Pregnancy hypertension

Pregnancy hypertensionIntroduction:

Gravitational hypertension is a group of conditions whose nature and mechanisms remain imperfectly classified.

This situation affects 10 to 15% of pregnant women, which is considerable. It also represents, in our climate, the leading cause of perinatal morbidity and mortality.

“Illness of Hypotheses”, preeclampsia was also a “disease of dogmas” and many successive and contradictory truths were defended with a rare passion, in defiance of any scientific support. Over time, however, pathophysiological research has made significant progress and essential achievements have been achieved. When we wrote the previous edition of this general review, some scattered pieces of a new puzzle could be presented. Today, the puzzle is certainly not complete, but an overall logic is emerging more and more clearly.

Definitions and classification:

The term of gestational hypertension brings together very disparate entities. We consider successively the clinical facts of observation, then the classifications that are proposed.


Hypertensive disorders of pregnancy revolve around two main symptoms, hypertension and proteinuria. The third classic symptom, oedemas, is today abandoned in classifications.


The definition of hypertension during pregnancy is not as clear as in other circumstances, since blood pressure (Pa) drops physiologically in early pregnancy. A diastolic Pa greater than or equal to 90 mmHg in at least two successive measurements separated by at least 4 hours is the usually accepted criterion. The old definition based on an increase of 30 mmHg or more in two successive exams is no longer used today. Systolic Pa, much more labile in pregnant women, is a fragile criterion. Nevertheless, the National High Blood Pressure Education Program’s (NHBPEP) latest recommendation, which a Task Force on Hypertension in Pregnancy released in 2000, states values ​​of 140 mmHg for systolic or 90 mmHg for the diastolic. We regularly refer to this recommendation to the extent that it is authoritative.

Measurements of Pa are delicate in pregnant women because of their lability (remember that the cardiac output is increased by 30%). It is essential to practice these measurements on a subject as relaxed as possible, and at a distance from the gynecological examination. The most used position is the sitting position, after a few minutes of calm and conversation. Endless debates concern the choice of phase IV or V of Korotkoff. The latter is currently favored, but not unanimous.

The tension figures are very variable in the same subject, for this reason measurements must be iterative. In this variability comes the stressor, whose participation can be roughly estimated by measuring the heart rate. But an important factor of variation is also introduced by the nycthemeral rhythm, very marked, but also inverted during hypertensions, with a nocturnal maximum.

Ambulatory Blood Pressure Measurement (ABPM) is not recognized as a diagnostic criterion. In some cases, it can nevertheless help to recognize the so-called “white coat” hypertension. Normality values ​​in pregnancy are almost established. No predictive value has been attributed to it so far.


Proteinuria is also defined very differently. In obstetrical practice (especially across the Atlantic), its quantification is often limited to a number of “crosses” to the strip, estimate tainted by many errors. A collection of 24 hours is unfortunately rarely done. Proteinuria is said to be “significant” if it exceeds 1 g / L on a sample or 0.3 g on urine of 24 hours.

Such proteinuria is superadded to hypertension in up to 10% of cases. It does not precede it, but follows it almost always, constituting the picture of preeclampsia. The few exceptions to this rule usually reveal previously unrecognized nephropathies. This proteinuria is glomerular and has a predominant albuminuria.


This third element of the symptomatic triad characterizing preeclampsia no longer enters a pathological definition today. In fact, edema occurs at one time or another in 80% of normal pregnancies.

Nevertheless, diffuse oedemas, affecting the lower limbs, but also the hands (sign of the ring) and the face, can be a warning sign, especially if they are major and brutal constitution.


The onset of hypertension during pregnancy is not unambiguous. A first rational classification was published in 1972 under the auspices of the American College of Obstetricians and Gynecologists (ACOG); it has subsequently undergone minor changes. Another classification was proposed in 1988 by a committee of the International Society for the Study of Hypertension of Pregnancy (ISSHP). The most recent one is the NHBPEP, of which we have already mentioned the recent report.

It is this that we summarize here, for reasons of actuality rather than reliability or novelty. In reality, all revolve around the same terms and have the same weaknesses.

NHBPEP classification:

This classification separates pregnancy hypertension into four broad categories.

– Chronic hypertension: this is hypertension that is present before pregnancy, or found before the 20th week of pregnancy. The threshold value of definition of hypertension is 140/90 mmHg. Any hypertension found during pregnancy that does not disappear postpartum falls under the same heading.

– Pre-eclampsia-eclampsia: it is a specific syndrome of the gravid state. It appears most often after the 20th week and associates hypertension and proteinuria, according to the threshold values ​​indicated above.

The authors recognize that threshold values, the specificity of this definition is mediocre. “Certainty level of diagnosis” is higher for systolic BP of 160 mmHg or greater, diastolic BP of 110 mmHg or greater, proteinuria of 2 g / 24 h or more, creatinine of 12 mg / L or greater, thrombocytopenia , headache or visual disturbances, epigastric bar pain.These criteria actually define the severe forms of preeclampsia, with a particularly high risk of maternal and / or fetal accidents.

Eclampsia consists of the occurrence, in a preeclamptic woman, of convulsions without any other individualizable cause.

– Pre-eclampsia added: it is the appearance of a significant proteinuria in a woman with chronic hypertension.

The prognosis then matches that of preeclampsia. The same diagnosis is admitted in the event of a sudden increase in previously safe hypertension, thrombocytopenia or hepatic cytolysis.

– Gestational hypertension (or pregnancy): this is a hypertension noted for the first time after the 20th week.

Since proteinuria can always appear secondarily, this diagnosis is definitively established only postpartum. If hypertension also regresses completely within 12 weeks after delivery, it is transient hypertension of pregnancy.

Finally, we mention the particular table (not included in this classification) of proteinuria without hypertension, or at least largely preceding it. Discrete proteinuria may be the only physiological increase in glomerular filtration.Proteinuria greater than 1 g / 24 h is most likely a result of autonomic nephropathy, discovered during pregnancy.

Confused situation:

If the definitions given above are supposed to be authoritative, they can not satisfy the clinician or the researcher. The literature makes the rest of widely divergent definitions that make studies difficult to compare with each other.Moreover, various national scientific societies have established their own classification and defined their own criteria, which generates considerable confusion.

It is true that the same term refers to situations whose gravity can be very different. Classifiers acknowledge that they have searched for “minimalist” definitions to avoid default diagnostic errors, even if they are excessive. This attitude suggests “fixes”.

Thus, we speak of “severe” hypertension if:

– the diastolic Pa is measured only once at 120 mmHg or more;

– or if it is measured at 110 mmHg or more on at least two separate occasions of more than 4 hours.

Similarly, preeclampsia is separated into two forms, “mild” and “severe”. The distinction is important because if the prognosis of the first is not too far from that of a simple gravidic hypertension, that of the second is extremely serious and requires immediate therapeutic measures, coupled with the use of a Level 3 maternity. Unfortunately, there are no strict criteria for drawing the line between these two forms, and everyone can therefore refer to their own personal judgment.

In this disorder, the NHBPEP report admits that there should be a definition reserved for research (taking into account only the major forms of unambiguous prognostic value), and another reserved for the clinic (admitting less severe and less severe forms). differentiated)!


In 1982, Weinstein described an essentially biological syndrome, which he termed HELLP, associating moderate intravascular hemolysis, elevated transaminases (usually moderate, two to four times normal), and progressively worsening thrombocytopenia. Clinical signs occur in the third trimester and are associated with general malaise (90%), epigastric bar pain, or limited to the right hypochondrium (90%), nausea and vomiting (50%). This syndrome is associated with a very poor fetal prognosis, even maternal and, despite some heroic therapeutic attempts (immunoglobulins [Ig], plasma exchange …), most authors agree to consider as a single issue a rapid termination of the disease. pregnancy.

This syndrome is mentioned here because it is often (but not always) associated with hypertension and proteinuria. It has been considered alternatively, either as a complication of preeclampsia, or as a symptomatic variant of it. His pathogenesis remains the object of speculation.


The incidence of pregnancy-induced hypertension is estimated to be between 10 and 15% of pregnancies. The frequency is similar in most European countries and in the United States, except for some studies which overestimate this frequency due to a more lax definition.

About 10% of these women (2 to 3% of the population) have pre-eclampsia (as defined above). The percentage of preeclampsia, and especially severe preeclampsia, is in fact much more variable among countries, with a much higher incidence in developing countries. Preeclampsia is associated with maternal mortality, varying according to the country, between 0.1 and 5 per 1000 cases, or even more. This mortality is largely concentrated in patients with HELLP syndrome. Even though eclampsia (seizure) has become a rare event (0.56 births), at least in our climate, it remains a particularly serious event. Maternal mortality of 5% has been reported in Australia in case of eclampsia.

Gestational hypertensions appear readily from the first pregnancy, the age of the latter being not fundamentally different from that of normal pregnancies. The classic “double hump” distribution (a peak among very young women under 20, a second peak over 37-40 years) is no longer observed in our climate, but the rest in some developing countries.

In France, the frequency of hypertension and preeclampsia does not differ among ethnic groups. More discordant data have been reported in the United States. The differences between socio-professional categories are modest and the so-called “disadvantaged” categories are no more exposed than others. In a French study, there was an excessive frequency among senior executives and liberal professions, but also among service personnel, as compared to clerical and commercial employees, or workers. The frequency appeared lower among the women “without profession”. These facts suggest that the risk of developing pregnancy-induced hypertension is higher in women who have significant physical or mental activity, and / or poor social coverage.

Obesity is a favorable factor found in all studies.

Similarly, the frequency of preeclampsia is lower among smokers. The explanation of this last fact is not known.

Clinical tables:

They are of different presentation and severity. It is clear from this table that hypertension, as long as it is isolated, is accompanied by a modest increase in fetal risk. The same is true of isolated proteinuria. It is the tables associating the maximum values ​​of these two parameters which involve a major fetal risk. This fetal risk is usually associated with the risk of maternal complications.


Isolated hypertension during pregnancy thus only modestly obtains its prognosis, with a relative risk ranging from 1 to 3. According to the classifications above, this hypertension may be “gravid” or “chronic”; the difference is not always easy to make at the moment, even if the usual 20 week criterion is usually used as a marker. Some studies assign a slightly more negative prognosis to gestational hypertension, others to chronic hypertension.

These hypertensions are almost always asymptomatic. It should be remembered however that this type of situation is not fixed, and that at any time proteinuria can come to complete the picture, thus substantially increasing the risk.


When significant proteinuria is associated with hypertension, the risk is much higher.

It remains modest when the blood pressure figures are moderately high and easily controllable, usually coexisting with proteinuria of less than 1 g / 24 h. In these cases, enhanced surveillance, both fetal and maternal, is needed. It is not unusual that a premature pregnancy outcome is indicated, either because of a slowing or stopping of fetal growth, or because of any threat to the maternal prognosis.


All different is the picture of “severe” preeclampsia.

Hypertension is then major, threatening, and remarkably insensitive to antihypertensive treatments. Proteinuria is several grams, even tens of grams per 24 hours, with a nephrotic syndrome. There is usually diffuse edema, infiltrating the upper and lower limbs, the loins, the face. Fetal growth slows down and stops. Patients are often cephalalgic and photophobic. It is in such cases that a HELLP syndrome often complements the picture, and thrombocytopenia, which is rapidly progressive, creates a major threat in the short term. In this situation, the only way out is the termination of pregnancy, almost always by cesarean section. This decision is relatively easy if the term is sufficiently advanced to allow a reasonable chance of survival of the newborn under acceptable safety conditions. In the opposite case, one may be tempted to delay to obtain a little more fetal maturity, but this delay is only at the cost of an increase of hypotrophy, and the risk of maternal complications is then very high. The extreme gravity of the situation can sometimes justify a cesarean section called “maternal rescue” on an unsustainable child.

It is of course in such cases that haemodynamic maternal complications (pulmonary edema …) or acute renal failure appear most readily; it is also in these cases that the maternal vital prognosis is most severely threatened.


In the case previously described, maternal or fetal accidents may occur, complicating a situation whose gravity was already obvious. There are other circumstances in which a pregnancy that seemed normal (or not very pathological) suddenly turns to drama when retroplacental hematoma (HRP) or eclampsia occurs, often accompanied by fetal death. It is then after the accident that hypertension, proteinuria, and the whole procession of maternal complications occur which will increase the gravity. Note also that about a third of HELLP syndrome and a quarter of eclampsia occur in the postpartum.


As we have just seen, the risk involved is both maternal and fetal. For the mother, it is the possible occurrence of HRP or eclampsia. Recall that they are often accompanied by a consumption coagulopathy (disseminated intravascular coagulation [DIC]), especially in case of HELLP syndrome, and can be followed by acute renal failure, or even cortical necrosis. This means that for rare that they have become, they keep a very serious, even dramatic, prognostic meaning. Thus, in a series of 442 pregnancies with HELLP syndrome, Sibai et al report a DIC in 21% of cases, HRP in 16%, acute renal failure in 7.7%, pulmonary edema in 6%; 55% of the patients required transfusions and 2% had laparotomy due to a bleeding syndrome. Maternal mortality was 1.1%.

For the fetus, the risk is that of a delay, or even a stop of the growth by default of perfusion, resulting in the maximum in utero death. We refer the reader to the treatise on obstetric gynecology for a detailed description of these complications.



Gravidic hypertension is almost never observed spontaneously in the animal kingdom and it is difficult to obtain an experimental model having some points in common with the disease observed in the human species.

Hypertension models:

In genetically hypertensive rats, pregnancy and parturition are unaffected by hypertension. It tends to fade during pregnancy. The placental and renal morphology of pregnant animals is normal, as is the weight of newborns.

Overall, experimental studies to create hypertension (mainly renal artery stenosis, or vasopressor infusion) in the gestational animal showed a normal continuation of pregnancy and the absence of consequences of hypertension on survival. or the birth weight of the young. Uteroplacental blood flow was not altered, or very transiently. These data remove some credit for the idea that elevation of blood pressure is the cause of placental dysfunction or fetal distress.It should be noted, however, that in a model of renal arterial stenosis, uterine blood flow was very dependent on the level of Pa, and very sensitive to a reduction of it by antihypertensive products.

Placental ischemia models:

Acute placental ischemia can easily be produced by ligation of the uterine arteries. This maneuver leads to hypertension, proteinuria and fetal death. Hypertension and proteinuria disappear immediately after parturition. Abitbol et al conducted a model of chronic placental ischemia in guenon and rabbit by narrowing of the infrarenal aorta using an inflatable clamp whose pressure can be adjusted from the outside. A 40% reduction in blood flow results in immediate high blood pressure, and proteinuria appears on the fifth day. The renal lesion observed is superimposable on the so-called “endotheliosis” lesion observed in human disease. This syndrome is reversible if the clamp is released after a sufficiently short period of time. Finally, it appears only in the pregnant animal, the same maneuver having no effect on the non-pregnant animal.

Inflammation and endotoxins:

An injection of bacterial endotoxin can reproduce in the animal closely the manifestations of preeclampsia. Very low doses must be injected, otherwise there will be shock and pregnancy will stop. Under these conditions, the pregnant animal shows an increase in Pa, proteinuria, coagulopathy and glomerular fibrinogen deposits with monocyte infiltration. There is simultaneous activation of circulating polynuclear cells, and the whole reproduces a complete model of inflammatory reaction. The same maneuver is ineffective in the non-pregnant animal.


Takimoto et al reported a fascinating model of preeclampsia in transgenic mice for components of the human renin-angiotensin system. The crossing of a female carrier of the transgene of angiotensinogen with a male carrying the transgene of renin (nothing happens if it is the opposite) leads to severe hypertension at the end of gestation, with proteinuria and loss frequent fetal The renal histological lesions are comparable to those described in the human species. Finally, everything is regressive after parturition.


Early stages of placentation: physiology

The recognition that the primum movens of hypertensive pathologies of pregnancy was a very early abnormality of placentation has stimulated considerably the research on the mechanisms and the possible anomalies of this one.

The so-called “hemochorial” placentation as it occurs in the human species requires a connection between the nascent placenta and the maternal vessels. The latter must also acquire a caliber sufficient to ensure the blood flow necessary for good quality exchanges. This connection is effected by an invasion of maternal structures by the trophoblast, which behaves a little like an invasive tumor. One of the peculiarities of this phenomenon is that it is normally self-limiting, which supposes powerful regulating factors. The main exceptions to this self-limitation are hydatiform moles and choriocarcinoma.

The resulting diploid mole for the paternal genome induces preeclampsia and aggressive, neoplastic trophoblastic invasion.

Just days after fertilization, the villous cytotrophoblast differentiates at the periphery of the blastocyst into a highly invasive syncytiotrophoblast, which allows penetration and anchorage of the blastocyst in the endometrium. Then the extravillous cytotrophoblast colonizes the syncytial mass and invades the decidua to the spiral arteries. This is the first interstitial phase of trophoblastic invasion. The second phase, later, is the endovascular invasion of the spiral arteries of the myometrium, which will go up to about a third of it. During this phase, the trophoblastic cells undergo a deep transformation conferring on them an endothelial type phenotype.

This invasion is a prerequisite for the establishment of a suitable maternal circulation.

The invasion is done through proteolytic enzymes, mainly metalloproteases. Its progression is initiated and controlled by various growth factors and cytokines. In all these phenomena, the oxygen tension as well as the NO production seem to play a major role, as well as perhaps direct hemodynamic factors.

The decidua is infiltrated by many cells. Although B and T lymphocytes are relatively rare, monocytes / macrophages and natural killer (NK) cells are particularly abundant. The extravillous trophoblast (and it alone) expresses a particular combination of human leukocyte antigen (HLA) class I, HLA C, E and G molecules (the HLA G is totally trophoblast specific). The NK cells that infiltrate the decidua are in close contact with the invasive trophoblast and contain receptors that recognize these HLA I antigens.

This interaction could be a key element of invasion regulation by modulating the cytolytic effect of NK cells. HLA G signals the presence of the placenta and protects the trophoblast by inhibiting the lytic effect of NK. Unlike the so-called adaptive immunity of T and B cells, which recognize the self of non-self, this “native” immunity recognizes the missing self since NK cells are cytotoxic only in the absence of HLA G.

It should also be noted that this phenomenon must take into account paternal components whose aggression must be avoided to prevent rejection of the fetal allograft. This could be the role of HLA C.

Monocytes, on the other hand, promote trophoblast apoptosis via tumor necrosis factor (TNF) a . This is certainly another essential regulatory element.

Nevertheless, the spiral arteries of the myometrium are colonized around 15 weeks by the trophoblast which replaces the endothelium (acquisition of specific cadherins) and destroys the muscular structures. These arteries are thus transformed into channels whose diameter is multiplied by 4 to 6, and which no longer have a resistive but only conductive function. This “transformation” of the spiral arteries is clearly a prerequisite for adequate irrigation of the placenta and fetus.

Anomaly of trophoblastic invasion:

The existence of an anomaly of this trophoblastic invasion was a major step in the physiopathological understanding of preeclampsia. It has been shown as early as the 1970s on placental bed biopsies that trophoblastic invasion is defective when preeclampsia is to occur in the third trimester, or during isolated fetal growth delays. This anomaly consists either in the absence of transformation of the spiral arteries or in incomplete transformation over an insufficient length.

This abnormality of placentation therefore precedes by several months the first manifestations of hypertension or proteinuria, but everything suggests that from this moment, the game is played. The vascularization of the placenta is insufficient, ischemia develops gradually, and it is only from a critical threshold of ischemia, reached much later, that appears hypertension.


Many arguments suggest that a moderate inflammatory response, involving the placenta but also other vascular structures of the maternal organism, would be present in normal pregnancy. This reaction appears considerably increased, and still more diffuse, in preeclampsia. The latter would in a way represent a “decompensation” of this inflammatory reaction due either to a too intense immunological stimulus or to an exaggerated maternal reaction. This inflammatory process is thought to be closely related to the cellular infiltration already mentioned in the placenta, and the abnormalities that contribute to the insufficiency of the trophoblastic invasion would be a powerful stimulus. It is admitted, without strong evidence, that the triggering factor of this inflammatory reaction is immunological.

Release of trophoblastic cells:

The placenta, both ischemic and inflammatory, releases into the maternal circulation a very large quantity of necrotic trophoblastic cells, possibly degraded and limited to vesicles; this fact is well acquired. In vitro, these vesicles are able to strongly inhibit the proliferation of endothelial cells and even to break the cell layer of the culture. It has been hypothesized that these cells or vesicles liberated in large excess by an ischemic placenta and in apoptosis would cause endothelial rupture, further increased by the activation of monocytes (and polynuclear cells, via TNF a ), triggering the cascade. classical vasoconstriction, activation of hemostasis, etc.

Lipid peroxidation and free radicals:

In this phenomenon of endothelial pain, an important role has been attributed to oxidative stress, whose manifestations appear both placental and systemic. The circulating rate of free fatty acids is very early increased before preeclampsia, and the incorporation of these fatty acids into the endothelial cells is increased. The serum of these patients has a high lipolytic activity. Maternal lipid abnormalities could potentiate the generation of free radicals.


Its mechanism is unlikely to be unambiguous. On the contrary, it is highly probable that it is at this stage that the diversity and heterogeneity of the disease “gestational hypertension” are expressed. The hypotheses envisioned below are therefore not exclusive of each other, and other hypotheses are likely to be formulated in the years to come.

Mechanical hypothesis:

In this hypothesis, the oldest and simplest of all, placental ischemia results from the mechanical compression of the aorta and / or uterine arteries by the uterus. The well-known promoting role of twinning and hydramnios would thus be easily explained. The direct arteriographic evidence of a significant reduction in the caliber of the infrarenal aorta during pregnancy has also been made in a few anecdotal cases.

Pre-existing vascular pathology:

Many patients with gestational hypertension have heavy genetic and / or metabolic risk factors for vascular disease.These patients have every reason to have vascular alterations prior to pregnancy. In fact, renal vascular lesions, sometimes impressive, were found histologically, even though the patients were normotensive. It is easy to imagine that such vascular lesions, probably ubiquitous, are a major obstacle to normal placentation. In this case, the repetition of accidents over successive pregnancies would be easy to understand.

Pre-existing thrombophilic pathology:

Dekker et al reported a very high incidence of thrombophilic diseases in young women with early and severe preeclampsia. These abnormalities were mainly a circulating anticoagulant or antiphospholipid, a protein C or S deficiency, an activated protein C resistance (known as Leiden Factor V mutation), or a hyperhomocysteinemia.

A gene mutation encoding prothrombin (factor II) was later added to the list. These data have been fairly widely cross-checked by various authors, and some admit that more than 50% of women with severe pre-eclampsia have at least one of these abnormalities.

While it seems likely that these abnormalities may be involved in the genesis of preeclampsia, at least as an aggravating factor, it must be remembered that the geographical distribution of these mutations is highly variable and in no way parallel. to that of preeclampsia.

Immunological factors:

The fetus, whose genetic capital is half of paternal origin, represents the equivalent of a semi-allogeneic transplant, the survival of which requires a state of maternal immune tolerance.

During pregnancy, there is recognition by the mother of paternal antigens and immunization against these antigens.

Thus, 20% of primiparas and 50% of multiparas have circulating antibodies directed against paternal HLA components. A humoral facilitation system was thus highlighted and extensively studied in the 1970s. This mechanism was found completely absent in cases of repeated abortions and greatly decreased in preeclampsia.

A second factor of tolerance would be the induction of suppressor T cells. An additional role could be played by the passage of fetal lymphocytes (probably T suppressors) into the maternal circulation. Finally, we have mentioned above the importance currently given to NK cells and their interaction with HLA I antigens carried by the trophoblast. HLA G, which is polymorphic and placenta-specific, would signal the presence of the placenta and inhibit cytotoxicity. HLA C would primarily translate an allogenic signal of paternal origin, and E would trigger the inhibitory effect of NK cells.Ultimately, cytotoxicity would depend on the balance and interaction between these three elements.

The defect of trophoblastic invasion, and therefore preeclampsia, could be related to immune aggression of the placenta. In the 1970s, the idea of ​​a lack of humoral immunological facilitation in case of a high degree of histocompatibility between father and mother was widely developed. This fact partly explains the finding that gestational hypertension occurring for the first time in a multiparous is often associated with a change of partner, and also that prior transfusions have been shown to have a protective effect against of gravid hypertension. Although this idea has gone out of fashion today, publications periodically remind us that the facts observed 20 or 30 years ago are still accurate.

The process of pediatric immunization is probably a little more subtle than was imagined at the time, but its presence and importance remain. The degree and mode of sperm exposure seem to play the predominant role. Robillard et al have shown that the risk of pre-eclampsia is higher in the case of early conception in a recent couple than in the case of later conception in a couple that has been established for a longer time, a phenomenon described little poetically as “duration of sexual cohabitation”. . Similarly, in case of artificial insemination, the risk of preeclampsia is higher if the sperm comes from a foreign donor rather than the spouse. The practice of oral sex, according to several authors, would be associated with better protection against preeclampsia than vaginal sex alone.

According to some authors, the use of barrier contraception such as condoms is associated with an increased incidence of preeclampsia.

Genetic aspects:

Some family aggregation of preeclampsia cases is classically accepted. In some patients who have had eclampsia, there are sisters, the mother, or a grandmother who had the same accident. A careful analysis of these families had previously suggested that this would be a monogenic transmission. Nowadays, eclampsia is rare. The disease “gestational hypertension” is much more heterogeneous than previously thought, and data from genetic studies appear less clear. On the contrary, everything suggests that various genes involved in the regulation of Pa, the regulation of plasma volume, vascular remodeling, and various more specifically placental factors, intervene in various ways as “susceptibility genes” of preeclampsia.

Cohort studies do suggest genetic transmission of preeclampsia. Thus, Cincotta and Brennecke studied 368 young primiparous. Eighteen of them had their mother or sister (or both) who had pre-eclampsia. Of these 18 women, five (27.8%) had pre-eclampsia, compared with 29 (8.3%) of those without a family history, which corresponds to a relative risk of 3.4 (interval). 95% confidence level: 1.5-7.6). This risk is even higher for “severe” preeclampsia. Arngrimsson et al studied 94 Icelandic families (very homogeneous population) over four generations in the offspring of women who had severe preeclampsia or eclampsia in the years 1931 to 1947.

The prevalence of preeclampsia was higher (23%) in girls than in daughters-in-law (10%). Predisposition was transmitted by both men and women.

Nevertheless, in a study of 99 couples of monozygotic twins, ten developed preeclampsia and none of their binoculars.In a larger series, the same authors estimated the genetic transmission of preeclampsia at 0% and that of gestational hypertension at 25%. Conflicting results have however been reported.

Few are the plausible candidate genes. An association between preeclampsia and the M235T variant of the angiotensinogen gene has been reported, but has not been found by all authors. Nevertheless, this mutation seems to be associated with less dilatation of the spiral arteries, which would establish a link between a genetic anomaly and the lack of trophoblastic invasion. We have discussed the thrombophilic abnormalities readily associated with preeclampsia.

Leiden factor V mutation was the most studied and the results were somewhat discordant. We will also note that this mutation, quite common in Europe, is virtually absent in other countries (Japan) where the incidence of preeclampsia is not lower.

Genomic studies have allowed quite a variety of suggestions.

The most consistent is the region of chromosome 7q36, encoding eNOS. Such a susceptibility gene would be physiologically very relevant.

At the border between immunology and genetics: the father

Preeclampsia is not just the problem of an individual, it is also that of a couple. The father can intervene in the genesis of this pathology in two ways: an immunological “conflict” between father and mother, or the paternal transmission of a gene (or other factor) responsible for the placental dysfunction.

Lie et al, based on a Norwegian birth registry of 1.7 million admissions, studied pregnancies following a pre-eclamptic pregnancy according to the individuals involved. When a pregnancy has been preeclamptic in a couple, a new procreation between the same father and a different woman practically doubles the risk of preeclampsia for the latter.The risk of preeclampsia is also increased in the same proportions in the half-sister of a woman who has herself had preeclampsia, if the two women are of the same father and mother different.

Other publications show that a man from a pre-eclamptic pregnancy increases the risk of preeclampsia for his wife.

Dizon-Townson et al found a high frequency of Leiden mutation in case of iterative miscarriages with placental necrosis.

The mutation was found more often in fetal deoxyribonucleic acid (DNA) than in maternal DNA, clearly indicating that in some cases the gene was of paternal origin.


Leaving aside the fetal consequences of placental insufficiency, we limit ourselves to the study of the mechanisms by which placental insufficiency is responsible for hypertension, renal disease, which is both anatomical and functional, and DIC.

Endothelial dysfunction:

The reduction of placental perfusion following a defective implantation is followed by a cascade of abnormalities that demonstrate an alteration of endothelial functions:

– an increase in sensitivity to pressor hormones: this is known for a very long time, manifested inter alia by the loss of the “refractory state” to angiotensin, which characterizes normal pregnancy;

– an activation of haemostasis: the frequency and extent of fibrin deposits in the placenta and in many organs made us suspect very early the role of haemostasis disorders in the manifestations of gestational hypertension. Preeclampsia has thus been assimilated to a state of CIVD, and it is the latter that would explain the polyviscular manifestations observed, especially in the kidney, liver (HELLP syndrome), it is also that would explain eclampsia. In fact, in the light of more recent work, a real CIVD seems rare, if any, in preeclampsia. In contrast, early platelet activation is certain.Such stimulation is compatible with early endothelial damage. It could lead to secondary activation of coagulation and fibrinolysis;

– Decreased prostacyclin production: an imbalance in the production of eicosanoids also exists very early. During a normal pregnancy, the production of prostacyclin and thromboxane A 2 are both strongly stimulated, although with a very favorable relationship to prostacyclin. This stimulation is evidenced by a considerable increase in the level of their metabolites, both in the serum and in the urine. This suggests that their stimulation is a global phenomenon in the body. In fact, the production of prostacyclin is increased in all the territories of the circulation, the renal production is also increased and the uteroplacental unit synthesizes abundant quantities. The mechanism of this stimulation is currently unknown and is part of a complex interregulation of all hormonal systems with direct or indirect vasomotor activity. However, the increased production of prostacyclin clearly plays a major role in the systemic and renal vasodilation that characterizes the hemodynamics of the pregnant woman. It largely counterbalances the vasoconstrictor and procoagulant effect of thromboxane.

In hypertensive pregnancies, thromboxane stimulation is substantially identical to that seen in normal pregnancies, whereas prostacyclin is little or not stimulated. The report is therefore in favor of thromboxane, that is to say of the vasoconstrictor and procoagulant element. This anomaly is probably indicative of a functional disorder of the endotheliums, which are primarily responsible for the production of prostacyclin;

– The appearance of biochemical markers: additional arguments in favor of this hypothesis are provided by the elevation of circulating levels of fibronectin and factor VIII, markers of endothelial lesion;

– an endothelial activator? Some authors have demonstrated in the plasma of preeclamptic patients a substance capable of inducing a high production of platelet derived growth factor (PDGF) in endothelial cells in culture, demonstrating an intense activation of these cells.


It is in this context of endothelial dysfunction that systemic vasoconstriction and the resulting hypertension should be integrated.

Hypertension is mainly due to the loss of vasodilatation characteristic of normal pregnancy and the appearance, on the contrary, of a vasoconstriction. Normally, pregnancy is characterized by a state refractory to pressurized hormones and especially angiotensin II; this situation disappears before the emergence of preeclampsia. An angiotensin test has even been used to predict preeclampsia.

The mechanism of vasoconstriction remains debated. The imbalance between prostacyclin and thromboxane certainly plays an important role. It is also possible that the vasoconstrictor potential of other substances (angiotensin, endothelin) is amplified by a decrease in NO synthase activity. The endothelial cells themselves can be altered by the action of proinflammatory cytokines (TNFα) and by increased oxidative stress.

Hemodynamic studies have been few and their results are contradictory. Normal pregnancy is accompanied by an increase of about 30% in cardiac output. In spite of this, the vasodilatation is such that Pa drops physiologically.

Cardiac output generally remains high in benign hypertensia, but declines in severe preeclampsia.

Plasma volume (normally increased by nearly 50%) is very low in severe forms, or even collapsed in so-called “toxemia gravidic” forms with significant proteinuria and fetal growth retardation. This volume contraction is directly correlated with the birth weight of the child. It could result from either vasoconstriction itself or a more subtle disorder of sodium excretion.

Still other factors may play a role in the genesis or maintenance of hypertension, sympathetic nervous system, natriuretic atrial factor, calciotropic factors, magnesium metabolism.


Functional data:

· Renal function:

The evolution of renal function during normal pregnancy has been the subject of many reviews, mainly those of Davison and Atherton. There is normally an increase of approximately 50% in the renal plasma flow. Glomerular filtration evolves in a substantially parallel manner, leading to a clearance of creatinine of the order of 180 ml / min.These two parameters are generally decreased in pregnancy-induced hypertension. The decrease is most often of the order of 25%, that is to say that the values ​​observed are still above those considered normal before pregnancy.

In the most severe forms, however, glomerular filtration may be much lower, and acute renal failure is a rare but usually extremely serious complication, severe preeclampsia or HELLP syndrome.

· Sodium balance:

As mentioned above, a retention of 900 to 1,000 mEq of sodium occurs during pregnancy. It is necessary for volemic expansion. It occurs despite the increase in glomerular filtration, which implies a major adjustment of sodium reabsorption. Volume expansion is defective or absent in severe forms of pregnancy-induced hypertension. This anomaly is sometimes, but not always, associated with a globally lowered extracellular volume. A disturbance of the vascular permeability can intervene, contributing to the formation of the edemas.

· Renal excretion of uric acid:

During normal pregnancy, serum uric acid drops by an average of 30%, while both clearance and fractional excretion of uric acid occur. Hyperuricemia is associated with severe forms of gestational hypertension. It is proportional to the severity of the renal anatomical involvement and represents a reputed index of fetal prognosis. In fact, there is a negative correlation between changes in serum uricemia and plasma volume, suggesting that the decrease in uric acid clearance reflects the physiological response of the kidney to hypovolemia.

Anatomical data:

The injuries found can be grouped under three headings.

· Glomerular endotheliosis:

This is the most ancient lesion described. It has been considered by most authors as specific to “preeclampsia”.

It is composed of swelling of glomerular endothelial cells, irregular thickening of basement membranes and epithelial pedicel fusion. Subendothelial deposits of fibrinogen can be observed. Some authors have demonstrated by immunofluorescence deposits of IgG or IgM. The essential feature of glomerular endotheliosis is its complete reversibility in a few weeks after delivery. At the very most, it can leave a few minute parietal irregularities or a slight thickening of the mesangium, whose pathological significance is doubtful.

· Vascular lesions:

They are probably less frequent, but it is certain that they have been largely underestimated in the past. It can be either a fibroelastic endarteritis, sometimes severe, affecting cortical arteries of medium caliber, or hyaline deposits, possibly occlusive, in the wall of arterioles. Overall, these lesions are very similar to those seen after several years of permanent hypertension. They are often in stark contrast to patients’ normotension and the brief hypertensive period that marked the end of pregnancy. In our experience, these lesions are very often a predictor of permanent hypertension at term of about 5 years.

· Nephropathies independent of pregnancy:

Various nephropathies, especially glomerular, can be discovered during pregnancy-related hypertension: segmental and focal hyalinosis, Berger’s disease, membranoproliferative glomerulonephritis, etc. These lesions are clearly independent of pregnancy, probably preceded it, and in any case survive it. However, some observations indicate that segmental and focal hyalinosis lesions may develop during preeclampsia itself. Apart from this very particular case, pre-existing nephropathies can be revealed by proteinuria of early discovery in pregnancy, often without hypertension.They are then easily suspected.

In many cases, however, the clinical picture is that of hypertension with proteinuria appearing only in the third trimester. The persistence of proteinuria several months after giving birth may then be the only suggestive symptom, yet it may be missed. It is understood that the diagnosis is often unrecognized in these cases, if the indication of a renal biopsy was not asked.


Thrombocytopenia is by far the most common hematologic abnormality in hypertensive pregnancy. It is modest in most cases; however, the drop in platelet counts below (sometimes much below) 100 000 / mm 3 is a mark of severe forms, usually requiring rapid intervention. It can be accompanied by the appearance of fibrin degradation products, or even all the stigmas of a CIVD.

The coexistence of decreased antithrombin III and increased fibronectin suggests that endothelial pain is associated with it.


Histological abnormalities of the liver (periportal hemorrhages, ischemic lesions and fibrin deposition) have been reported in autopsy series well before the description of HELLP syndrome.

Hepatic abnormalities range from moderate cytolysis to subcapsular hematoma, or even liver rupture, which need not be emphasized.


Eclampsia (convulsive phase of preeclampsia) remains a major complication. It is most often attributed to focal ischemia by fibrin deposition and / or vasoconstriction. Classical cerebral edema or hypertensive encephalopathy are much more improbable mechanisms, especially since many eclampsia appear with a very modest hypertension, even without hypertension. Various aspects have been described since the use of CT or magnetic resonance imaging (MRI). The often posterior location of these lesions would explain the frequency of visual precursor disorders.

Surveillance of a hypertensive pregnant woman:

Although the fetal prognosis of hypertensive pregnancy has improved steadily over the last twenty years, the accuracy of surveillance is certainly a key factor. The purpose of this monitoring is to obtain a prediction, as far as possible, of the risk of complications, before any clinical element.

Thus, therapeutic decisions can be made in optimal conditions.


Increased clinical surveillance is required in these patients because of the risk of maternal complications and the therapeutic decisions that may result. The idea of ​​a multidisciplinary surveillance, associating (or alternating) gynecologist-obstetrician and specialized doctor, has gradually imposed itself in all the centers preoccupied by this pathology.

Given the frequency of pregnancy-related hypertension and its very unequal gravity, the idea of ​​a hierarchy of care has also been imposed, leading the patient, according to the degree of risk estimated, to centers that are increasingly equipped. and specialized.

We will not dwell here on the monitoring of Pa and various cardiovascular parameters. We focus on the prognostic elements of pregnancy, whose value is decisional.


Symptoms of chronic fetal distress are of paramount importance in the surveillance of pregnancy-induced hypertension. It is obvious that a stagnation of the uterine height or a slowing down of the fetal growth during the echographic controls are heavy consequences, as well diagnostic as therapeutic.


The emergence of Doppler surveillance in pregnant women has been a major step forward. The exploration at the level of the umbilical artery makes it possible to appreciate the blood velocity in the fetal compartment. This reflects fairly well the “fetal well-being”. Its alteration is always associated with fetal growth retardation and often announces a severe accident. The exploration of uterine arteries gives information on the maternal side of the circulation. Its alteration is usually the witness of a deficient maternal vascular state. Such a situation would be associated with an increased risk of HRP. Finally, the exploration of the cerebrovascular areas of the fetus highlights self-regulating defense reflexes. Its alteration is evidence of acute fetal distress and often an imminent complication.

Also, this last examination is often a decision-making element for the obstetrician.


Proteinuria remains the main biological stigma to hunt down.

This monitoring is done routinely with test strips, but laboratory quantification becomes essential as soon as the research is positive.

Hemostasis monitoring, at least for platelets, is usually done. Thrombocytopenia is always an element of poor prognosis, and can lead to urgent therapeutic decisions.

Regular monitoring of transaminases allows the early detection of HELLP syndrome.

Creatinine should be periodically checked, especially if proteinuria has appeared.

The measurement of plasma volume in the prediction of fetal hypotrophy has been supplanted by Doppler and is hardly used today. A simple hematocrit gives a rough estimate. A hematocrit greater than 38% regularly indicates hemoconcentration with hypovolemia. This situation can nevertheless be masked by anemia from another cause.

Uricemia has long been another major marker of the risk of short-term complications. For most authors, the value of 350 μmol / L represents a critical threshold beyond which the risk of death in utero increases almost linearly, approaching 100% from 600 μmol / L. In case of doubt about the previous values, it is essentially the gradient of serum uric acid over successive dosages that must be taken into consideration. In fact, if uricemia is a marker of high specificity, its sensitivity is low and therefore the value attributed to it has greatly diminished. Again, Doppler examination data provides more sensitive and reliable information.


The therapy of gravid hypertension is not the least debated point and it is certainly the most disappointing. The most controversial issue is the desirability and modalities of antihypertensive therapy.


Physical and psychic rest is one of the few measures whose usefulness is beyond doubt. Resting in bed, preferably in the left lateral decubitus, lowers blood pressure, is often associated with a decrease in serum uricemia and appears to be beneficial for fetal growth. The explanation given is the clearance of the aorta and inferior vena cava, which would increase uterine blood flow and cardiac output. This therapeutic mode is obviously dependent on the material possibilities of the patient (housing conditions, presence of other children …).


If we refer to what has been said above about the initiating role of placental ischemia, whose hypertension is only a consequence, it is not clear that antihypertensive treatment is beneficial to the placenta, nor to fetal growth. On the contrary, it may be suspected that a lowering of the pressure within a resistive circuit leads to a decrease in the flow rate, which would be the opposite of the goal sought.

Animal data:

Experimentally, various antihypertensive drugs have been used by Brinkman and Assali in pregnant hypertensive animals.

Abrupt reduction of Pa, as obtained with diazoxide injection, is accompanied by an impressive drop in uterine blood flow. It is the same after the intravenous injection of furosemide. On the other hand, progressive lowering of Pa with methyldopa does not alter the uterine flow rate very much. If this lowering of pressure is accompanied by an increase in cardiac output, as is the case with hydralazine, the uterine flow rate is completely respected, or even slightly improved.

Chronic Hypertension or Moderate Arterial Hypertension:

These are situations in which the obstetric prognosis is most often favorable. Antihypertensive therapy in these situations provides no benefit.

A meta-analysis of these studies was conducted by Magee et al.

It shows that overall, the treatment has some positive effects in the mother: less than 160/100 hypertensions, and fewer hospitalizations. On the other hand, it has no effect on the prognosis of pregnancy and on the fetal prognosis in particular. On the contrary, there is a tendency for a higher incidence of fetal hypotrophy under treatment. Von Dadelszen et al have stated in another meta-analysis that there is a significant correlation between the decrease in Pa and the percentage of hypotrophic children. This fact had already been found in some individual studies where the treatment in question was a beta-blocker. Magee et al confirmed this by specifically reviewing these studies.

The effect of beta-blocker pressure changes has occasionally been documented in the short term by umbilical Doppler.Finally, the comparison between different classes of antihypertensives showed no decisive advantage of one class over another.

One isolated study indicates that the use of a beta-blocker in patients with very high cardiac output could have a beneficial effect and even prevent preeclampsia.

The conclusion is that antihypertensive treatment provides a very modest maternal benefit whose practical value is not obvious.

Excess hospitalizations in the absence of treatment are not due to an objective complication but only to medical concern. Antihypertensive treatment does not improve the fetal prognosis, but may be responsible for hypotrophy if it is too intense. However, it must be agreed with Sibai that the numbers of studies have never been sufficient for an effect on fetal death or HRP (incidence of the order of 2%) to be demonstrated. In fact, a 50% reduction in one of these accidents would require a staff of 2,000 patients per group. No studies have reached such numbers, and meta-analysis is not necessarily an infallible method to overcome this deficiency. Even if this margin of uncertainty is admitted, the antihypertensive treatment in these indications is obviously not a very interesting therapeutic act.

Severe hypertension:

The case here is much less simple in that there have been no controlled studies, for obvious reasons. The reasoning by analogy with other hypertensions indicates that the benefit of a treatment for a short-term hypertension in a young woman is probably not negligible, even if it is not major. This treatment is likely to prevent maternal complications, first and foremost pulmonary edema. The classic assertion of the risk of stroke is hardly credible in this same reasoning by analogy. Indeed, cases are rare and the accountability of blood pressure figures has never been adequately supported. Nevertheless, the general practice is to treat these hypertensions when the figures regularly exceed 160 to 180 and / or 110 mmHg. It is certainly as important as before, if not more so, to act sensitively, and not to reduce the numbers below 140 and 90 mmHg.

Which antihypertensive drugs?

Diuretics, widely used in a time, are now completely abandoned. Indeed, they decrease the plasma volume, already often deficit, and can thus aggravate chronic fetal pain. They decrease placental perfusion and many clinical studies have shown that they are associated with lower birth weight.

Central antihypertensives (methyldopa, clonidine) have been widely used in pregnancy. These are certainly the products for which the experience is the greatest and the longest decline. Their efficacy is adequate and their safety seems largely established. They are also the only ones for which pediatric surveillance over years is available, demonstrating the absence of long-term adverse effects in children, both in terms of growth and intellectual and academic performance.

Hydralazine has a comparable decline. Its efficacy is remarkable in fairly high doses and it has the theoretical advantage of not crossing, or very little, the placental barrier.

Unfortunately, the counterpart of this efficacy is a poor clinical tolerance (palpitations, intense headaches that may impose a threat of eclampsia), due to the increase in cardiac output already high in these patients, and that limits the use.

Prazosin is also used. She even enjoys a certain favor in the United States. Its antihypertensive efficacy is good and its tolerance without problem. Like hydralazine, prazosin has a strong protein binding and its transplacental passage is weak.

Beta-blockers are widely used in pregnancy.

Contrary to what one might fear, they do not increase uterine motility. As they cross the placenta, they normally carry a risk of neonatal hypoglycemia, bronchospasm and bradycardia. In fact, over the years, this risk has appeared more theoretical than real, and the benefits of beta-blocking treatment seem to outweigh this risk. This reassuring data does not change anything that has been said above about the risk of hypotrophy if the treatment is too intense. It is also clear that very careful neonatal surveillance of children born with beta-blockers, especially if they are premature and hypotrophic, should be ensured.

Inhibitors of the converting enzyme are responsible in animals for an increased frequency of fetal deaths. This risk did not appear in the human observations reported, but these remain anecdotal. On the other hand, neonatal complications have been reported, especially anurias, many of which have been fatal. These products are contraindicated in the last two trimesters of pregnancy. Note, however, that no teratogenicity was observed. No particular concern is therefore justified when pregnancy begins under a drug of this class. Finally, to date, all indications are that angiotensin II receptor antagonists have the same effects as angiotensin converting enzyme inhibitors and share the contraindication.

Calcium blockers are widely used, at least in France, in pregnant women. Yet their record is remarkably poor. There is little certainty about their lack of teratogenicity.

Their tocolytic action, which is invaluable in the event of a threat of premature labor, can be a source of difficulty during childbirth or even postpartum. Only strong studies, which are still lacking today, could give them a reasonable level of proof.


The desodized diet has been widely used for several decades and it has even been seen in a time a panacea. The proof of its uselessness and even its harmfulness was brought in 1958 by a remarkable study of Robinson, and after innumerable procrastinations, the international scientific community has definitively banished it from the panoply of useful measures in a pregnant woman at the beginning. in fact, it limits the volemic expansion and therefore risks increasing fetal suffering; it also has no preventive effect of preeclampsia as had been expected in a time. The doubt that still exists in some people after 40 years of proof is therefore difficult to understand!

Most other dietary manipulation attempts were unsuccessful and were abandoned in their turn. The interest of an increased calcium intake remains debated, but keeps convinced supporters. It is probably useful at least in populations with deficient calcium intake.


We do not dwell on this subject, which is in fact specialized resuscitation units. The usual severity of hypertension makes its treatment indisputable. This one is usually parenteral. The number of usable drugs is here more limited.

While hydralazine remains the preferred treatment in the United States, other products are more widely used in Europe. A recent meta-analysis has shown no superiority of hydralazine over other parenteral drugs. Labetalol has been the subject of many studies of good quality, and its effectiveness as well as its safety can be held for some.Urapidil has been less studied, but seems to compare favorably with hydralazine. Nicardipine, a favorite in France, has not given rise to any acceptable controlled study. Finally, the rapid forms of nifedipine, proposed in a time, are currently contraindicated in any antihypertensive treatment according to all recommendations, French and international.

This treatment should be conducted gently despite the seriousness of the situation. A plateau should be reached in a few hours to a diastolic not less than 100 mmHg. Decay at around 90 mmHg should only be done secondarily and more slowly. Aggressive treatment exposes both maternal complications and rapid fetal death.

Some extra measures have been proposed in very severe forms. Their effectiveness is difficult to judge because they are applied late, in indications where the prognosis is generally very pejorative. Thus, heparin therapy and plasma volume expansion have been used with varying degrees of success. Their indications must be carefully weighed in a specialized environment.


All that has just been stated clearly indicates that the medical treatment of pregnancy-induced hypertension is most often disappointing.

It does not change the forms whose prognosis is spontaneously benign, and makes it possible to gain very little ground in the severe forms. Although the maternal and fetal prognosis in gestational hypertension has improved significantly over the last two decades, it is not he who can be credited with it, but the steady progress that has been made in surveillance and tactics. obstetric. In fact, stopping pregnancy is the only measure that puts an end to maternal hypertensive and proteinuric manifestations. It is therefore this decision which must be taken without hesitation in serious forms when fetal distress is announced, without placing in medical treatment a hope that is likely to be disappointed. To do so, however, is only possible at a sufficiently advanced time for the neonatal risk to be acceptable. And on this point, the steady progress of neonatology has made it possible to approach fetal extractions with almost inconceivable terms until recently.

It is before this term that all medical resources must be put into play, in order to gain some precious weeks of fetal maturity. We do not dwell here on obstetric methods, nor do we consider the treatment of eclampsia and HRP.

Preventive treatments:

If the primum movens of gravid hypertension is placental ischemia, the disappointment brought by the medical treatment is not surprising conceptually, since it is a symptomatic treatment, acting downstream of the motor phenomenon. To act on this phenomenon is conceivable only as a precautionary measure, before the irreversible placental lesions are formed and when the symptoms which are their consequence appear.


Ideally, preventive treatment should:

– be instituted very early, that is to say when the anomalies due to the defective trophoblastic invasion begin to appear;

– have an antithrombotic action, maybe even anti-inflammatory;

– restore the correct balance between prostacyclin and thromboxane, by a relatively selective inhibition of the latter.

Low dose aspirin is a consistent solution to the problem. It exerts on the placental arteries in vitro an action comparable to that shown in other systems-inhibition of thromboxane synthesis with relative respect to that of prostacyclin. In vivo, low doses of aspirin cause, in pregnant women, a reduction in the urinary elimination of thromboxane B 2 , without modification of the elimination of 6-ketoprostaglandin F 1 a . The production of thromboxane is also inhibited in the fetus.


We reported in 1985 a first controlled study of aspirin treatment in “high-risk” pregnancies because of a history of pathological obstetrics. This pilot study showed almost complete prevention of preeclampsia and fetal growth retardation. In the years that followed, several other controlled studies came to corroborate ours.

These studies all had in common that they involved, despite very different inclusion criteria, high-risk patients. This is evidenced by the high rate of preeclampsia and fetal accidents in the control series.

A second series of studies, on a very large scale, was undertaken, on patients this time not or very seldom selected.Two US studies, involving unselected primiparas, have confirmed significant prevention of preeclampsia. Two other studies did not show any effect. They involved patients selected on the basis of a risk qualified as “average”, with rather vague criteria. Patients considered to be at high risk were excluded because, at the time, there was a belief that aspirin therapy provided them with a very real benefit. Testimony of this “counter-selection” the particularly favorable prognosis of pregnancies, including in the control series. On the other hand, these studies have provided extremely reassuring data on the safety of aspirin for both mother and child.

These latest studies, in particular the gigantic CLASP study with its almost 10,000 patients, have nevertheless thrown doubt into the minds and the usefulness of aspirin has been questioned. The publication of Caritis et al completed the disorder because it involved high-risk patients (hypertensive, diabetic, twin pregnancies, history of preeclampsia …).The incidence of preeclampsia was about 20% in both the control and aspirin groups. After this publication, some authors (mostly the most enthusiastic before) considered that the case was closed and aspirin ineffective.


Without going into the details of the argument, it emerged that the recent negative studies had suffered from a very heterogeneous selection, delay of treatment initiation up to 32 weeks and doses of aspirin too low (in 60 mg / day).

It appears that overall, in spite of the negative studies of considerable numbers, the treatment remains active on the fetal growth (a last meta-analysis practiced after the publication of Caritis et al shows that this situation is always unchanged). If the dose of aspirin is at least 100 mg / day, the efficacy appears much higher, and even a significant effect on perinatal mortality is observed, which no individual study had shown, because of the the happy rarity of this complication.

Efficacy is also greatly enhanced if treatment is started before 17 weeks.

In any case, the odds ratios appearing in this meta-analysis speak for themselves against the accusation of inefficiency or marginal efficiency of which aspirin is today the object. A retrospective study of patients who received aspirin in our department confirmed the critical importance of early treatment and showed that longer bleeding time with aspirin was also an important factor in the success of this treatment.

By analogy with other situations where aspirin has been shown to be effective, an increasingly precocious or even preconceptional treatment could be considered. The addition of low doses of corticosteroids is another possibility.These attitudes are for the moment either anecdotal observations or short series, and therefore can not be recommended on a larger scale until more consistent evidence has been provided.

The combination of aspirin and heparin, or the substitution of aspirin with heparin, is also discussed, with a level of evidence that is still far below the desirable minimum. Nevertheless, these different hypotheses under test suggest the possibility of serious changes in strategy in the next decade.


The fact of having a preventive treatment raises the problem of its indications. The need for a very early treatment, largely prior to any maternal symptom, centers the question on an early prediction. This problem is not resolved at this time.

The knowledge of the patient’s antecedents has shown good effectiveness, but on the one hand it remains relatively empirical, on the other hand it is applicable only after accidents have already occurred, which is not satisfactory. . We have no reliable biochemical marker at such an early stage. Some studies give hope that a doppler study could have a good discriminative value between primiparous people who will have preeclampsia or not. This discrimination, if it seems to be confirmed, remains at a later date than the desirable end of treatment.This earlier prediction remains one of the main challenges in the years to come.

The long-term future:


It is a tradition that patients who have had isolated and / or early hypertension during pregnancy are exposed to almost systematic recurrence over subsequent pregnancies. This fact, although inconstant, is easily explained by the fact that these patients have a high vascular risk and a family site of hypertension. This does not necessarily imply a worsening of the prognosis of these pregnancies, these hypertension remaining generally benign.

More debated is the meaning that must be given to the “pure” pre-eclampsia of the primipara. We studied subsequent pregnancies in 221 patients who were in this case. It can be seen that less than 30% of pregnancies are normal. Half of them are marked by isolated hypertension, and the recurrence of preeclampsia itself is not exceptional. Moreover, in these patients, the occurrence of major accidents (HRP, fetal death, growth retardation) is also a frequency much higher than the norms. It is therefore clearly not a specific pathology of the first pregnancy and does not run any risk for subsequent pregnancies. It is necessary to take the greatest account for the management of these pregnancies.

Sibai et al studied the subsequent pregnancies of 406 women with severe pre-eclampsia of the primipara.Preeclampsia occurred in the second pregnancy in 46% of cases (compared to 7.6% in a matched control population).The same authors studied 223 patients who had eclampsia; 22% of subsequent pregnancies were complicated with preeclampsia, 1.9% with eclampsia, 2.5% with HRP, 2.7% with fetal death. In this study, the risk for pregnancy was significantly higher in women whose eclampsia had occurred before 30 weeks. For our part, we did not notice this difference between early and late preeclampsia.

Other, smaller studies have confirmed that women with pre-eclampsia during first pregnancy are at high risk of later complicated pregnancies.


The occurrence of hypertension during pregnancy is, in fact, unlikely to be independent of the baseline vascular risk, and therefore the vascular future of these young women is included.

We studied the blood pressure future of 941 patients who had a report 3 months after hypertension in pregnancy.Among them, 33% remained hypertensive thereafter (26% if we limit ourselves to those known as normotensive before pregnancy). This figure is in large excess over that of the control population, which accounts for 1.79% of hypertensive women in this age group. Of the patients who remained hypertensive, 78% were primiparous and 40% had hypertension in the third trimester, with (20%) or without (20%) proteinuria. The different risk factors mentioned above were distributed identically between primiparous and multiparous, and between early and late hypertension. Prolonged follow-up showed that 20% of patients who remained normotensive after pregnancy became hypertensive in the years that followed.

Fisher et al. Studied the incidence of subsequent hypertension in patients with “pure” pre-eclampsia, a clinical diagnosis confirmed by the absence of any other lesions in renal histology. This frequency is similar to that observed in the female control population of the same age. On the other hand, if one isolates from this control population the women having one or more pregnancies, all normotensives, the frequency of the hypertension is extremely low. This means that in this age group, hypertensives are women who have had pregnancy-induced hypertension or who have not had pregnancies.

Sibai et al also showed a very high incidence of permanent hypertension in women with preeclampsia, especially if it was early (before 30 weeks) or recurrent.

Other long-term retrospective studies have confirmed these facts. It is impossible here not to mention the large longitudinal study of Chesley, unique in its duration (more than 40 years!), Which has shown that the cardiovascular mortality of women who had eclampsia in the past was much higher, even higher. when it was a multipare than when it was a primiparous.

In other words, it can be taken for granted that hypertension during pregnancy (preeclamptic or not, first pregnancy or not) unmasks in a very large number of cases a hypertensive tendency which will prove to be more or less long term.This fact is essential for the subsequent medical follow-up of these young women.


Hypertension in pregnancy is common and remains a major cause of maternal and fetal mortality and morbidity. It is by far the leading cause in developed countries and the third in developing countries (after infection and hemorrhage).The knowledge of its physiopathology, still incomplete, is currently in rapid progress and a certain overall logic is gradually emerging.

To sum it up in two words, the base is a very early disorder of the trophoblastic invasion, compromising the blood supply to the fetoplacental unit, resulting in a diffuse pathology of the endothelium. The latter is responsible for intense vasoconstriction and a thrombotic tendency which conditions the various visceral manifestations or complications.

The symptomatic treatment of hypertension does not improve the prognosis of these pregnancies, but can protect mothers against acute accidents if the hypertension is particularly severe. Preventive treatments, especially low-dose aspirin, if used very early in pregnancy, can significantly improve the prognosis. Finally, many of the patients who have suffered from this condition will have recurring accidents over the course of pregnancy, and are future hypertensives.