A systematic review of clinical data and reporting quality in NMDAR-antibody encephalitis and pregnancy

239 words Key messages: 182 words Word count: 3354 words Tables/Illustrations: 1 Table, 5 Figures

34 Supplementary material: 6 Tables, 4 Figures All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint

Background NMDAR antibody encephalitis (NMDAR-Ab-E) can have an onset during, after, or prior to a pregnancy. In animal models, transplacental NMDAR-IgG transfer can affect neurodevelopment. In contrast, clinical reports of mothers affected by NMDAR-Ab-E typically are reassuring. We systematically reviewed maternal, infant, and childhood clinical data pertaining to NMDAR-Ab-E with an onset before, during, or after pregnancy and compared this to our single autoimmune neurology centre experience.

After pre-registration on PROSPERO (CRD42023408447), we searched PubMed and Scopus for NMDAR-Ab-E case reports/series with an onset before, during, or after pregnancy (last search 19/10/2023). We extracted maternal, neonatal, and childhood outcomes using an idealised checklist to derive summary statistics.

After quality control we identified 66 pregnancies in 61 women from 48 reports or series. 72% of women recovered with minimal or no neurological deficits, comparable to non- pregnancy associated NMDAR-Ab-E. Likewise, 80% of pregnancies resulted in livebirths with a single neonatal death reported. Data on neonatal outcome measures were frequently unreported and childhood follow-up in only 60%. Our centre’s experience is consistent: 3/4 mothers recovered with no functional deficits and 7/8 children without evidence of compromise at median of two years follow-up.

Current evidence does not overall suggest unfavourable maternal, fetal, or childhood outcomes after NMDAR-Ab-E. However, the available sample is small, predominantly single case reports with modest follow-up, lacks standardisation, and data are often incomplete. Future approaches should address these caveats; developing multi-centre collaboration towards an international registry. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Key messages What is already known on this topic Some animals models of NMDAR-IgG transplacental transfer show adverse effects on brain development. However, caveats include species differences and potentially non-physiological exposures. Moreover, although some case reports identify adverse maternal and fetal outcomes, previous systematic reviews and single centre summaries of clinical data have been more reassuring. What this study adds We update and expand upon previous systematic reviews by including cases of NMDAR-Ab- E in the postpartum period and cases of pregnancy after recovery, as well as reporting the experiences of our autoimmune neurology centre. Additionally, we also focus on childhood outcomes and have contacted authors of published case reports for further follow-up. These data show generally good outcomes for mothers and children but reporting is patchy and not standardised. How this study might affect research, practice or policy To overcome these shortcomings in reporting we recommend collaboration amongst the autoimmune neurology clinical-research community to consolidate experience. This could include establishing an international registry to foster reporting standardisation and improve understanding of interactions between the illness, pregnancy, and potential effects on neonatal and childhood outcomes. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint

N-methyl-D-aspartate receptor antibody encephalitis (NMDAR-Ab-E) is an autoimmune neurological disorder predominantly affecting women of reproductive age (1,2). Mediated by IgG autoantibodies against the NR1 (GluN1) subunit of the NMDA receptor (NMDAR-IgG), NMDAR-Ab-E presents with combinations of acute psychiatric disturbance, movement disorders, seizures, dysautonomia, hypoventilation, and altered level of consciousness. Increasingly, this condition has been identified during pregnancy or in the postpartum period (3). Additionally, many who recover from the illness have yet to start or complete their family. They and their clinicians require clarity on potential risks for both mother and baby. NMDAR-IgGs are typically of the IgG1 subclass. IgG1 autoantibodies can cross the placenta and induce congenital disease including in the nervous system. For example, in myaesthenia gravis autoantibodies against fetal acetylcholine receptor isoforms can cause fetal acetylcholine receptor antibody-related disorders, a spectrum of disorders ranging from milder myopathic presentations to arthrogryposis multiplex congenita (4). In these cases, immunomodulation, particularly early in pregnancy, has been shown to improve survival and reduce complications for the developing foetus. In an era of autoantibodies against central nervous system targets a similar question has been posed of fetal brain development. In animal models, CASPR2 and NMDAR-autoantibodies have been shown potentially to affect neurodevelopment (5,6). Furthermore, NMDAR-IgG seropositivity often persists despite clinical remission (7,8) and so syncytiotrophoblastic neonatal FcRn receptors could mediate transfer of the dominant IgG1 sub-class autoantibodies (9,10). Nonetheless, real world clinical outcomes have been more reassuring. For example, a previous systematic review found 10/13 livebirths with 8/10 healthy neonates (3) and an experienced autoimmune All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint neurology centre reported 10/11 neonates healthy at birth (11). Moreover, in cases where there has been proven NMDAR-IgG transfer with sub-optimal neonatal outcomes, potential confounders have included maternal condition, medication, and placental factors (12,13). Here, we aimed to assess maternal, fetal/neonatal, and childhood outcomes with a focus on reporting quality to inform recommendations on future standards. We deployed an idealised checklist of features in pregnancy and developmental features to systematically review literature-reported cases and compare with experience from our own autoimmune neurology centre.

We pre-registered the study protocol with NIHR PROSPERO on 17/3/2023 (CRD42023408447) and followed Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidance. Search strategy We searched two databases (PubMed and Scopus) without language or date restriction using the search terms (“anti-NMDA receptor” OR “anti-NMDAR” OR “anti-N-methyl-D- aspartate receptor encephalitis” OR “NMDAR-antibody encephalitis” OR “NMDAR-Ab-E” OR “NMDAR encephalitis” OR NMDARe) AND (pregnancy OR postpartum OR post-partum OR puerperal OR puerperium OR foetus OR fetus OR gestation OR birth OR neonate OR infant OR child OR perinatal). We screened the reference lists of included papers for additional publications. The search was repeated twice to identify any papers published prior to the final analysis. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Eligibility criteria We included case reports and series which reported on patients with an onset of NMDAR- Ab-E before (‘non-pregnancy-associated’), or during or after pregnancy (‘pregnancy- associated’), as well as reports of children born to these patients. We planned to restrict to cases that strictly met the 2016 consensus criteria for definite anti-NMDAR encephalitis (14). However, our initial search yielded seven cases, including four published prior to these criteria, which did not fully meet definite classification due to not measuring CSF NMDAR- IgG. Yet being highly typical for the illness they met probable criteria, and given the modest sample size and valuable clinical information therein, we chose to include these cases. We initially defined postpartum onset as within 42 days as per WHO (15). However, only two of eight postpartum cases occurred within this period. Further aiming to maximise the inclusion of clinically-relevant information, we extended the postpartum definition to include cases where the presenting disorder was classified as postpartum in onset, which here was a maximum of 11 months postpartum. Outcome measures The full template for data collection including all extracted outcomes is provided in Supplementary Table 1. As primary outcomes we aimed to ascertain maternal morbidity, mortality, and functional status, pregnancy complications, and morbidity, mortality, and functional status in neonates ( baby <28 days old), and where available, later childhood developmental progress. We defined preterm birth as before 37 weeks and low birth weight at term as 500ml as per RCOG (17). We defined normal CSF parameters as protein concentration 15-40 mg/dL All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint and white cell count 0-5/mm3 (18). As secondary outcomes, we noted whether NMDAR-IgG autoantibodies were reported in cord or neonatal blood samples alongside maternal serology. Data extraction We removed duplicate papers to produce a final list of abstracts for screening. Two authors (SH, AAD) independently compared a representative sample (n=21) of the abstracts and reached consensus on inclusion with full agreement. The remaining abstracts were screened yielding 48 papers eligible for inclusion. SH performed the extraction which was then independently cross-checked (AAD, DS, and HF). Any differences were resolved by discussion. Where data was insufficient, we contacted the report authors to supplement the available published data. Quality Assessment Studies were assessed for quality using the tool for evaluating the methodological quality of case reports and case series (19). We made project-specific modifications to prioritise whether there was sufficient information to: 1) confirm the diagnosis of NMDAR-Ab-E and 2) allow basic evaluation of neonatal outcomes (Supplementary Table 1). Maternal outcome data was not used to determine inclusion as we did not wish to exclude records of children born secondary to pregnancies complicated by NMDAR-Ab-E which may not report maternal outcome. Those providing information on childhood outcome and with follow-up of at least one year were considered good quality. Local case series To contextualise the global experience from the systematic review we reported cases from our autoimmune neurology service who satisfied the same eligibility criteria. All are All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint participants in the Immune Factors in Neurological Disease research study (REC 16/YH/0013) and gave informed consent in accordance with the Declaration of Helsinki. If an assenting participant lacked capacity to consent for themselves then there was a next of kin declaration. Additional publication-specific consent was obtained for de-identified detailed individual participant data including offspring. Data analysis Data were tabulated with Excel version 16.83 (Microsoft). Statistical analyses and visualisation were conducted with Prism version 10.2.1 (GraphPad). Fisher’s exact test was used to compare between pregnancy groups and the binomial or Chi-square tests to compare observed to expected results. Statistical significance was inferred where P <0.05.

Identification of records Our initial search identified 1587 records (733 PubMed; 854 Scopus, Fig. 1). Later repeated searches identified an additional 107 records (49 PubMed, 58 Scopus). 638 duplicate records were removed leaving 1056 for abstract screening. This was then refined to 60 eligible records including two identified through abstract screening (Supplementary Table 2). Two systematic reviews and two papers with insufficient data were removed. We then quality-assessed the remaining 56 records, finding 48 of sufficient quality for inclusion (n=8 inadequate, n=33 adequate, n=15 good; Fig. 1). There were 53 NMDAR-Ab-E cases associated with pregnancy including 45 with an onset during pregnancy involving 43 individual women with two cases of relapse during subsequent pregnancies. There were eight cases after pregnancy. There were 13 non-pregnancy-associated cases occurring before a All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint pregnancy including two women who had NMDAR-Ab-E during a previous pregnancy, and one woman who had two pregnancies after recovery. NMDAR-Ab-E description and treatment The average age of illness onset was consistent across the groups (Fig. 2). The overall ovarian teratoma rate in the pregnancy-associated cases was 43% (23/53) comparable to the literature-reported rate (20). This comprised 17/45 (38%) during and 6/8 (75%) after (non- significant; P = 0.065, Fisher’s exact test). The majority of cases that occurred during pregnancy were early with only four in the third trimester (9% (4/45), P = 0.001, Chi-Square test). The overall clinical profile of the cases associated with pregnancy versus non-pregnancy associated cases differed in autonomic dysfunction (40% vs 10%, P = 0.145, Fisher’s exact test), reduced consciousness (70% vs 10%, P = 0.0006, Fisher’s exact test), and hypoventilation (49% vs 10%, P = 0.034, Fisher’s exact test) (Supplementary Fig. 1A). The rates of these features in the pregnancy-associated group is broadly in keeping with a recent large meta-analysis of 1550 predominantly female patients, where reduced level of consciousness was reported in 55% and autonomic dysfunction or central hypoventilation in 43% (20). Investigation and treatment profiles were also broadly similar across case types (Supplementary Fig. 1B-D). Pregnancy outcomes For cases with an illness onset during pregnancy, most resulted in live births (33/45, 73%) (Fig. 3A). However, 18/33 (55%) were preterm (median gestational age 33 weeks, range 27- 36) of which most were iatrogenic, i.e. either induced or involved a caesarean section (14/18, 78%; Fig. 3A). Fourteen women had a livebirth at term (14/33, 42%) and 29% (4/14) of these All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint babies were born by caesarean section. We plotted available birth weights against week of delivery and these clustered at or below the normative median with 1/12 >97 th centile and 2/12 <3rd centile (Fig. 3C). One paper reported a birthweight of 408g at 33 weeks which is close to the limit of viability and appeared implausibly low (21). Attempts to contact the authors to clarify were unsuccessful therefore we elected to exclude this data point from this analysis (retained and compared for reference in Supplementary Fig. 2). Six pregnancies ended spontaneously (four miscarriages and two stillbirths) and another six were terminated (Supplementary Table 3). In the non-pregnancy-associated before group there was one termination and the rest were livebirths (12/13, 92%) with two (17%) caesarean sections. All the cases with illness onset after pregnancy were livebirths with no sections reported. Generally, antenatal and delivery outcomes were rarely reported (Supplementary Fig. 3). Maternal outcomes Most women recovered fully or with minimal neurological deficit ( before 8/10, 80%; during 31/43, 72%; after 5/8, 63%). However, follow-up duration was modest, with only 12/40 (30%) of the during cases reporting maternal follow-up for more than a year. Across all the cases there were four maternal deaths (4/61, 6.6%; Fig. 4A – top and Supplementary Table 4). These were predominantly secondary to sepsis, a relatively common cause of maternal death that accounted for 10% of maternal deaths in the UK between 2019-2021 (22). Neonatal outcomes Overall, while there was little evidence for poor neonatal outcomes (Fig. 4A – middle), reporting of specific neonatal outcomes was generally sparse (Fig. 4B). This was especially All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint true for after cases where there was only one case of preterm birth reported, secondary to placental abruption (Fig. 4A – middle). In three of these cases no specific statement on neonatal health was given, but we could infer the neonate was alive because the mother was breastfeeding, or the delivery was described as normal. Overall, 7/53 (five during cases and two before cases; 13%) neonates were identified as compromised (Supplementary Table 5). Of these seven, five were tested for NMDAR-IgG, of which four (80%) were positive. Five healthy neonates were tested with only one positive (Fig. 5A-B). There was one death reported of an already compromised neonate. The mother had previously recovered from NMDAR-Ab-E, although the interval to the pregnancy was relatively short with the illness onset preceding delivery by 18 months (13). Whilst there were concerns regarding an encephalitis relapse, the patient was unaware of her pregnancy and presented with unmodified hypertension and deranged liver function consistent with pre- eclampsia, indicative of a likely confounding aetiology. Childhood outcomes Of the 33 cases in which illness began during pregnancy ending with a live birth, childhood outcomes were provided in 22 (Fig. 4A – bottom). The level of detail was largely restricted to a general statement in most cases that the children were healthy and/or meeting developmental milestones (Fig. 4C). However, the duration of follow-up and therefore opportunity to identify more complex neuro-developmental outcomes was limited. For cases with an illness onset during pregnancy 8/22 (36%) were followed-up longer than a year, whilst this figure was 44% for those born to mothers who had recovered from NMDAR-Ab-E prior to pregnancy (4 of 9 cases providing childhood follow-up, Fig. 4D). There was no childhood data provided for cases with an illness onset after pregnancy. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Amongst during cases, one child had global developmental delay and in this case the mother’s illness was severe and she died secondary to infection (12). The baby’s serum was positive for NMDAR-IgG at birth but negative by one year. This child was identified as compromised in the neonatal period, but the other three compromised neonates for whom data was available, went on to develop normally. For the before cases only one child had reported medical diagnoses, which were torticollis and strabismus (23). This child was not identified as compromised at birth and the one surviving compromised child was described as developmentally normal. Maternal, neonatal, and child outcomes are summarised in Supplementary Fig. 4. Given the number of potential childhood outcomes left unreported and limited length of reported follow-up we contacted authors to ascertain if further follow-up was available. 7 of 27 authors contacted responded, of which three were able to provide further follow-up. No new diagnoses were made to alter the existing reported literature. Oxford autoimmune neurology experience In addition to a group summary (Table 1), where possible we obtained specific consent to report de-identified individual participant data according to our checklist (Supplementary Table 6). We have not encountered any patients with a postpartum onset of the illness but two cases with an onset during pregnancy, a relapse and first illness. Both neonates were born premature but live (one spontaneous delivery and one emergency C-section secondary to non-reassuring fetal heartbeat). Both were admitted to the special care baby unit and have developed along normal trajectories with a median follow-up of 1.5 years. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Beyond this, most of our experience is non-pregnancy-associated NMDAR-Ab-E, occurring and resolving before pregnancy (seven pregnancies from three mothers). All mothers had recovered without residual deficit, and six pregnancies resulted in term livebirths, with one miscarriage. Four of the six neonates were born in good condition. One had raised respiratory rate at birth and was treated for possible sepsis. Heel prick blood from this neonate was positive for NMDAR-IgG, but there have been no developmental concerns with five years of follow-up. The other was treated for sepsis and jaundice in the context of preterm premature rupture of membranes (PPROM). Here, development has been largely as expected but an assessment for potential neurodiverse needs is awaited. All other children were achieving normal milestones at most recent follow-up (median 3 years old, range 0.5-5).

Overall, we have found a relatively small and still developing literature. The evidence was of sufficient quality to synthesise, but compared to our idealised checklist, there was considerable missing data. While our conclusions are largely reassuring, the strength of the evidence is low and should be considered provisional. However, given the prevalence of case reports (45/48, 94%) a format that is intrinsically potentially biased towards atypicality and concern, their low frequency offers a degree of reassurance. Maternal outcomes did not differ significantly from a systematic review of the disease overall, with 36/51 (71%) of pregnancy-associated cases either fully recovering or with minimal deficit versus 918/1284 (72%) with mRS 0-2. The rate of maternal deaths was also similar (4/51, 8%, versus 81/1284, 6%) (20). The NMDAR-Ab-E cases with onset before pregnancy were generally less severe. This could plausibly reflect a selection bias of sufficient recovery to allow subsequent pregnancy. With regard to pregnancy outcome in All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint cases with an onset during pregnancy, three miscarriages occurred during the first trimester, and one in the second trimester, broadly in keeping with spontaneous fetal loss (15% in the first trimester and 1-2% in the second) (24,25). While the stillbirth rate was higher than

(5.7% (2/35) vs 0.4%; P = 0.0171, binomial test) (26), this is lower than has been reported for pregnant women admitted to intensive care (9.9%) (27). Moreover, the small sample size of this rare sub-group of a rare disease influenced by reporting bias considerably caveats this comparison. It is encouraging that the literature and our own experience find cases of pregnancy following resection of ovarian teratoma to treat acute NMDAR-Ab-E. Despite published reports of ovary-preserving surgery (28), given the risk of residual teratoma tissue driving ongoing disease or relapse, oophorectomy remains common. Therefore, consideration of preserving oocytes in young women who have yet to start a family is important. Our multi-disciplinary approach includes a specialist gynaecologist with expertise in both teratoma resection and ovarian cryo-preservation to discuss options with patients and their next of kin (29). Despite the evidence from animal models, in the available published data we found little positive evidence of developmental disorders in children born to mothers in any of the three sub-groups. We found one reported case of developmental delay reported in 53 (2%) livebirths, broadly in keeping with the frequency amongst children under five in the general population (1-3%) (30). Moreover, whilst this infant’s serum was positive for NMDAR-IgG, the infant was also born prematurely secondary to uteroplacental insufficiency and the mother had severe illness and died of secondary infection (12). Thus, the potential specific effects of transplacental transfer of NMDAR-IgG in this case are challenging to disentangle from other relevant factors. Similarly, in our case series of children born to mothers with All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint either active or previous NMDAR-Ab-E, only one is being evaluated for neurodiverse needs and there was no detectable maternal seropositivity during the pregnancy. Conversely, the quality of the data we have identified makes it impossible to fully exclude an increased risk of neurodevelopmental conditions. Firstly, specific neonatal health outcomes are often missing from reports (Fig. 4C). Secondly, the data on long term outcomes are very limited and human neurodevelopment manifests over years-decades. Moreover, outside of our re-contacts data, duration of follow-up has not been updated in the literature. Additionally, the type of cohort study design needed to truly determine the effect of autoantibody transfer to be adequately powered and control sufficiently for confounding variables would require multi-centre co-ordination. Pregnancies occurring after full disease remission may be amenable to this but those complicated by NMDAR-Ab-E during or after are by definition heavily confounded by the effects of the disease on maternal and placental condition as well as by the multiple supportive medical, interventional, and immunotherapeutic interventions needed to survive and recover from the illness. Finally, testing of trans-placental autoantibody transfer remains rare. Compromised neonates were disproportionately likely to be tested, making up five of ten (50%) cases tested despite only seven of the 53 (13%) cases reporting a compromised infant. Four of the five comprised infants tested were positive for NMDAR-IgG, but two went on to meet their developmental milestones at one year of age, whilst one had global developmental delay, and the other died during the neonatal period (Supplementary Table 5). Both the latter had relatively high titres at 1:320 and 1:450 respectively, substantially higher than the 1:20 titre of one of the infants who developed normally but comparable to the other where the titre was 1:400. Furthermore, one healthy infant tested positive for serum NMDAR-IgG (31). Autoantibody testing was All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint available for one neonate in our cohort, and whilst compromised at birth they too have subsequently developed normally. Thus, the clinic-pathologic sequelae of transplacental NMDAR-IgG in humans remains to be fully elucidated. Certainly, NMDARs are important in the developing fetal brain (32). Evidence from mouse models indicates that transplacental transfer of patient derived NMDAR-IgG can result in reduced survival rates in the postnatal period, with reduced brain volume and neurodevelopmental abnormalities in adulthood (5). However, human and murine neurodevelopment differ with an established blood-brain barrier forming postnatally in mice and between 22 and 32 weeks’ gestation in humans (33). Overall, the reliance on reports introduces a reporting bias and unsystematic reporting. The development of a global, confidential registry to systematically record disease presentation, treatments and disease course, in addition to obstetric, maternal, neonatal and childhood outcomes could improve evidence quality. This has precedent in obstetric practice with caesarean scar pregnancies and within neurology for multiple sclerosis (34,35). Our idealised checklist could serve as a starting point, with further input from existing international clinical-research networks including multi-disciplinary expertise supported by patient advocacy organisations. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint

1. Dalmau J, Tüzün E, Wu H, Masjuan J, Rossi JE, Voloschin A, et al. Paraneoplastic anti-N-methyl-D-aspartate receptor encephalitis associated with ovarian teratoma. Ann Neurol. 2007;61(1):25–36. 2. Irani SR, Bera K, Waters P, Zuliani L, Maxwell S, Zandi MS, et al. N-methyl-D- aspartate antibody encephalitis: temporal progression of clinical and paraclinical observations in a predominantly non-paraneoplastic disorder of both sexes. Brain. 2010;133:1655–67. 3. Shi Y-C, Chen X-J, Zhang H-M, Wang Z, Du D-Y. Anti-N-Methyl-d-Aspartate receptor (NMDAR) encephalitis during pregnancy: Clinical analysis of reported cases. Taiwan J Obstet Gynecol. 2017;56(3):315–9. 4. Allen NM, O’Rahelly M, Eymard B, Chouchane M, Hahn A, Kearns G, et al. The emerging spectrum of fetal acetylcholine receptor antibody-related disorders (FARAD). Brain. 2023;146(10):4233–46. 5. Jurek B, Chayka M, Kreye J, Lang K, Kraus L, Fidzinski P, et al. Human gestational N /i1methyl/i1d/i1aspartate receptor autoantibodies impair neonatal murine brain function. Ann Neurol. 2019; 86(5):656–70. 6. Coutinho E, Menassa DA, Jacobson L, West SJ, Domingos J, Moloney TC, et al. Persistent microglial activation and synaptic loss with behavioral abnormalities in mouse offspring exposed to CASPR2-antibodies in utero. Acta Neuropathol. 2017;134(4):567–83. 7. Makuch M, Wilson R, Al /i1Diwani A, Varley J, Kienzler A, Taylor J, et al. N/i1methyl/i1D/i1aspartate receptor antibody production from germinal center reactions: Therapeutic implications. Ann Neurol. 2018;83(3):553–61. 8. Al-Diwani A, Theorell J, Damato V, Bull J, McGlashan N, Green E, et al. Cervical lymph nodes and ovarian teratomas as germinal centres in NMDA receptor-antibody encephalitis. Brain. 2022;145(8):2742–54. 9. Tüzün E, Zhou L, Baehring JM, Bannykh S, Rosenfeld MR, Dalmau J. Evidence for antibody-mediated pathogenesis in anti-NMDAR encephalitis associated with ovarian teratoma. Acta Neuropathol. 2009;118(6):737–43. 10. Palmeira P, Quinello C, Silveira-Lessa AL, Zago CA, Carneiro-Sampaio M. IgG placental transfer in healthy and pathological pregnancies. Clin Dev Immunol. 2012;985646. 11. Dalmau J. Pregnancy, N /i1Methyl/i1D/i1Aspartate Receptor Antibodies, and Neuropsychiatric Diseases. Ann Neurol. 2020;87(2):324–5. 12. Jagota P, Vincent A, Bhidayasiri R. Transplacental transfer of NMDA receptor antibodies in an infant with cortical dysplasia. Neurology. 2014;82(18):1662–3. 13. Chourasia N, Watkins MW, Lankford JE, Kass JS, Kamdar A. An Infant Born to a Mother With Anti-N-Methyl-d-Aspartate Receptor Encephalitis. Pediatr Neurol. 2018;79:65–8. 14. Graus F, Titulaer MJ, Balu R, Benseler S, Bien CG, Cellucci T, et al. A clinical approach to diagnosis of autoimmune encephalitis. Lancet Neurol. 2016;15(4):391– 404. 15. WHO Technical Consultation on Postpartum Care. 2010. 16. WHO recommendations for care of the preterm or low-birth-weight infant. 2022. 17. RCOG. Prevention and Management of Postpartum Haemorrhage (Green-top Guideline No. 52)e. 2016. 18. Hrishi AP, Sethuraman M. Cerebrospinal Fluid (CSF) Analysis and Interpretation in Neurocritical Care for Acute Neurological Conditions. Indian J Crit Care Med. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint 2019;23(Suppl 2):S115–9. 19. Murad MH, Sultan S, Haffar S, Bazerbachi F. Methodological quality and synthesis of case series and case reports. BMJ evidence-based Med. 2018;23(2):60–3. 20. Nosadini M, Eyre M, Molteni E, Thomas T, Irani SR, Dalmau J, et al. Use and Safety of Immunotherapeutic Management of N -Methyl- d -Aspartate Receptor Antibody Encephalitis. JAMA Neurol. 2021;78(11):1333. 21. Ogawa M, Matsuda Y, Kanda E, Konno J, Mitani M, Makino Y, et al. Survival rate of extremely low birth weight infants and its risk factors: case-control study in Japan. ISRN Obstet Gynecol. 2013;2013:873563. 22. MBRRACE-UK. Lessons learned to inform maternity care from the UK and Ireland Confidential Enquiries into Maternal Deaths and Morbidity 2019-21. 2023. 23. Magley J, Towner D, Taché V, Apperson ML. Pregnancy outcome in anti-N-methyl- D-aspartate receptor encephalitis. Obstet Gynecol. 2012;120(2 Pt 2):480–3. 24. Shaker M, Smith A. First Trimester Miscarriage. Obstet Gynecol Clin North Am. 2022;49(3):623–35. 25. Cullen S, Sobczyk K, Elebert R, Tarleton D, Casey B, Doyle S, et al. Second-trimester miscarriage: a review of postnatal investigations and subsequent pregnancy outcomes. Irish J Med Sci. 2023;192(4):1757–60. 26. Births in England and Wales - Office for National Statistics. 27. Ramlakhan KP, Gommers D, Jacobs CERM, Makouri K, Duvekot JJ, Reiss IKM, et al. Women of reproductive age in a tertiary intensive care unit: indications, outcome and the impact of pregnancy-a retrospective cohort study. BMC Womens Health. 2021;21(1):248. 28. Jones BP, Rees R, Saso S, Stalder C, Smith JR, Yazbek J. Ultrasound-guided laparoscopic ovarian preserving surgery to treat anti-NMDA receptor encephalitis. BJOG. 2017;124(2):337–41. 29. Lakhoo K, Davies J, Chakraborty S, Berg S, Tennyson R, Fowler D, et al. Development of a new reproductive tissue cryopreservation clinical service for children: the Oxford programme. Pediatr Surg Int. 2019;35(11):1271–8. 30. Mithyantha R, Kneen R, McCann E, Gladstone M. Current evidence-based recommendations on investigating children with global developmental delay. Arch Dis Child. 2017;102(11):1071–6. 31. Zengin E, Kharisova I, Emechebe D, Anziska Y. Concurrent NMDAR and GFAP Antibody Encephalitis During Pregnancy. BMJ Case Rep. 2023;16(7). 32. Ewald RC, Cline HT. NMDA Receptors and Brain Development. Biology of the NMDA Receptor. CRC Press/Taylor & Francis; 2009. 33. Semple BD, Blomgren K, Gimlin K, Ferriero DM, Noble-Haeusslein LJ. Brain development in rodents and humans: Identifying benchmarks of maturation and vulnerability to injury across species. Prog Neurobiol. 2013;106–107:1–16. 34. Agten AK, Monteagudo A, Timor-Tritsch IE, Thilaganathan B. Cesarean Scar Pregnancy Registry: an international research platform. Ultrasound Obstet Gynecol. 2020;55(4):438–40. 35. Butzkueven H, Chapman J, Cristiano E, Grand’Maison F, Hoffmann M, Izquierdo G, et al. MSBase: an international, online registry and platform for collaborative outcomes research in multiple sclerosis. Mult Scler. 2006;12(6):769–74. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Tables Table 1 – Group summary of single centre experience of NMDAR-Ab-E and pregnancy with maternal, neonatal, and childhood outcomes NMDAR-Ab-E onset relative to pregnancy Number of patients, n Number of pregnancies, n Trimester at NMDAR-Ab- E onset Miscarriage, n Live births, n Birth timing years post illness onset, median (range) Preterm birth, n T erm birth, n SCBU admission, n Years maternal follow-up since episode, median (range) Maternal outcome mRS 0-1, n Yea rs chil d follow-up, median (range) Child outcome within normal limits, n Before 3 7 N/A 1 6 3.5 (1.5-6) 0 6 1* 7 (2-12) 3 1.75 (0.5-5) 5 During 2 2 1 st and 2nd 0 2 N/A 2 0 2 1.5 (1-2) 1 1.5 (1-2) 2 After 0 0 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A *Phototherapy for neonatal jaundice in context of PPROM Abbreviations: mRS=modified Rankin score, NMDAR-Ab-E=NMDAR-antibody encephalitis, n=number, N/A = not applicable, SCBU=special care baby unit All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Figure legends Figure 1: PRISMA flow diagram The combined findings of the initial search conducted on the 21 st of March 2023 and subsequent searches on the 17th of July and 19th of October 2023 are shown. After screening, removal of duplicates, and quality control, 48 studies were identified as eligible for inclusion. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Figure 2: Onset of NMDAR-antibody encephalitis relative to maternal age and pregnancy Cases of NMDAR-antibody encephalitis are plotted according to stage of pregnancy and maternal age. They are split into pregnancy associated (illness onset during or after pregnancy; left) and non-pregnancy associated (illness onset before pregnancy; right). For during cases, 41 of the 45 cases are represented since four cases did not give a specific onset time in gestational weeks but were all within the first trimester and two were associated with a teratoma. One teratoma-associated before case is not shown since the time interval was not clearly stated. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Figure 3: Pregnancy outcomes A – Cases of NMDAR-antibody encephalitis during pregnancy are shown according to illness onset in relation to gestational time and pregnancy outcome (gestational week at presentation = triangles; end of pregnancy = circle; termination = black; spontaneous fetal loss = grey; spontaneous livebirth = blue; iatrogenic livebirth = red). Vertical dotted lines represent the end of the first and second trimesters and the dashed line at week 37 demarcates term. Four first trimester cases of fetal loss (two termination and two miscarriage) are not shown as specific timings were unavailable. Three spontaneous livebirths are not shown since specific delivery time was not described (36-38). B – In cases with an illness onset during pregnancy, pie charts summarise the proportion and absolute number of indications for caesarean section (preterm deliveries, top; term deliveries, bottom). C – Twelve available birthweights are plotted by gestational week (onset during pregnancy pink, onset before pregnancy black; uncompromised=circle; compromised=triangle) with a line of best fit (grey) in the context of normative birth weight ranges (blue lines; as per Nicolaides et al Ultrasound Obstet Gynecol 2018; 52: 44–51). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Figure 4: Maternal, neonatal and childhood outcomes A – Maternal, neonatal and childhood outcomes are shown divided by onset of NMDAR- antibody encephalitis in relation to pregnancy. For maternal outcomes, cases where deficit was reported as “minimal” or modified Rankin score (mRS)=1-2 are shown as mild deficit. Cases are defined as moderate-severe deficit for mRS 3-5 or clear functional disability described. For neonatal outcomes, infants are specified as compromised if this was specifically stated or as indicated by Apgar scores. B – Positively reported neonatal outcomes are plotted as percentages for pregnancies with an onset of NMDAR-Ab-E occurring during (pink) or before (black). C – Reported childhood outcomes are plotted as percentages for pregnancies with an onset of NMDAR-Ab-E occurring during (pink) or before (black). General statement refers to a descriptor of adequate progress such as “healthy”, “met all developmental milestones”, “developing normally”. Also plotted are statements of a specific condition and whether duration of follow-up was stated. D – Where available, specific duration of childhood follow-up of global development is plotted as a bar chart for pregnancies with an onset of NMDAR-Ab-E occurring during (left) or before (right). Offspring in whom an abnormality was reported are shaded grey. Abbreviations: APGAR=Appearance, Pulse, Grimace, Activity and Respiration, mRS=modified Rankin score, NMDAR-Ab-E=NMDAR-antibody encephalitis, NICU= neonatal intensive care unit. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Figure 5: Neonatal NMDAR-autoantibody testing A – The absolute number of infants tested for NMDAR IgG autoantibodies are plotted according to the test result. Positive results are divided according to whether the infant was clinically healthy or compromised at the time of testing. The bars are sub-divided according to whether NMDAR-Ab-E occurred during (pink) or before (black) the associated pregnancy. B – Results of NMDAR-IgG assays are plotted according to corresponding bio-fluid tested. Dotted lines note samples connected within a maternal-neonatal pair. Abbreviations: CSF=cerebrospinal fluid, NMDAR-IgG=Immunoglobulin G autoantibody against N-Methyl D-Aspartate receptor All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Authors contributions Conceptualisation – SLH, DS, HF, AAD Data collection and curation – SLH, SNMB, PC-W, AAD Formal analysis – SLH, AH, AAD Writing - original draft – SLH, AAD Writing - review & editing – All Funding sources SLH and PC-W declare no funding. DS is funded by a NIHR Clinical Lectureship and Academy of Medical Sciences. Starter Grant for Clinical Lecturers (SGL029\1038). SNMB is funded by a National Institute for Health Research (NIHR) Clinical Lectureship. MIL is funded by the UK National Health Service (Myasthenia and Related Disorders Service and National Specialised Commissioning Group for Neuromyelitis Optica, UK) and by the University of Oxford, Oxford, UK. She has been awarded research grants from the UK association for patients with myasthenia, Myaware, and the University of Oxford. She has received speaker honoraria or travel grants from Biogen Idec, Novartis, argenx, UCB, and the Guthy-Jackson Charitable Foundation. MIL serves on scientific or educational advisory boards for UCB Pharma, argenx, and Viela/Horizon. SRI declares funding by a senior clinical fellowship from the Medical Research Council (MR/V007173/1), Wellcome Trust Fellowship (104079/Z/14/Z) and the NIHR Oxford Biomedical Research Centre (BRC). AEH declares funding by the Medical Research Council (MR/X022013/1), Oxford Health Biomedical Research Centre (BRC), MyAwa re, and UCB Pharma. AAD is funded by a NIHR Clinical Lectureship, Academy of Medical Sciences Starter Grant for Clinical Lecturers (SGL027\1016), and Oxford Health Biomedical Research Centre (BRC). The All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. The funders played no role in the study. Conflicts of interest SRI is a co-applicant and received royalties on patent application WO/2010/046716 (Neurological Autoimmune Disorders) and has filed two other patents regarding autoantibody diagnostic algorithms. None of these are felt to be of direct relevance to the current manuscript. The remaining authors declare no commercial conflicts of interest. Data availability statement Further data supporting the study are available from the corresponding authors upon reasonable request. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint Records Identified PubMed: 782 Scopus: 912 Duplicates Removed: 638 Abstracts screened: 1056 Excluded: 998 Full text-article assessed for eligibility:60 Number of full text articles excluded • Systematic review: 2 • Insufficient information in text: 2 • Failed Quality Assessment: 8 Number of studies included: 48 IdentificationScreeningEligibilityIncluded

Screening: 2 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint 10 20 30 40 50 After n=8 Delivery 0 10 20 30 40 10 15 20 25 30 35 40 45 50Age (years) During n=45 1st trimester 2nd trimester 3rd trimester Term 0-10 11-20 21-30 31-40 41-50 0 1 2 3 4 5 6 7 10 15 20 25 30 35 40 45 50Age (years) Before n=13 Onset of NMDAR-antibody encephalitis relative to pregnancy 0 5 10 15 20 25 Frequency 0-10 11-20 21-30 31-40 0 5 10 15 20 25 Frequency 0-1 2-3 4-5 6-71-2 5-63-4 No Years between initial onset and deliveryWeek of pregnancy Weeks postpartum Yes Ovarian teratoma Pregnancy-associated n=53 Non-pregnancy-associated All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint A B C 0 10 20 30 40Term Week of pregnancy Termination n=6 Spontaneous n=6 Spontaneous n=15 Iatrogenic n=18 Kumar et al. Case 2 Liu et al. first Kokubun et al. Liu et al. second Kim et al. Zhang et al. second Reisz et al. Keskin et al. Joubert et al. Case 2 McCarthy et al. Joubert et al. Case 4 Dono et al. Fredrich et al. Kalam et al. Joubert et al. Case 5 Mathis et al. Mizutamari et al. Kumar et al. Case 3 Ito et al. Bansal et al. Shahani L. Scorrano et al. Jagota et al. Joubert et al. Case 6 Crowley et al. Kumar et al. Case 1 Sperling et al. Zengin et al. Demma et al. Lamale-Smith et al. Joubert et al. Case 3 Ueda et al. Jung et al. Joubert et al. Case 1 Tailland et al. Xiao et al. Duan et al. Liao et al. Pregnancy Outcomes - During Pregnancy 1st trimester 2nd trimester 3rd trimester Term Fetal loss n=12 Onset Delivery Live birth n=33 Emergency Elective Unclear Caesarean section (preterm) n=14 reported 7 4 2 Fetal heart rate =5 Spontaenous preterm labour = 1 Maternal arrhythmia = 1 Surgical exploration = 1 Large for dates (GDM) = 1 Maternal condition = 1 Maternal condition = 1 Indication not stated = 3 3 Caesarean section (term) n=4 reported 1 2 Failed ECV = 1 Surgical exploration = 1 Indication not stated = 21 Emergency Elective Unclear 2 30 35 40 0 1 2 3 4 5 Delivery (Weeks pregnant) Birth weight (kg) Birth weight n=12 reported During (uncompromised) Before (uncompromised) Before (compromised) During (compromised) Estimated birth weight (Nicolaides et al 2018) Term Linear regresion (all) Median 97th centile 3rd centile Pre-term All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint n = 43 14 17 3 8 1 n = 10 44 1 1 Maternal outcomes Neonatal outcomes Childhood outcomes Onset of NMDAR-Ab-E relative to pregnancyA n = 12 4 3 1 1 3 n = 9 8 1 n = 8 2 3 1 1 1 n = 33 12 15 2 3 1 7 1 n = 8 n = 22 21 1 Developmental Delay Other Medical Condition Healthy Pregnancy-associated Non-pregnancy-associated B C General statement Specific condition Follow-up duration 0 20 40 60 80 100Percent reporting (%) Childhood outcome reporting D Birth weight APGARsCord gasNICU stay CoolingInfectionSeizures 0 20 40 60 80 100Outcome reported (%) Neonatal outcome reporting Onset relative to pregnancy Before DuringDuring After Before Healthy, Dates Unknown Healthy, Term Healthy, Pre-term Compromised, Term Compromised, Pre-term Unknown Condition, Pre-term Neonatal Death Unknown Full Recovery Mild Deficit Moderate-Severe deficit Death 0-11-22-33-44-55-66-7 0 5 10 15 Number of offspring Global development Years of follow-up During Before 0-11-22-33-44-55-66-77-8 Normal Abnormal Onset relative to pregnancy Before During All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint 0 2 4 6 8 10Number of Infants Neonatal NMDAR-IgG testing n=10 reported Negative Positive Healthy Positive Compromised A B Maternal serumAmniotic fluidUmbilical cordNeonate serumNeonate CSF NMDAR-IgG across compartments DetectableUndetectable Onset relative to pregnancy Before During All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted December 1, 2024. ; https://doi.org/10.1101/2024.11.28.24317822doi: medRxiv preprint