=0.001) (figure 1A), without differences between adult and paediatric patients (online supplemental figure 1A). Titres from remission samples in patients with a relapsing course were higher (median 1:640, IQR 1:160–1:5120) compared with monophasic patients (median 1:480, IQR 1:180–1:1280, p=0.02; figure 1B). No differences were detected among relapsing and monophasic patients in attack titres (p=0.38, figure 1B). However, this analysis did not take into consideration the number of sample available per patient and was exposed to autocorrelation bias. Median MOG-IgG attack titres did not differ according to MOGAD clinical manifestations at onset, nor according to the degree of recovery from the onset attack (online supplemental figure 1B,C).
Supplemental material
MOG-IgG titres in attack and remission samples. (A,B) Attack and remission titres considering all samples (A), divided according to clinical course (B); (C,D) Remission titres in monophasic and relapsing patients, considering only the first remission titre available (C) or the highest remission titre available (peak remission titre) (D) in each patient. The red bars represent the medians with IQRs. The black dotted line represents the MOG-IgG cut-off set at 1:160. MOG, myelin oligodendrocyte glycoprotein.
Onset versus remission paired sample MOG-IgG titres in individual patients
Onset samples were available for 42 patients, while all patients had at least one remission sample (as per inclusion criteria). Onset sample was collected after a median of 2 days after symptoms onset (IQR 1.0–8.5). An onset versus remission paired sample analysis was performed and showed a decline in titres in 38 of 42 patients, an increase in 3 (2 children, 1 of them with ADEM-like disease and 1 with TM, and 1 adult with ON) and stable titre in 1 patient (p<0.001, figure 2A). The degree of titre reduction was similar for adult and paediatric patients (p<0.001 in both groups, figure 2B) but more pronounced in monophasic (p<0.001) versus relapsing patients (p=0.02, figure 2C). Titres at onset did not correlate with the highest disability measured with the EDSS during the attack (online supplemental figure 2), with disability at the end of follow-up (not shown) or with a relapsing course (figure 3A). No differences in titre changes were detected in relation to the type of acute phase treatment received (not shown).
Paired onset and remission MOG-IgG titres. Paired attack–remission titre analysis in the 42 patients with available onset samples (A), divided according to age at onset (B) and clinical course (C). The remission sample considered is the first remission sample available after onset. MOG, myelin oligodendrocyte glycoprotein.
Relapse risk and onset and remission MOG-IgG titres. Kaplan-Meyer survival curves for onset (A) and remission (B) titres in all patients and remission titres and relapse in adult and paediatric patients (C,D). T0 is the disease onset in (A) and the time of collection of the first remission sample in (B–D). MOG: myelin oligodendrocyte glycoprotein.
Remission titres in individual patients and relapse risk
We then considered the first remission titre available for each patient, whose correlated serum sample was collected after a median time of 11 months from onset (IQR, 6–31). The first remission sample was obtained within 6 months after onset in 31 patients (30.4%), between 6 months and 12 months in 28 patients (27.5%), between 12 months and 24 months in 14 patients (13.7%), and after more than 24 months in 29 (28.4%). Among relapsing patients, the first remission titre was obtained before the first relapse in only nine patients, while in 12 of 44 patients, the sample was obtained after the first relapse, but before subsequent relapses. In 23 of 44 cases, the sample was obtained after the last documented relapse. In patients with first remission sample collected before a relapse, the median time from sampling to relapse was 3 months (IQR 1.0–6.5). First, remission titres were considerably higher in relapsing versus monophasic patients (p=0.001) (figure 1C). Similar results were obtained when evaluating the highest remission titre available (peak remission titre, p=0.004; figure 1D).
We identified the ideal discriminatory cut-off at 1:2560,26 and high titres at first remission sample were associated with an increased risk of subsequent relapses in survival analysis (figure 3B; log rank p<0.001; patients with first remission samples collected after the last relapse were not included in these analyses). Interestingly, this finding was confirmed, also performing the analysis separately for adult (figure 3C, log-rank p<0.001) and paediatric patients (figure 3D, log-rank p=0.002). No differences were detected when analysing onset titres (figure 3A) both in adult and paediatric patients (online supplemental figure 3).
In a multivariate Cox model considering variables at the time of first remission sample collection, high titres were the only independent predictor of relapses (p<0.001, HR 6.874, 95% CI 2.435 to 19.407; table 2).
Relapse MOG-IgG titres
Relapse samples were available for 16 attacks from 15 patients. Relapses were associated with titres ≥1:160 in 14/15 patients. Only one patient experienced a relapse with a below the cut-off sample (titre, 1:40), while no relapses occurred with samples negative at 1:20. In 4/6 patients with pre-attack samples available (collected 3–6 months before relapse), relapse titres were higher than those on pre-attack samples (figure 4A). Seven of nine samples collected after relapse showed a titre reduction (figure 4A).
MOG-IgG titres and relapses. (A) MOG-IgG titres in the nine patients with relapse titres available. Six patients had prerelapse and postrelapse samples (collected within 3–6 months from relapse); three additional patients had only relapse and postrelapse samples. (B) Longitudinal titre kinetics in selected patients with MOGAD with samples available every 3 months for 24 months. Relapses (marked as ‘R’) occurred during both in correspondence (top two panels) and independently (bottom two panels) from MOG-IgG titres sudden increases (titre spikes). The red dotted line represents the MOG-IgG cut-off set at 1:160. MOG, myelin oligodendrocyte glycoprotein; MOGAD, myelin oligodendrocyte glycoprotein associated disorder; Rtx, rituximab.
Four patients had titres of samples collected prospectively every 3 months over a period of more than 1 year, which included at least one relapse. Their individual titre kinetics showed that relapses were associated with titre spikes only in two patients. In addition, titre spikes also took place independently from relapses (figure 4B).
MOG-IgG titres and treatments
Most patients (96.1%) received acute phase treatment with steroids, intravenous immunoglobulins (IVIg), PLEX or their combination (table 1). Treatment strategies differed between adult and paediatric patients. Intravenous steroids were more commonly used in paediatric patients(p<0.001), while second-line treatments were common in adults (p=0.001). In 51 of 102 patients with paired attack (onset or relapse) and remission samples available (median interval between attack and remission sample 8.5 months, IQR 4–14), titres decreased after acute phase treatment administration (median 1:1280, IQR 1:640–1:5120, vs median 1:320, IQR 1:80–1:640; p<0.001), with no relevant differences in relation to specific treatments. Ensuing immunosuppressive/immunomodulating treatments were administered to 38 patients (after a median time from onset of 16 months, IQR 8–29) and included azathioprine (AZA; n=22), rituximab (n=12), mycophenolate mophetil (MMF; n=4), methotrexate (n=1), ocrelizumab (n=2), monthly intravenous IVIG (n=2) and tocilizumab (n=1) (6 patients received more than one treatment).
Both a pretreatment sample and a post-treatment sample (collected within 6 months before and after treatment initiation) were available for 13 of 38 patients receiving immunosuppressive drugs including rituximab (n=5), AZA (n=7, in one patient followed by MMF) and MMF (n=1). In patients receiving AZA and/or MMF, half of the population experienced a titre increase, and the other half a titre reduction. On the other hand, in all patients receiving rituximab, titres either decreased (n=3) or remained stable (n=2) (online supplemental figure 4). In addition, preimmunosuppression peak remission titres were higher compared with postimmunosuppression peak remission titres collected at least 6 months after treatment initiation (p=0.038; samples from 14 patients were available for this analysis; online supplemental figure 5).
MOG-IgG positivity status modification over time and relapse risk
Overall, 61 patients (59.8%) remained persistently positive, while 41 (40.2%) descended below the cut-off at least once during follow-up, with no differences between monophasic and relapsing patients (p=0.55) or adult and paediatric patients (p=0.59) (table 1 and figure 5). Median time between onset and last available sample was similar in patients that seroconverted to negative (28 months, IQR 13–45) and in those who remained persistently positive (28 months, IQR 16–56; p=0.94). Among the 27 patients with titres that seroconverted to negative with subsequent follow-up samples available, 17 later became positive again at least once, and 10 remained persistently below the cut-off. Only 3 of 27 patients (11.1%, one paediatric and two adults) experienced relapses at respectively 1, 8 and 9 months after seroconversion to negative, and all of them became positive again at subsequent follow-up testing (figure 6A). Using a random effect Poisson regression model that included preseroconversion and postseroconversion observation time (median 7 months, IQR 1–18), we found that the incidence of relapses was reduced by 95% after MOG-IgG titres seroconverted to negative (incidence rate ratio: 0.05, 95% CI 0.0 to 0.14; p<0.001) (figure 6B). No clinical, radiological or laboratory factor was associated with seroconversion to negative.
Relapses and MOG-IgG serostatus in monophasic and relapsing patients. Each row represents a patient. Red circles represent relapses, and rhombi represent samples (red=positive; yellow=negative, titre below the cut-off between 1:20 and 1:80; green=negative at 1:20). The coloured lines refer to the serostatus according to the previous sample available. MOG, myelin oligodendrocyte glycoprotein.
Relapse risk and MOG-IgG seroconversion to negative over time. (A) Frequency of attacks in patients with relapsing MOGAD that underwent MOG-IgG seroconversion below the cut-off over time. Attacks are represented as black dots, and the red dotted line represents the seroconversion time. Black horizontal lines represent the follow-up time for each patient. (B) Poisson regression analysis for MOG-IgG seroconversion below the cut-off and relapse risk. MOG, myelin oligodendrocyte glycoprotein; MOGAD, myelin oligodendrocyte glycoprotein associated disorder.
Discussion
In this study, we retrospectively assessed the clinical relevance of longitudinal MOG-IgG titres in a large cohort of paediatric and adult patients with a definite diagnosis of MOGAD.
The diagnostic relevance of MOG-IgG antibodies in acquired demyelinating syndromes of the central nervous system has been well established. Indeed, MOG-IgG predicts a non-MS disease phenotype in both adults and children with such syndromes and defines the spectrum of MOGAD.5 14 27 In addition to the diagnostic information of a positive versus negative result, MOG-IgG titres can be important to increase the positive predictive value of the test.24 28
However, the clinical and prognostic relevance of MOG-IgG titration, especially at disease onset, is still debated. In our study, we confirmed that titres tend to be higher during attacks compared with remission disease phases but did not find any association between titres and clinical presentation, attack severity or response to treatment.5 12 We also showed that, as found in other studies, high titres are not a clear-cut hallmark of relapse.19 Indeed, relapses can also occur with low titres and, as was the case of one of our patients, even with titres below the positivity cut-off. Similarly, a sudden increase in titres (titre ‘spikes’) was found during stages of clinical stability. These data suggest that MOG-IgG titre should not be taken as an absolute indicator of the disease phase and that titre kinetics in individual patients should be interpreted carefully.
MOGAD can be a relapsing disease, and relapses are considered one of the main factors contributing to long-term disability. However, as opposed to AQP4-IgG-associated NMOSD, up to half of the patients can show a monophasic course, especially in children.4 5 10–13 29 Hence, the identification of prognostic biomarkers that could help to stratify the relapse risk is extremely important to guide therapeutic strategies. Previous studies found that age <10 years, higher disability at onset, male sex and spinal cord involvement were associated with a lower relapse risk and with a longer time to relapses in patients with MOGAD.29 30 In addition, higher onset titres were initially associated with a relapsing course,11 but this was not confirmed in further studies.12 19 In our study, we did not identify any correlation between disease course and onset MOG-IgG titres. Periventricular lesions on the brain MRI, treatment with intravenous steroids and a complete recovery after the first attack were more frequent in relapsing patients, but only the complete recovery was confirmed as predictor in a multivariate analysis. This finding has no clear explanation. Notably, 6 of 12 patients who showed no improvement after the first attack had optic neuritis with very low EDSS scores (1–2) that remained unchanged, possibly due to a low sensitivity of EDSS in detecting a mild improvement.
As a novelty, we found that remission titres of >1:2560 assessed at least 3 months after onset were the only relapse predictor in a multivariate model and were associated with a shorter time from remission sample collection to relapse in survival analysis. This suggests that, despite the variable timing of remission sample collection, patients with high titres are nonetheless exposed to a higher relapse risk. Moreover, survival analysis remained significant even when analysing separately adult and paediatric patients, demonstrating that our results can be applied to both populations.
The role of MOG-IgG in the pathogenesis of MOGAD is still uncertain, but some studies suggested a possible direct role of MOG-IgG in contributing to oligodendrocyte damage, for example, through complement activation.31 The presence of high levels of MOG-IgG, even during remission phases, could identify patients with persistent immune activation and pathogenic autoantibody production, and therefore more prone to relapses. Our findings add a useful tool in MOGAD relapse risk stratification and warrant further confirmation in prospective studies.
Another relevant factor contributing to relapse risk reported in the literature are changes in the MOG-IgG serostatus, especially in children, where early conversion from positive to negative is associated with a monophasic course.17 30 ,32 However, seroconversion may be only transient in some cases (as also observed in the present study) in both adult and paediatric patients, reflecting an either spontaneous or often treatment-induced temporary decline in titres.19
In our study, seroconversion of previously MOG-IgG-positive patients occurred in 40% of the patients, and in some cases, MOG-IgG titres fluctuated between negative and positive, with no relevant differences according to age or disease course. This percentage is in line with other reports, with slight differences that could be explained by the different assay cut-offs.4 16 19 31 33 Importantly, we showed that conversion to negative MOG-IgG serostatus over time, independent of its timing from onset, was associated with a significantly lower relapse risk (figure 4). Interestingly, seroconversion only occurred after the administration of immunosuppressive treatment in all relapsing patients. This finding confirms what has been reported in other studies30 and demonstrates that MOG-IgG seroconversion to negative is another important factor to take into consideration in evaluating future relapse risk and treatment efficacy.
The strengths of our study are the large number of patients and samples included in the analysis, the fact that it was conducted in a ‘real-life setting’ and the involvement of referral labs that used the current gold standard diagnostic technique for MOG-IgG testing.
However, this study has several limitations. It was a retrospective study where samples were collected at different timepoints, sometimes with wide gaps, and selection biases cannot be excluded. Sample availability was limited to less than half of the patients for those collected at onset of the disease and to only a few during relapses. This prevented us from investigating the individual longitudinal titre kinetics, and we might have missed changes in both remission and acute phases. Our finding that high remission titres predict relapses could also be biased by this limitation, as remission samples were not collected at a specific timepoint and in some cases several months after onset. Prospective studies will have to confirm this finding and define the ideal timepoint for sample collection in patients with MOGAD. We did not set a minimum follow-up to define patients as monophasic. Even though median follow-up time did not significantly differ between monophasic and relapsing patients (table 1), minimum follow-up among monophasic patients was 6 months, and up to 20% had a follow-up of <12 months. It is therefore possible that some of these patients would be reclassified as relapsing with a longer follow-up. As for the risk of including patients with MS in our cohort, our evaluations could not be supported by brain and spinal cord MRI scans, which were not revised specifically for this study. On the other hand, all our patients were included in the study only if seropositive on highly specific, live CBA detecting MOG antibodies of the IgG1 subclass.24 Such seropositivity strongly support a non-MS diagnosis.
In conclusion, persistent MOG-IgG positivity and high remission titres are associated with an increased relapse risk. Longitudinal MOG-IgG titres could be useful to stratify patients to treatment with long-term immunosuppression. Prospective studies are needed to confirm such findings and to determine the optimal time window for antibody testing.
Data availability statement
Data are available in a public, open access repository. Raw data are available at the Zenodo repository (DOI: 10.5281/zenodo.6631485)
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by IRCCS, San Matteo, Pavia, Italy (code p-20200039541). The participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank Professor M Reindl for providing the myelin oligodendrocyte glycoprotein plasmids.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Collaborators Francesca Rossi (Neurology Unit, Mater Salutis Hospital, Legnago, Italy, francesca.rossi{at}aulsslegnago.it, patient recruitment and data collection), Zulliani Luigi (Multiple Sclerosis Centre, S. Bortolo Hospital, Vicenza, Italy, luigi.zuliani{at}aulss8.veneto.it, patient recruitment and data collection), Edoardo Mampreso (Headache Centre, Neurology - Euganea, Padua Health Unit, Padova, Italy, edoardo.mampreso{at}aulss6.veneto.it, patient recruitment and data collection), Silvia Gambara (Unit of Child Neurology and Psychiatry, ASST Spedali Civili, Brescia, Italy, silvia.gambara{at}asst-spedalicivili.it, patient recruitment and data collection), Sabrina Squilini (Department of Neuropsychiatry, Children Hospital "G. Salesi", Ospedali Riuniti, Ancona, Italy, sabrina.siliquini@ospedaliriunioni.marche.it, patient recruitment and data collection), Sara Matricardi (Department of Neuropsychiatry, Children Hospital "G. Salesi", Ospedali Riuniti, Ancona, Italy, sara.matricardi{at}yahoo.it, patient recruitment and data collection), Valentina Torri Clerici (Neuroimmunology and Neuromuscolar Diseases Unit, IRCCS Foundation Carlo Besta Neurological Institute, Milano, Italy, Valentina.Torri{at}istituto-besta.it, patient recruitment and data collection), Anna Nunzia Polito (Department of Woman and Child, Neuropsychiatry for Child and Adolescent Unit, General Hospital "Riuniti" of Foggia, Foggia, Italy, annanupolito{at}gmail.com, patient recruitment and data collection), Gabriella Di Rosa (Unit of Child Neurology and Psychiatry, Department Human Pathology of the Adult and Developmental Age, Messina, Italy, gdirosa@unime.it, patient recruitment and data collection), Roberta Bedin (Department of Neurosciences, Ospedale Civile, Azienda Ospedaliero-Universitaria, Modena, Italy, bedin.roberta@aou.mo.it, patient recruitment and data collection), Francesca Vitetta (Ospedale Civile Azienda Ospedaliero-Universitaria di Modena, Modena, Italy, vitetta.francesca{at}aou.mo.it, patient recruitment and data collection), Antonio Varone (Department of Neurosciences, Division of Neurology, Santobono-Pausilipon Children's Hospital, Napoli, Italy, antoniovarone{at}live.com, patient recruitment and data collection), Davide Maimone (Multiple Sclerosis Center, Garibaldi Hospital, Catania, Italy, dmaimone{at}tiscali.it, patient recruitment and data collection), Alvino Bisecco (I Division of Neurology, Department of Advanced Medical and Surgical Sciences (DAMSS), University of Campania “Luigi Vanvitelli”, Naples, Italy, Alvino.bisecco{at}unicampania.it, patient recruitment and data collection), Elena Di Sabatino (Section of Neurology, Department of Medicine, University of Perugia, Perugia, Italy, elenadisab{at}gmail.com, Patient recruitment and data collection), Sara Scannapieco (Department of Medicine, Surgery and Dentistry "Scuola Medica Salernitana", University of Salerno, Salerno, Italy, sarascannapieco1{at}gmail.com, patient recruitment and data collection), Claudia Papi (Department of Neuroscience, Università Cattolica del Sacro Cuore, Rome, Italy, claudia.papi13{at}gmail.com, patient recruitment and data collection), Simone Guerrieri (Neurology Unit, IRCCS San Raffaele Scientific Institute, Milano, Italy, guerrieri.simone{at}hsr.it, patient recruitment and data collection), Francesco Pisani (Child Neuropsychiatry Unit, Neuroscience Division, Medicine & Surgery Department, University of Parma, Parma, Italy, francesco.pisani{at}unipr.it, patient recruitment and data collection), Andrea Praticò (Unit of Rare Diseases of the Nervous System in Childhood, Department of Clinical and Experimental Medicine, Section of Pediatrics and Child Neuropsychiatry, University of Catania, Catania, Italy, andrea.pratico{at}unict.it, patient recruitment and data collection), Chiara Po' (Pediatric Neurology and Neurophysiology Unit, Department of Women's and Children's Health, Treviso, Italy, chiarapo.ped{at}gmail.com, patient recruitment and data collection), Luisa Grazian (Pediatric Unit, ULSS 2 Marca Trevigiana, Ca' Foncello Hospital, Treviso, Italy, luisa.grazian{at}aulss2.veneto.it, patient recruitment and data collection), Alice Passarini (Child Neuropsychiatry Unit, ASST Grande Ospedale Metropolitano Niguarda, Milano, Italy, alice.passarini{at}ospedaleniguarda.it, patient recruitment and data collection), Stefania Bergamoni (Child Neuropsychiatry Unit, ASST Grande Ospedale Metropolitano Niguarda, Milano, Italy, STEFANIA.BERGAMONI{at}ospedaleniguarda.it, patient recruitment and data collection), Sasha Rasia (Multiple Sclerosis Center, Spedali Civili di Brescia, Presidio di Montichiari, Montichiari, Italy, Sasha.rasia{at}gmail.com), patient recruitment and data collection), Roberto Bergamaschi (IRCCS Mondino Foundation, Pavia, Italy, roberto.bergamaschi{at}mondino.it, patient recruitment and data collection), Enrico Marchioni (IRCCS Mondino Foundation, Pavia, Italy, enrico.marchioni{at}mondino.it, patient recruitment and data collection).
Contributors MG: original study design, data analysis and manuscript drafting. This authors is responsible for the overall content as guarantor. TF: patient recruitment and data collection and contribution to study design. GG and ER: patient recruitment and data collection, data analysis, figure drafting and manuscript revision. SSc: laboratory data. SM: manuscript revision. SF, SM, LB, MM,TG, DF, MDC, AG, MDF, LB, GN, MV, PB, RI, LM, ET, SSar, MN, MR, SSav, EC and SM: patient recruitment and data collection. EB and SJ: manuscript revision and data analysis. DF: original study design and manuscript revision.
Funding The present study was supported by the Italian Ministry of Health, 'Ricerca Corrente' Grant to the IRCCS Mondino Foundation (code RC22012B).
Competing interests MG has received honoraria as a speaker and for the partecipation to Advisory boards from Roche, UCB and Alexion.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.