Prospective study needed for confirmation
He offered a hypothetical case. A 32-year-old woman with prior severe preeclampsia who delivered at 32 weeks asks you, “Should I be treated with LMWH in my next pregnancy?” What should you tell her? To answer this question, Dr. Rodger and his associates conducted a study-level meta-analysis of six randomized controlled trials that included 848 pregnant women with prior placenta-mediated pregnancy complications (Blood. 2014;123[6]:822-8). The primary objective was to determine the effect of LMWH in preventing placenta-mediated pregnancy complications in women with prior late placenta-mediated pregnancy complications. This included patients with or without thrombophilia who were treated with or without LMWH. The primary outcome was a composite of preeclampsia, birth of an SGA newborn, placental abruption, or pregnancy loss greater than 20 weeks. Overall, 67 (18.7%) of 358 of women being given prophylactic LMWH had recurrent severe placenta-mediated pregnancy complications, compared with 127 (42.9%) of 296 women with no LMWH (relative risk reduction, 0.52; P = .01, indicating moderate heterogeneity). They identified similar relative risk reductions with LMWH for individual outcomes, including any preeclampsia, severe preeclampsia, SGA below the 10th percentile, SGA below the 5th percentile, preterm delivery less than 37 weeks, and preterm delivery less than 34 weeks with minimal heterogeneity. They concluded that LMWH “may be a promising therapy for recurrent, especially severe, placenta-mediated pregnancy complications, but further research is required.”
At the meeting, Dr. Rodger noted that the positive studies in the analysis were single-center trials, “which are generally acknowledged to be of a lesser methodologic quality, and the majority of patients in these single-center trials are from a small area in the south of France. Multicenter trials don’t show an effect, so is it single-centeredness or is it something else? The other feature that’s distinct is that the positive trials recruited patients with prior severe complications only, while the negative trials included patients with nonsevere complications. So maybe LMWH works in patients who have a very strong phenotype that have had very bad prior complications. We can’t tease that out with a study-level meta-analysis because we’re getting average effects over heterogeneous groups of patients.”
To expand on the study-level meta-analysis, Dr. Rodger and his associates conducted a systematic review and individual patient data meta-analysis of eight randomized trials of 963 patients conducted between 2000 and 2013 of LMWH to prevent recurrent placenta-mediated pregnancy complications (Lancet. 2016;388:2629-41). “In this approach you get individual patient data from the trials, and you create a new randomized, controlled data set,” he explained. “That way we could tease out the patients who have had the prior severe complications and whether their mild or severe outcomes are being prevented or not.”
The study’s composite primary outcome was one or more of the following: early-onset or severe preeclampsia, SGA newborn below the 5th percentile, late pregnancy loss (over 20 weeks), or placental abruption. Dr. Rodger and his associates found that LMWH did not significantly reduce the risk of recurrent placenta-mediated pregnancy complications, compared with patients who did not receive LMWH (14% vs. 22%, respectively; P = .09). In subgroup analyses, however, LMWH in multicenter trials reduced the primary outcome in women with previous abruption (P = .006) but not in any of the other subgroups of previous complications. “There were small numbers of patients in this subgroup, though, so I would use caution,” Dr. Rodger said. Two recent randomized, controlled trials from separate investigators further support the overall null findings of the individual patient data meta-analysis (Obstet Gynecol. 2016;128[5]:1053-63 and Am J Obstet Gynecol. 2017 Mar;216[3]:296.e1-296.e14).