Anti-GAD antibody-associated epilepsy in pregnancy

Introduction

Anti-GAD antibody-associated epilepsy (AAE) is a rare neurological condition caused by autoantibodies against the intracellular enzyme glutamic acid decarboxylase (GAD) 65. ¹ Seizures are associated with aberrant GABAergic transmission, and are driven by prolonged autoimmune responses and resultant structural damage to the brain parenchyma. ² The pathophysiology of anti-GAD AAE remains disputed, with evidence supporting possible contributions from both cell-mediated and humoral immune responses. ³ The condition most commonly manifests in women of childbearing age, ⁴ and is notoriously difficult to treat, with its symptoms often persistent and refractory to anti-seizure medications (ASM) and immunotherapy. ⁵

There are fewer than 200 case reports of anti-GAD AAE in the literature, ¹ with none in pregnancy to our knowledge.

Case report

A 33-year-old nulliparous female first presented at 7 weeks’ gestation. She had temporal lobe anti-GAD AAE diagnosed in 2010 and was managed pre-pregnancy with levetiracetam 1500 mg twice daily, lamotrigine 300 mg twice daily, azathioprine 125 mg daily and intravenous immunoglobulin (IVIG) infusions every 10 weeks. Her epilepsy was drug-resistant, and had previously been refractory to prednisolone, mycophenolate mofetil, tocilizumab and rituximab. Before pregnancy, her condition was relatively well-controlled; she was usually seizure-free in the first 6–7 weeks after IVIG, with symptoms then recurring before the next infusion. She had predominately focal impaired-awareness episodes, with occasional secondary generalised tonic-clonic seizures and a degree of memory impairment. She intermittently used the anticonvulsant clobazam to control her seizures. Her brain MRI showed stable appearances before pregnancy. She was otherwise well, with a past medical history of well-controlled asthma. She received pre-pregnancy counselling from our service two years before conception.

Throughout pregnancy, the woman was regularly followed up in a joint obstetric/neurology clinic. Though no changes were made to her medications at booking, she was informed that metabolism changes may warrant increasing doses of ASMs as the pregnancy progressed; levels were monitored at least monthly and doses adjusted accordingly.

At 19 + 3 weeks, she reported a single nocturnal generalised tonic-clonic seizure (her first in over a year). ASM levels were in the normal range: lamotrigine 5.1 mg/L (normal 3–15 mg/L), levetiracetam 33 mg/L (normal 12–46 mg/L). As her lamotrigine level had been higher when not pregnant (13 mg/L), was at the lower end of normal and was anticipated to fall further, the dose was increased to 350 mg twice daily. By 23 weeks, she was experiencing worsening focal impaired-awareness episodes, so the frequency of her IVIG was increased to every 8 weeks. Between 23 and 29 weeks, her ASM levels progressively declined, leading to increases in lamotrigine doses to 400 mg twice daily and levetiracetam to 1750 mg twice daily. She remained seizure free until 34 weeks when seizures recurred, precipitating a further increase in her levetiracetam dose to 2000 mg twice daily.

At 37 weeks of gestation, she presented with severe itch and elevated bile acids of 60µmol/L, leading to a diagnosis of intrahepatic cholestasis of pregnancy (ICP), a recognised adverse effect of azathioprine. A decision was made to proceed with induction of labour, and she delivered a healthy male infant by vaginal delivery at 38 weeks. She experienced a cluster of impaired-awareness episodes during labour, and a tonic-clonic seizure and two impaired-awareness episodes a few minutes after delivery; both were treated with clobazam. She expressed concern about experiencing impaired-awareness episodes whilst holding her baby, so was administered IVIG before discharge, approximately 7 weeks after her last infusion. At 3 weeks postpartum, her ASM doses started to be reduced and the IVIG changed to every 12 weeks. Her bile acids normalised following birth.

Early neonatal anti-GAD antibody testing and baseline neurological examination were undertaken, given the uncertain significance of passively transferred maternal antibodies. Cord blood testing at birth showed markedly elevated titres of anti-GAD antibodies (1400 units/ml, normal <5) and TPO antibodies (28.78IU/ml, normal <5), although neonatal examination was normal. These titres gradually decreased over time, with anti-GAD antibodies falling to 200 units/ml and TPO antibodies to normal levels within 6 months. The baby remains under paediatric immunology follow-up and reassuringly continues to be well, with no clinical or developmental concerns identified thus far at 1 year of age.

Discussion

This case is the first case report of AAE in pregnancy, and outlines the challenges in optimising disease control. Seizures increase in frequency as pregnancy progresses, due to physiological changes that alter the absorption, metabolism and excretion of ASMs. ⁶ Lamotrigine and levetiracetam levels can fall by as much as 40–60% in pregnancy, stressing the importance of regular dose adjustments as necessary to maintain good seizure control. ⁶ Given the severity of this patient's focal epilepsy, complete seizure freedom was not an expected outcome. Indeed, despite frequent adjustment of ASM doses and increasing the frequency of administration of IVIG according to the woman's clinical symptoms and ASM levels, her seizures were not fully controlled. She had several seizures in pregnancy and labour, including 2 generalised tonic-clonic seizures (at 19 weeks, and shortly after delivery). Seizure deterioration in pregnancy is associated with worse maternal and fetal outcomes, including an increased risk of preterm birth, caesarean section, small-for-gestational age baby and neurodevelopment delay in the offspring.^7,8

The Royal College of Obstetricians and Gynaecologists does not recommend routine monitoring of maternal ASM levels in pregnancy based on evidence from the Empire study, ⁹ but suggests that individual circumstances should be considered. ¹⁰ Given the complexity of this patient's epilepsy, frequent monitoring of medication levels in this case was appropriate. More frequent monitoring was limited by the 2 weekly turnaround time for obtaining results. The availability of in-house testing may have facilitated earlier dose adjustments and prevented increases in seizure frequency.

Immunomodulatory therapy also played a key role in disease management in this case. IVIG has been established as a relatively safe and effective treatment for several autoimmune disorders in pregnancy, supported by reassuring maternal and fetal safety data from observational studies. ¹¹ Concerns include thrombosis and meningism, particularly with repeated doses. The woman had no other risk factors for venothromboembolism and therefore did not require thromboprophylaxis. More frequent doses of IVIG could potentially have improved her seizure control; her last dose was given at 31 weeks’ gestation. An additional dose prior to labour may have prevented the peripartum seizures that she experienced.

Her weight-based azathioprine dose was not changed in pregnancy. In 2025, the UK Medicines and Healthcare Products Regulatory Authority (MHRA) warned of emerging evidence from case reports of an association between azathioprine and ICP, with cholestatic features typically developing earlier in pregnancy than idiopathic ICP and often refractory to treatment with ursodeoxycholic acid. ¹² Dose reduction or discontinuation of azathioprine has been associated with an improvement in liver function tests. ¹² Her ICP was late in onset, and rather than reduce the dose of azathioprine and risk further destabilising her epilepsy, a decision to expedite delivery was made in keeping with national recommendations for timing of birth with ICP. ¹³

An important area of uncertainty in this case relates to the potential impact of maternal anti-GAD antibodies on the neonate. It is well-established that transplacental transfer of IgG occurs from the second trimester onwards, ¹⁴ but the clinical significance of autoantibodies on the fetus in AAE remains unclear. Whilst certain maternal autoimmune conditions such as connective tissue diseases associated with anti-Ro and anti-La antibodies ¹⁵ and myasthenia gravis ¹⁶ can result in well-described neonatal syndromes attributed to transplacental transfer of maternal antibodies, no similar neonatal effects have yet been reported in association with AAE. Although the neonate in this case was clinically well at birth and at 1 year of age, ongoing neurodevelopmental follow-up has been arranged to ensure early detection of any potential abnormalities that may arise.

Conclusion

This is the first reported case of AAE in pregnancy to our knowledge. It highlights the complexity of managing anti-GAD AAE in pregnancy, and reinforces the importance of proactive, tailored multidisciplinary management in optimising outcomes. Regular patient review and measurement of ASM levels is a critical step in the management.