|
For an outline of the Australian categorisation of drugs in pregnancy, and explanatory qualifications, see Definitions of categories for drug use in pregnancy. This section also contains a table of drugs with their pregnancy categorisation and a recommendation about use during breastfeeding (see Drugs and their categories in pregnancy and breastfeeding).
Practical issues relating to drugs used in rheumatology are discussed in more detail below.
NSAIDs
Nonsteroidal anti-inflammatory drugs (NSAIDs) are one of the commonest classes of drugs used in the childbearing population, particularly when one includes the over-the-counter preparations available (eg aspirin, ibuprofen). They are classified as C on the basis of their metabolic and placental blood flow effects in the third trimester.
However, recent data has suggested a 5- to 7-fold increased risk of miscarriage in women taking NSAIDs around the time of conception. This and other studies have shown no increase in malformations or any other obstetric complication in the women who had an ongoing pregnancy. There are older studies of women with rheumatoid arthritis taking NSAIDs that show fertility rates the same as the general population, but these studies did not focus so clearly on NSAID use at the time of conception nor ultrasound confirmation of gestational dates, and may have underestimated the risk. These newer data alert us to the advisability of stopping NSAIDs, where possible, if a woman intends to become pregnant. However, once pregnancy is established, or if pregnancy occurs despite their use, they can be continued into the second trimester. NSAIDs should be ceased before the third trimester due to the possibility of early closure of the fetal ductus arteriosus or oligohydramnios via an effect on fetal renal function.
There are insufficient data to comment on the use of COX-2 selective drugs in early pregnancy, and the prudent action is to switch from a COX-2 to either a nonselective NSAID or simple analgesics if a pregnancy occurs or is planned.
Specific obstetric use of NSAIDs
Many studies have looked at low-dose aspirin in the first trimester for the prevention of pre-eclampsia and treatment of antiphospholipid syndrome, and have found no increase in fetal malformations.
Anti-inflammatory doses of NSAIDs are still used to inhibit uterine contractions in preterm labour, but always in a situation where fetal monitoring is available and delivery can be expedited if necessary.
Corticosteroids
If it is anticipated that anti-inflammatory drugs will
be required throughout most of pregnancy, changing from NSAIDs to prednis(ol)one
during the second trimester is recommended.
There are important differences between types of corticosteroid that determines
their safety in pregnancy. Hydrocortisone and prednisolone are largely
(90%) metabolised by placental dehydrogenase, but fluorinated corticosteroids
(eg betamethasone) and dexamethasone are not
Animal studies show a significant risk of cleft palate, and reference to this still occurs in the human literature though there are no human data to support it. Maternal doses of prednis(ol)one up to 40 mg have not been associated with teratogenicity, but are associated with obstetric complications such as gestational diabetes mellitus, pregnancy-induced hypertension, premature membrane rupture, and pre-eclampsia. The largest reported series, a cohort study, followed 311 women who had used cortisone in the first trimester, and found no statistically significant increase in major malformations compared with controls, and, in particular, there was no case of cleft palate.
Fish oils
Fish body oils containing omega-3 fatty acids can be used in pregnancy, and their mild anti-inflammatory activity may make it possible for patients who wish to become pregnant to cease NSAIDs.
Hydroxychloroquine
Human studies of hydroxychloroquine have failed to confirm the ototoxicity observed with high-dose use of chloroquine in rats. The tissue half-life of hydroxychloroquine is 1 to 2 months, so patients taking it for rheumatoid arthritis need to plan their pregnancy if they intend to cease it. As inflammatory joint pains tend to improve during pregnancy in patients with rheumatoid arthritis this is a reasonable strategy, though the risk of increased postpartum joint pains off treatment should be discussed with the patient.
On the other hand, lupus flares are not uncommon during pregnancy, the strongest predictor of flare being the state of activity of the lupus in the 6 months up to the time of conception. Therefore it is generally recommended that hydroxychloroquine be continued through pregnancy in lupus patients.
Sulfasalazine
There is a large literature on the safety of sulfasalazine in women during pregnancy. It is used predominantly by women with inflammatory bowel disease, but can be continued in women with rheumatoid arthritis when necessary. Men using sulfasalazine should be advised of the possibility of oligospermia and reduced sperm motility, which may affect fertility. These effects are reversible on cessation of sulfasalazine.
Methotrexate
The clearance of methotrexate varies, dependent both on renal function and dose. With increased assay sensitivity, a half-life of up to 52 hours has been measured. Detectable levels can leach out of tissues for up to several weeks.
Methotrexate interferes with the biosynthesis of purines via the enzyme dihydrofolate reductase. A teratogenic syndrome (the aminopterin syndrome, consisting of skeletal abnormalities of skull and limbs, microcephaly, and dysmorphic face) was described in women being treated for malignancies on higher doses of methotrexate than are used in rheumatoid arthritis (ie 10 to 25 mg/week). Animal studies show a widely varying species susceptibility to this effect.
Several small studies of human use of methotrexate suggest increased risk of abnormality after exposure during a critical period from 8 to 10 weeks gestation. A recent small study of early accidental pregnancy exposure (mean maternal methotrexate dose 10.5 mg/week) suggests no increased risk of fetal abnormality provided methotrexate is ceased before the critical period of gestation. Termination of pregnancy therefore may be avoided when methotrexate is ceased early after pregnancy is detected, and if gestational dates are confirmed with ultrasound and mothers have used folic acid supplements regularly.
In humans, exposure to dihydrofolate reductase inhibitors (like triamterene or trimethoprim) or other folate antagonists (like anticonvulsants) at between 6 to 12 weeks of gestation has been reported to be associated with a 2- to 3-times increase in the risk of cardiovascular defect or cleft palate. Concurrent use of folic acid reduced the differences to statistical insignificance.
Clearly, safe contraception and ceasing methotrexate before conception are recommended. A recommendation has been made to cease use of methotrexate 3 months before conception. Although there are no data indicating any adverse effects on spermatogenesis, most agencies recommend that potential fathers also cease methotrexate 3 months prior to planned conception.
Methotrexate is excreted in breast milk in low concentration and may be associated with neutropenia and immunosuppression in the infant, and is therefore not recommended with breastfeeding.
Leflunomide
Leflunomide is a pyrimidine synthesis inhibitor. Animal studies have shown its association with fetal death and teratogenicity. Its very long half-life makes just ceasing it insufficient to ensure elimination from plasma, so an elimination protocol with cholestyramine is recommended (cholestyramine 8 g 3 times daily for 11 days). Delaying pregnancy for 3 months following the washout treatment is also recommended. See also Leflunomide in 'Getting to know your drugs'.
Azathioprine
Although azathioprine is widely used for the treatment of systemic lupus erythematosus (SLE), the greatest pregnancy experience comes from its use in transplants, inflammatory bowel disease and multiple sclerosis. Azathioprine crosses the placenta but the fetal liver lacks the phosphorylase to convert it to its active metabolites.
Many human studies report a trend to increased fetal abnormalities; some fail to reach statistical significance, however, others do so. The increased risk is small and the abnormalities are the same as seen in the general population, so linking these abnormalities to the drug is difficult. All authors are unanimous that the risks are small, and that the severity of maternal disease should be the major determining factor in choosing whether or not to use azathioprine
Cyclosporin
Most data concerning cyclosporin use comes from organ transplantation. Cyclosporin crosses the placenta and is detected in fetal blood at levels up to 50% of maternal levels. Some trials comparing pregnancy outcome in organ transplant patients with and without cyclosporin use have reported an increased risk of fetal abnormalities, but others have not. The recommendation is that cyclosporin can be used in pregnancy if maternal disease justifies it.
Mycophenolate mofetil
Mycophenolate mofetil crosses the placenta. Its effects on pregnancy are unknown, and women should be changed over to azathioprine if pregnancy is planned.
bDMARDs, including tumour necrosis factor inhibitors
Antibodies directed against the cytokines tumour necrosis factor (TNF) and interleukin-1 are examples of the growing class of biological disease-modifying antirheumatic drugs (bDMARDs). To date, the greatest clinical experience has been with TNF inhibitors.
Human TNF inhibitors don't cross-react with TNF of other species, so safety data from animals would be irrelevant. However, mouse studies with murine anti-TNF have not shown embryo toxicity or teratogenic effects. The largest series to date of human pregnancy data in the literature relates to infliximab use in Crohn's disease. No significant difference in rate of fetal malformations was found between patients with Crohn's disease treated with conventional therapy compared to TNF inhibitor or to the general population, and the type of abnormalities seen was the same in all groups. One series of 29 pregnant patients with rheumatoid arthritis treated with etanercept showed no difference in major malformation rate compared to rheumatoid arthritic and normal controls.
As the amount of pregnancy data is limited, and there is no published data on breastfeeding, bDMARD use should be avoided where possible.
Other drugs
Cyclophosphamide and chlorambucil
are both alkylating agents, cross-linking DNA and disrupting cell division.
There is an increased risk of congenital malformations with these drugs,
and women of childbearing years should be counselled about adequate contraception
before beginning them. Some women diagnosed with cancer later in pregnancy
have chosen to continue pregnancy despite treatment with cyclophosphamide
and developmentally normal babies have resulted
There are insufficient data on the use of injected gold salts (thiomalate and thioglucose) in pregnancy to counsel on their safety but they have been shown to cross the placenta. They should be discontinued before or as soon as possible after pregnancy is detected. Gold has been detected in small amounts in breast milk and trace amounts in fetal blood but the significance of this is unknown.
Penicillamine is associated with cutis laxa, a rare fetal condition that may be fatal. It is now rarely used for treatment of rheumatological conditions and, although still used for the treatment of Wilson's disease, it should be avoided in pregnancy.
None of the drugs discussed in this section should be used during breastfeeding.
Use of antirheumatic drugs in potential fathers
Apart from the reversible effect of sulfasalazine on sperm motility and therefore fertility, little is published concerning the impact of paternal drug use on fertility or pregnancy outcome. Because of damaging effects on dividing cells, men are usually counselled to cease methotrexate, leflunomide, and cyclophosphamide at least 3 months before planning a pregnancy. A review of the clinical experience of methotrexate in males is available at www.motherisk.org. Azathioprine is known to decrease spermatogenesis and sperm viability in rats. Its effects on male fertility in humans are not known.
Revised March 2006. ©Therapeutic Guidelines Ltd (etg31demo, July 2010)