Mood Stabilizers and Pregnancy
From MindLinc Wiki
Question: What are the issues in treating patients with bipolar disorder in pregnancy? Specifically, what are the risks involved in continuing mood stabilization with Depakote or Tegretol during pregnancy?
- Finnerty et al (1996) report on a woman who was hospitalized at 26 weeks of pregnancy, 4 weeks after she and her psychiatrist decided in the outpatient setting to d/c her Depakote because she informed him that she was 5 months pregnant. The patient immediately de compensated, becoming hostile at home and ceasing to control her diabetes. She was admitted to the hospital for stabilization and was tried on haldol and benzodiazepines, which even at high doses (40 haldol, 14 ativan) was not effective. She eventually underwent ECT; however, at 29th week of pregnancy, on ECT, she experienced premature rupture of the membranes and required C section. Post-partum, she deteriorated and required psychiatric hospitalization and stabilization—tried on VPA, ativan and haldol, finally required ECT.
The article brings up a number of interesting issues about the problems inherent in bipolar disorder and pregnancy:
- Patients need to be informed of the risk of possible problems in pregnancy in the presence of mood stabilizing medications;
- A manic patient is not exercising good judgment and is more likely to have an unplanned pregnancy;
- The same patient is also likely not to receive good prenatal care;
- Medical illness in the same patient will not be well controlled, which can cause problems for the developing fetus (e.g., diabetes);
- Most exposure to medications/drugs happens early in the pregnancy, before the physician is aware of pregnancy;
- Bioavailability of medications during pregnancy is altered and drug levels may be irrelevant due to different ratios of protein binding and altered distribution;
- The post-partum period is a high-risk period for relapse into depression or mania;
- ECT is thought by psychiatrists to be safe in pregnancy, although there are some reports of premature rupture of the membranes in ECT which are questionably attributable to ECT;
- The legalities of committing a woman based solely on the potential for harm to her fetus are questionable (most states do not allow commitment for this reason).
- Eller et al. (1997) examined the implications of anticonvulsant therapy during pregnancy by conducting a review of the literature available through ~1996. Since the 1980’s, there have been reports in the literature about specific patterns of malformations in children born to women receiving depakote (VPA) and tegretol (CBZ).
Notably, all the women studied in the literature for these data have taken these medications for the treatment of seizure disorder, in which there is a recognized increased risk of birth defects in uncontrolled seizure disorder; there is also the question of whether or not any of these abnormalities were inherited rather than acquired because of the medications.
Recommendations have been to treat women who take VPA or CBZ as if they have had children with neural tube defects in the past—e.g., folic acid 4 mg PO qD (usual prenatal dose 400 mcg). There is some possibility, however, that this dose of folic acid might induce CP450 enzymes and alter metabolism of the medications, requiring close monitoring. Further, there is no good evidence that this intervention would work (it is extrapolated from other, widely known data about the efficacy of folic acid in preventing neural tube defects). There are a number of theories about why VPA and CBZ are associated with neural tube defects, which seem to boil down to glutathione and SAM metabolism (animal studies).
There are also recommendations that women taking VPA or CBZ should take vitamin K 10 mg PO qD for the last 1-2 months of pregnancy, and the infant should receive a 1 mg injection of vitamin K at birth in order to address the increased risk of intracranial hemorrhage.
- Finn et al. (2000) conducted a prospective study of neonates born to mothers who had been taking valproic acid with the goal of determining the incidence of withdrawal symptoms and hypoglycemia.
Objective: To define the risk of hypoglycemia in term infants exposed to valproic acid, and to describe the withdrawal symptoms.
Design: Prospective study in Denmark on a case-control basis.
Cases: 22 infants selected from 24,500 infants born in the North Jutland County in Denmark between 7/93 and 2/97 whose mothers were taking valproate during pregnancy. Exclusion criteria were preterm (<37 weeks), or the second twin. 19 of the mothers had primary generalized epilepsy, and three had complex partial epilepsy. 20 were taking VPA alone and 2 were taking VPA + CBZ.
Controls: 233 infants born at term to mothers not taking VPA or having epilepsy during the same time period provided one blood sample each (at either 1,2,4,6,12,24,48,72 or 96 hours post-partum) for glucose measurements; 22 additional infants from this hospital (term, normal pregnancies, no epilepsy in mothers) had their cord blood analyzed for glucose, insulin, proinsulin and C peptide.
Method: Women with epilepsy taking VPA were seen monthly by a neurologist during pregnancy and had blood levels of VPA monitored. All patients had U/S; 11 had amniocentesis, and one had a chorionic villus biopsy. The women either abstained from EtOH or drank small amounts.
Glucose was measured at birth (cord blood), then at 1,2,4,6,12,18,24 hours and then every 8 hours till the 5th day of life.
Insulin, proinsulin, and C peptide were determined, as well as LFT’s.
The infant’s VPA level was determined at days 2 and 4 of life.
Infants were observed with q8-hourly records of withdrawal symptoms (irritability, jitteriness, hypertonia, seizures, and vomiting).
Measures: Hypoglycemia was defined as <1.8mMol/L (32 mg/dL) based on previous studies by this same group.
Results: 14 boys, 8 girls. 13 of 22 infants in the case group had episodes of hypoglycemia. 7 occurred at 1 hour, 3 occurred at 2 hours, 1 occurred at 4 hours, one at 12 hours, and one at 67 hours. All infants developing hypoglycemia were placed on glucose drips and all were asymptomatic. One infant had 8 episodes; one had three, two had two, and nine had one episode of hypoglycemia. At one hour, the case infants had significantly lower average glucose (2.1) than the control infants (2.9), p<0.01. At two hours, there was a trend (p=0.06). A logistical regression showed that the mother’s VPA level in the third trimester was significantly and inversely correlated with infants’ plasma glucose concentrations (p<0.01).
10 of 22 infants in the case group demonstrated w/d symptoms beginning at 12-24 hours and lasting until 2-7 days. There was one infant who had a seizure. One infant had an elevated AST, but had also suffered asphyxia; the rest of them had normal LFT’s. A proportion of the infants required formula supplementation of breast mild because of w/d symptoms.
There were no major malformations among the case infants; four infants did have minor abnormalities, including one infant who met criteria for fetal VPA syndrome.
- Moore et al. (2000) reviewed 57 children with fetal anticonvulsant syndromes (mostly VPA) in order to further understand the clinical and behavioral phenotype of fetal anticonvulsant syndromes.
Cases: 52 children located through a parents’ support group (National Fetal Anticonvulsant Syndrome Association) and 5 others found in a clinic.
Method: Retrospective analysis via medical and family history questionnaire at a clinic.
Measures: Questionnaire; karyotypes (51 cases); DNA testing for Fragile X (48 cases); review of photos by two experts who rated the children on whether or not they had FACS; ophthalmologic exams (46 cases)
Results:
- 46 of the 59 mothers were taking VPA (34 VPA alone, 12 in combination); 4 were taking CBZ, 4 were taking phenytoin. There were 15 women taking more than one medication.
- 11 (19%) had experienced neonatal w/d—4 required NGT, 3 were floppy, 6 were jittery, 2 had seizures (and one of these were found to be hypoglycemic in nature).
- No cleft palates were reported.
- Myopia (16/46), astigmatism (11/46) and strabismus (8/46) were prevalent in this group of children.
- 44 (77%) had developmental delays or learning difficulties; 28 of 38 school-aged children (74%) were in special schools or receiving special education in mainstream schools. 10 of 38 school aged children (26%) were in normal mainstream education; three of these were in their first year. 41 of 53 children older than 2 (77%) needed to have speech therapy. 32 of the children (56%) had gross motor delays (13 severe enough to require OT).
- 46 (81%) had aberrant behavior; a formal diagnosis of behavior d/o was made in 9 (16%) cases. 22 (39%) were rated as hyperactive; 4 as autistic; 2 as having Asperger’s syndrome. Only four children had neither developmental delays nor behavior problems.
Disadvantages: Selection bias! The children included were those whose parents had sought a support group for their abnormalities. Retrospective. Self-report (questionnaires). Possibility that the children’s anomalies were inherited but, because of maternal anticonvulsant use, were being attributed to the medications.
Summary of the findings about the risk of VPA and CBZ in pregnancy from the four references below.
| VPA | CBZ |
|---|---|
| 11.1% risk of malformation | 5.7% risk of malformation |
| Neural tube defects 1-2% (lumbosacral) | Neural tube defects 1% (lumbosacral) |
| Fetal VPA syndrome—brachycephaly, high forehead, shallow orbits, ocular hypertelorism, small nose and muth, low-set, posteriorly rotated ears, long overlapping fingers and toes, hyperconvex fingernails (Eller et al) | Epicanthic folds, infraorbital groove, medial deficiency of eyebrows, flat nasal bridge, short nose with anteverted nares, smooth/shallow philtrum, long thin upper lip, thick lower lip, small and downturned mouth (Moore et al) |
| Cranial facial defects, fingernail hypoplasia, developmental delay (Eller et al) | Epicanthic folds, short nose, long philtrum, upward slanting palpebral fissures, hypoplastic nails (Moore et al) |
| Risk of hemorrhage b/c impaired vitamin K dependent clotting factors | Risk of hemorrhage b/c impaired vitamin K dependent clotting factors |
| Risk of acute liver failure in the infant | Risk of transient neonatal hepatic dysfunction |
| Risk of developmental delay and behavioral problems (Moore et al) | - |
References Eller DP, Patterson CA, Webb GW. Prescribing in pregnancy: Maternal and fetal implications of anticonvulsive therapy during pregnancy. Obstetrics and Gynecology Clinis 24(3): 523-532, September 1997.
Finnerty M, Levin Z, Miller LJ. Acute manic episodes in pregnancy. American Journal of Psychiatry 153(2): 261-263, February 1996
Presented by Jane Gagliardi, MD, 11/27/00
Comments:
