The risk for severe and complicated malaria is increased during pregnancy. It is therefore even more important to provide pregnant women with safe and effective chemoprophylaxis.
All pregnancies carry risks. Approximately 15% to 20% end in spontaneous miscarriage. The incidence of congenital malformations among live births is approximately 5% to 6% after long‐term follow‐up. 1–3 Approximately half of these are diagnosed shortly after birth. Thus, when prescribing an antimalarial to a pregnant woman, there is always a substantial risk for adverse outcome even after intake of a fully safe drug. Avoiding travel is the easy way out but in many situations there is a definite need or a strong wish to visit endemic areas despite pregnancy. In addition, some women become pregnant while traveling and using malaria prophylaxis thus exposing the fetus to potentially toxic drugs.
Unfortunately, it is very difficult to show that a drug is safe during pregnancy; extremely large numbers of pregnancies have to be studied and the offspring have to be followed for many years to provide some measure of comfort. Even then, the constraints and limitations of such studies implicate that subtle adverse effects might be overlooked. Our current methods of safety surveillance are crude, including those undertaken by the pharmaceutical industry. Most information is based on observational studies or post‐marketing studies. Ideally, one should rather talk of a risk–benefit ratio than true safety for any prophylactic drug which is further complicated by the fact that there are in general only crude estimates of the actual risk of contracting Plasmodium falciparum malaria in different parts of the world. The only recommended prophylactic regimens for any traveler to highly malarious areas at present are atvaquone/proguanil, mefloquine, and doxycycline.
Atovaquone–proguanil (Malarone, GlaxoSmith Kline, Rixensart, Belgium) contains a combination of proguanil and atovaquone. Proguanil is considered to be safe during pregnancy but the experience is still limited for atovaquone. The combination is therefore either not recommended during pregnancy 4 or should only be considered “if the expected benefit to the mother outweighs any potential risk to the foetus.” 5 Post‐marketing surveillance data are essential but scarce and not available to us.
In a small randomized trial in Thailand, 81 pregnant women with a mean gestation of age 21 weeks were given Malarone and artesunate or quinine for treatment of uncomplicated P falciparum malaria. None of them were in the first trimester. Three congenital abnormalities and one stillbirth was observed. 6
Opinion differs on whether mefloquine can be recommended during the first trimester of pregnancy. The manufacturer of Lariam (Roche, Basel, Switzerland) holds the view that “women of childbearing potential should be advised to practice contraception during malaria prophylaxis with Lariam and 3 months afterwards.” 7 The World Health Organisation (WHO) provides no guidance, “There is very limited information on the safety and efficacy of most antimalarials in pregnancy, particularly during the first trimester.” 8 The Centers for Disease Control and Prevention (CDC) and others in the USA recommend use of mefloquine during the whole pregnancy period. 9–11 All agree that the drug can be given safely for prophylaxis during the second and third trimesters. The diverging opinions are due to remaining insecurity about possible teratogenicity in humans.
In a post‐marketing survey up to September 1996, a total of 1,526 pregnant women taking mefloquine (95.3% as prophylaxis) were followed. 12 Almost all women (97.7%) were exposed to mefloquine within 2 months before conception and/or during the first trimester. Only 646 resulted in deliveries while the rest were still pregnant at the time of survey (n = 192), had aborted (n = 325), or were lost to follow‐up (n = 363). There were 26 congenital malformations among the deliveries (4%). In a subset of 476 children who were exposed during the first trimester, malformations were noted in 24 of them, ie, 5.4%. No specific pattern of malformation was seen. The authors concluded that previous animal data, which suggested that teratogenicity was observed at high doses, cannot be applied to humans.
An update to October 2005 adds the number of exposed women to 2,216 of which 975 delivered. Of of these 975 children, 42 had congenital malformations (4.3%). The total number of women exposed in the first trimester is not shown. 13
During therapy for malaria, increased risk for still births was reported in one study from Thailand. 14 There was no increased risk during mefloquine therapy followed by prophylaxis in another study in Malawi but medication was then only initiated after the first antenatal visit. 15
Tetracyclines form a stable calcium complex in bone‐forming tissue. When used during tooth development, which takes place during the last half of pregnancy in humans, discoloration of the primary teeth might occur. The permanent teeth are not affected. Doxycycline is a tetracycline and might carry the same risk but according to a review no tooth staining has been documented in humans with this compound. 16 There is no knowledge on potential impact of the growing fetus on metalloproteinase inhibition which might in theory be harmful with a calcium chelating drug. Further studies are needed.
Tetracyclines can be used as therapy for a variety of infectious diseases for a short period of 1 to 2 weeks and during longer periods for therapy of acne (usually with tetracyklines other than doxycycline). In both cases, the recommended dosages are similar to malaria prophylaxis, ie, 100 mg of doxycycline each day. The risk for discoloration is not exposure dependent, ie, the potential risk is the same, regardless of whether doxycycline is used for short or long periods of time. 17
Tetracyclines are considered to be contra‐indicated during the whole pregnancy by most bodies including WHO and CDC. In contrast, it was concluded that tetracyclines are only contraindicated after the fourth month of pregnancy in an extensive review on tetracyclines and fetal and neonatal risks. 18 Similarly, doxycycline could be used during the first half of pregnancy according to the latest Swedish Summary of Product Characteristics (SPC). 19 As doxycycline should be continued for 4 weeks after leaving an area endemic for malaria, doxycycline can still only be considered for women leaving an endemic area, at the latest, at the end of the first trimester.
In a retrospective case‐control study from Hungary, women were asked for doxycycline use during pregnancy. 20 Among 32.804 women who had infants without defects, 0.19% had been treated with doxycline compared with 0.30% among women who had offspring with congenital abnormalities. The difference was significant but there was no significant relation between malformation and intake during the second and third month of gestation and recall bias might have had an influence. The authors concluded that doxycycline presented very little, if any, risk to the fetus and if treatment is necessary during pregnancy, there would appear to be no contraindication. Similarly, in a recent review, 16 the teratogenic potential of doxycycline was considered unlikely and the drug placed in the same category as amoxicillin.
In a non‐peer‐reviewed surveillance study of Medic aid recipients, data on 1,795 children exposed to doxycycline during pregnancy did not support an association between the drug and any of six specific malformations (cardiovascular defects, oral clefts, spina bifida, polydactyly, limb reduction defects, and hypospadias). 21
The Swedish medical birth register administered by the Swedish National Board for Health and Welfare contain data from 1973 and onward. According to this register, a total of 1,809 women were exposed to tetracyclines (the majority probably to doxycyline) during early pregnancy. In a detailed follow‐up of the malformations during the period 1996 to 2005, 980 children were monitored. Malformations were found in 52. Compared to a control group with no exposure OR was 1.13, 95% CI 0.85 to 1.49. The interpretation by the leading Swedish expert was that tetracyclines do not have a teratogenic effect.
Even if the present information on possible teratogenic effects of antimalarials is crude and no direct comparison between drugs have been undertaken, the percentage of children with any kind of malformation is virtually the same, regardless if the child has been exposed to mefloquine or doxycycline during prophylaxis.
We, and other members of the Swedish national group for recommendations on malaria prophylaxis, 22 therefore consider doxycycline at least as safe as mefloquine for use as malaria prophylaxis during early pregnancy. This will add doxycycline as a choice for pregnant women, especially for those who do not tolerate mefloquine or who travel to areas with resistance to mefloquine.
Declaration of interests
The authors state that they have no conflicts of interest to declare.