The question, “Doctor, is it safe for me to become pregnant?” is one of the more charged issues confronting a woman with idiopathic thrombocytopenic purpura (ITP) and her physician. Maternal platelet counts can fall during even uncomplicated pregnancies. Some of the drugs commonly used to manage ITP either are contraindicated in pregnancy, increase the incidence of gestational complications, or are unstudied, and splenectomy is both technically difficult and can precipitate premature labor. In addition, immunoglobulin G (IgG) antibodies that cause ITP readily traverse the placenta and can cause neonatal thrombocytopenia of sufficient severity to predispose to intracerebral hemorrhage in the neonate. There is little correlation between maternal and fetal platelet counts and no way to predict neonatal outcome that is either noninvasive or less dangerous than ITP itself.
Studies from the group at McMaster University over the past decade have changed the way we manage pregnancy in women with ITP. The study by Webert and colleagues in this issue of Blood (page 4306) describes the largest cohort of pregnant patients with ITP from North America reported to date (119 pregnancies in 92 women) and provides valuable insight into many of the important clinical issues involved in the management of the mother and fetus. In general, pregnancy was safe for both parties, although about one-third of the women ultimately required treatment for thrombocytopenia or bleeding. It is notable and disturbing, in this regard, that fewer than half of the mothers responded to such therapy, generally involving prednisone and intravenous immunoglobulin (IVIG). Fortunately, vaginal delivery was generally uncomplicated, even in the 7 women with severe thrombocytopenia (platelet count < 20 × 109/L). Only 5.5% of the neonates were born with severe thrombocytopenia, and only 1 death from intracerebral hemorrhage (ICH) occurred among the 123 neonates in this series. As has been noted by others, there were no maternal predictors of neonatal outcome other than a prior affected pregnancy.
Many of these findings reinforce the results of a recently published study of 286 neonates born to 284 mothers with ITP by Fujimura et al,1 including the proportion of diagnoses made during gestation, the fraction of women requiring treatment, the safety of vaginal delivery even at very low platelet counts, and the low incidence of ICH. However, in the study by Fujimura et al, maternal response to prednisone and IVIG was comparable with that seen in nonparous patients. One potential reason for the difference between the 2 studies lies in the difficulty in distinguishing ITP from incidental thrombocytopenia of pregnancy when there is no antecedent history. In the study by Webert et al, 34 patients were diagnosed with ITP based on a maternal platelet count less than 70 × 109/L or thrombocytopenia that developed in the first or early second trimester rather than by documenting persistent thrombocytopenia after delivery, whereas a response to corticosteroids was required for inclusion in the series from Fujimura et al1 when there was no antenatal history of ITP. However, both groups make the cogent observation that there were no substantive differences in the clinical course related to the timing of the diagnosis. Additional studies will therefore be needed to validate the criteria for diagnosing ITP in the absence of a pregestational history. This is important when counseling women with presumed ITP and severe thrombocytopenia as to the likelihood that their disease can be managed effectively throughout pregnancy.
The study by Webert et al leaves us with 2 important messages. On the one hand, we can reassure our patients that there is high probability of a favorable outcome for mother and infant. On the other hand, it is humbling that despite the strides that have marked the past 20 years in the diagnosis and treatment of ITP, neonatal ICH still occurs and can be neither predicted nor prevented.