Lithium for Bipolar Disorder in Pregnancy
From MindLinc Wiki
LITHIUM FOR BIPOLAR DISORDER IN PREGNANCY
Bottom line: Both studies seemed to conclude that Lithium may not be a major human teratogen with Zalzstein's showing a maximum risk of Ebstein's as 1:700 live births, and Jacobson's as 1:6000 births. However, both studies may have missed sub clinical cases and other factors such as under powering of study and small sample size should be noted.
Background: Several studies have linked lithium exposure in pregnancy with teratogenicity since the 1970's. The most often association has been with Ebstein's anomaly (tricuspid valve dysplasia with downward displacement of valve into right ventricle. The rate of Ebstein's has been documented to be 1:20,000 live births.
Search terms: Lithium, pregnancy, birth defects, perinatal morbidity, and teratogenicity
Clinical Case: 32 yo aaf brought to the ED with SI and depressive symptoms after a recent manic episode one month prior. Pt has a history of Bipolar Disorder Type I and was discharged in Dec, 2002 on Lithium 300mg tid, zoloft 50mg qd and trazodone 50mg qhs. Pt followed up once at the Durham center after discharge after she found out she was pregnant, and was told not to take lithium, but to continue zoloft and trazodone.
Question: In a pregnant female with Bipolar disorder, does the use of lithium in the first trimester increase the risk of perinatal morbidity?
Methods: Study Sources: MEDLINE 1990-1998, 42 articles found, included 2 most recent studies
| Zalzstein | Jacobson | |
|---|---|---|
| Study Design | Case-control retrospective | Prospective multicenter study |
| Setting | Tertiary referral center of 6 million in Canada | Voluntary retrospective reporting center, 4 teratogen information centers in USA and Canada |
| Patient Population | Inclusion criteria 59 cases of Ebstein's anomaly born between 1971-1988 168 babies with neuroblastoma (all cases btwn Jan 1971-Oct 1986) served as control group 1 mother had Ebstein's anomaly herself In the control group- 2 women had depressive disorder, I manic with Li treatment during pregnancy Dx of Ebstein's made- displacement>8mm/mm^2 of tricuspid septal cusp with 2 dimensional echo, confirmed postnatally Ebstein anomaly diagnosed at mean age of 128 days(I day to 9yrs) Most children had other cardiac anomalies 168 cases of neuroblastoma diagnosed at mean age of 204 days (I day to 3 yrs) 1 case of Intra-uterine lithium exposure in the pairs with neuroblastoma Mean age of mothers 26 Exclusion manic depressive psychosis, use of lithium carbonate before or during pregnancy | Inclusion Criteria 148 women (mean age 30 yrs) SD 15-40, using Li during 1st trimester of pregnancy who consulted 4 teratogen information centers in Canada and USA 148 controls with exposure to non-teratogenic compounds Screened 225 cases reported by 1983 25(11 %) had major congenital malformations 18(72%) had cardiac anomalies-1/3rd had Ebstein's anomaly All pregnant women who called, lithium exposures as early as 3 weeks, serum concentration can still be present 2 weeks after cessation of treatment All pt's were offered fetal echo's at 18 weeks gestation, if pre-natal echo not done, postnatal echo done Controls similar age +/- 2yrs 1 control had VSD 1 had congenital hip dislocation 1 had cerebral palsy and torticollis |
| Exposure being tested | Association between 1st trimester exposure to lithium and Ebstein's anomaly | Pregnancy outcome after lithium exposure (mean daily dose 927 mg) during the 1st trimester |
| Analysis | Rates of exposure- Fischer's exact test Means of values- Student's t test | Chi Square to compare frequency of malformations in study and control groups Risk ratios 95% Cl(Taylor series) stats obtained 1" for all congenital malformations, then for cardiac malformations and Ebstein's Birthweight, gestational age, developmental milestones compared with Sudent's t test Correlation between values- least-square regression analysis |
| Outcomes | None of the children born with Ebstein's were exposed to lithium in utero Rates of Lithium exposure in pregnancy did not differ between the study and control groups 95% Confidence limit (0.04-26.13) Maximum risk of Ebstein's anomaly after exposure to lithium during I" trimester- 28 fold (1:700 live births) Conclusion: The risk of Ebstein's anomaly is much lower after maternal exposure to lithium than was previously thought(500 fold risk) | Outcomes did not differ between pt's and controls as far as # of births, frequency of major anomalies, spontaneous or therapeutic abortions, ectopic pregnancy, and prematurity were concerned l case of Ebstein's anomaly in 137 subjects in the exposed group, and no cases in the 148 unexposed subjects 2 major congenital malformations occurred in each group 2.8% of live births in Li group and 2.4% in controls 10 pregnancies which final outcome not known were included in analysis 4 sets of twins in lithium group Conclusion: Lithium is not a major human teratogen and can be continued during pregnancy |
| Follow-up | 59pt with Ebstein's- 45 survived, 14 died at time of analysis | All of the women were followed through to delivery or termination 10 of the 148 pt were lost to follow-up postnatally; but included in analysis |
Validity
| Pt's and control similar? | No demographic data of control group | Lithium group had more smokers |
| Outcomes and Exposures measured similarly? | No, control group babies-no echo's | No, control group mothers-no echo's done |
| Follow-up long and complete? | No, only observed live births | |
| Temporal relationship? | Yes | Yes |
| Dose-response gradient | Unknown | Unknown |
Results
| Strength of association between exposure and outcome | Cannot say, increased risk of no more than 28 fold(1:700) is compared to previously reported risk of 500 fold | Relative risk for Ebstein's 3.5 1:6000 births compared to spontaneous rate of 1:20,000 Relative risk for all cardiac anomalies in the lithium exposed group 1.1 |
| How precise are the risk estimates for exposure | No data | 95% Cl(0.1-8.5) very wide intervals |
| Absolute level of risk from this exposure | Odds ratio- 0 | Relative Risk-not calculable (denominator is 0). |
Comments: Both articles seem to conclude that the risk of acquiring Ebstein's anomaly in babies from lithium exposure during the 1' trimester may not be as much as previously thought, and that lithium is not a major human teratogen.
