Safe Prescribing for Women of Reproductive Age
Safe Prescribing for Women of Reproductive Age
Teratogenic medications can be characterized as 'high-risk teratogens' (e.g., isotretinoin, which is thought to affect approximately 20–30% of exposed fetuses) or 'moderate-risk teratogens', which increase the risk of a birth defect to a smaller degree. Potential fetal effects after exposure to teratogenic medications include spontaneous abortion, malformations, visible developmental impairment, small size for gestational age, impaired intellectual development, carcinogenesis and increased risk of genetic mutations. These fetal effects are modified by a variety of factors, including medication dose and duration of exposure, gestational age at exposure and genetic susceptibility. Adequate nutrition (e.g., folate intake may be able to decrease the risk of teratogenicity of some medications; e.g., in animal studies, folic acid attenuates the risk of valproic acid-related birth defects).
Unfortunately, the most significant adverse effects of medications occur early in pregnancy, before many women realize that they are pregnant. From conception to 2 weeks after fertilization, teratogens increase the risk of early fetal death and spontaneous abortion. Organogenesis occurs during weeks 3–8, and during this time the fetus is at risk of developing major morphologic abnormalities. During later weeks of gestation, teratogenic exposures increase the risk for more subtle morphologic abnormalities and can produce biochemical, behavioral or reproductive abnormalities. Of note, however, the consequences of later exposures can still be of great clinical significance (e.g., renal impairment from angiotensin-converting enzyme inhibitor use, which can be fatal for the fetus).
Balanced against the risks that many medications pose to a fetus are the risks that medical conditions can pose to both the mother and fetus. Poorly controlled diabetes, cardiovascular, renal and thyroid disease all have potentially adverse effects on a fetus. Depression during pregnancy can affect a woman's participation in prenatal care, make her more likely to use tobacco and alcohol during pregnancy and may affect fetal growth and infant behavior.
Since 1979, the FDA has provided a pregnancy risk classification system to guide decision-making about medication use (Table 1). These pregnancy classes are determined by the existence of human or animal data regarding the pregnancy safety of a medication, the documentation of adverse fetal outcomes in humans or animals and also by the perceived risk:benefit ratio for women. Medications in Class A have been studied in humans and have not been demonstrated to cause fetal harm. Class B is used to identify medications for which animal studies have not shown risk, but there are no adequate studies of use by pregnant women. For medications in Class C, animal risk has been demonstrated, but adequate studies in humans are not available. Drugs in Classes D and X are identified as potential teratogens. In particular, Class D medications are drugs for which there is evidence of fetal risk but "the potential benefits from the use of the drug in pregnant women may be acceptable despite its potential risks". Class X medications are contraindicated in women who are or may become pregnant and generally include medications whose potential benefits in pregnant or potentially pregnant women were deemed to be not greater than the risks. Table 2 shows commonly prescribed Class D and X medications.
The limitations of the FDA pregnancy categories have been widely acknowledged and exemplify the challenges of translating pregnancy safety data into clinical recommendations. First, the data underpinning the labeling is limited. Animal studies have limited ability to predict adverse effects in a human fetus because of variations between mammalian species and also the different doses that are administered in animal studies. Few premarketing drug trials include pregnant women for ethical reasons and, moreover, most such studies would be too small to detect differences in the incidence of birth defects. Postmarketing surveillance is thus an essential source of information. Types of postmarketing data sources include epidemiologic studies such as cohort studies, pregnancy registries (i.e., cohort studies in which pregnant women exposed to a certain drug are enrolled and maternal and infant outcomes are followed through time and sometimes compared with a control group) and case–control studies. These types of studies can be very helpful, but are often limited by bias (e.g., recall bias for case–control studies, and self-referral bias for pregnancy registries) and the challenge of selecting an appropriate comparison group (for pregnancy registries). Pregnancy registries are often initiated by pharmaceutical companies and are drug-specific. In addition, since birth defects are rare, cohort studies are often not adequately powered to detect an increased rate of events with exposure to a medication. Lastly, reporting of birth defects (e.g., by patients and by maternal providers) is often inaccurate, partly because some abnormalities may only be detected later in an infant's life. Case reports and adverse-event reporting systems such as the FDA's have also been used in raising concerns about medication safety and may also influence pregnancy safety labeling, but are vulnerable to underreporting and reporting bias, and cannot establish causality.
As consumers and clinicians felt that the FDA categories were overly simple and often misleading, suggesting to some a gradation in risk from A to X and including medications with substantially different pregnancy-risk profiles in different categories would be useful. Starting in 2008, the FDA added more detailed labeling to these categories with the plan to eventually exclude letter categories and instead provide more information on risk:benefit calculations, including clinical considerations and a more in-depth presentation of supporting data. However, the final version of these regulations has not yet been published. It is hoped that these will offer a more nuanced portrait of the risks associated with medications in pregnancy. Given the limitations of the underlying data, even with the new labeling, physicians and their patients will probably continue to face clinical situations fraught with uncertainty, which will require highly individualized decision-making.
Defining Pregnancy Risk
Teratogenic medications can be characterized as 'high-risk teratogens' (e.g., isotretinoin, which is thought to affect approximately 20–30% of exposed fetuses) or 'moderate-risk teratogens', which increase the risk of a birth defect to a smaller degree. Potential fetal effects after exposure to teratogenic medications include spontaneous abortion, malformations, visible developmental impairment, small size for gestational age, impaired intellectual development, carcinogenesis and increased risk of genetic mutations. These fetal effects are modified by a variety of factors, including medication dose and duration of exposure, gestational age at exposure and genetic susceptibility. Adequate nutrition (e.g., folate intake may be able to decrease the risk of teratogenicity of some medications; e.g., in animal studies, folic acid attenuates the risk of valproic acid-related birth defects).
Unfortunately, the most significant adverse effects of medications occur early in pregnancy, before many women realize that they are pregnant. From conception to 2 weeks after fertilization, teratogens increase the risk of early fetal death and spontaneous abortion. Organogenesis occurs during weeks 3–8, and during this time the fetus is at risk of developing major morphologic abnormalities. During later weeks of gestation, teratogenic exposures increase the risk for more subtle morphologic abnormalities and can produce biochemical, behavioral or reproductive abnormalities. Of note, however, the consequences of later exposures can still be of great clinical significance (e.g., renal impairment from angiotensin-converting enzyme inhibitor use, which can be fatal for the fetus).
Balanced against the risks that many medications pose to a fetus are the risks that medical conditions can pose to both the mother and fetus. Poorly controlled diabetes, cardiovascular, renal and thyroid disease all have potentially adverse effects on a fetus. Depression during pregnancy can affect a woman's participation in prenatal care, make her more likely to use tobacco and alcohol during pregnancy and may affect fetal growth and infant behavior.
Since 1979, the FDA has provided a pregnancy risk classification system to guide decision-making about medication use (Table 1). These pregnancy classes are determined by the existence of human or animal data regarding the pregnancy safety of a medication, the documentation of adverse fetal outcomes in humans or animals and also by the perceived risk:benefit ratio for women. Medications in Class A have been studied in humans and have not been demonstrated to cause fetal harm. Class B is used to identify medications for which animal studies have not shown risk, but there are no adequate studies of use by pregnant women. For medications in Class C, animal risk has been demonstrated, but adequate studies in humans are not available. Drugs in Classes D and X are identified as potential teratogens. In particular, Class D medications are drugs for which there is evidence of fetal risk but "the potential benefits from the use of the drug in pregnant women may be acceptable despite its potential risks". Class X medications are contraindicated in women who are or may become pregnant and generally include medications whose potential benefits in pregnant or potentially pregnant women were deemed to be not greater than the risks. Table 2 shows commonly prescribed Class D and X medications.
The limitations of the FDA pregnancy categories have been widely acknowledged and exemplify the challenges of translating pregnancy safety data into clinical recommendations. First, the data underpinning the labeling is limited. Animal studies have limited ability to predict adverse effects in a human fetus because of variations between mammalian species and also the different doses that are administered in animal studies. Few premarketing drug trials include pregnant women for ethical reasons and, moreover, most such studies would be too small to detect differences in the incidence of birth defects. Postmarketing surveillance is thus an essential source of information. Types of postmarketing data sources include epidemiologic studies such as cohort studies, pregnancy registries (i.e., cohort studies in which pregnant women exposed to a certain drug are enrolled and maternal and infant outcomes are followed through time and sometimes compared with a control group) and case–control studies. These types of studies can be very helpful, but are often limited by bias (e.g., recall bias for case–control studies, and self-referral bias for pregnancy registries) and the challenge of selecting an appropriate comparison group (for pregnancy registries). Pregnancy registries are often initiated by pharmaceutical companies and are drug-specific. In addition, since birth defects are rare, cohort studies are often not adequately powered to detect an increased rate of events with exposure to a medication. Lastly, reporting of birth defects (e.g., by patients and by maternal providers) is often inaccurate, partly because some abnormalities may only be detected later in an infant's life. Case reports and adverse-event reporting systems such as the FDA's have also been used in raising concerns about medication safety and may also influence pregnancy safety labeling, but are vulnerable to underreporting and reporting bias, and cannot establish causality.
As consumers and clinicians felt that the FDA categories were overly simple and often misleading, suggesting to some a gradation in risk from A to X and including medications with substantially different pregnancy-risk profiles in different categories would be useful. Starting in 2008, the FDA added more detailed labeling to these categories with the plan to eventually exclude letter categories and instead provide more information on risk:benefit calculations, including clinical considerations and a more in-depth presentation of supporting data. However, the final version of these regulations has not yet been published. It is hoped that these will offer a more nuanced portrait of the risks associated with medications in pregnancy. Given the limitations of the underlying data, even with the new labeling, physicians and their patients will probably continue to face clinical situations fraught with uncertainty, which will require highly individualized decision-making.