Gris and colleagues report on a randomized controlled trial comparing 100 mg aspirin to 40 mg enoxaparin started at 8 weeks gestational age in women with prior fetal loss later than 10 weeks and thrombophilia (factor V Leiden, prothrombin gene variant, protein S deficiency).1 The authors report an astounding 57% absolute risk reduction with enoxaparin compared with aspirin. This absolute risk reduction translates into a number needed to treat of 1.7; that is, 1.7 women with thrombophilia and a loss later than 10 weeks need to be treated with enoxaparin throughout pregnancy to prevent one fetal loss between 8 weeks and term compared with aspirin. Late fetal loss is a very common problem occurring in 2.3% of pregnancies.2 This dramatic finding may have a large impact on the care of women with prior late fetal loss. The case for thrombophilia screening and prophylaxis of these women is strengthened.
However, the authors, the editors, and the reviewers of this article failed to ensure that essential information was included in the article that would permit the reader to determine the strength of the internal and external validity of the study. Table 1 fails to display baseline characteristics according to treatment groups (aspirin versus enoxaparin) as is suggested in the Consolidated Standards of Reporting Trials (CONSORT) guidelines.3 By failing to display this information, the reader is unable to determine whether known and important prognostic factors (eg, maternal age) were equally balanced between the treatment groups. The absence of a figure to display flow of trial participants as suggested by the CONSORT guidelines does not permit the reader to determine generalizability of the findings and assess important internal validity issues. When were these patients recruited (ie, over what time frame)? Where and how were patients screened for enrollment? How many screened patients were excluded and for what reasons? How many consenting patients received the intended treatment (ie, treatment as allocated by randomization)? The method of allocation concealment was not described. How many complied completely with the intended treatment? How many were lost to follow-up? How many were analyzed for the primary outcome? Was the analysis done by intention to treat? Were there any interim analyses?
By not reporting these important study characteristics we cannot be reassured that bias is not driving these dramatic results, nor can we be sure that these results apply to our patients.4,5
The authors should be congratulated on completing this study. Performing drug intervention trials in pregnancy is extremely challenging, fraught with medicolegal challenges, regulatory hurdles, and an absence of interest from the pharmaceutical industry. However, prior to adoption of this study's findings, this important information needs to be revealed.
Thromboprophylaxis for first fetal loss
We would like to thank Dr Rodger for his comments on our study, a nested trial in the case-control study “NOHA big first,” the results of which are not yet published.
The aim of the Nîmes Obstetricians and Haematologists (NOHA) big first, which began on January 1, 1999, was to clarify the risk factors for a first pregnancy loss.
Obstetricians and gynecologists from the southern French region of Languedoc sent us patients who experienced a pregnancy loss during their first or second pregnancy. Our multidisciplinary staff selected these cases of unexplained pregnancy loss, and 97.8% of them were investigated at the outpatient department of hematology. Patients with an obvious medical etiologic factor (1.15%) or thromboembolic antecedents (3.4%) were excluded, and 98.8% of the residual patients agreed to join the case-control study (n = 3496). There were 186 patients who were heterozygous for the factor V Leiden mutation (FVL), 127 for the prothrombin 20210G>A mutation (F2M), and 31 with a protein S insufficiency (PS). Of them, 69% (who were negative for classical antiphospholipid antibodies and hyperhomocysteinemia) were potentially eligible for our therapeutic trial (FVL: n = 127; F2M: n = 85; PS: n = 25; total: n = 237).
The therapeutic trial began January 1, 2001, and ran until January 1, 2003. Participation in this trial was proposed to the aforesaid patients with pregnancy loss from the 10th week (n = 94) who had not initiated a new pregnancy (n = 88), and to the 96 consecutive new eligible patients. Informed consent to participate was obtained from 180 of these patients; 174 conceived during the 3 following months. The 160 women with an ongoing pregnancy at the eighth week were definitively included.
Previous testing for protein Z and anti-protein Z antibodies helped to define, with the 3 main thrombophilic factors, 12 subgroups, and to stratify the 180 informed patients accordingly to them. Pairs of patients were defined as 2 consecutive patients, belonging to the same subgroup. Treatments were randomly and blindly proposed by an independent statistician to the first patient of a given pair and the second patient received the alternative treatment.
No differences, between the 2 treatments, concerning the age of the patients (median, 26 years; range, 18-36 years in the low-molecular-weight heparin [LMWH] group vs 26 years; range, 18-35 years), the weight (median body mass index [BMI], 24.2, range, 21-32 vs 24.1; range, 20.8-32.1), women older than 30 years (12/80 vs 10/80), or obese women (BMI 30 or greater, 5 vs 3) could be evidenced.
All treated patients received the specified treatment, with an excellent compliance, as evaluated by regular questioning, counting of syringes that had been used, or blisters. No patients were lost to follow-up. The primary outcome of the study was analyzed in all included patients. We did not perform any interim analyses.
Finally, we are convinced that roughly 10% of the women with a first pregnancy loss from the 10th week should be treated with prophylactic LMWH and folate intake.
Correspondence: Jean-Christophe Gris, Laboratoire d'Hematologie, Centre Hospitalier Universitaire (CHU), Groupe Hospitalo-Universitaire Caremeau, Place du Pr. Robert Debré, F-30029, Nîmes Cédex g, France; e-mail: jcgris@chu-nimes.fr.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal