Conflicts of interest: The author has received travel assistance and an honorarium for a talk from Nippon Boehringer Ingelheim, Co., Ltd.
Department of Gynecology, Meirikai Chuo General Hospital,
3-2-11 Higashi Jyujo, Kita-ku,
Tokyo 114-0001, Japan.
Many women suffer from allergic rhinitis (AR). The disease is often pre-existing and sometimes coincidental during pregnancy, and can worsen, improve, or stay the same during pregnancy. Besides ameliorating the detrimental effects of AR on the patient's quality of life, correct treatment is important for controlling concomitant asthma. If possible, it is important to highlight the risks of not taking such medications at a pre-conception visit. Although most medications for AR readily cross the placenta, there are several choices of treatment for controlling the symptoms during pregnancy. The choices may be varied depending on the disease course and symptoms, and inhaled corticosteroids are considered to be the first-line medical treatment. In addition, either a first-generation antihistamine, such as chlorpheniramine, or a second-generation antihistamine, such as cetirizine or loratadine, can be prescribed as the second-line medical treatment. As an alternative, intranasal cromolyn can be prescribed safely. Some of the leukotriene receptor antagonists and nasal decongestant sprays can only be prescribed when other methods are no longer valid and strict benefits can be expected. It is considered safe to continue immunotherapy during pregnancy.
Numerous pregnant women suffer from allergic rhinitis (AR). The disease is often pre-existing and sometimes coincidental during pregnancy , and may worsen, improve, or stay the same during pregnancy . Besides ameliorating the detrimental effects of AR on the patient's quality of life (QOL), correct treatment is important for controlling concomitant asthma .
Discussions with pregnant women about their treatment are particularly important for compliance and concordance by eliminating unnecessary anxiety. The primary concerns of pregnant women are focused on the risks of taking medications. Consequently, it is important to discuss these issues thoroughly, and to highlight the risks of not taking the medication before prescription as well as the potential hazards of the medication for the fetus. When the disease is pre-existing, adequate discussions are required before conception. Even if the patient is not planning to become pregnant, such discussions are required because of the high chance of unexpected pregnancy. Health care providers are therefore requested to train their skills for proper risk communications concerning medical treatments during pregnancy.
The purpose of this review is to summarize the recent data regarding the risks associated with drugs prescribed for AR. I will also discuss the drug safety information service for pregnant women in Japan.
General principles of prescription in pregnancy
It can be difficult and stressful for clinicians to prescribe drugs for pregnant women, especially because of the lack of evidence-based information. In particular, data on the long-term effects of drug exposure during the fetal period are almost completely lacking. For example, there is reasonably sufficient data for the teratogenicity of first-generation antihistamines, but data regarding their effects on postnatal physical and psychological development and learning ability are few. Recently, the intrauterine environment has been proposed to have a permanent impact on the determination of the fundamental processes of life, such as hypertension . We still have very limited knowledge about the long-term effects of drug exposure during the fetal period.
With these points in mind, no treatment and non-specific measures should be evaluated and discussed thoroughly before starting medical treatment. For AR, avoiding known irritants or triggers, such as pollen, dust mites, pets, and smoking, is important. An external nasal dilator or humidification are also among the possible non-specific measures.
If no treatment and non-specific measures are not effective, and the patient's QOL is highly impaired and limited, there is no need to hesitate before prescribing drugs for AR. However, it is preferable to avoid newly developed drugs, such as leukotriene receptor antagonists, because of the lack of available human data.
The drug effects on a fetus can be divided into several types which relate to the exposure time during pregnancy (Fig. 1) . First, until about 2 weeks after conception (i.e., until about 4 weeks of pregnancy, since pregnancy weeks are calculated using the first day of the last menstrual period as week 0 of pregnancy), exposure to a teratogen produces an all-or-none effect. This means that a conceptus usually either does not survive or survives without anomalies even when detrimental effects occur, and as a result, no fetal malformation can be caused by exposure to drugs during this period. After 4 weeks of pregnancy, organogenesis of the main fetal organs, such as the neural tube, heart, limbs and so on, starts and continues until about 7 weeks of pregnancy. During this critical period, teratogenic drugs may cause malformations of the organs developing at the time of drug exposure. For a few weeks after this period, the palate and external genitalia still continue to differentiate. Corticosteroids during this period may cause cleft palate. After the second trimester of pregnancy, fetal malformations do not tend to be caused by drugs, but fetotoxicity, such as increased fetal and infant mortality, impaired physiological function, alterations to growth, increased preterm or post-term birth, and perinatal symptoms of drug intoxication or withdrawal, may be caused by drugs. Even after the second trimester, abnormal morphologies of fetal organs may be caused by drugs through breakdown of normally developed organs. This mechanism is known as disruption, and may be caused by drugs that can damage the fetal vascular system. Thus, the susceptibility of a fetus to an agent and the subsequent outcome of exposure to the agent vary with the developmental stage at which the exposure occurs, meaning that the timing of exposure is another important factor for consideration when prescribing drugs for pregnant women. For example, drugs with teratogenic effects on the fetus should preferably be avoided during the first trimester.
When prescribing drugs during pregnancy, one should keep in mind that almost all drugs are able to pass through the placenta, with the exception of those with molecular weights of > 1,000 g/mol, such as insulin and heparin. As shown in Table 1, the molecular weights of all the drugs mentioned in this review are below 1,000 g/mol . This means that the drugs prescribed for the mother's benefit may be delivered to the fetus without any direct benefit.
Table 1. Molecular weights of the drugs described in this review, taken from 
Molecular weight (g/mol)
Leukotriene receptor antagonists
Besides the placental passage of drugs, the route of administration, i.e. systemic or topical, is sometimes important, because any drug can adversely affect fetal development by passing through the placenta from the maternal blood circulation. In other words, the maternal blood concentrations of a drug are important. Topical administration of a drug with low absorption to the affected site can maintain a significantly lower maternal blood concentration than systemic administration. It is therefore important to select agents that are poorly absorbed into the maternal circulation, when the same benefit can be expected. Fortunately for pregnant women, the affected sites in AR are the nasal mucosa and surrounding paranasal sinuses where drugs can be delivered topically.
Dose is another important point for consideration, as there is a dose-response relationship between the dose of a drug and its teratogenicity or fetotoxicity. In general, the lowest effective dose of a single agent for the shortest necessary period is recommended.
Owing to differences in the regulations among countries, it is advisable to conform to the resident country's regulations. In Japan, measures for prescription during pregnancy are written on the drug information sheets provided by each manufacturer.
The contents of risk communications should differ for drugs before prescription and drugs already prescribed to women who are unaware of pregnancy. In the former case, the maximum benefits and minimal risks are required, while in the latter case, possible harmful effects of drugs on the fetus should be compared with the 2–3% background risk of malformations.
Drug safety information service for pregnant women in Japan
In April 1988, a pioneering clinic was opened at Toranomon Hospital, located in the centre of Tokyo, for counselling of patients who had been exposed to drugs during pregnancy. The majority of the attending women visited the clinic with excessive anxiety relating to the teratogenic effects of drug exposure while they were unaware of their pregnancy. The clinic was operated by the Department of Obstetrics and Gynecology and Department of Pharmacology. Several trained obstetricians and pharmacists were engaged in the project.
The mothers were usually referred to the clinic by attending physicians or visited directly. Initially, they were asked to fill out questionnaires regarding the name, dose, and duration of medications as well as their history of past and current pregnancy. The obstetricians and pharmacists reviewed evidence-based reports, case reports, expert opinions, and animal studies. The staff tried to establish a database for the risks of individual drugs, and tried to classify the teratogenic risks into six grades (0–5). The drugs with epidemiological studies showing no increase in the incidence of malformations and no animal studies indicating fetal damage were classified into grade 0. On the other hand, when a drug was reported to significantly increase the incidence of fetal malformations in epidemiological studies, the drug risk per se was classified as 5. The data were updated and accumulated with every new consultation. The risks concerning gestational age of drug exposure were also classified into six grades (0–5). When the drug was taken during the period of 28–50 days after the first day of the last menstrual period, representing the critical period for major malformations, the gestational age-related risk was assigned as 5. And the risk was assigned as 0 for 0–27 days, representing the period of all-or-none effect. In addition, the risks were assigned as 3 for 51–84 days, 2 for 85–112 days, and 1 for 113 days to delivery, since the potential risk for malformation was reduced as pregnancy progressed after the critical period for major malformations.
By multiplying the two risk numbers assigned as 0–5, scores of 0–25 could be obtained. These scores were assigned as the overall risk scores. Risk communications between the staff and the mothers were carried out on a face-to-face basis for the overall risk. When the overall risk score was above 20, the mother was informed about the possibility of increased incidence of malformations. In all cases, the baseline incidence of malformations without drug exposure, the possible harmful effects on the fetus by the mother's illness per se, and the benefits of the drugs were described in the same setting.
Up to March 2007, i.e. for 20 years, a total of 8,982 women visited the clinic . Table 2 shows the top 20 categories of consulted drugs and the numbers of cases. As consultations for anti-epileptic drugs, which are known to be teratogens, were rather rare and not among the top 20 categories, they are not shown in Table 2. Therefore, on suspicion, the attending physicians had to explain about the risks and benefits of anti-epileptic drugs in advance of prescription, so that pregnant women felt secure about taking these drugs. In contrast to anti-epileptic drugs, the drugs potentially prescribed for AR, such as anti-allergic agents, antihistamines, and corticosteroids, were among the top 20 consulted drugs. It was suggested that women visited the clinic not because of the risk of drug exposure per se but instead because of inadequate or insufficient information from their attending physicians. Consequently, adequate and thorough information and discussions before prescription are particularly important for pregnant women.
Table 2. Top 20 categories of consulted drugs at Toranomon Hospital clinic for pregnancy and drugs from April 1988 to March 2007 
Top 20 categories of drugs
No. of cases
Adapted from Drugs and Pregnancy, 2nd edn. Tokyo: JIHO, 2010: 43–6 (in Japanese) with permission.
Those might be prescribed for allergic rhinitis. OTC, over-the-counter.
This pioneering clinic pointed out the needs and importance of such a service and encouraged the establishment of a system for information delivery. The Japan Drug Information Institute in Pregnancy, National Center for Child Health and Development, was established in October 2005 as a national center for information delivery, and 18 hospitals have been registered as cooperative hospitals to date.
Besides the national centre, a training system was started for Board Certified Pharmacists in Pharmacotherapy during Pregnancy and Lactation and Board Certified Pharmacy Specialists in Pharmacotherapy during Pregnancy and Lactation by the Japan Society of Hospital Pharmacists in 2008.
According to our experience, thorough discussions and complete risk communications before prescription are essential for successful pharmacological intervention.
Synthetic corticosteroids administered to mothers are known to be teratogenic in animals and humans. In humans, the odds ratio in case-control studies examining oral clefts was reported to be significant (3.35; 95% CI: 1.97, 5.69) . Betamethasone and dexamethasone cross the placenta to the fetus, and are prescribed for the fetus through the mother to reduce the incidence of neonatal respiratory distress syndrome and intracranial haemorrhage and, as a result, to increase the survival of premature infants . Besides the positive actions for the fetus, many adverse effects other than teratogenicity, such as intrauterine growth restriction, altered fetal brain development , and altered cortisol response to a stressor , have been reported. For systemic treatment of the mother, prednisolone or prednisone is recommended because the placenta can oxidize these drugs to inactive forms  and reduce their adverse effects on the fetus.
For AR, inhaled corticosteroids such as budesonide, fluticasone, and beclomethasone are known to be extremely effective because of their expected high concentrations at receptor sites in the nasal mucosa, with minimal absorption into the systemic blood flow . All drugs can affect the fetus through the maternal blood and placenta. Thus, minimal absorption into the maternal systemic blood flow means minimal effects on the fetus. As a result, no adverse effects of inhaled corticosteroids on the fetus were reported for congenital malformations and development [14, 15].
Based on their high efficacy and safety, inhaled corticosteroids are thought to be the first-line choice for the treatment of AR in pregnancy.
Two types of antihistamines are prescribed for AR, referred to as first-generation and second-generation antihistamines. The first-generation antihistamines include diphenhydramine, chlorpheniramine, clemastine, promethazine, hydroxyzine, and dimenhydrinate. Promethazine was reported to have some possible teratogenic effects in humans , but the findings were not conclusive. The other first-generation antihistamines are suggested to be safe as far as teratogenicity is concerned. However, their long-term effects on the psychological and physical development of newborns have been poorly studied.
The second-generation antihistamines were developed to be unable to cross the blood–brain barrier to reduce the side-effects of the first-generation antihistamines on the central nervous system, such as sedation and drowsiness. Cetirizine, desloratadine, fexofenadine, and loratadine are among the second-generation antihistamines. Desloratadine is a major metabolite of loratadine. No teratogenic effects have been reported for cetirizine and loratadine/desloratadine, while no human data are available for fexofenadine .
For the treatment of nasal symptoms caused by AR, first-generation or second-generation antihistamines need to be administered systemically. This means that antihistamines work through the maternal blood flow reaching the nasal sites. Therefore, all of the antihistamines have the possibility of crossing the placenta and reaching the fetal blood. The molecular weights of the first-generation antihistamines were far below 1,000 g/mol, and some of them were reported to cross the placenta . The molecular weights of the second-generation antihistamines are similar to or slightly higher than those of the first-generation antihistamines, but still low enough that passage to the fetus should be expected. No human data for placental passage are available for the second-generation antihistamines.
Human data start to be accumulated from the first marketing of a drug even if only case reports have been published . This means that the evidence of safety for individual drugs depends upon their marketing history. The first-generation antihistamines have more than 50 years of marketing history, while the second-generation antihistamines have around 20 years. Some of the antihistamines were marketed as over-the-counter medications with expected wide exposure to pregnant women. Many women with AR are considered to purchase antihistamines without the intervention of a physician or pharmacist.
For pregnant women who cannot obtain sufficient benefits from inhaled corticosteroids as the first-line treatment, which antihistamines should be recommended as the second-line treatment, first-generation or second-generation antihistamines? Some clinicians prefer to choose the first-generation antihistamine chlorpheniramine for pregnant women because it has been extensively studied. Other clinicians may prescribe second-generation antihistamines for working pregnant women to avoid the side-effect of drowsiness.
When prescribing an oral antihistamine, the provider is recommended to choose chlorpheniramine as a first-generation antihistamine, or cetirizine or loratadine as a second-generation antihistamine, because these drugs are classified as FDA pregnancy category B .
Leukotriene receptor antagonists
Pranlukast, montelukast, and zafirlukast are among the leukotriene receptor antagonists. For AR in non-pregnant women, leukotriene receptor antagonists are another treatment choice. However, it is preferable to avoid leukotriene receptor antagonists during pregnancy, because of the little available experience and data for human pregnancy and their effects on the fetus. In particular, pranlukast should be avoided because of the lack of human data. The molecular weights of these drugs are higher than those of antihistamines, but are still below 1,000 g/mol, and they have long elimination half-lives. Fetal exposure through the placenta should be expected, although it is not known whether the drugs in this class cross the human placenta.
Sympathomimetic vasoconstrictors may be prescribed for the short-term relief of nasal congestion in AR. Pseudoephedrine and phenylephrine are commonly used oral decongestants. Although teratogenicity of oral sympathomimetic vasoconstrictors is not suspected in humans , they are teratogenic in animals and may cause constriction of the uterine vessels, thereby reducing the uterine blood flow. Therefore, it is preferable to avoid oral decongestants during pregnancy.
Naphazoline and oxymetazoline are used topically in long-acting nasal decongestant sprays. It is reasonable to expect that the topical use of decongestants is safer than oral use. However, the use of nasal decongestant sprays should be limited to patients for whom inhaled corticosteroids and antihistamines are ineffective. Topical vasoconstrictors should not be used continuously because they can induce pharmacological rhinitis.
Other medical treatments
Intranasal cromolyn is not absorbed through the mucosal surface and is thought to have excellent safety. It was previously considered to be a first-line therapy for AR. However, inhaled corticosteroids have recently been considered to have similar safety but more efficacy than intranasal cromolyn. For patients who are not candidates for corticosteroids, intranasal cromolyn is a good alternative.
Allergen-specific immunotherapy is clinically useful for improving the symptoms of AR by inducing a state of immunological tolerance. Although it is considered safe to continue immunotherapy during pregnancy, increasing the dose or starting immunotherapy during pregnancy is not recommended because systemic reactions may occur .
Although AR is not a life-threatening disease, it may have detrimental effects on the patient's QOL. In addition, correct treatment of AR is important for controlling concomitant asthma. Inhaled corticosteroids, antihistamines, leukotriene receptor antagonists, nasal decongestant sprays, intranasal cromolyn, and immunotherapy have not been associated with increased incidences of human malformations. However, the long-term physical, psychological, and developmental effects on children have not been well studied. Inhaled corticosteroids are recommended as a first-line therapy during pregnancy because of their therapeutic effectiveness, low absorption into the maternal circulation, lack of reported adverse effects, and long marketing history.