Current Guidelines for Anticoagulation in Pregnant Women with MHVs
The 2008 American College of Cardiology/American Heart Association (ACC/AHA) guidelines state that there are insufficient grounds to make definitive recommendations about optimal antithrombotic therapy in pregnant patients with MHVs, because properly designed studies have not been performed [Bonow et al., 2008]. Generally, both the ACC/AHA and the European Society of Cardiology (ESC) [Regitz-Zagrosek et al., 2011] guidelines recommend discussing the risks of available anticoagulation regimens with the pregnant patient. Antithrombotic preventive therapy options during pregnancy include continuation of VKAs throughout the second trimester of pregnancy as well as dose-adjusted sc or iv UFH between the 6th and the 12th week or throughout pregnancy with an activated partial thromboplastin time (aPTT) at least twice the control level. The ACC/AHA guidelines include the option of LMWH instead of UFH with peak anti-Xa factor levels between 0.7 and 1.2 U/ml 4 h after administration. The American College of Chest Physicians Conference (ACCP) on antithrombotic and thrombolytic therapy concluded that it is reasonable to use one of the following three regimens: (i) either LMWH or UFH between 6 and 12 weeks and close to term only, with warfarin used at other times; (ii) aggressive dose-adjusted UFH throughout pregnancy; or (iii) aggressive adjusted-dose LMWH throughout pregnancy aiming to attain a peak anti-Xa levels of 0.7–1.2 U/ml at 4–6 h after administration [Bates et al., 2012]. The authors recommend that the decision has to be completely individualized. Women of childbearing age and pregnant women MHVs should be counseled about potential maternal and fetal risks associated with various anticoagulant regimens. In women at very high risk of thromboembolism in whom concerns exist about the efficacy and safety of UFH or LMWH, the guidelines suggest VKAs throughout pregnancy with replacement by UFH or LMWH close to delivery rather than one of the regimens above (Grade 2C). However, those women who place a higher value on avoiding fetal risk than on avoiding maternal complications are likely to choose LMWH or UFH over VKAs. In addition to VKAs or heparins, low dose aspirin is recommended for high-risk group of pregnant women with MHVs.
UFH or LMWH throughout pregnancy is not recommended by the recent ESC guidelines, considering continuation of VKAs throughout pregnancy when the warfarin dose is <5 mg daily [Regitz-Zagrosek et al., 2011]. Discontinuation of VKAs and a switch to UFH or LMWH is recommended between weeks 6 and 12 under strict dose control and supervision. When a higher dose of VKAs is required, discontinuation of VKAs between weeks 6 and 12 and replacement by adjusted-dose UFH (aPTT ≥ 2 times the control, in high-risk patients applied as an intravenous fusion) or LMWH twice daily (dose adjusted according to weight) is recommended (the anti-Xa level should be maintained between 0.8 and 1.2 U/ml 4–6 h after application).
Published Data and Recommendations on Anticoagulation Regimens in Pregnant Patients with Prosthetic MHVs
In the absence of controlled clinical trials, current recommendations are based on limited, observational data. Unfortunately, only a few, mostly small, series [Iturbe-Alessio et al., 1986; Sbarouni and Oakley, 1994; Vitale et al., 1999; Al-Lawati et al., 2002; Geelani et al., 2005] comprise the basis from which current guidelines and recommendations are derived [Bonow et al., 2008; Regitz-Zagrosek et al., 2011]. Maternal mortality in patients with MHVs remains the most devastating complication and even contemporary series confirm that mortality and complications may not necessarily be avoided [Born et al., 1992; Sbarouni and Oakley, 1994; McGehee, 1998]. Chan et al.  evaluated maternal and fetal outcomes according to the type of anticoagulation used during pregnancy: VKAs alone, VKAs with UFH during the first trimester, UFH throughout pregnancy, and antiplatelet agents alone. Rates of maternal thromboembolic complications in women who received UFH alone, VKAs with heparin substitution during the first trimester, and warfarin alone were 33.3%, 9.2%, and 3.9%, respectively. However, the rates of congenital fetal anomalies were 0%, 3.4%, and 6.4%, respectively. These results suggest that heparin alone is insufficient to prevent thromboembolism among pregnant women with MHV compared with VKAs regimens. Compared with regimens using warfarin alone, substitution with UFH during the first trimester was associated with a reduction in embryopathy from 6.4% to 3.4% but at the same time with an increase in maternal thromboembolic risk from 3.9% to 9.2%. Vitale et al.  suggested that the risk of fetal damage was reduced although not eliminated, if the daily warfarin dose was below 5 mg. Several reports however described the development of warfarin embryopathy and fetal loss even with low dose warfarin [Meschengieser et al., 1999; Sadler et al., 2000; Finkelstein et al., 2005; Khan, 2007]. For this reason, one cannot guarantee any fetal complications even to women who are well anticoagulated on a <5 mg/day of warfarin during pregnancy. The risk of fetal malformation and other effects associated with use of VKAs throughout pregnancy suggests that these drugs should only be considered in women with a very high risk of thrombosis, such as highly thrombogenic MHVs or a history of thromboembolic complications on a therapeutic dose of heparin [Elkayam and Bitar, 2005; Bates et al., 2012]. Supporting the published review by Chan et al. , Sillesen et al.  more recently described 79 women who had 155 pregnancies after valve replacement. Two women died during pregnancy, one from heart failure and one from postpartum bleeding. There were four thromboembolic episodes in the early study period in women with a mitral prosthesis on UFH. Detailed information on the UFH regimen and monitoring results at the time of the episodes were not available. Owing to lack of information related to the level of anticoagulation and its monitoring, these reports might suggest resistance to moderate doses of UFH in high-risk women with old-generation MHVs. For this reason, UFH should be avoided if all possible. If no other choice is available in a woman who prefers not to use warfarin, a high dose UFH should be preferably administered as an intravenous continuous infusion and at high dose [Elkayam and Bitar, 2005; Bonow et al., 2008] and adjusted to achieve an aPTT ratio of >2.5 control value with very careful maintenance of the central line to prevent infection and a risk of endocarditis. If continuous iv infusion is not feasible, a high dose sc UFH (7500 to 20 000 U every 12 h) should be used aiming to achieve a mid level (6 h) of aPTT ratio of >2.5 control value.
Therapy with LMWH in pregnancy is an attractive and convenient alternative to VKAs and UFH. Substantial evidence shows the efficacy and safety of LMWH in prevention and treatment of thromboembolism during pregnancy in patients with evidence of deep vein thrombosis (DVT) and thrombophilia [James, 2009], and there has been an increasing experience with the use of this therapy in women with MHVs. Earlier published data on the use of LMWH in women with MHVs during pregnancy was described by Elkayam et al.  and was limited to small groups of patients or to isolated reports with several of these cases complicated by valve thrombosis and even death. A careful review of the reported cases, however, indicated that most, if not all, of these cases were associated with an inadequate dose, lack of monitoring, or subtherapeutic anti-Xa levels [Arnaout et al., 1998; Berndt et al., 2000; Lev-Ran et al., 2000; Rowan et al., 2001; Consensus Reports, 2002; Aventis Pharmaceuticals, 2004; Oran et al., 2004; Bates et al., 2012]. A more recent review by Oran et al.  comprised of 81 pregnancies with MHVs in whom LMWH was used reported 10 thromboembolic events in women with mechanical mitral valves, of which nine occurred in the 30 pregnancies with a fixed LMWH dose and only one in the 51 pregnancies with adjusted LMWH dose. Rowan et al.  reported on their experience in 14 pregnancies in women with MHVs who were treated with LMWH [Rowan et al., 2001], valve thrombosis was described in one patient who had a St Jude mitral valve and had stopped warfarin 3 months before conception. She presented at 8 weeks' gestation, on no anticoagulation treatment, with transient ischemic attacks and suspected thrombus on transesophageal echocardiography (TEE). The patient was started on enoxaparin, with apparent resolution of the thrombus on TEE but re-presented at 20 weeks' gestation after further transient ischemic attack due to subtherapeutic level of anticoagulation (peak anti-Xa level was 0.62 U/ml). Abildgaard et al.  reported on 12 pregnancies with MHVs treated with LMWH, in which thromboembolism occurred in two women with aortic MHVs. Both events were attributed to subtherapeutic doses of LMWH during the initial 3 weeks of pregnancy. Quinn et al.  conducted a prospective audit of the use of adjusted dose LMWH in 12 pregnancies with MHV. LMWH +/− low-dose aspirin was started at therapeutic-dose with monitoring of anti-Xa levels to achieve a target level of 1.0–1.2 U/ml. This necessitated a mean increase in the dose of LMWH of 54.4% over initial dose. One nonfatal valve thrombosis occurred at 26 weeks gestation associated with subtherapeutic anti-Xa levels, and three patients experienced major bleeding. McLintock et al.  reported thromboembolic complications in seven out of 47 pregnancies, of which five were thought to be associated with the use of enoxaparin therapy. Similar to reports by other investigators described above, poor compliance with therapy and subtherapeutic peak anti-Xa levels was an issue in all cases. No thromboembolic complications occurred in the 20 pregnancies when enoxaparin was commenced before 6 weeks' gestation, a group that was compliant with medication and monitoring of peak anti-Xa levels. Based on the published data, there is increasing experience with the use of LMWH for anticoagulation in pregnant women with MHVs, most, if not all, the cases reported to develop thromboembolic complications were related to poor compliance with therapy, inadequate monitoring and subtherapeutic levels of anticoagulation. The pharmacokinetics of LMWH are altered during pregnancy with lower plasma concentrations, probably related to higher clearance and volume of distribution [Casele et al., 1999]. A recent review by McLintock  has summarized the maternal and fetal complications in 92 women from five prospective cohort studies treated with dose-adjusted LMWH throughout pregnancy. Nine episodes of valve thrombosis were reported and were attributed to poor compliance or suboptimal LMWH doses in the majority of the cases. However, Yinon et al.  reported a case of valve thrombosis in a woman with a Medtronic Hall aortic valve replacement who presented with a transient ischemic attack at 24 weeks' gestation in spite of guideline recommended levels of peak anti-Xa levels and a history of compliance with therapy. Trough levels of anti-Xa however, were not measured. This woman had a new generation mechanical aortic prosthesis and achieved therapeutic peak anti-Xa levels (1.0–1.4 U/ml) and was treated with warfarin until 5 weeks of gestation and then was switched to LMWH and aspirin. At 24 weeks of gestation she presented with a transient ischemic event (peak anti-Xa level was 0.99 U/ml). Echocardiography demonstrated an elevated mean gradient across her aortic valve and on TEE no thrombus was seen, but the aortic valve leaflets were not optimally visualized. The LMWH dose was increased, but at 26 weeks of gestation, the patient was admitted with cardiac arrest and died. The autopsy demonstrated aortic valve thrombosis. This case demonstrates the potential limitations of reliance on guidelines recommended peak anti-Xa levels and the need to also assure a therapeutic trough level as well. This suggestion was initially made by Barbour et al. , who evaluated 138 peak and 112 troughs anti-Xa levels in 13 pregnancies in 12 patients. With peak levels of 0.63, 0.70, and 0.69 U/ml, at the 1st, 2nd and 3rd trimesters, respectively, mean trough level were 0.21, 0.30, and 0.40 U/ml with only 9% of the measurements >0.5 U/ml. Even when peak levels were between 0.75 and 1.0 U/ml, only 15% of trough levels were >0.5 U/ml. [Barbour et al., 2004]. Similarly, in a recent series of 15 pregnant women at different gestational ages, a subtherapeutic anti-Xa level was demonstrated in 20% of the peak levels and 73% of the trough levels despite “therapeutic” enoxaparin administration [1 mg/kg bid; Friedrich and Hameed, 2010]. In an unpublished study on 26 pregnant women who received anticoagulation with LMWH for various indication including nine patients with MHV s/c q12 h, we analyzed both through and peak anti-Xa levels throughout pregnancy for a total of 177 determinations [Fan et al., 2011]. Adjusted dose LMWH achieving guidelines recommended peak anti-Xa between 0.7 U/ml and 1.2 U/ml were associated with subtherapeutic trough anti-Xa levels in about 50% of cases. On the other hand, therapeutic trough anti-Xa levels 0.6 U/ml–0.8 U/ml were rarely associated with excessive peak anti-Xa levels.
Supporting the use of both peak and trough for adequate anti-Xa level monitoring, Vijayan and Rachel  recently published a longitudinal observational study of 13 pregnancies in a cohort of five women with MHVs (four mitral and one aortic, Sorin, St Jude and Caromedics) from 2006 to 2010 comparing outcomes of pregnancies on warfarin and LMWH managements [Vijayan and Rachel, 2012]. Eight pregnancies prior to 2007 were anticoagulated with warfarin that was given at 14 through 36 weeks gestation followed by heparin. Dose of warfarin had to be increased in pregnancy and apart of one woman who perpetually required high prepregnancy doses of 7 mg, all others required dose increments of 40–50% in pregnancy, and all case treated with warfarin ended with fetal loss. The five pregnancies since 2007 anticoagulated with the LMWH (enoxaparin) using peak and trough anti-Xa level monitoring resulted in livebirths after normal vaginal deliveries. Coagulation profiles were monitored on a weekly basis until therapeutic targets were achieved. A trough anti-Xa level of 0.7/ml and a peak level of 1.0–1.2/ml were targeted for enoxaparin and monitoring was continued 2–3 weekly once targets were achieved. All women were also prescribed low dose aspirin of 75 mg daily as an adjunct. In addition to this series, Schwartzenberg et al.  described a case of a 33-year-old high-risk pregnant woman with MHV and four previous miscarriages and endocarditis on previous warfarin and heparin treatments who presented with mitral valve thrombosis during early pregnancy. Resolution of the thrombus and eventually resumption of normal prosthetic mitral valve function and normal delivery was obtained through LMWH treatment with close peak and trough anti-Xa level monitoring. The available data [Fan et al., 2011; McLintock, 2011; Goland and Elkayam, 2012; Vijayan and Rachel, 2012; Schwartzenberg et al., 2013] in addition to the documented risk of valve thrombosis with subtherapeutic predose anti-Xa levels suggest the importance of routine measurement and maintenance of trough levels at therapeutic range of anti-Xa ≥0.6 in low risk and ≥0.7 U/ml in high-risk patients (Table 1) [Elkayam and Goland, 2012]. Because of possible bleeding complications [Quinn et al., 2009], peak levels should also be monitored to prevent excessive anticoagulation (anti-Xa levels >1.5 U/ml), in which case, an 8-hourly rather than a12-hourly dosing should be used. To ensure patient compliance and adequate prophylaxis, anti-Xa activity should be measured once weekly for the first 4 weeks and at least once every 2 weeks for the duration of treatment. Catheter placement for epidural anesthesia is not advisable within 10–12 h of the last dose, because of longer half-life of LMWH [Elkayam et al., 2004]. For this reason, and to prevent spinal or epidural hematoma, LMWH should be withdrawn 18–24 h before an elective delivery and substituted with intravenous UFH. Because of the potential added benefit, a small dose of aspirin (75–100 mg/day), which is safe during pregnancy [Elkayam et al., 2004; Bonow et al., 2008], might be added in high-risk patients to further reduce the incidence of thromboembolism.
Table 1. Recommended Approach to Anticoagulation Therapy for Women with Mechanical Prosthetic Heart Valve
|Higher risk||Lower risk|
|Old-generation MPHV in mitral position, MPHV in tricuspid position, atrial fibrillation, history of TE on heparin||New-generation MPHV in mitral position and MPHV in aortic position|
|Warfarin (INR 2.5–3.5) for 35–36 weeks followed by IV UFH (aPTT > 2.5) to parturition + ASA 81–100 mg/day||LMWH SQ Q12 h (trough anti-Xa ≥0.6 IU/ml, peak anti-Xa <1.5 IU/ml) to 35–36 weeks, then UFH IV (aPTT >2.0) to parturition|
|LMWH SQ Q12 h (trough anti-Xa ≥0.7 IU/ml, peak anti-Xa <1.5 IU/ml) or UFH SQ Q12 h or IVa(mid interval aPTT >2.5) for 12 weeks, followed by warfarin (INR: 2.5–3.5) to 35–36 weeks, then UFH IV (aPTT >2.5) to parturition + ASA 81–100 mg/day.||LMWH SQ Q12 h (trough anti-Xa ≥0.6 IU/ml, peak anti-Xa <1.5 IU/ml) or UFH SQ Q12 h or IVa(midinterval aPTT >2.0) for 12 weeks followed by warfarin (INR: 2.5–3.0) until 35–36 weeks, then UFH IV (aPTT >2.0) to parturition|