IN DEPTH

Hypercoagulability and Stillbirth - Explaining the Unexplained?

Richard M. Pauli, M.D., Ph.D.

Introduction

When stillborns are assessed in the usual manner, less than half will have an identifiable cause for their death. What about the others? Over the past four years an extraordinary story has begun to unfold that suggests a common cause for many of those instances of intrauterine death that before now remained unexplained. It appears that specific, genetic predispositions to hyper-coagulability may predispose to (and at least in that sense cause) many stillbirths. While the data are preliminary, I suspect that soon all of us will need to modify our thinking about those instances where no malformational or obvious cord or placental process accounts for death, and about what assessment is warranted following intrauterine death.

Hemostasis 101

Clotting is a complicated process. It involves not only coagulation generated by soluble proteins within the blood, but also, of course, the function of platelets and interactions with the endothelium of blood vessels. However, for our purposes the most important part of hemostasis is the clotting cascade. This cascade (Figure 1 shows a simplified representation) is a beautiful example of a multi-step process that allows for various feedback control and, through that, exquisite balance between clotting essential for hemostasis and the prevention of abnormal thrombosis. Each coagulation factor circulates in an inactive form but is activated as shown in Figure 1 (where the active factors are followed by a lower case ‘a’).

Fig. 1, THE CLOTTING CASCADE

Less well known (at least to those who went to school as long ago as did I) are the specific inhibitors of the coagulation cascade. Three of these are well delineated and are important in discussing hypercoagulability states. Antithrombin III inhibits the function of factors IXa, Xa and thrombin through direct binding to these proteins. Protein C inactivates both factor VIIIa and factor Va, while protein S acts as a co-factor in this action of protein C (Figure 2). One might anticipate that defects in the function of any of these three could result in hypercoagulable states, since one of the balancing components has been lost. That is, indeed, the case.

Figure 2, SITES OF ACTION CAUSING HYPERCOAGULABILITY

Genetic mutations that can predispose to hypercoagulability

Abnormalities of Protein C, protein S or antithrombin III were first imputed as being of importance in adults with recurrent deep vein thrombosis and, in particular, in families in which multiple members were so affected. Over time is has become apparent that partial deficiency of these three proteins does impart increased risk for thrombotic phenomena (while complete deficiency secondary to homozygosity can result in much more severe problems). What is imparted, however, is a predisposition. All of these mutations resulting in partial deficiency are quite common. Indeed best estimates of prevalence are, for protein C about 1/200, for protein S about 1/80, and for antithrombin III about 1/600. They are all more common in those experiencing recurrent thromboses; yet some with these abnormalities remain completely healthy.

In addition to these deficiencies of anti-thrombophilic proteins, there are two mutations of proteins of the clotting cascade that increase their activity and thereby also can predispose to thrombophilia and recurrent thrombosis. One is called Factor V Leiden (obviously a change in function of factor V) and is extraordinarily common — being present in about 1/12 people. The other is a specific mutation of prothrombin which isn’t quite so frequent, but is common nonetheless, present in around 1/50. Translating all of these numbers into another more digestible form, for every 1000 people hypercoagulability on these genetic bases will be found:

in 80 secondary to Factor V Leiden
in 20 secondary to Prothrombin mutation
in 12 because of protein S deficiency
in 5 because of protein C deficiency
and in 2 with antithrombin III deficiency; or, nearly 1 in 8 overall!

Fascinating is the recent demonstration of the interaction of these risk factors in predisposing to thrombosis. That is, people with more than one deficiency (as might be expected given how frequent each is) have much greater risk than those with only one. Likewise, these risk factors interact with complicated environmental predispositions.

Another interactive genetic factor has also recently been identified (and will be of some import in understanding thrombophilia associated stillbirth). An enzyme called methylenetetrahydrofolate reductase (MTHFR) is important in homocysteine metabolism. Deficiency of this enzyme causes a rise in homocysteine levels in blood. And, there is unequivocal evidence that elevation of homocysteine in blood is correlated with thrombotic cardiovascular complications (mitigated by supplementation with folic acid). More recently an exceedingly common thermolabile variant of MTHFR was described; it is this thermolabile mutation that may have some relevance to the contribution of thrombophilia to intrauterine fetal deaths.

Environmental factors that can contribute to hypercoagulability

A number of environmental characteristics are known to predispose to thrombosis — immobility, increasing age, obesity etc. More important in this discussion are three additional factors: Pregnancy, per se, results in hypercoagulability. Low folate levels are imputed in increased coagulability too, and folate demands increase dramatically in the latter stages of pregnancy. Antiphospholipid antibodies are a third factor that seems of particular importance in pregnancy-associated thrombotic events.

Why might hypercoagulability cause stillbirth?

For a long time it has been recognized that stillbirth can be associated with various placental changes that suggest a hypercoagulable state. Sometimes, at least, thrombosis almost certainly does lead to abnormal placental perfusion and directly causes intrauterine death. Placental markers of particular importance include, grossly, multiple infarctions, and, histologically, thrombosis of spiral arteries, ischemic necrosis and perivillous fibrin deposition.

Such changes are markers, it is assumed, of insufficient perfusion which, in turn, can precipitate at its most extreme a life-taking series of events (perfusion changes in the fetus, hypoxia, and, probably ultimately, cardiac failure as the proximate cause of death).

Why such placental changes arose often remained enigmatic — until now!

Genetic mutations and the causes of stillbirth

Recently, two very important research articles have appeared that suggest that mutations related to the coagulation cascade may be central in explaining a large proportion of currently unexplained intrauterine deaths.1,2 The first was published in January, the second in June. Each has its strengths and weaknesses, but they are both so important that I want to review each in considerable detail.

The first, by Kupferminc et al.1, actually addresses a variety of pregnancy complications (abruption, pre-eclampsia, intrauterine growth retardation) in addition to stillbirth. All of these have been linked to poor placental perfusion, and so the authors postulate that each might be associated with a maternal predisposition to thrombophilia. Thus, they assessed women with any of these complications for abnormalities of protein S, protein C or antithrombin III, for factor V Leiden, for the prothrombin activating mutation and for homozygosity for the thermolabile mutation of MTHFR. An appropriate, equal sized control group was also assessed. Overall 71 of 110 women with some obstetrical problem had at least one abnormality compared with 20 of 110 who had had normal pregnancies — a highly significant difference. Each of the thrombophilic factors was statistically significantly associated with these complications (taken as a whole). Assessment of their contribution to fetal death was limited since only 12 of the 110 in the cohort had had stillborns (here defined as > 23 weeks gestation). Seven of 12 had a thrombophilic abnormality. While too small to lead to any sure conclusions, this rate (0.58) is much greater than that in the control group (0.18). Factor V Leiden was the most common in women with a prior stillbirth, present in three.

Obviously this study is limited by the small number of stillborns included. There are some other potential problems. First, this study, done in Israel, included only Jewish women; so whether this can be generalized is a concern. Secondly, although described as consecutive, I find it hard to imagine that some women would not decline participation — making me suspect that the character of the non-participants is simply not addressed.

The second study2 is considerably more exciting. This French investigation specifically addresses thrombophilic changes associated with late fetal loss. This, too, was a case-control study. Here 232 instances of women experiencing stillbirth (> 22 weeks gestation) and 464 controls were evaluated in a manner quite similar to the previous study. Not including antiphospholipid studies (also assessed here), the following was found:

Thrombophilic Abnormalitly	Patients	Controls

Antithrombin III	0 %	0 %
Protein C Deficiency	0.4 %	0.4 %
Protein S Deficiency	4.7 %	0.2 %
Factor V Leiden	6.5 %	1.3 %
Prothrombin Activating Mutation	0.9 %	1.0 %
MTHFR Thermolabile [homozygous]	15.7%	21.1%

Independently, then, only protein S deficiency and factor V Leiden were correlated with stillbirth. However, most interestingly, while MTHFR frequency alone was no greater, in those with homozygosity for the MTHFR mutation plus an additional thrombophilic risk factor, and in the absence of supplementation with folic acid, the risk for stillbirth was 100% (28/28)!

This study also confirmed that those with the assessed risk factors and stillbirth virtually always showed maternal vascular markers when the placenta was thoroughly evaluated. That finding is important in thinking about a protocol for assessment.

This investigation, too, has some substantial problems. These are referred patients, not really as stated, a consecutive series. The bias this results in is evidenced by the fact that 84 of 232 women (36%) had more than one stillbirth — a huge proportion.

So, how convincing are these studies? It seems unquestionable that genetic thrombophilic disorders substantially increase the risk of stillbirth. Combined with a host of other recent studies3-14 there is now convincing evidence that such factors are, in fact, very important contributors to unexplained fetal death. However, no accurate estimate of the proportion of unexplained stillbirth explicable on this basis is yet available. To me the summed experience suggests that perhaps as much as 20-25% of all stillbirth may have this basis.

What should be done?

Should we begin testing for the six thrombophilia risk factors in all women who deliver a stillborn? The answer is not yet really clear. First, association (as demonstrated in the summarized studies) does not necessarily imply causality. Secondly, these abnormalities are sufficiently frequent in the general population (particularly the MTHFR mutation) that false positives are inevitable. Third, no good data yet exist regarding recurrence risk implications for those in whom different abnormalities and combinations of abnormalities are discovered. Finally, while certain therapy (below) might be considered, efficacy is uncertain.

Further study is needed. We are currently considering incorporating such a study (at no cost to patients) into the stillbirth protocol; others, I am sure, will pursue various approaches to answer these questions, too. In the meantime it is probably premature to recommend testing after all stillbirth.

How might these problems be treated?

In the near future, if such testing becomes routine, or now, if one elects to undertake such evaluation, certain abnormalities probably will commend certain therapies in a next pregnancy.

First, in those with homozygosity for the MTHFR mutation and any other predisposing factor (protein S deficiency, protein C deficiency, antithrombin III deficiency, the thrombophilic prothrombin mutation, factor V Leiden, or antiphospholipid antibodies), high dose folate supplementation beginning preconceptually may have a protective effect. Further, in those who have had one or more stillbirths and are found to have a thrombophilic mutation (not including MTHFR homozygosity in isolation) heparin and low dose aspirin may well be appropriate. Time (but probably not a lot of time) will tell.

Vol. 6 No.1 and 2

December, 1999