Intended for healthcare professionals

Rapid response to:

Letters

Post-marketing surveillance should include effects during pregnancy

BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7486.308-a (Published 03 February 2005) Cite this as: BMJ 2005;330:308

Rapid Response:

Post-marketing surveillance and pregnancy : the MADRE project

Editor,

In her recent letter (BMJ 2005;330:308, doi:10.1136/bmj.330.7486.308)
Patricia Yeargin rightly stresses the importance of including the effects
during pregnancy in post-marketing surveillance of medications. Physicians
and women can benefit when they know more about the reproductive toxicity
of medications, and, crucially, when the evidence does not support claims
or fears of teratogenicity. This knowledge can help prevent unnecessary
abortions when fears arise about medications taken before pregnancy is
recognized, as well as help use effective medications safely in pregnant
women.

We would like to share with the readers of the journal some findings
of one such ongoing post marketing surveillance that is being carried out
since the late 1980’s by the International Clearinghouse for Birth Defects
Surveillance and Research [ICBDSR previously known as ICBDMS] (1). The
Clearinghouse is a network of registries that exchange information and
track over time the occurrence of major birth defects, including multiple
congenital anomalies, chromosomal conditions, and teratogen-associated
phenotypes.

As part of such collaborative exchange, the Clearinghouse has
implemented the MADRE (MAlformations DRug Exposure) project, which tracks
first trimester use of medications among pregnancies whose outcome was a
birth with major congenital anomalies (2). MADRE uses such data which is
routinely collected by selected registries around the world as a screening
tool for potentially teratogenic medications.

To screen for potential teratogenicity the MADRE data is assessed
with an «exposed case-only » design. The study population consists of
cases (infants or terminations of pregnancy with birth defects) all of
which were exposed to some medication during the first trimester of
pregnancy. Medications are cross-tabulated against birth defects. In the
simplest approach, for each combination of medication and malformation, a
two-by-two table is constructed and evaluated using an odds ratio,
stratified by registry. In this table, a case is a subject (infant or
fetus) if it has the malformation in question and a control otherwise
(really an affected control because it has other defects). A subject is
defined as exposed if the mother used during the first trimester the
medication under study, and unexposed otherwise (although exposed to other
medications). If medications and malformations are not specifically
associated, then they should be randomly distributed in the sample,
otherwise deviations from this distribution are signalled by increased
odds ratios for the corresponding tables. We use such deviations and
corresponding odds ratios to screen for associations that deserve further
study.

The MADRE database now includes over 15,000 cases exposed to first
trimester medications, reported from France, Israel, Italy, Japan,
Northern Netherlands and Latin America. Such international setting and
large sample improves statistical power and provides, to some extent,
internal indicators of consistency. For example, an association is more
likely to be due to factors other than chance if it occurs simultaneously
in different registries. Moreover, because all subjects are affected,
differential maternal recall of medication use is less of a concern
compared to when unaffected controls are utilized. Conducting these
analyses as ongoing surveillance allows also for further confirmation over
time.

As an example of findings, we report selected associations. Of note,
classic associations were confirmed: phenobarbital with oral clefts and
heart defects; valproic acid with spina bifida, hypospadias, and heart
defects; and carbamazepine with heart defects (3). Novel associations were
also detected and are considered as hypotheses to be tested on further
samples. For example, the association between corticosteroids and
orofacial clefts was initially noted in 1994 (2), was later confirmed as a
weak association by three epidemiological studies (4-6), and further
confirmed in 2003 using MADRE data (7).

Another use of MADRE data database is to test associations as they
appear in the literature. For example, we confirmed the suggested
association between valproic acid and craniosynostosis (mainly
trigonocephaly) (8), and were unable to replicate the reported association
between trimethoprim and malformations such as neural tube and cardiac
defects (9). Another example is the putative association of thyroid
therapy and renal anomalies. A case report noted bilateral renal agenesis
in a girl born to a hyperthyroid mother who received methimazole in early
pregnancy (10), and experimental data in rats suggests that pharmacologic
hypothyroidism during a critical window might affect renal development (11
-12). In the MADRE database, we identified a significant association of
thyroid therapy with cystic kidney [23 exposed cases, odds ratio 1.9, 95%
confidence interval 1.2 to 2.9], and with unilateral kidney agenesis [18
exposed cases, odds ratio 2.3, 95% confidence interval 1.4-3.7]. Potential
associations such as these warrant further testing in other samples. These
are only selected examples and other can be found in the Annual Report of
the Clearinghouse (13).

We think that the MADRE approach, which leverages the strength of
existing birth defect registries and is relatively inexpensive, can be a
useful adjunct tool and resource for postmarketing surveillance of the
pregnancy effects of medications, and should be expanded. Moreover,
physicians and researchers who wish to test selected associations between
first trimester use of medications and structural malformations are
welcome to contact us.

Pierpaolo Mastroiacovo1, Elisabeth Robert-Gnansia2, Lorenzo Botto3

1International Centre on Birth Defects, Rome, Italy icbd@icbd.org

2Institut Européen des Génomutations, Lyon, France
elisabeth.robert@ieg.asso.fr

3 Department of Pediatrics, University of Utah School of Medicine,
and Utah Birth Defects Network, Salt Lake City (Utah), United States
lorenzo.botto@hsc.utah.edu

The MADRE project was created after an idea from B Källén (Lund) and
would not be possible without the ongoing effort and collaboration of the
directors and staff of participating birth defects registries, which we
gratefully acknowledge: France Central East (E Robert-Gnansia), France
Paris (C De Vigan), Israel (P Merlob), Italy IMER (G Cocchi), Italy BDRCAM
(G Scarano) Japan JAOG (Y Sumiyoshi), Latin America ECLAMC (E Castilla),
Northern Netherlands (H de Walle).

We thank A Lisi (ICBD, Rome) and M Atkinson and D Gambrell (CDC, USA)
for help in the statistical analysis of data from the MADRE project.

References

1. ICBDMS. Congenital Malformations worldwide. Elsevier 1991, 220pp.

2. Robert E, Vollset SE, Botto L, Lancaster PAL, Merlob P,
Mastroiacovo M, et al. Malformation surveillance and maternal drug
exposure: the MADRE project. Risk & Safety Med 1994:75-118.

3. Arpino C, Brescianini S, Robert E, Castilla EE, Cocchi G, Cornel
MC, et al. Teratogenic effects of antiepileptic drugs: use of an
International Database on Malformations and Drug Exposure (MADRE).
Epilepsia. 2000;41:1436-43.

4. Carmichael SL, Shaw GM. Maternal corticosteroid use and risk of
selected congenital anomalies. Am J Med Genet 1999;86:242-4.

5. Edwards MJ, Agho K, Attia J, Diaz P, Hayes T, Illingworth A, et
al. Case-control study of cleft lip or palate after maternal use of
topical corticosteroids during pregnancy. Am J Med Genet 2003;120A:459-63.

6. Park-Wyllie L, Mazzotta P, Pastuszak A, Moretti ME, Beique L,
Hunnisett L, et al. Birth defects after maternal exposure to
corticosteroids: prospective cohort study and meta-analysis of
epidemiological studies. Teratology 2000;62:385-92.

7. Pradat P, Robert-Gnansia E, Di Tanna GL, Rosano A, Lisi A,
Mastroiacovo P. First trimester exposure to corticosteroids and oral
clefts. Birth Defects Res Part A Clin Mol Teratol 2003;67:968-70.

8. Lajeunie E, Barcik U, Thorne JA, Ghouzzi VE, Bourgeois M, Renier
D. Craniosynostosis and fetal exposure to sodium valproate. J Neurosurg
2001;95:778-82.

9. Hernández-Diaz S, Werler MM, Mitchell AA. Teratogen update :
trimethoprim teratogenicity. Birth Defects Res. Part A 2004(70 (S10)):276
(abstract).

10. Rodriguez-Garcia R. [Bilateral renal agenesis (Potter's syndrome)
in a girl born to a hyperthyroid mother who received methimazole in early
pregnancy] Ginecol Obstet Mex. 1999;67:587-9.

11. Tan JP, Seidler FJ, Schwinn DA, Page SO, Slotkin TA. A critical
period for the role of thyroid hormone in development of renal alpha-
adrenergic receptors. Pediatr Res. 1997;42:93-102.

12. Ali M, Clos J.Ontogenesis of the kidney in the congenital
hypothyroid rat. Biochemical and anatomical parameters of general
development. Biol Neonate. 1986;49:158-67.

13. International Clearinghouse for Birth Defects Surveillance and
Research. Annual Reports since 1980 available from International Centre on
Birth Defects, Rome. (www.icbd.org)

Competing interests:
None declared

Competing interests: No competing interests

10 March 2005
Pierpaolo Mastroiacovo
Director
Elisabeth Robert-Gnansia, Lorenzo D. Botto
International Center of Birth Defects - Via Pilo Albertelli, 9 - 00195 Rome - Italy