Lee Cohen

www.womensmentalhealth.org

Historically, lithium has been a mainstay of treatment for bipolar disorder. However, over the last decade, anticonvulsant drugs such as sodium valproate and lamotrigine (Lamictal) have become more widely used to treat this disorder.

The use of lithium in the first trimester is associated with a 0.05%–0.1% risk for Ebstein's anomaly, a well-described and frequently serious cardiac malformation. But data from the North American Antiepileptic Drug (NAAED) Pregnancy Registry and other international registries indicate that first-trimester exposure to sodium valproate is associated with an 8%–10% risk of major congenital malformations, notably neural tube defects and cardiac malformations.

As a result, many clinicians have been relieved to have the option of lamotrigine, which is an effective treatment for bipolar disorder and for which there had been extremely reassuring reproductive safety data over the last 5–7 years.

And until recently, several global teratovigilance programs had not found any indication that first-trimester use of this medication was associated with an increased risk for major congenital malformations.

In what is an important development, recent data from the NAAED registry note a prevalence rate of 2.7% for overall major malformations; however, five infants (8.9/1,000) had oral clefts. (See accompanying article.)

The baseline incidence of oral clefts in the general population has been calculated to be between 0.5 and 2.16 per 1,000 births; thus the data from the NAAED registry suggest at least a fourfold increase in the risk of cleft lip and palate or an absolute risk of approximately 0.9%. Interestingly, in five other registries surveyed, the frequency of oral clefts was 2.5 per 1,000 births, far less than reported by the NAAED Registry.

So how is the clinician to understand these new data, which suggest a signal of teratogenic risk, and how do the data inform the clinical care of patients who rely on the medication for control of chronic relapsing illnesses such as epilepsy or bipolar illness?

While stopping medication for the first trimester may appear to be an option for patients with bipolar disorder, unfortunately, a significant proportion of bipolar patients who do so will relapse.

Pregnancy does not appear to protect women with bipolar disorder against relapse if the mood stabilizer they are using is discontinued: In both a retrospective and prospective study, approximately 50% of patients relapsed during the first 6 months of pregnancy following discontinuation of mood stabilizer. It is also noteworthy that women with bipolar disorder are already at a fivefold increased risk for postpartum depression, compared with the general population, a risk that increases further if they relapse during pregnancy.

Therefore, many women with bipolar disorder who want to conceive are caught between a rock and a hard place, because many compounds used to treat bipolar disorder are either known teratogens, or are agents for which the available reproductive safety data are extremely sparse, such as the atypical antipsychotics, i.e., olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), and aripiprazole (Abilify).

Clinicians need to work collaboratively with patients to make treatment decisions, making every effort to minimize risk of relapse and fetal risk, realizing that some patients may have to assume some risk if they are to sustain affective well-being during pregnancy. For women who are on lamotrigine and are planning to conceive, the patients and prescribing clinician should now discuss the increased risk for oral clefts.

Patients who require treatment with a mood stabilizer, particularly those with recurrent disease, may consider a trial of lithium, which, while a teratogen, is associated with an extremely small risk for a cardiovascular malformation.

Certainly, the risk associated with lamotrigine is dramatically more modest than the risk associated with first-trimester exposure to sodium valproate, and many patients may elect to continue lamotrigine.

While it may seem intuitive to consider one of the atypical antipsychotics as an alternative to lamotrigine or lithium, given their efficacy in bipolar illness, the total absence of systematically derived data regarding the reproductive safety of atypicals makes them a less attractive alternative, and frankly the last resort, as compared with medications with known reproductive safety data.

When drug choice during pregnancy is considered, proceeding with a drug with known small risks as opposed to one with totally unknown risks is advantageous, particularly if the known risk is a modest one, which is the case with lamotrigine and lithium.

Ultimately, the clinician is left having to make decisions on a case-by-case basis, in collaboration with the patient, realizing that no decision is absolutely risk free. But decisions can be made that minimize morbidity associated with recurrence of bipolar illness, as well as prenatal exposure to any potentially harmful compound.

When presented with the options, women may make very different decisions. Some women in fact may decide to assume a small risk of oral cleft over a 0.05% risk for a heart malformation because they feel that oral clefts can be repaired more easily, while the morbidity and mortality of Ebstein's anomaly is high, even though the risk is exceedingly small. That is why these decisions have to be made individually, because such decisions will be made not based on relative risk or even absolute risk but rather on each patient's perception of risk.

www.womensmentalhealth.org

Historically, lithium has been a mainstay of treatment for bipolar disorder. However, over the last decade, anticonvulsant drugs such as sodium valproate and lamotrigine (Lamictal) have become more widely used to treat this disorder.

The use of lithium in the first trimester is associated with a 0.05%–0.1% risk for Ebstein's anomaly, a well-described and frequently serious cardiac malformation. But data from the North American Antiepileptic Drug (NAAED) Pregnancy Registry and other international registries indicate that first-trimester exposure to sodium valproate is associated with an 8%–10% risk of major congenital malformations, notably neural tube defects and cardiac malformations.

As a result, many clinicians have been relieved to have the option of lamotrigine, which is an effective treatment for bipolar disorder and for which there had been extremely reassuring reproductive safety data over the last 5–7 years.

And until recently, several global teratovigilance programs had not found any indication that first-trimester use of this medication was associated with an increased risk for major congenital malformations.

In what is an important development, recent data from the NAAED registry note a prevalence rate of 2.7% for overall major malformations; however, five infants (8.9/1,000) had oral clefts. (See accompanying article.)

The baseline incidence of oral clefts in the general population has been calculated to be between 0.5 and 2.16 per 1,000 births; thus the data from the NAAED registry suggest at least a fourfold increase in the risk of cleft lip and palate or an absolute risk of approximately 0.9%. Interestingly, in five other registries surveyed, the frequency of oral clefts was 2.5 per 1,000 births, far less than reported by the NAAED Registry.

So how is the clinician to understand these new data, which suggest a signal of teratogenic risk, and how do the data inform the clinical care of patients who rely on the medication for control of chronic relapsing illnesses such as epilepsy or bipolar illness?

While stopping medication for the first trimester may appear to be an option for patients with bipolar disorder, unfortunately, a significant proportion of bipolar patients who do so will relapse.

Pregnancy does not appear to protect women with bipolar disorder against relapse if the mood stabilizer they are using is discontinued: In both a retrospective and prospective study, approximately 50% of patients relapsed during the first 6 months of pregnancy following discontinuation of mood stabilizer. It is also noteworthy that women with bipolar disorder are already at a fivefold increased risk for postpartum depression, compared with the general population, a risk that increases further if they relapse during pregnancy.

Therefore, many women with bipolar disorder who want to conceive are caught between a rock and a hard place, because many compounds used to treat bipolar disorder are either known teratogens, or are agents for which the available reproductive safety data are extremely sparse, such as the atypical antipsychotics, i.e., olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), and aripiprazole (Abilify).

Clinicians need to work collaboratively with patients to make treatment decisions, making every effort to minimize risk of relapse and fetal risk, realizing that some patients may have to assume some risk if they are to sustain affective well-being during pregnancy. For women who are on lamotrigine and are planning to conceive, the patients and prescribing clinician should now discuss the increased risk for oral clefts.

Patients who require treatment with a mood stabilizer, particularly those with recurrent disease, may consider a trial of lithium, which, while a teratogen, is associated with an extremely small risk for a cardiovascular malformation.

Certainly, the risk associated with lamotrigine is dramatically more modest than the risk associated with first-trimester exposure to sodium valproate, and many patients may elect to continue lamotrigine.

While it may seem intuitive to consider one of the atypical antipsychotics as an alternative to lamotrigine or lithium, given their efficacy in bipolar illness, the total absence of systematically derived data regarding the reproductive safety of atypicals makes them a less attractive alternative, and frankly the last resort, as compared with medications with known reproductive safety data.

When drug choice during pregnancy is considered, proceeding with a drug with known small risks as opposed to one with totally unknown risks is advantageous, particularly if the known risk is a modest one, which is the case with lamotrigine and lithium.

Ultimately, the clinician is left having to make decisions on a case-by-case basis, in collaboration with the patient, realizing that no decision is absolutely risk free. But decisions can be made that minimize morbidity associated with recurrence of bipolar illness, as well as prenatal exposure to any potentially harmful compound.

When presented with the options, women may make very different decisions. Some women in fact may decide to assume a small risk of oral cleft over a 0.05% risk for a heart malformation because they feel that oral clefts can be repaired more easily, while the morbidity and mortality of Ebstein's anomaly is high, even though the risk is exceedingly small. That is why these decisions have to be made individually, because such decisions will be made not based on relative risk or even absolute risk but rather on each patient's perception of risk.

www.womensmentalhealth.org

Historically, lithium has been a mainstay of treatment for bipolar disorder. However, over the last decade, anticonvulsant drugs such as sodium valproate and lamotrigine (Lamictal) have become more widely used to treat this disorder.

The use of lithium in the first trimester is associated with a 0.05%–0.1% risk for Ebstein's anomaly, a well-described and frequently serious cardiac malformation. But data from the North American Antiepileptic Drug (NAAED) Pregnancy Registry and other international registries indicate that first-trimester exposure to sodium valproate is associated with an 8%–10% risk of major congenital malformations, notably neural tube defects and cardiac malformations.

As a result, many clinicians have been relieved to have the option of lamotrigine, which is an effective treatment for bipolar disorder and for which there had been extremely reassuring reproductive safety data over the last 5–7 years.

And until recently, several global teratovigilance programs had not found any indication that first-trimester use of this medication was associated with an increased risk for major congenital malformations.

In what is an important development, recent data from the NAAED registry note a prevalence rate of 2.7% for overall major malformations; however, five infants (8.9/1,000) had oral clefts. (See accompanying article.)

The baseline incidence of oral clefts in the general population has been calculated to be between 0.5 and 2.16 per 1,000 births; thus the data from the NAAED registry suggest at least a fourfold increase in the risk of cleft lip and palate or an absolute risk of approximately 0.9%. Interestingly, in five other registries surveyed, the frequency of oral clefts was 2.5 per 1,000 births, far less than reported by the NAAED Registry.

So how is the clinician to understand these new data, which suggest a signal of teratogenic risk, and how do the data inform the clinical care of patients who rely on the medication for control of chronic relapsing illnesses such as epilepsy or bipolar illness?

While stopping medication for the first trimester may appear to be an option for patients with bipolar disorder, unfortunately, a significant proportion of bipolar patients who do so will relapse.

Pregnancy does not appear to protect women with bipolar disorder against relapse if the mood stabilizer they are using is discontinued: In both a retrospective and prospective study, approximately 50% of patients relapsed during the first 6 months of pregnancy following discontinuation of mood stabilizer. It is also noteworthy that women with bipolar disorder are already at a fivefold increased risk for postpartum depression, compared with the general population, a risk that increases further if they relapse during pregnancy.

Therefore, many women with bipolar disorder who want to conceive are caught between a rock and a hard place, because many compounds used to treat bipolar disorder are either known teratogens, or are agents for which the available reproductive safety data are extremely sparse, such as the atypical antipsychotics, i.e., olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), and aripiprazole (Abilify).

Clinicians need to work collaboratively with patients to make treatment decisions, making every effort to minimize risk of relapse and fetal risk, realizing that some patients may have to assume some risk if they are to sustain affective well-being during pregnancy. For women who are on lamotrigine and are planning to conceive, the patients and prescribing clinician should now discuss the increased risk for oral clefts.

Patients who require treatment with a mood stabilizer, particularly those with recurrent disease, may consider a trial of lithium, which, while a teratogen, is associated with an extremely small risk for a cardiovascular malformation.

Certainly, the risk associated with lamotrigine is dramatically more modest than the risk associated with first-trimester exposure to sodium valproate, and many patients may elect to continue lamotrigine.

While it may seem intuitive to consider one of the atypical antipsychotics as an alternative to lamotrigine or lithium, given their efficacy in bipolar illness, the total absence of systematically derived data regarding the reproductive safety of atypicals makes them a less attractive alternative, and frankly the last resort, as compared with medications with known reproductive safety data.

When drug choice during pregnancy is considered, proceeding with a drug with known small risks as opposed to one with totally unknown risks is advantageous, particularly if the known risk is a modest one, which is the case with lamotrigine and lithium.

Ultimately, the clinician is left having to make decisions on a case-by-case basis, in collaboration with the patient, realizing that no decision is absolutely risk free. But decisions can be made that minimize morbidity associated with recurrence of bipolar illness, as well as prenatal exposure to any potentially harmful compound.

When presented with the options, women may make very different decisions. Some women in fact may decide to assume a small risk of oral cleft over a 0.05% risk for a heart malformation because they feel that oral clefts can be repaired more easily, while the morbidity and mortality of Ebstein's anomaly is high, even though the risk is exceedingly small. That is why these decisions have to be made individually, because such decisions will be made not based on relative risk or even absolute risk but rather on each patient's perception of risk.

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (INTERNAL MEDICINE NEWS, Nov. 15, 2005, page 24). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does one counsel women of reproductive age who have major depression? And what is the best option for patients being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in women who are actively trying to get pregnant or plan to in the future.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (INTERNAL MEDICINE NEWS, Nov. 15, 2005, page 24). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does one counsel women of reproductive age who have major depression? And what is the best option for patients being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in women who are actively trying to get pregnant or plan to in the future.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (INTERNAL MEDICINE NEWS, Nov. 15, 2005, page 24). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does one counsel women of reproductive age who have major depression? And what is the best option for patients being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in women who are actively trying to get pregnant or plan to in the future.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (FAMILY PRACTICE NEWS, Nov. 1, 2005, page 41). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does the clinician then counsel women of reproductive age who suffer from major depression? And what is the best option for patients who are being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in these women.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (FAMILY PRACTICE NEWS, Nov. 1, 2005, page 41). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does the clinician then counsel women of reproductive age who suffer from major depression? And what is the best option for patients who are being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in these women.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

www.womensmentalhealth.org

Multiple studies over the past decade have been supportive of the reproductive safety of the selective serotonin reuptake inhibitors (SSRIs) when used during the first trimester; these studies include one recent metaanalysis and other extensive reviews. Particularly reassuring have been the prospective data on fluoxetine (Prozac) and citalopram (Celexa). As a result, clinicians have been relatively reassured about the absence of teratogenic risk associated with the SSRIs.

New concerns were recently raised about the reproductive safety of paroxetine by a presentation at the Teratology Society annual meeting that reported an increased risk of omphalocele associated with first-trimester exposure. This report was based on preliminary, unpublished data from the National Birth Defects Center, which I reviewed in a recent column (FAMILY PRACTICE NEWS, Nov. 1, 2005, page 41). A weaker association was also found between omphalocele and other SSRIs.

A Food and Drug Administration public health advisory about paroxetine followed in December, describing preliminary results of two other unpublished studies indicating that paroxetine exposure in the first trimester may increase the risk of congenital malformations, particularly cardiac malformations. At the FDA's request, paroxetine manufacturer GlaxoSmithKline has changed the pregnancy category label for paroxetine from C to D.

It is surprising that the FDA's recommendation and advisory are based on preliminary analyses from several recent, unpublished, non-peer-reviewed epidemiologic studies, as these are data that should be considered, at least at this point, inconclusive.

Using data from the Swedish National Registry, one study found a 2% rate of cardiac defects among infants exposed during the first trimester to paroxetine vs. 1% among all registry infants. But a previous study using registry data that was based on a slightly smaller number of children exposed to paroxetine did not report this association (J. Clin. Psychopharmacol. 2005;25:59–73).

Another study, using data from a U.S. insurance claims database, found the rate of cardiovascular malformations was 1.5% among infants exposed to paroxetine during the first trimester vs. 1% among infants exposed to other antidepressants. The majority were atrial or ventricular septal defects, which are common congenital malformations.

The modest increases in relative risk of a common anomaly, when derived from a claims database with inherent methodologic limitations, make interpretation of these data problematic. Unfortunately, the language in the FDA advisory, suggesting that “the benefits of continuing paroxetine may outweigh the potential risk to the fetus,” may get lost in the information patients receive.

Although there are not as many published studies on the teratogenic risk of paroxetine as for other SSRIs, it is noteworthy that prospective studies have not identified a higher rate of congenital or cardiac malformations associated with prenatal exposure to paroxetine.

How does the clinician then counsel women of reproductive age who suffer from major depression? And what is the best option for patients who are being treated with paroxetine who want to get pregnant or who have an unplanned pregnancy? Until the issue is clarified with more rigorously obtained and conclusive data, it is reasonable to avoid paroxetine in these women.

For those with major depression who are antidepressant-naive, it may be most prudent to prescribe an SSRI or an SNRI for which there are no unfavorable data to date, such as fluoxetine or citalopram/escitalopram, or an older tricyclic antidepressant such as nortriptyline.

What makes sense for those who have failed to respond to one of those medications previously, as in the all-too-common scenario of nonresponse to multiple SSRIs and response only to paroxetine? In this situation, the use of paroxetine in women who are planning to conceive or who are already pregnant should not be considered absolutely contraindicated.

If the medication is discontinued before or during pregnancy, it should be done gradually, as is consistent with standard clinical practice.

Until the data are peer-reviewed and published, decisions about use of this medicine in women who are planning a pregnancy or are pregnant will have to be made on a case-by-case basis. But we need to keep in mind that nothing is more critical than sustaining euthymia during pregnancy. Untreated depression in pregnancy is associated with compromised fetal well-being as well as increased risk for postpartum depression.

The FDA advisory is available online at www.fda.gov/cder/drug/advisory/paroxetine200512.htm

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). Long-term neurobehavioral data have been somewhat sparse, but no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania indications as well. They are also being used widely across psychiatric disease states, including anxiety, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports, which have inherent biases with respect to overreporting of adverse outcomes.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information. Clinicians have been in a bind with respect to use of the atypicals during pregnancy.

A study published last April, the first prospective study of the reproductive safety of the atypicals in the literature, provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample of 151 patients. Investigators from the Motherisk Program in Toronto prospectively followed these women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All of the women had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate in the general population; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in rates of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9). This is the first prospective study that complements spontaneous reports from the manufacturers.

Among the 242 reports of olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of the 523 clozapine exposed pregnancies reported, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

Obviously, if a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and decisions have to be made on a case-by-case basis weighing relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and who is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. However, we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring.

Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies that would rapidly provide the amount of cases we need to reliably estimate reproductive risks. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). Long-term neurobehavioral data have been somewhat sparse, but no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania indications as well. They are also being used widely across psychiatric disease states, including anxiety, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports, which have inherent biases with respect to overreporting of adverse outcomes.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information. Clinicians have been in a bind with respect to use of the atypicals during pregnancy.

A study published last April, the first prospective study of the reproductive safety of the atypicals in the literature, provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample of 151 patients. Investigators from the Motherisk Program in Toronto prospectively followed these women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All of the women had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate in the general population; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in rates of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9). This is the first prospective study that complements spontaneous reports from the manufacturers.

Among the 242 reports of olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of the 523 clozapine exposed pregnancies reported, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

Obviously, if a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and decisions have to be made on a case-by-case basis weighing relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and who is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. However, we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring.

Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies that would rapidly provide the amount of cases we need to reliably estimate reproductive risks. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). Long-term neurobehavioral data have been somewhat sparse, but no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania indications as well. They are also being used widely across psychiatric disease states, including anxiety, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports, which have inherent biases with respect to overreporting of adverse outcomes.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information. Clinicians have been in a bind with respect to use of the atypicals during pregnancy.

A study published last April, the first prospective study of the reproductive safety of the atypicals in the literature, provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample of 151 patients. Investigators from the Motherisk Program in Toronto prospectively followed these women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All of the women had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate in the general population; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in rates of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9). This is the first prospective study that complements spontaneous reports from the manufacturers.

Among the 242 reports of olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of the 523 clozapine exposed pregnancies reported, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

Obviously, if a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and decisions have to be made on a case-by-case basis weighing relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and who is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. However, we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring.

Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies that would rapidly provide the amount of cases we need to reliably estimate reproductive risks. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with maternal use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to a number of SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded.

They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3). Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter also includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Many clinicians who prescribe SSRIs may be confused with the volley of new reports that suggest some potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of far greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being. That is an unacceptable outcome, which can be stated absolutely.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with maternal use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to a number of SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded.

They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3). Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter also includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Many clinicians who prescribe SSRIs may be confused with the volley of new reports that suggest some potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of far greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being. That is an unacceptable outcome, which can be stated absolutely.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with maternal use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to a number of SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded.

They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3). Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter also includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Many clinicians who prescribe SSRIs may be confused with the volley of new reports that suggest some potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of far greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being. That is an unacceptable outcome, which can be stated absolutely.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to several SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded. They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3).

Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Clinicians may be confused about SSRIs with the volley of new reports that suggest potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to several SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded. They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3).

Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Clinicians may be confused about SSRIs with the volley of new reports that suggest potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being.

www.womensmentalhealth.org

Over the past few years, several published studies have addressed the reproductive safety of the selective serotonin reuptake inhibitors. Recent studies have focused on the risk for neonatal discontinuation syndrome or symptoms of perinatal jitteriness associated with use of SSRIs during the latter portions of pregnancy.

Estimates of risk of first-trimester exposure to SSRIs derive from data accumulated over the last 15 years, which support the absence of major congential malformations associated with first-trimester exposure.

Data on the teratogenicity of SSRIs come from relatively small cohort studies and larger, international teratovigilance programs, and they have cumulatively supported the reproductive safety of fluoxetine (Prozac) and certain other SSRIs. These include a Scandinavian-based registry study of 375 women exposed to citalopram (Celexa) in the first trimester, which failed to indict SSRI as a teratogen.

A recent metaanalysis conducted by researchers at the Motherisk Program in Toronto supported the absence of teratogenicity associated with first-trimester exposure to several SSRIs.

Another recent report from the Swedish Medical Birth Registry failed to identify higher rates of congenital malformations associated with prenatal exposure to a number of SSRIs, including fluoxetine, citalopram, paroxetine (Paxil), and sertraline (Zoloft).

But at the Teratology Society's annual meeting in June, investigators from the University of British Columbia, Vancouver, reported an increased risk of omphalocele and craniosynostosis associated with first-trimester exposure to SSRIs. Using data from the National Birth Defects Prevention study, they compared data on 5,357 infants with selected major birth defects with 3,366 normal controls and interviewed mothers about exposures during pregnancy and other possible risk factors. Children with chromosomal anomalies or known syndromes were excluded. They found an association between exposure to any SSRI during the first trimester and omphalocele (odds ratio of 3).

Paroxetine accounted for 36% of all SSRI exposures and was associated with an odds ratio of 6.3 for omphalocele. Use of any SSRI during the first trimester was also associated with having an infant with craniosynostosis (odds ratio of 1.8). No association was noted between SSRI use and the other classes of major malformations studied.

This preliminary unpublished report is also described in a letter to physicians from GlaxoSmithKline, which markets paroxetine as Paxil. The letter includes additional data from an uncontrolled study of SSRI use during pregnancy, which noted a twofold increased risk in overall congenital malformations and cardiovascular malformations (most were ventricular septal defects) in offspring exposed to paroxetine, compared with other SSRIs. These data were derived from an HMO claims database.

Clinicians may be confused about SSRIs with the volley of new reports that suggest potential teratogenic risk associated with this class of compounds. Indeed, previous reports fail to describe such an association. Many of the more recent findings derive from either retrospective data sets taken from HMO claims data or from case-control studies, which also have certain methodologic limitations, compared with prospective cohort studies.

These recent findings of increased risk with prenatal SSRI exposure are inconsistent with earlier published findings. Nevertheless, large case-control studies can uncover an association not previously identified because of the inadequate statistical power of previous cohort studies, which were not large enough to detect an infrequent anomaly.

Even if we assume the associations from the new case-control study are true and that they are indeed causal, an odds ratio of 6.4 is associated with an absolute risk for omphalocele of only 0.18%. Absolute risk is of greater clinical value than relative risk and should be taken into account before patients are arbitrarily counseled to discontinue antidepressants during pregnancy.

The new findings are not necessarily a cause for alarm. Patients who are planning to conceive and are at significant risk for depressive relapse associated with antidepressant discontinuation may benefit from a switch to an antidepressant for which there are the greatest amount of data supporting reproductive safety. These antidepressants include the SSRIs fluoxetine, citalopram, escitalopram (Lexapro), as well as the older tricyclics.

However, for women who present when pregnant and still taking SSRIs, including paroxetine, discontinuation of the medication should not be arbitrarily pursued. Abrupt discontinuation of antidepressants can threaten maternal affective well-being.

www.womensmentalhealth.org

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). While long-term neurobehavioral data have been somewhat sparse, no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania as well.

But they are also being used widely across psychiatric disease states, including anxiety, agitation in the elderly, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information.

A study published in April—the first prospective study of the reproductive safety of the atypicals in the literature—provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample. Investigators from the Motherisk Program in Toronto prospectively followed 151 women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in the rate of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9).

As the authors point out, the sample was relatively small, the study was statistically underpowered, and long-term neurobehavioral outcomes were not evaluated.

The authors included the number of spontaneous reports of pregnancy exposures to atypicals, provided by the respective manufacturers, with the exception of newer agents. Among 242 olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of 523 clozapine exposed pregnancies, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

If a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and these decisions have to be made on a case-by-case basis weighing the relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. But we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring. Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

www.womensmentalhealth.org

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). While long-term neurobehavioral data have been somewhat sparse, no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania as well.

But they are also being used widely across psychiatric disease states, including anxiety, agitation in the elderly, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information.

A study published in April—the first prospective study of the reproductive safety of the atypicals in the literature—provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample. Investigators from the Motherisk Program in Toronto prospectively followed 151 women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in the rate of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9).

As the authors point out, the sample was relatively small, the study was statistically underpowered, and long-term neurobehavioral outcomes were not evaluated.

The authors included the number of spontaneous reports of pregnancy exposures to atypicals, provided by the respective manufacturers, with the exception of newer agents. Among 242 olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of 523 clozapine exposed pregnancies, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

If a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and these decisions have to be made on a case-by-case basis weighing the relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. But we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring. Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

www.womensmentalhealth.org

The reproductive safety of the older typical antipsychotics, such as haloperidol, is supported by extensive data that have accumulated over the past 40 years, at least with respect to teratogenic risk. Much of the data come from their use in treating nausea, particularly with prochlorperazine (Compazine). While long-term neurobehavioral data have been somewhat sparse, no particular indications of risk have been raised in more than 4 decades of use.

We have far less reproductive safety data on the newer “atypical” class of antipsychotics that have become widely used over the past decade because they lack some of the long-term side effects associated with the typical antipsychotics. These drugs—olanzapine (Zyprexa), risperidone (Risperdal), quetiapine (Seroquel), aripiprazole (Abilify), ziprasidone (Geodon), and clozapine (Clozaril)—are approved for schizophrenia; several are approved for acute mania as well.

But they are also being used widely across psychiatric disease states, including anxiety, agitation in the elderly, generalized anxiety disorder, and obsessive-compulsive disorder, and as adjunctive treatment of depression.

Because reproductive safety data on the atypicals have been sparse, clinicians are again faced with the difficult situation where a relatively new class of medicine is being used frequently. What data are available have been largely limited to manufacturers' accumulated case series or spontaneous reports.

To date, such information has not suggested any “signals” with respect to specific concerns regarding their use during pregnancy, but we can make only limited conclusions based on such information.

A study published in April—the first prospective study of the reproductive safety of the atypicals in the literature—provides some reassuring data regarding the risk of malformations, albeit in a relatively small sample. Investigators from the Motherisk Program in Toronto prospectively followed 151 women who took olanzapine, risperidone, quetiapine, or clozapine during pregnancy. All had taken one of these agents during the first trimester, and 48 were exposed throughout pregnancy. A total of 151 pregnant women who had taken a nonteratogenic drug also were followed.

In the atypical-exposed group, one child was born with a major malformation (0.9%), a rate lower than the 1%–3% background rate; compared with two (1.5%) babies in the control group, an insignificant difference.

Differences between groups in the rate of spontaneous abortions, stillbirths, or gestational age at birth were not statistically significant. Women taking atypical antipsychotics did have significantly higher rates of low-birth-weight babies (10% vs. 2%) and therapeutic abortions (10% vs. 1%) (J. Clin. Psychiatry 2005;66:444–9).

As the authors point out, the sample was relatively small, the study was statistically underpowered, and long-term neurobehavioral outcomes were not evaluated.

The authors included the number of spontaneous reports of pregnancy exposures to atypicals, provided by the respective manufacturers, with the exception of newer agents. Among 242 olanzapine-exposed pregnancies, there was no increase of major malformations or other abnormal outcomes above baseline. Of 523 clozapine exposed pregnancies, there were 22 “unspecified malformations.”

Of the 446 quetiapine-exposed pregnancies, 151 outcomes were reported, of which 8 were different congenital anomalies. Eight malformations were reported among the approximately 250 reports of pregnancies and lactation exposed to risperidone, but no pattern of abnormalities was noted.

If a patient can do without the medication, then it would be appropriate to discontinue it, but this is frequently not the case and these decisions have to be made on a case-by-case basis weighing the relative risks versus benefits.

For a patient planning a pregnancy who has a severe psychiatric illness and is maintained on an atypical antipsychotic to sustain functioning, switching to a typical antipsychotic may be prudent. But we often see women who present when they are already pregnant and on an atypical agent. At this point a switch may not be the wisest decision, if she is at a risk of relapse. For those women, the Motherisk data are not a guarantee of safety but do provide information that is at least moderately reassuring. Although this small study is encouraging, given the prevalence of reproductive-age women on these agents, it would be ideal if the industry performed postmarketing surveillance studies. Such studies may soon be mandated by the Food and Drug Administration in this post-Vioxx era, with increased emphasis on the safety of marketed drugs.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and two grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, what we're seeing is not acute withdrawal, like we see with heroin or methadone during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link. When considered with other case series in the literature, this study may indicate the potential risk for a perinatal syndrome.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and the possibility that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach in pregnant patients with depression. The data in this study may be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this case, we have more fog obscuring any guidance for the clinician than we have clarification of an already complicated situation.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and two grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, what we're seeing is not acute withdrawal, like we see with heroin or methadone during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link. When considered with other case series in the literature, this study may indicate the potential risk for a perinatal syndrome.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and the possibility that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach in pregnant patients with depression. The data in this study may be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this case, we have more fog obscuring any guidance for the clinician than we have clarification of an already complicated situation.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and two grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, what we're seeing is not acute withdrawal, like we see with heroin or methadone during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link. When considered with other case series in the literature, this study may indicate the potential risk for a perinatal syndrome.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and the possibility that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach in pregnant patients with depression. The data in this study may be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this case, we have more fog obscuring any guidance for the clinician than we have clarification of an already complicated situation.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns who were exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of SSRIs and SNRIs.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and 2 reports of grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several notable limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications among reproductive age women. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, certainly what we're seeing is not acute withdrawal, like we see with heroin or methadone use during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link.

When considered with other case series in the literature, this study may indicate the potential risk for some type of perinatal syndrome associated with the use of these medications, particularly around the acute peripartum period.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach. The data in this study may, in fact, be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this particular case, we have more fog obscuring any type of guidance for the clinician than we have clarification of an already complicated situation.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns who were exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of SSRIs and SNRIs.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and 2 reports of grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several notable limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications among reproductive age women. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, certainly what we're seeing is not acute withdrawal, like we see with heroin or methadone use during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link.

When considered with other case series in the literature, this study may indicate the potential risk for some type of perinatal syndrome associated with the use of these medications, particularly around the acute peripartum period.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach. The data in this study may, in fact, be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this particular case, we have more fog obscuring any type of guidance for the clinician than we have clarification of an already complicated situation.

www.womensmentalhealth.org

Multiple articles over the past several years have cited perinatal symptoms in newborns whose mothers were taking an antidepressant late in pregnancy, including transient restlessness, jitteriness, tremulousness, and difficulty feeding.

There have now been enough reports to suggest that certain vulnerable children or subgroups of newborns who were exposed in utero may be at a slightly increased risk for this syndrome. Indeed, last year the Food and Drug Administration required the addition of related information to the labels of SSRIs and SNRIs.

The results of a recent study of 93 cases worldwide (including 64 associated with paroxetine) from a World Health Organization adverse event reporting database do not represent new findings. The reports include descriptions of nervousness, agitation, abnormal crying, and tremors, which the authors consider a “signal” for perinatal or neonatal toxicity. The study also refers to 11 reports of neonatal convulsions and 2 reports of grand mal seizures, with no further description of the cases (Lancet 2005;365:482–7).

Although the report of neonatal convulsions is relatively new, the study itself has several notable limitations. It is difficult to interpret these results because they are from a spontaneous adverse event reporting system, where typically adverse outcomes are overreported and do not provide adequate information on when the drug was used, the duration of illness, or whether the woman was depressed during pregnancy. And the absence of a controlled sample makes it difficult to estimate the incidence of this type of problem, which likely is very low, considering the wide use of these medications among reproductive age women. Moreover, depression in the mother has been associated with many of the newborn symptoms reported.

The use of the term “withdrawal” syndrome is a dicey clinical call at best. Based on what we know about the kinetics and placental passage of these medications, certainly what we're seeing is not acute withdrawal, like we see with heroin or methadone use during pregnancy. The main metabolites of the drugs remain in the baby's circulation for at least days to weeks, so to see something so early and so transient, even for paroxetine (which has a shorter half-life than the other SSRIs), is not consistent with the pharmacokinetics of the compounds being described.

I don't disagree with these findings. Acknowledging the probable biases involved with collecting and reporting these cases, the report provides another data set that calls attention to the possibility of some type of perinatal syndrome associated with SSRI exposure later in pregnancy, which may not necessarily be a causal relationship. The authors suggest their findings are more of a “signal” that a problem may exist, rather than a definitive causal link.

When considered with other case series in the literature, this study may indicate the potential risk for some type of perinatal syndrome associated with the use of these medications, particularly around the acute peripartum period.

What is of concern, however, is the impact this report may have on appropriate prescribing of these drugs to pregnant women, and that patients, as well as physicians, will uniformly and arbitrarily avoid these drugs during pregnancy.

The article falls profoundly short in terms of helping the clinician. While the results indicate that more vigilance is necessary during the peripartum period in cases of SSRI use, the data do not imply that any particular SSRI should be avoided in women of reproductive age. The authors conclude that the signal is stronger for paroxetine, which they say should either not be used during pregnancy or used at the lowest effective dose. I certainly would not rule out using paroxetine in women of reproductive age on the basis of this report, with the possible exception of a woman with immediate plans to become pregnant or a woman with recurrent disease.

A reduction in the appropriate use of these drugs in depressed pregnant women would be a serious problem because relapse of recurrent depression during pregnancy is exceedingly common, and depression during pregnancy is the strongest predictor of risk for postpartum depression. Reducing the dose or discontinuing the antidepressant around the time of labor and delivery increases the risk of relapse, although some women may tolerate this approach, particularly if the drug is reinstituted immediately post partum.

Physicians should remain vigilant and carefully plan their treatment approach. The data in this study may, in fact, be a signal that a problem exists. But a signal should be some kind of beacon that guides the clinician. In this particular case, we have more fog obscuring any type of guidance for the clinician than we have clarification of an already complicated situation.