> Skip repeated content

What has been the experience with anti-TNF agents during pregnancy? Should they be discontinued prior to pregnancy and, if so, how far ahead of a planned pregnancy?

Ask the Expert

Biologic medications directed against the inflammatory effects of tissue necrosis factor (TNF) in rheumatoid arthritis have had a powerful effect on our ability to treat rheumatoid arthritis (RA).  As with any new medication utilized by women of child-bearing age, questions arise as to pregnancy-related effects and safety. Unfortunately, there are no easy answers. Although clinical information is being collected as increasing numbers of pregnant women are treated (by the Organization of Teratology Information Services, or OTIS, among others), placebo-controlled studies are unlikely to be conducted for obvious reasons. The short answer to the question of pregnancy safety, then, is "we don't know." The longer answer below tries to put this into context and to provide reasonable practice guidelines.

Etanercept is a fusion protein of the TNF-alpha receptor binding domain and the IgG1 antibody constant region; infliximab is a chimeric monoclonal antibody, and adalimumab, a humanized monoclonal antibody, directed against the TNF-alpha receptor. All three inhibit the pro-inflammatory actions of TNF-alpha. Etanercept and adalimumab are self-administered as subcutaneous injections; infliximab is infused intravenously every four to eight weeks.

Pharmacology studies on laboratory animals are routinely conducted to look for evidence of embrytoxicity or teratogencity. All three TNF-alpha inhibitors are rated FDA pregnancy category B, meaning that animal studies reveal no fetal harm but no adequate human studies are available (the case here), or that animal studies do show harmful fetal effects but adequate human studies do not. Etanercept did not cause developmental toxicity in animals at up to 100 times human doses. Similarly, studies of adalimumab in monkeys at levels up to 373 times the recommended human dose revealed no evidence of fetal harm. Infliximab does not cross react with TNF-alpha in species other than humans and chimps, but no embryotoxicity or teratogenicity occurred in mice treated with an analogous antibody that inhibits functional activity of murine TNF-alpha.

No adequate or controlled studies in pregnant women have been done for any of the TNF-alpha inhibitors, although a very limited number of case reports in humans have not identified developmental problems. Antoni et al, reported two complications out of 36 live births in women treated with infliximab for Crohn's disease or RA for which they had outcome data: one neonatal death at three weeks, and one infant with tetralogy of Fallot (consistent with expected outcomes in a healthy national cohort)[1]. One problem of extrapolating from Crohn's patients to RA, however, is that infliximab is often given only intermittently to patients with Crohn's disease (as opposed to the more regular administration for RA), and is not necessarily given during the first trimester. A single recent case report describes ovulation induction, intrauterine insemination, and normal pregnancy and delivery following chronic anti-TNF-alpha therapy with etanercept.

It is well recognized that remission occurs during pregnancy in up to two-thirds of patients with RA. This may obviate the need for medication in some cases, but remission is neither automatic nor instantaneous, and early pregnancy is usually the most vulnerable risk period for medication effects. Should TNF inhibitors be continued while patients are trying to conceive? The manufacturers recommend stopping medication at least several months before conceiving. This is not possible for many women.

As a practical compromise in those for whom stopping medication is not an option, it seems reasonable to decrease the frequency of administration during this period, even holding anti-TNF agents, if possible, from time of ovulation until menses occurs. Once a patient is pregnant, low dose prednisone (which has limited placental transfer) may be preferable to TNF-alpha inhibitors for the first trimester. Second and third trimester use of TNF blockers may present less risk.

If deciding on a specific TNF-alpha inhibitor in a woman with pregnancy plans in the near future, it makes sense to institute a drug that has a shorter half life, such as etanercept. Note that TNF-alpha inhibitors are often co-administered with methotrexate: concomitant use of methotrexate is not an option, however, for patients trying to conceive. It must be stopped three to six months before conception, as it is teratogenic (FDA pregnancy category "X").

Therapy with anti-TNF agents during breast-feeding is not recommended, although little data is available. A single case report suggests that infliximab is not excreted in breast milk. It is not known if etanercept or adalimumab are excreted in human milk or absorbed systemically after ingestion. Given the known immunosuppressive effects of these medications, the unknown risk of transfer to the neonate, and the available alternative of infant formula, avoidance of breastfeeding while taking TNF-alpha inhibitors seems logical.

TNF-alpha is expressed in normal endometrial, placental and fetal tissues and may have complex biological effects, both positive and negative, during pregnancy. There is increased expression of TNF-alpha in pre-eclamptic placentas, for example, and increased levels of TNF-alpha predict insulin resistance in pregnancy. Berman, et al. recently suggested that TNF-alpha release is necessary for induction of fetal loss by antiphospholipid antibodies[2]. Some speculate that elevated levels of TNF-alpha from uterine NK cells may play a role in immunological pregnancy loss, and several reproductive immunologists propose therapy with etanercept, although there are no convincing data as yet to support this practice.

What are clinical rheumatologists advising, despite the limited data available, as more young women are treated with these agents? In a recently published survey of rheumatology practice patterns, fewer than one-half of the 175 responding rheumatologists felt that pregnancy is contraindicated with use of etanercept or infliximab (38.6% and 46.5%, respectively); in contrast, most agreed that pregnancy is contraindicated in women taking leflunomide (92.8%) or methotrexate (95%)[3].

In making decisions regarding pregnancy therapy in RA, it is important to balance functional status and long-term joint preservation with fetal safety. Discussion with the patient (and her partner) is critical, as decisions need to be individualized to the particular patient. Certain RA medications are felt to be relatively safe in pregnancy, including sulfasalazine and hydroxychloroquine. Of the more potent DMARDs, however, methotrexate and leflunomide are clearly contraindicated. For those patients with active severe disease, anti-TNF agents may be the most reasonable choice for treatment immediately before and during pregnancy. Frequent updates of pregnancy outcome in treated patients is an effective way to reassure RA patients contemplating pregnancy in the future.

[1] Antoni CE, Furst D, Manger B, Lichtenstein GR, Keenan GF, et al. Outcome of Pregnancy in Women Receiving Remicade (infliximab) for the Treatment of Crohn's Disease or Rheumatoid Arthritis. Arthritis Rheum 2001;44:S152.

[2] Berman J, Giardi G, Salmon JE. Tumor necrosis factor-alpha is an Important Effector in Antiphospholipid-induced Pregnancy Loss in Mice. Arthritis Rheum 2003;48:S446.

[3] Chakravarty EF, Sanchez-Yamamoto D, Bush TM. The Use of Disease-modifying Antirheumatic Drugs in Women with Rheumatoid Arthritis of Childbearing Age: A Survey of Practice Patterns and Pregnancy Outcomes. J Rheumatol 2003 Feb; 30(2):241-6.


Headshot of Lisa R. Sammaritano, MD
Lisa R. Sammaritano, MD
Associate Attending Physician, Hospital for Special Surgery
Associate Professor of Clinical Medicine, Weill Cornell Medical College

Related Professional Content