Neonatal seizures represent a significant health burden on the term and preterm neonatal population and are linked to poor long-term neurodevelopmental outcomes. Currently, there are no US Food and Drug Administration–approved antiepileptic drugs for neonates, and authors of the medical literature have yet to reach a consensus on the most adequate approach to neonatal seizures. Topiramate is readily used in the adult and older pediatric population for the management of migraines and partial-onset seizures. Topiramate continues to gain favor among pediatric neurologists who often recommend this medication as a third-line treatment of neonatal seizures. We report our recent experience with 4 preterm neonates, born between 2015 and 2017, who developed radiographic signs of necrotizing enterocolitis after receiving topiramate for seizures. Each was given oral topiramate for the treatment of electrographic and clinical seizures and developed the subsequent diagnosis of necrotizing enterocolitis, with abdominal distention, hemoccult-positive stools, and radiographic signs of intestinal distention and pneumatosis. More research regarding the risk factors of topiramate use in premature infants is needed.
Case Series
Seizures represent a significant health burden on the term and preterm neonatal population and are strongly associated with poor neurodevelopmental outcomes.1,2 The management of neonatal seizures is riddled with issues pertaining to diagnosis, monitoring, and treatment. The etiology of seizures in premature and term neonates is multifactorial, including perinatal asphyxia, intracranial bleeding, cerebrovascular lesions, central nervous system malformations, infections, metabolic disturbances, and drug withdrawal. At present, there are no medications for the treatment of seizures in preterm or term infants <1 month of age that are approved by the US Food and Drug Administration. The broad spectrum of pathologies and the relative paucity of treatment options make for difficult medical management of neonatal seizures.
Since the early 1980s, phenobarbital has been the most commonly used medication for neonatal seizures. Negative long-term neurodevelopmental outcomes of phenobarbital compared with newer antiepileptic drugs (AEDs) such as levetiracetam have shifted the pharmacotherapy for neonatal seizures.3,4 In cases of uncontrolled seizures with dual AED therapy, the addition of topiramate as a third agent is controversial. Topiramate was first isolated in 1978 and has been approved for the treatment of epilepsy and migraines in the adult population since the late 1990s. Recently, the compound has received US Food and Drug Administration approval as monotherapy for the management of epilepsy in children >10 years of age and as an adjunctive therapy for the treatment of seizures in children >2 years of age. Fifty percent of pediatric neurologists recommend topiramate as a tertiary medication in neonates.5
Despite several animal studies, the evidence and potential consequences of the use of topiramate in neonatal seizures is unknown. The literature has limited information on the neonatal side effects of topiramate and the neurodevelopmental implications of this treatment in neonatal seizures.6 We reviewed 10 cases of preterm infants treated with topiramate for neonatal seizures, 4 of whom developed necrotizing enterocolitis (NEC).
Methods
This is a case series of 10 preterm infants (born before 37 weeks of gestation) who received oral topiramate for neonatal seizures in the level III NICU at Holtz Children’s Hospital between September 2015 and September 2017. The diagnosis of seizures was made by using continuous video electroencephalography. Seizures were monitored electrographically and clinically, and a seizure-free state was considered as achieved after an EEG demonstrated the resolution of seizures. The length and frequency of EEG monitoring was based on the discretion of the pediatric neurologists and by the clinical fragility of the patient. Treatment with topiramate was initiated at the following dosages: 10 mg/kg per day per os on day 1, followed by 5 mg/kg per day on day 2 (with the exception of case 1, in which a patient received 2 mg/kg of topiramate daily).7,8 Topiramate drug levels were not clinically available. The diagnosis of NEC was made on the basis of clinical signs (abdominal distention, bloody stools, etc) and radiographic findings of intestinal distention and pneumatosis intestinalis.
Results
Ten infants received topiramate for the diagnosis of neonatal seizures at our institution between 2015 and 2017. The patients were all born prematurely (23–36 weeks), with birth weights between 440 and 2100 g, and were equally divided between boys and girls (Table 1). There were no profound differences in neonatal comorbidities between the 10 infants.
Clinical Characteristics of the Patients Who Received Topiramate for Neonatal Seizures
| Case | Race and/or Ethnicity | Sex | GA, weeks | Birth Weight, kg | Birth Weight, % | Apgar Score 1 Min | Apgar Score 5 Min | Wt at Topiramate, kg | Wt at Topiramate, Percentile | PCA at Topiramate |
|---|---|---|---|---|---|---|---|---|---|---|
| 1a | African American | Female | 24 | 0.44 | 3 | 3 | 6 | 0.55 | 4 | 27 |
| 2a | Hispanic | Male | 23 | 0.58 | 55 | 2 | 5 | 0.70 | 6 | 28 |
| 3a | African American | Female | 23 | 0.54 | 51 | 2 | 4 | 0.77 | 73 | 25 |
| 4a | White | Female | 36 | 2.21 | 18 | 2 | 3 | 2.24 | 3 | 38 |
| 5 | African American | Male | 32 | 1.76 | 47 | 9 | 9 | 2.88 | 27 | 38 |
| 6 | African American | Male | 25 | 0.61 | 16 | 8 | 8 | 1.82 | 2 | 38 |
| 7 | African American | Female | 25 | 0.89 | 94 | 8 | 8 | 6.50 | 27 | 59 |
| 8 | African American | Male | 25 | 0.77 | 59 | 1 | 4 | 1.24 | 31 | 30 |
| 9 | African American | Male | 27 | 1.19 | 94 | 1 | 1 | 1.21 | 59 | 29 |
| 10 | African American | Female | 24 | 0.72 | 87 | 7 | 9 | 0.72 | 32 | 26 |
| Case | Race and/or Ethnicity | Sex | GA, weeks | Birth Weight, kg | Birth Weight, % | Apgar Score 1 Min | Apgar Score 5 Min | Wt at Topiramate, kg | Wt at Topiramate, Percentile | PCA at Topiramate |
|---|---|---|---|---|---|---|---|---|---|---|
| 1a | African American | Female | 24 | 0.44 | 3 | 3 | 6 | 0.55 | 4 | 27 |
| 2a | Hispanic | Male | 23 | 0.58 | 55 | 2 | 5 | 0.70 | 6 | 28 |
| 3a | African American | Female | 23 | 0.54 | 51 | 2 | 4 | 0.77 | 73 | 25 |
| 4a | White | Female | 36 | 2.21 | 18 | 2 | 3 | 2.24 | 3 | 38 |
| 5 | African American | Male | 32 | 1.76 | 47 | 9 | 9 | 2.88 | 27 | 38 |
| 6 | African American | Male | 25 | 0.61 | 16 | 8 | 8 | 1.82 | 2 | 38 |
| 7 | African American | Female | 25 | 0.89 | 94 | 8 | 8 | 6.50 | 27 | 59 |
| 8 | African American | Male | 25 | 0.77 | 59 | 1 | 4 | 1.24 | 31 | 30 |
| 9 | African American | Male | 27 | 1.19 | 94 | 1 | 1 | 1.21 | 59 | 29 |
| 10 | African American | Female | 24 | 0.72 | 87 | 7 | 9 | 0.72 | 32 | 26 |
GA, gestational age.
Patients who displayed radiographic signs of NEC after topiramate administration.
Case 1
A girl 22 days old, 24 weeks’ postconceptual age (PCA), was born via cesarean delivery with Apgar scores of 3 and 6 at 1 and 5 minutes. She developed respiratory distress syndrome and a dependence on mechanical ventilation. The patient had a large patent ductus arteriosus (PDA) that remained untreated and systemic hypotension with chronic vasopressor dependence. An abdominal abscess was diagnosed at a PCA of 25 weeks, requiring Penrose drainage. The patient had no intraventricular hemorrhage (IVH) but was diagnosed with seizure disorder at 25 weeks’ PCA. EEG showed frequent epileptiform discharges originating from the central brain regions and seizure activity. She was started on levetiracetam with no improvements in seizure activity on EEG; phenobarbital and fosphenytoin were then added. When the patient then failed to have seizure control after a pyridoxine challenge, topiramate was started at 27 weeks’ PCA. At that time, the patient was tolerating 40 mL/kg per day of expressed breast milk (EBM) and weighed 550 g. After 2 days on topiramate, she developed abdominal distention, bloody stools, leukocytosis of 33 000/μL, acidosis (pH of 7.15), C-reactive protein elevation to 12.9 mg/dL, and radiographic evidence of NEC (Fig 1). Despite this strong temporal association between the administration of topiramate and the infant’s diagnosis of NEC, it is important to entertain the possibility of a recurrence of the patient’s previous intra-abdominal infection.
Anterior posterior (A) and left lateral decubitus (B) views of case 1 taken after two days of topiramate administration. This film shows asymmetric, distended, and featureless loops of bowel; air-fluid levels are seen on the lateral radiograph (B). These films are suggestive of NEC.
Anterior posterior (A) and left lateral decubitus (B) views of case 1 taken after two days of topiramate administration. This film shows asymmetric, distended, and featureless loops of bowel; air-fluid levels are seen on the lateral radiograph (B). These films are suggestive of NEC.
Case 2
A 23-week premature neonate was born after clinical chorioamnionitis, with Apgar scores of 2 and 5 at 1 and 5 minutes. Antibiotics were administered for 48 hours for empirical early-onset sepsis coverage with negative cultures. Hospital course was complicated by IVH grade 3, septicemia, Gram-positive cocci in both eyes, and respiratory failure with mechanical ventilation dependence. At 4 weeks of age, an EEG was obtained secondary to jerky movements of the arms and legs and showed electrographic correlation with associated multifocal sharp waves. Levetiracetam was started, followed by phenobarbital because of inadequate clinical and electrographic response. Because of further seizure activity and respiratory depression, phenobarbital was switched to fosphenytoin. When the EEG continued to show frequent seizures, he was loaded with topiramate at 27 weeks’ PCA. At that time, the infant was tolerating 100 mL/kg per day of EBM and weighed 700 g. The following day, the infant was found to have abdominal distention, and NEC was confirmed on abdominal radiograph (Fig 2).
Abdominal radiograph, (A) anterior posterior view and (B) left lateral decubitus, 24 hours after the administration of one dose of topiramate showing fixed asymmetric moderately dilated loops of bowel concerning for developing NEC.
Abdominal radiograph, (A) anterior posterior view and (B) left lateral decubitus, 24 hours after the administration of one dose of topiramate showing fixed asymmetric moderately dilated loops of bowel concerning for developing NEC.
Case 3
A 4-week-old infant was born at 22 6/7 weeks of gestation, with an early hospital course remarkable for sepsis, PDA, and bilateral grade 4 IVH. At 27 weeks’ PCA, the infant developed multifocal myoclonic status epilepticus and was loaded with levetiracetam and phenobarbital. Continuous video EEG monitoring demonstrated improvement of myoclonic seizures after several loading doses of phenobarbital and the initiation of topiramate. At that time, the infant was tolerating 110 mL/kg per day of EBM and weighed 770 g. Seven days later, this infant developed abdominal distension with NEC changes on abdominal radiograph, necessitating small bowel resection with a right hemicolectomy and ileostomy (Fig 3).
Abdominal radiograph, (A) anterior posterior view and (B) left lateral decubitus, 24 hours after the discontinuation of 12 days of topiramate showing moderate to severely distended and elongated bowel loops with pneumatosis, diagnostic of NEC. LLD, left lateral decubitus.
Abdominal radiograph, (A) anterior posterior view and (B) left lateral decubitus, 24 hours after the discontinuation of 12 days of topiramate showing moderate to severely distended and elongated bowel loops with pneumatosis, diagnostic of NEC. LLD, left lateral decubitus.
Case 4
A premature neonate was born with intrauterine growth restriction at 36 weeks, initially requiring mechanical ventilation and was found by echocardiography to have a PDA with mitral and tricuspid regurgitation. The patient had multiple congenital malformations, including hypotonia, microcephaly with hypoplastic vermis, and thinning of the corpus callosum, congenital cataracts, dysmorphic facies, and pancytopenia. A microarray revealed some regions of homozygosity, without any other findings of clinical relevance.
At 39 weeks’ PCA, the patient developed seizures, and levetiracetam was started. Despite the addition and increase in fosphenytoin dosing, seizure activity continued, and topiramate was begun at 40 weeks’ PCA, at which time the infant was tolerating 160 mL/kg per day of EBM and weighed 2240 g. Six days into treatment, the infant developed bloody stools and abdominal radiograph revealed diffuse colonic distention and pneumatosis, suggestive of NEC. The infant was changed to nothing by mouth and completed 10 days of antibiotics, during which oral topiramate was held and seizures were managed with levetiracetam and phenobarbital. Feeds were restarted, and once full feeds were reached, topiramate (43 weeks’ PCA) was restarted. The infant was advanced to full oral feeds, but 6 days later, developed diffuse colonic distension and pneumatosis suggestive of recurrent NEC (Fig 4).
A, Abdominal radiograph on the 7th day of this infant's first course of topiramate showing mottled lucencies and pneumatosis, consistent with NEC. B, Abdominal radiograph on the 4th day of this infant's second course of topiramate, showing pneumatosis in the ascending rectosigmoid colon.
A, Abdominal radiograph on the 7th day of this infant's first course of topiramate showing mottled lucencies and pneumatosis, consistent with NEC. B, Abdominal radiograph on the 4th day of this infant's second course of topiramate, showing pneumatosis in the ascending rectosigmoid colon.
Discussion
We reviewed the cases of 10 premature infants, 40% of which, after a short time interval of treatment with topiramate for seizures, developed NEC. This contrasts with our institution’s baseline rate of NEC of 5.8% for 2015–2016. The temporal association between the usage of topiramate and NEC has not been reported previously. Glass et al7 published a case series of 6 term neonates who received topiramate for neonatal seizures; none of them developed similar gastrointestinal findings observed in our cases. In reports on the use of topiramate in older children and adolescents, authors describe adverse side effects such as anorexia, vomiting, and flu-like symptoms.9,10
Gastrointestinal side effects of topiramate in adults have been widely reported. In a single retrospective study from Spain, authors found a significant association between topiramate therapy and microscopic colitis.11 Other reported adverse events include weight loss, anorexia, nausea, vomiting, and gastroenteritis. Such side effects were proposed to relate to the structure of topiramate, an intermediate molecule in the synthesis of fructose-1,6-diphosphate (FDP). Because topiramate is a precursor of FDP, it is not unreasonable to assume that topiramate, like FDP, could increase anaerobic carbohydrate metabolism and promote the formation of lactic acid within the gastrointestinal tract. In addition, the molecular resemblance of topiramate to acetazolamide could hypothetically explain the presence of metabolic acidosis in adults, secondary to renal bicarbonate losses.12,13 Interestingly, no infant in this cohort developed significant metabolic acidosis while on topiramate.
Unfortunately, data on neonates are not provided in the published safety profile of topiramate. This case series represents the only report of topiramate use in premature infants for the treatment of seizures and is the first study in which the clinical and radiographic findings of NEC after the administration of oral topiramate in preterm neonates is reported. Yet, in light of the patients’ multiple risk factors for NEC, topiramate may only be associated but not causal for the development of NEC.
Greater clinical data collection and analysis of infants treated with topiramate are warranted given topiramate’s increased popularity and use in the NICU. From this case series, hypothesis-driven, randomized controlled trials in which researchers look at the safety and efficacy of topiramate relative to other AEDs in the management of neonatal seizures is needed.
- AED
antiepileptic drug
- EBM
expressed breast milk
- FDP
fructose-1,6-diphosphate
- IVH
intraventricular hemorrhage
- NEC
necrotizing enterocolitis
- PCA
postconceptual age
- PDA
patent ductus arteriosus
Drs Courchia and Kurtom drafted the initial manuscript; Drs Del-Moral, Pensirikul, and Buch were involved in the care of the infants described in this report and were involved in reviewing and revising the submitted manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
Acknowledgment
We thank Ilene Sosenko, MD, for her helpful suggestions and review of the article.
Comments
RE: Topiramate in preterm infants and NEC
Courchia et al. describe ten preterm infants (born at 22-36 weeks gestation) given topiramate for seizures at postconceptual (PCA) ages of 26-59 weeks.1 Four of these preterm infants (born at 22,23,24, and 36 weeks PCA and given topiramate at 27, 27, 27, and 40 weeks PCA), went on to develop necrotizing enterocolitis (NEC). The authors propose there may be an association between usage of topiramate in preterm infants and development of NEC.
Outcomes reported by Courchia et al contradict our institutional experience with development of NEC after topiramate treatment of neonatal seizures. We obtained IRB approval (study # 2018-4498) to review outcomes of neonates/infants treated with topiramate for seizures. From 2011-2018 in our tertiary hospital’s neonatal intensive care unit (NICU) we treated 35 infants with topiramate for seizures. Topiramate was compounded by dissolving drug powder in 1:1 mixture of Ora-Sweet® and Ora-Plus® yielding a solution of 1.698 mOsm/gm osmolarity.2 Twelve of these 35 infants were born prematurely at 26-36 weeks postmenstrual age (PMA). Of these 35 infants, 2 developed NEC, however both infants developed NEC before they ever received topiramate. The first infant with NEC was a male born premature at 27 weeks PMA due to maternal sepsis. He suffered from respiratory failure, unilateral grade IV intraventricular hemorrhage, ventriculomegaly, and developed NEC at 32 weeks PMA. At 33 weeks PMA he began to seize and at 34 weeks PMA topiramate was added to his seizure medications. His family withdrew care due to the severity of his illness and he passed away at 35 weeks PMA. The second infant with NEC was a male born premature at 26 weeks PMA. He had bilateral grade IV intraventricular hemorrhage, ventriculomegaly, respiratory failure, anemia, thrombocytopenia, esophageal perforation, pneumothorax, and developed NEC at 29 weeks PMA. His seizures began at 34 weeks PMA, and topiramate was started at 36 weeks PMA. There was no recurrence of NEC in either of these patients after topiramate was started.
On average our 35 patients had been treated with 3.4 seizure medications prior to starting topiramate, though number of prior seizure medications ranged from zero to seven. We note that topiramate for our preterm patients was started at 34-67 weeks PMA, whereas Courchia et al reported starting topiramate at as early as 26 weeks PCA. Three of Courchia et al’s four premature infants with NEC after starting topiramate had started it at 27 weeks PCA. Of note, American Academy of Pediatrics policy3 states that postconceptual age (PCA) is an improper/inaccurate term, which is how Courchia’s manuscript reports ages. Correcting to the recommended terminology of postmenstrual age (PMA) can be accomplished by adding 2 weeks to the PCA. Our data are reported as PMA.
Stronger evidence is required before any association between topiramate treatment for neonatal seizures and NEC can be confirmed. Our institution’s data (35 neonates, 12 of whom were preterm) did not show any who developed NEC after starting topiramate. We agree with Courchia et al that more studies are needed.
References:
1)Courchia B, Kurtom W, Pensirikul A, Del-Moral T, Buch M. Topiramate for Seizures in Preterm Infants and the Development of Necrotizing Enterocolitis Pediatrics Jul 2018, 142 (1) e20173971; DOI: 10.1542/peds.2017-3971
2)Nahata MC, Pai VB, and Hipple TF. Pediatric Drug Formulations, 5th ed., Cincinnati, OH: Harvey Whitney Books Co., 2004 and Potency & Stability Testing with ARL.
3)Policy Statement: Age Terminology During the Perinatal Period. Pediatrics. 2004;114(5):1362–1364. Reaffirmed July 2014
RE: Putative relationship between topiramate and NEC
Sirs:
You present 4 cases of NEC following Topiramate administration. Your discussion postulates “Such side effects were proposed to relate to the structure of topiramate, an intermediate molecule in the synthesis of fructose-1,6-diphosphate (FDP). Because topiramate is a precursor of FDP, it is not unreasonable to assume that topiramate, like FDP, could increase anaerobic carbohydrate metabolism and promote the formation of lactic acid within the gastrointestinal tract.”
There are alternate possibilities. You gave the medication by mouth; however an oral solution/suspension does not seem to be readily available. The Web site https://www.uspharmacist.com/article/topiramate-20-mg-ml-oral-suspension reports that there is no liquid preparation. Rosemont Pharmacy makes a special preparation of topiramate, (https://www.rosemontpharma.com/products/central-nervous-system/topiramat...) but restricts access to the information. Dr. Katherine White and I hypothesized that hyperosmolar preparations of oral medications would predispose to NEC. (White, K. and Harkavy, K.L.: Hypertonic formula resulting from added oral medication, Amer. J. Dis. Child. 136: 931-933, 1982.) More recently, Akinmboni et al suggested that excipients are the etiology of NEC. (T O Akinmboni, et al, Excipient exposure in very low birth weight preterm neonates, Journal of Perinatology volume 38, pages 169–174 (2018).)
Would you please describe the solution you used to administer topiramate and the osmolality or the excipient content of the solution? If you do not have the osmolality, the percent excipients can be used to estimate the osmolality of the solution and of any mix with formula.
Thank you
Kenneth Harkavy, MD, MBA, FAAP
RE: NEC and oral topiramate in preterm infants
This recent article by Courchia, et al. describes a cohort of preterm infants that developed necrotizing enterocolitis (NEC) after receiving enteral topiramate for seizures. The authors describe ten preterm infants who received oral topiramate in their Neonatal Intensive Care Unit (NICU), four of which developed NEC. We are surprised that the authors did not describe the method by which the topiramate was compounded for this vulnerable population, as it is not commercially available in an oral liquid formulation. This detail is of upmost importance in this case, as hyperosmolar medications administered to the gastrointestinal tract of preterm neonates have been previously associated with development of NEC 1-4.
The four infants described in the case series have significant underlying risk factors for NEC. Three of the infants were extremely preterm with significant comorbidities and risk factors, such as a hemodynamically significant patent ductus arteriosus (PDA), hypotension/vasopressor use, sepsis, and intraabdominal infections. The fourth infant was significantly growth restricted with a hemodynamically significant PDA, both risk factors for NEC. As the article pointed out, a previous case series of six term infants treated with topiramate published by Glass, et al. 5did not describe this association. However, the infants in that series were all full-term, putting them at a much lower baseline risk of NEC.
Caution should be exercised in interpreting the association between topiramate and development of NEC in critically-ill preterm neonates, as key details regarding compounding and preparation for administration are not discussed in this article. The vehicles used in the preparation of these compounds may vary between institutions and osmolality of the compound must be taken into account for any medication being delivered to this high-risk population.
Sincerely,
Jennifer Burnsed, M.D., M.S. Assistant Professor of Pediatrics and Neurology, Division of Neonatology, University of Virginia, Charlottesville, Virginia
Kelly Jane Lunsford, PharmD, BCPPS. Clinical Pharmacist in Pediatrics, University of Virginia, Charlottesville, Virginia and Clinical Assistant Professor, Virginia Commonwealth University School of Pharmacy, Richmond, Virginia
Howard P. Goodkin, M.D., PhD. Professor and Chair of Neurology and Director, Division of Pediatric Neurology, University of Virginia, Charlottesville, Virginia
References
1. Chandran S, Chua MC, Lin W, Min Wong J, Saffari SE, Rajadurai VS. Medications that increase osmolality and compromise the safety of enteral feeding in preterm infants. Neonatology. 2017;111(4):309-316. doi: 10.1159/000454667 [doi].
2. White KC, Harkavy KL. Hypertonic formula resulting from added oral medications. Am J Dis Child. 1982;136(10):931-933.
3. Radmacher PG, Adamkin MD, Lewis ST, Adamkin DH. Milk as a vehicle for oral medications: Hidden osmoles. J Perinatol. 2012;32(3):227-229. doi: 10.1038/jp.2011.83 [doi].
4. Srinivasan L, Bokiniec R, King C, Weaver G, Edwards AD. Increased osmolality of breast milk with therapeutic additives. Arch Dis Child Fetal Neonatal Ed. 2004;89(6):F514-7. doi: 89/6/F514 [pii].
5. Glass HC, Poulin C, Shevell MI. Topiramate for the treatment of neonatal seizures. Pediatr Neurol. 2011;44(6):439-442. doi: 10.1016/j.pediatrneurol.2011.01.006 [doi].