Are controlled clinical trials truly unfeasible in pregnant women?
An impairment of the placental circulation is considered the common basis of such different obstetrical complications as recurrent miscarriage, stillbirth, pre-eclampsia and placental detachment . Recurrent pregnancy loss (RPL), including miscarriages occurring during the first trimester and fetal losses during the second or third trimester are the most widely investigated obstetrical complications with respect to the putative mechanistic role of thrombophilia. Unfortunately, the studies available so far lack power because of insufficient sample size, so that the association between thrombophilia and RPL is not definitely established. Two recent meta-analyses showed a two- to threefold increased risk of RPL in carriers of factor V Leiden or prothrombin mutations and an eightfold increased risk of late events (pregnancy losses in the second and third trimester) in carriers of factor V Leiden [2,3]. Hence, despite the absence of conclusive studies, the limitations of meta-analyses because of publication bias, the multiple etiology of RPL and the general difficulties encountered to carry out controlled trials of drugs in pregnancy, there is some rationale for antithrombotic prophylaxis in pregnant women with thrombophilia and RPL. In the year 2000, Brenner et al. carried out a small intervention study on 50 women with thrombophilia who were selected for their unfavorable obstetrical history (20% of previous live birth rate), and were given two different doses of the low-molecular-weight heparin (LMWH) enoxaparin throughout pregnancy . The choice of 40 or 80 mg per day of enoxaparin was made on the basis of the presence of one or more thrombophilic abnormality in each woman. The major limitation of that study was the absence of a control group, i.e. women selected with the same criteria but receiving no treatment. Recently in this journal, Brenner et al.  reported the results of a similar study – the LIVE-ENOX – carried out in 180 women who were treated with 40 or 80 mg day−1 of enoxaparin, independent of the number of thrombophilic abnormalities. The results of these two studies of Brenner et al. [4,5] are consistent in that both report a live birth rate of approximately 80%, with no difference between the two therapeutic regimens. In the LIVE-ENOX study, the most frequent obstetrical complication was early (first trimester) RPL. We know from general epidemiological data that the probability of three consecutive pregnancy losses is 24% and of four is 33% . This means that without any treatment, 76% and 67% of women end their third and fourth pregnancy successfully, respectively. These rates of successful pregnancies are not dramatically smaller than the 80% rate found by Brenner et al. [4,5]. Hence, in the absence of a control group, no conclusion on the efficacy of an anticoagulant prophylaxis in thrombophilic women is warranted. I am aware that conducting a controlled trial is extremely difficult in this setting, for at least three reasons: (i) financial support; (ii) randomization; (iii) selection of study population. To date, in Europe it is almost impossible to find financial support for clinical trials outside the pharmaceutical industry, that usually does not accept to fund these trials . The Italian Ministry of Health has recently voiced the need for European health authorities to promote not-for-profit clinical trials  to address relevant questions that are not tackled by the industry. I am convinced that controlled clinical trials in pregnant women with unfavorable obstetrical history are an important priority. Another concern is the issue of randomization, i.e. how to inform a woman that, depending upon randomization, she may or may not be administered a drug that might improve the outcome of her actual pregnancy. On the contrary, I believe that focused counseling on what is evidence-based (live birth rate after RPL without treatment, pros and cons of LMWH administration for several months) and what is not (limitations of the studies published so far), should help to reassure women and to obtain informed consent. Finally, the selection of the study population is crucial. There are two challenging issues: the type of obstetrical complications and the type of thrombophilia. Putting together all the various obstetrical complications according to the ‘thrombosis theory’ may lead to uncertain results, because they do indeed differ in terms of etiology. On the contrary, subgroup analysis would require a study of very large sample size. Focus on the most widely investigated obstetrical complication, i.e. RPL as defined by Brenner et al. , or on the most common one, i.e. pre-eclampsia , is perhaps a valid approach. With respect to the type of thrombophilia, there is little evidence that the homozygous MTHFR C677T polymorphism is a proven cause of thrombophilia and a risk factor for RPL . In addition, the role of acquired APC resistance in the absence of factor V Leiden (during pregnancy) as a risk factor for RPL is still debated . It happens that the homozygous C677T MTHFR polymorphism and acquired APC resistance in the LIVE-ENOX study represented 18% and 17% of the total number of thrombophilic abnormalities, respectively. In conclusion, many more attempts to truly explore the feasibility of independent, well-designed, randomized controlled trials should be pursued, before LMWH prophylaxis is established as the treatment of choice for the clinical management of pregnant women who had obstetrical complications. These trials would benefit our patients and render recommendations on treatment of women with RPL more relevant than the grade 2C evidence scored by LMWH plus low-dose aspirin in the last ACCP consensus conference .