Abstract
The pregnancy of a 38-year-old woman with Sturge-Weber syndrome and epilepsy is described here, with safe outcome for mother and baby despite considerable controversy about peripartum care. Literature review reveals seven case reports of pregnancy in women with Sturge-Weber syndrome and there is little to guide clinicians in the management of these complex cases. A care pathway for women with Sturge-Weber syndrome that are planning pregnancy or are pregnant is proposed.
Introduction
Sturge-Weber syndrome (SWS) is a rare congenital neurocutaneous condition occurring in approximately 1 in 20,000 to 50,000 births. It is characterised by a facial port wine stain in the ophthalmic branch of the trigeminal nerve in association with cerebral vascular abnormalities and occasional involvement of the eye, with choroidal haemangioma and glaucoma. Only seven women with SWS in pregnancy are reported in the English literature and as a result there is little to guide management. Described here is a pregnancy in a woman with SWS and epilepsy. Information to guide management is limited and has led us to propose a preconception and pregnancy care plan for these women that may be of benefit to obstetricians, neurologists and anaesthetists.
Case
A 38-year-old woman was diagnosed with SWS at birth when she presented with the classical facial port wine stain and neonatal seizures. Imaging confirmed the cerebrovascular abnormality and recent images are shown in Figure 1 demonstrating the typical calcification of abnormal capillaries on MRI and CT. She had frontal lobe epilepsy characterised by focal onset seizures with impaired awareness, becoming bilateral tonic-clonic events. She had three distinct seizure events: (A) focal onset with impaired awareness, characterised by speech and behavioural arrest lasting less than 1 min every one to two days, (B) self-limiting bilateral tonic-clonic seizures lasting for less than 3 min, and (C) Focal motor onset atonic seizures or ‘drop attacks’ lasting seconds, occurring several times per year. Epilepsy control had been suboptimal with anti-epileptic drugs (AEDs) but improved with the insertion of a vagal nerve stimulator (VNS) nine years previously. Her medical history included low Von Willebrand factor, asthma and depression. Her father had diabetes, an unprovoked DVT and was heterozygous for Factor V Leiden. She was not a carrier for this mutation.
Cerebral imaging of index case. Panels A and B FLAIR and FSE T2-weighted MRI from 2009; showing susceptibility artifact from calcification of cortical venous capillaries (arrows). Panels C and D CT brain 3 mm cuts showing calcification of venous capillaries in brain and bone windowed slices (arrows).
She booked for antenatal care with an unplanned pregnancy at 14 weeks’ gestation. Medications at conception included lamotrigine 250 mg twice daily, zonisamide 250 mg twice daily, folic acid 5 mg daily and buccal midazolam 10 mg when required. She worked as a receptionist and reported daily focal unaware seizures as described above with her last convulsion being more than three years previously. Antenatal care was shared with her neurologist and lamotrigine levels were monitored each trimester. She received education about epilepsy and pregnancy and information about feeding and caring for the newborn. Fetal anomaly and echocardiography scans were normal. Polyhydramnios and gestational diabetes was diagnosed at 29 weeks of gestation and treated with diet ary modification and insulin. Further admissions with threatened preterm labour occurred at 31 and 36 weeks and a tonic-clonic seizure was witnessed during the latter admission. Lamotrigine was increased to 275 mg twice daily and clobazam 10 mg three times a day was commenced. The dose of clobazam is 30% higher than the usual rescue regimen for escalating seizures, but there was strong concern, based on prior history, that her seizures may escalate to status epilepticus. The decision was made to keep her in hospital and induce labour at 37 weeks of gestation, because of this convulsive seizure.
The peripartum care plan was discussed at a Maternal Medicine MDT. Haematology confirmed normal VWF levels in the third trimester and no haematological contraindication to regional anaesthesia. Six weeks postnatal, LMWH was advised because of the family history of thrombosis. There was variation in opinion amongst anaesthetists about the safety of regional anaesthesia because of the theoretical risk of spinal vascular abnormalities. The consensus opinion from the perinatal anaesthetic department was that regional anaesthesia should be avoided if possible. Difficult intubation was not anticipated. Concern was raised about a risk of seizures, intracranial bleeding and the need for general anaesthetic administration should an acute emergency occur in labour. The consensus at that meeting was that this woman should have an elective caesarean section under general anaesthesia.
The VNS was deactivated and caesarean section was performed under general anaesthesia, resulting in the delivery of a healthy baby weighing 2.9 kg. The procedure was uncomplicated but frequent focal unaware seizures occurred in recovery followed by short tonic-clonic movements. She was given clobazam, fentanyl and clonidine sequentially and was transferred to the high dependency unit for monitoring. Patient controlled analgesia was initiated. Regular medications were recommenced and subcutaneous tinzaparin was prescribed. The VNS was reactivated the following morning but there were recurrent unaware seizures over the next 24 h that responded to activation of the VNS. She remained well thereafter and breast-fed her baby.
Discussion
This case was particularly challenging for a number of reasons. There is little information in the literature to guide decision-making about anaesthesia and mode of delivery in SWS. There was a background of refractory epilepsy, requiring multiple AEDs and VNS, with some deterioration in epilepsy close to term. Pregnancy increases the risk of death 10-fold in women with epilepsy, mainly as a result of sudden unexpected death in epilepsy. 1 Contributing factors include worsening of seizure control in pregnancy, medication adherence, altered drug bioavailability, sleep deprivation, exhaustion and psychological stress.
AEDs increase the risk of fetal anomalies. Some studies have suggested an increased risk of 3.5% when lamotrigine dose is greater than 400 mg daily, although the evidence for this is sparse. 2 There is less information about zonisamide but it may be associated with lower infant birthweight. 3 Epilepsy appears to be associated with a small increase in fetal growth restriction, induction of labour, postpartum haemorrhage and admission of the neonate to the neonatal intensive care unit. 4
VNS is a palliative stimulation technique that can be considered when seizures are refractory to AEDs and surgical resection of the epileptogenic focus is not an option. Stimulating leads are placed around the left vagus nerve in the carotid sheath and connected to an infraclavicular subcutaneous programmable pacemaker. The device is set to come on and off in a programmed way and the duty cycle in this woman was 36%. The newer devices have an auto-stimulation trigger linked to a tachycardia, the most common seizure-related arrhythmia. Stimulation can be activated also on demand by the woman using an external magnet that is set to turn the device on at a higher out for one cycle. There is reassuring data about the use of VNS in pregnancy but insufficient numbers yet to comment on maternal and fetal safety. 5
This woman had known SWS with an underlying cerebrovascular arteriovenous malformation that was seen on MRI with characteristic calcification of the abnormal venous capillary formations. The imaging did not show the typical large draining vein (Figure 1).
SWS is caused by a somatic mutation in the GNAQ gene. The gene product is a protein called GAQ that has a role in cell function including the regulation of blood vessels. The cerebrovascular abnormalities that result are capillary-venous but can result in hypoxia, ischaemia, thrombosis, venous occlusion, infarction and calcification of affected areas of the brain. Intracranial calcification can be detected on X-ray or CT of brain. Magnetic resonance imaging (MRI) with gadolinium is the optimal way to assess vascular change using susceptibility-weighted imaging (SWI) as this is good for the detection of venous abnormalities. Other MRI modalities can raise concerns about AVM when there is prominent venous malformation alone. 6 Single-photon emission CT scanning (SPECT) is useful in conjunction with other imaging.
The risk of subarachnoid or intracranial haemorrhage is thought to be low because the vascular abnormalities are capillary venous, but note that some authors 7 advise caution to avoid any increase in i pressure when anaesthetising patients with SWS because of the risk of intracranial haemorrhage (ICH). The incidence of ICH related to hypertension is simply unknown, particularly in pregnancy. Dolkart and Bhat reported a case of epilepsy recurring in a woman with SWS antenatally and postnatally with normal non-contrast MRI, but a small left intraventricular haemorrhage on CT brain. 8 In another case report, 9 a woman with SWS presented with stroke before delivery and imaging showed a small cavernous angioma in the right occipitoparietal region. Chabriat et al. 10 report a case of severe hemiplegia, hemianopia and aphasia in the third trimester. Imaging did not show a bleed but left hemispheric pial angiomatosis. PET scan 10 weeks later showed focal oligemia and there was gradual recovery over three months with some residual dysphasia. Stroke-like episodes or ‘transient episodes’ are documented to occur in SWS but the underlying mechanism is unclear. Hypotheses include: depression of postictal activity, or ischaemia secondary to cerebral hypoperfusion due to impaired venous drainage in the affected area. 11 Of interest, these episodes are associated with full recovery, and in a relatively short time period (under 24 h), but some cases can take months. These data are derived from a retrospective cohort of children and young adults and does not provide information on the rate and recovery of these events in adults with SWS. The influence of the hormonal and cardiovascular effects of pregnancy on the neurological manifestations of SWS is unknown. There is a suggestion that low-dose aspirin may reduce the frequency of stroke-like events in SWS but this hypothesis has not been tested prospectively. 12 It would not be unreasonable to prescribe low-dose aspirin in pregnancy in women with SWS because of potential benefit and known safety profile.
The decision about the safest anaesthetic technique to use for delivery in women with SWS is challenging. With general anaesthesia, one must consider the possible increases in intracranial and intraocular pressure during airway manoeuvres with resultant risk of rupture of intracranial angiomas. 7 The presence of facial and airway haemangiomas may make ventilation and intubation challenging and may predispose to obstructive sleep apnoea. Finally, identification of seizures in those with poorly-controlled epilepsy is more challenging under general anaesthesia, while careful consideration needs to be given to avoiding any medications that may precipitate seizures. 13
The principal concerns in considering regional anaesthesia include the possible existence of spinal haemangioma in women with overlap syndromes, e.g. Klippel-Trenaunay syndrome (KTS) or Parkes Weber syndrome (PWS). Regional anaesthesia has been performed successfully in pregnancy9,14 and is the preferred technique by many. 15 Epidural anaesthesia also has benefits for labouring parturients in avoiding increases in intracranial pressure and allowing for an assisted second stage of labour if required. 15 If regional anaesthesia is planned, neurology assessment should clearly indicate that the diagnosis is SWS and that there is no clinical suspicion of an overlap syndrome. The neurological advice in this case was that while KTS and PWS have capillary abnormalities, the occurrence of spinal involvement is not the same as the brain involvement in SWS. Similarly, there was no skin involvement outside of the trigeminal distribution and therefore spinal imaging was not required. If an overlap syndrome is suspected, then spinal imaging should be carried out, ideally in advance of pregnancy. 16
Whichever technique is chosen, consideration should be given to thorough preoperative airway assessment, identification of coexisting cardiac pathology and optimisation of AEDs prior to surgery. Intraoperatively normotension should be maintained with vigilant intraoperative monitoring and the minimisation of manoeuvres that may cause a rise in intracranial or intraocular pressures.
The optimal mode of delivery for women with SWS is unclear and should be individualised following multidisciplinary discussion. The nature, and extent, of cerebrovascular lesions, neurological symptoms and seizure control should be considered. Ocular involvement can include retinal detachment, retinal vessel varicosities and glaucoma and mandate avoidance of hypertension in pregnancy and minimisation of active pushing in the second stage of labour. Cardiac abnormalities are rare in SWS but septal defects, valvular lesions and transposition of the great vessels have been reported, and should be ruled out. 15 In case reports of SWS in pregnancy, the predominant mode of delivery (when the information is given) is by CS9,10,17–19 often prompted by neurological deterioration or concerns, but influenced by maternal request in one case. 18 In one woman who had a small intraventricular haemorrhage four days post-delivery, labour had been induced, followed by a CS at full dilatation. 8 Another woman with SWS is reported to have two vaginal deliveries and uncomplicated epidurals, but she had a hemispherectomy performed as a child with no residual leptomeningeal angiomatosis (with non-contrast imaging). 14
Care pathway for women with Sturge-Weber syndrome.
AEDs Antiepileptic drugs; MRI Magnetic Resonance Imaging; SWI Susceptibility Weighed Imaging, SPECT Single photon emission CT; ECG Electrocardiogram; MDT Multidisciplinary Team.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical approval
Coombe Women and Infants University Hospital.
Informed consent
Written informed consent was obtained from the patient for their anonymised information to be published in the article.
Guarantor
BB.
Contributorship
BB was involved in reviewing the literature, drafting and writing the manuscript. RFOC and CD proof read the manuscript and contributed their expertise to the discussion and proposed care plan.
Acknowledgement
We wish to acknowledge the willingness of the patient to allow publication of her case report.
