Tải bản đầy đủ - 0 (trang)
1 Most Investigated Causes and Risk Factors of Recurrent Miscarriage

1 Most Investigated Causes and Risk Factors of Recurrent Miscarriage

Tải bản đầy đủ - 0trang


Most Investigated Causes and Risk Factors of Recurrent Miscarriage


abnormalities can be detected. In the remaining cases, numerical CAs were found

(Desjardins and Stephenson 2012; Kochhar and Ghosh 2013). In light of these facts,

these authors question the necessity of routine chromosomal testing of parents, in

accordance with other researchers (Branch et al. 2010; Tang and Quenby 2010).

Therefore, it is likely that structural chromosomal abnormalities are rarely the cause

of repeated, consecutive abortions.


Anatomical Factors

Malformations resulting from the impaired fusion of the Mullerian duct (the most

common forms are subseptated, septated and bicornuate uterus) occur in approximately 4 % of patients with normal obstetric history whereas in 8 % of patients with

recurrent miscarriage (Sugiura-Ogasawara et al. 2010). In view of the structural

anomalies of the uterus, their role in the occurrence of abortions seems almost evident, although clinical experiences appear to contradict this. Even together with

these developmental abnormalities, a ratio of 60–70 % successful births can be

achieved in patients with several preceding spontaneous abortions (Acien 1993;

Kirk et al. 1993), and a higher success rate (60–75 %) is not even achieved after

metroplasty (Ayhan et al. 1992; Acien 1993; Fedele et al. 1993). Moreover, after

surgery, infertility emerged in 10–25 % of patients (Fedele et al. 1993). SugiuraOgasawara et al. (2010) found that in women with 2–12 abortions, the live-birth rate

of the first pregnancy following examination was lower than in recurrent miscarriage patients with normal uterine cavities (59.5 % and 71.7 %), although the difference in cumulative birth rates was not significant after three pregnancies (70.8 and

85.5 %). In patients with uterine cavity alterations whose pregnancies ended with

abortion or birth, it was mainly the D/C ratio (septum length/remaining uterine cavity length) that showed significant differences, as other authors have also stated

(Salim et al. 2003; Sugiura-Ogasawara et al. 2010). It is also against the primary

causal role of septated uterine cavities that after the complete, controlled hysteroscopic resection of the septum, the rate of abortions hardly changes (34.1 %) and

conception occurs only in 65 % of patients over 3 years (Paradisi et al. 2011). In

mature births after septum resection, the weight of the newborn is significantly

lower than in women with normal uterine cavities (Agostini et al. 2009).

Out of 1000 couples, 50 (5 %) are likely to suffer from RM. Therefore, in 50

women with recurrent miscarriage, the above-mentioned anomalies are expected to

develop in four (8 %) and in 38 (4 %) of the remaining 950 couples. Thus, the question arises: why only approximately every 10th woman with uterus alteration will

develop recurrent miscarriage? It seems plausible that causal factors other than uterine cavity abnormalities are required for the development of habitual abortion.

Moreover, in the presence of uterine abnormalities, 85 % of women with recurrent

miscarriage usually give birth in their third or fourth pregnancies (Jaslow et al.

2010). After their fourth pregnancy, only a small proportion did not give birth,

which is another factor that makes the primary causal role of these alterations doubtful in RM.



Recurrent Miscarriage and Folliculo-Luteal Function

Acquired anatomical conditions, myoma and Asherman’s syndrome only rarely

cause RM. In cases of myoma, the size and position determine possible surgical

treatment. In cases of uterine polyp, removal is recommended.

The exact causal role of cervical incompetence is also unclarified, although it is

primarily regarded to be a consequence of surgical intervention. The diagnosis is

most often based on the occurrence of pregnancies aborted in the second trimester

(Rock and Murphy 1986; Gardó 1993). The efficacy of cerclage operations has not

been clearly demonstrated either. According to randomised, multicentred trials, its

efficacy could be confirmed in 1 out of 25 patients in a high-risk group containing

1200 patients (MacNaughton et al. 1993).



Alterations causing thrombophilia can be detected in 20 % of women who had complication-free births (Factor V Leiden mutation, activated C-protein resistance, protein S deficiency, etc.). The causal role of these conditions in the development of

recurrent miscarriages was anticipated earlier. However, a similar percentage of

these patients give birth successfully as the group with unexplained abortion (Clark

et al. 2010) and together with low-dose aspirin and heparin treatment in a similar

ratio as the placebo group (Visser et al. 2011). There are no randomised, placebocontrolled studies available that would confirm the causal role of thrombophilia in

RM (Rodger 2011, 2013). Therefore, heparin treatment is only considered justified

in case of history with thromboembolism (Branch et al. 2010; Tang and Quenby



Immunological Factors

From the immunological causes, only the causal role of antiphospholipid syndrome

(APS) is debated in RM. While some authors have failed to demonstrate a difference between the pregnancy outcome in case of APS and in unexplained recurrent

miscarriage (Cohn et al. 2010), others have found the abortion rates to be lower

under heparin treatment (Tang and Quenby 2010). Further investigations are

required to confirm the efficacy of heparin treatment (Tang and Quenby 2010).

Prednisone and aspirin treatment have been proven to be ineffective in APS (Laskin

et al. 1997; Rai and Regan 2006). Furthermore, the prevalence of gestational diabetes, hypertension and preterm birth increases during prednisone treatment, which

means its application is contraindicated (Branch et al. 2010). Pregnancies conceived

in women with APS require intensive monitoring because of the higher occurrence

of thromboembolic complications.

The prevalence of APS is 3–5 % in the general obstetric patient population and

5–15 % in women with recurrent miscarriage (Branch et al. 2010). On this basis it is

difficult to interpret why only every 9th–10th woman with APS will have recurrent

miscarriage, and after two abortions and approximately 85 % out of them will give


Most Investigated Causes and Risk Factors of Recurrent Miscarriage


birth without treatment in their third or fourth pregnancy (Jaslow et al. 2010), not to

mention those who have uncomplicated births despite their APS-characteristic laboratory results. This observation argues against the primary causal role of APS in

RM. It also contradicts the causal role of APS and that the presence of antiphospholipid antibodies does not impact the outcome of pregnancies during IVF (ASRM

2006a). Other authors have also found favourable pregnancy outcomes in APS

cases, excluding patients with SLE and thromboembolism (Bowman et al. 2015).

Other immunological causes (HLA, immunity against the father’s leukocytes or

antibodies synthesized against other alloantigens) were considered as contributing

factors in the development of habitual abortion, and immunotherapeutic attempts

were made to treat them by immunisation of donor cells or cells originating from

the father (e.g. white blood cells or thrombocytes). It has now become evident that

these attempts were in vain and they do not improve the rate of successful pregnancy compared to the placebo group (Tang and Quenby 2010). At the same time,

the level of alloantibodies does not show any correlation with pregnancy outcomes

(Bartel et al. 2011).

Many authors attribute a role to the elevated number of NK cells and the abnormal cytokine levels of the endometrium (e.g. Th1/Th2 level) in the development of

recurrent miscarriage. Szekeres-Barthó et al. have confirmed in numerous publications the fundamental role of progesterone in the immunoprotection of pregnancy,

mainly via the activity of progesterone-induced blocking factor (PIBF). PIBF

favourably influences the number of NK cells and the cytokine levels and ratios

(Szekeres-Barthó 1991: Szekeres-Barthó and Balasch 2008; Halász and SzekeresBarthó 2013). P after adequate preceding oestradiol level plays an indispensable

role in ensuring the optimal circumstances for implantation. Low E2 and then low

P levels make not only the implantation inadequate but, based on the above, the

proper immunoprotection of pregnancy also. The mother’s body strives to “get rid”

of the foetus, which appears foreign to it, thus immunological processes definitely

have a role in the abortion of pregnancies with insufficient P level. However, P

administered in an early phase of pregnancy can no longer influence implantation.

With the mere administration of an appropriate amount of E2 and then P, an

adequate endometrium can be developed, which is suitable for physiological

implantation and successful childbearing, for example, with oocyte donation

(Paulson 2011; Young 2013). This implicates that the physiologic production of NK

cells and cytokines are essentially determined by these two hormones.

Immunoprotection of the pregnancy is mainly provided by the correct P level

(Druckmann and Druckmann 2005; Szekeres-Bartho et al. 2008), which is proven

by the above. However, this immunoprotection is expected to be less secure if the P

level falls. According to some observations, immunological factors might underlie

placentation abnormalities of different severities (Matthiesen et al. 2012). Yet this

could be the result of the weakened maternal immunoprotection stemming from the

low level of P that causes the mother’s body to strive to “get rid” of the foetus that

appears to it as foreign through immunological mechanisms. (Arck et al. 2008)

found significantly lower levels of P and PIBF in pregnancies that were aborted

later compared to women with habitual abortion who had successful births. Maternal



Recurrent Miscarriage and Folliculo-Luteal Function

immune response can thus be a natural consequence of the weakened immunoprotection and it probably cannot be assigned to specific immunological alterations.

Based on the suspected immunological causes, earlier attempts were made to use

IVIG (intravenous immunoglobulin) treatment, although it has now become evident

that IVIG treatment is ineffective in any form of recurrent miscarriage as it results

in the same successful pregnancy rates as in the placebo group (Branch et al. 2010;

Stephenson et al. 2010; Tang and Quenby 2010). Ata et al. (2011) summarised the

results of six randomised, placebo-controlled trials (N = 272) and failed to demonstrate any differences between the successful births rates in the treated and control



Hormonal Causes

Many conditions were presumed candidates as causes underlying RM, yet their

direct role seems minor. In polycystic ovary syndrome (PCOS), the role of abnormal sugar load or insulin resistance also emerged as a possibility, but metformin

treatment proved to be ineffective to improve successful birth rates in RM (ESHRE

and ASRM 2007, 2012). Well-compensated diabetes does not impact the development of habitual abortion, similarly to well-compensated thyroid disease. The role

of low progesterone level produced during the luteal phase was also suspected, but

progesterone replacement did not fulfil its promise, although physiological values

have not yet been established (Branch et al. 2010; Tang and Quenby 2010; Ata et al.

2011). The effect of P is essentially determined by the preceding oestradiol value.

If the level of oestradiol is low during the follicular phase, the developed endometrium that is less thick and poor in P-receptors cannot be normalised by P administered after ovulation (only in borderline cases); moreover, it can no longer influence

the maturation process of the oocyte.

In luteal insufficiency, inadequate secretory transformation can be recognised by

histological examination of the endometrium in approximately 35–45 % of RM

cases, but P levels obtained from a single sample in the middle of the luteal phase

exhibit low values (<10 ng/ml) only in half or one-third of the cases (Li et al. 2002).

This is usually explained by the decreased reactivity of the endometrium. Another

possible explanation might be that the actual physiological level P is higher than we

currently think. We found significantly higher physiological average luteal P levels

typical in births (21.0 ng/ml) during the 6-day plateau observed in the luteal phase.

In an unselected group of women with recurrent miscarriage, we found P levels

under 10 ng/ml in only 16 % of the cases (Siklósi et al. 2012) which corresponds to

the occurrence rates by other authors (15–17 %) (Li et al. 2002; Jaslow et al. 2010).

The episodic nature of P secretion was confirmed long ago (Filicori et al. 1984) and

we measured its extent as ±40.3 % (95 % CI) (Siklósi et al. 1984). A single value of

10 ng/ml means that the actual level of P is between 6 and 14 ng/ml.

The role of folliculo-luteal insufficiency is indicated by the proliferative endometrium being significantly less thick in RM than in physiological cycles (Bromer

et al. 2009). The plasma level of luteal P is significantly decreased in RM than in


Most Investigated Causes and Risk Factors of Recurrent Miscarriage


verified fertility, and the P content of the endometrium is 200 times lower (Salazar

and Calzada 2007). In RM, the luteal P values in pregnancies ending with abortion

are significantly lower than in pregnancies ending in birth (Arck et al. 2008; Siklósi

et al. 2012). Moreover, the level of corticotropin-releasing factor is significantly

higher (Arck et al. 2008), and blood flow of the uterine artery and the secretory

endometrium is also decreased (Nakatsuka et al. 2003; Chen et al. 2011).


Psychological Factors

The negative effects of stress – and especially mental stress – on reproductive functions have already been confirmed (Chap. 3). Several authors demonstrated elevated

psychosocial stress levels in RM (Sugiura-Ogasawara et al. 2002; Arck et al. 2008; Li

et al. 2012). The fact that the cortisol and corticotropin-releasing factor (CRF) levels

are significantly higher in aborting women compared to that shown in successful birth

also supports the role of increased stress (Arck et al. 2008). It is likely that tender loving care also affects stress relief (Sugiura-Ogasawara et al. 2002; ASRM 2012).


Infectious Origin

The role of chronic or repeated infections in RM was investigated, especially concerning mycoplasma, toxoplasma and listeria infection cases. Mycoplasma hominis

and Ureaplasma urealyticum can be detected in almost 50 % of women without any

symptoms, and we do not know of any report that would have identified them in

women with recurrent miscarriage (Byrn and Gibson 1986). Although Toxoplasma

gondii infections can lead to severe congenital disorders, its recurrent occurrence

could not be observed in the pregnancies of over 400 women (Byrn and Gibson

1986). While bacterial or viral infections may cause sporadic abortion, the existence

of any infection that would cause RM is unknown (ASRM 2012).


Unknown Origin

In about half of women with recurrent miscarriage, none of the above-mentioned

causes or risk factors can be detected. Despite this, their pregnancy outcomes are

almost identical to those cases in which their presence is confirmed, and the prevalence of numerical CAs is around 50 % in case of abortions and in the presence of

the listed causes and risk factors alike. If neither the occurrence rate of abortion nor

the prevalence of numerical CAs differs substantially between the two groups, the

question arises whether these causes actually have any role in RM or whether the

primary cause of this condition is a currently unrecognised common factor (or


Many treatments were used in unexplained RM, but none of them proved to have

a positive effect. In a randomised, double-blinded trial, neither aspirin nor heparin



Recurrent Miscarriage and Folliculo-Luteal Function

treatment improved pregnancy outcomes (Clark et al. 2010; Kaandorp et al. 2010),

and IVIG and progesterone therapy were also ineffective (Tang and Quenby 2010).

These factors illustrate that it is quite challenging to clarify the actual role of a

given cause in RM. For example, without treatment, abortion occurs in 45 % of

cases and birth in 55 % of cases after three previous abortions. When we achieve

birth by any treatment after three abortions, we consider it to be the result of the

treatment itself. However, any given treatment has a real effect only if it causes a

significantly higher proportion of patients than 55 % to successfully give birth.

Therefore, the actual effect of a treatment can only be confirmed by randomised,

placebo-controlled studies. When summarising six studies from recent years regarding IVIG treatment (in 272 patients), no difference was found between the birth

rates of the placebo and the IVIG treatment group (ATA et al. 2011). Based on this

and on other similar studies, it became clear that IVIG treatment is ineffective for

treating recurrent miscarriage. The situation is also similar for other suspected

causes: whereas the studies later failed to confirm the role of numerous suspected

causes, treatment procedures based on them became accepted practice over time.


Why the Above Enlisted Causes Cannot Be the Real

Cause of Recurrent Miscarriage

Based on the well-known epidemiological data, it seems obvious that the causes

mentioned above cannot be the real cause of RM (or only in rare cases).

1. It is a common observation that the probability of recurrent abortions increases

with the number of abortions in the anamnesis of RM. The occurrence of abortion is typically 33 % after two, 45 % after three, 55 % after four, 65 % after five

and 70 % after six abortions (average occurrence rates) (Branch et al. 2010; Tang

and Quenby 2010; ASRM 2012). The prevalence of preterm birth and intrauterine growth retardation (IUGR) increases in proportionally with the number of


Although the number of miscarriages in the anamnesis makes women more

and more susceptible to miscarriage in the future – which justifies and explains

the worsening fear of another abortion the patient (and the doctor) feels – the

number of patients without children continuously lessens in the patient population, based on the number of births. Based on average recurrence probabilities

and rounded patient numbers, 67 % of pregnancies end with birth without any

treatment, in 100 women with habitual abortion after two abortions in their medical history. Out of the remaining 33 women, after three abortions, 18 of pregnancies end with birth, and 15 will miscarry. Out of these 15 women with four

abortions in the anamnesis, 7 of pregnancies end in birth and 8 in pregnancy loss.

These eight women who already have five miscarriages give birth after their next

pregnancy in three and five will have a spontaneous abortion. Finally, the remaining five women will have successful birth in two after six spontaneous abortions

and three will lose pregnancy (and still there is a 15–20 % chance for giving birth


Why the Above Enlisted Causes Cannot Be the Real Cause of Recurrent Miscarriage


Table 5.1 Pregnancy outcomes during the third to seventh pregnancies of 100 untreated patients

after two miscarriages

Preceding no. of miscarriages







No. of patients

Birth N (%)






67 (67 %)

18 (55 %)

7 (45 %)

3 (35 %)

2 (30 %)

97 (97 %)

Miscarriage N (%)

33 (33 %)

15 (45 %)

8 (55 %)

5 (65 %)

3 (70 %)


after seven or more miscarriages). Altogether, in 100 patients with two previous

abortions, 97 will give birth over five consecutive pregnancies. Out of the pregnancies following two abortions, 40 % will end in miscarriage and out of those

giving birth, 40 % will have preterm birth (Table 5.1).

Based on epidemiological data, merely 3 % of patients fail to give birth without treatment between their third and seventh pregnancies. Therefore, up to 3 %

of patients with RM may have a permanent cause underlying pregnancy loss,

which would lead to the abortion of all further pregnancies. This observation

seriously calls into question – or rather excludes – the actual role of immunological factors, thrombophilia, subseptated or septated uterine cavity, etc. in

RM, as the spontaneous cessation of these causes can hardly be expected. The

failure of treatments based on these presumed causal factors also seems to support this conclusion.

We certainly do not present these results in order to advocate neglecting therapeutic interventions, but solely because of the logical conclusions they imply

with regards to the cause of RM. A majority of patients do not get as far as seven

abortions. Some of them “surpass” the reproductive age in time, which is further

facilitated by the fact that in 15 % of patients, miscarriage is followed by a minimum 2-year infertile period and a significant proportion of them give up trying.

2. According to the observations of several authors (Coulam 1991; Jaslow et al.

2010), the prevalence of the above listed causal or risk factors – immunological

causes, thrombophilia, anatomical conditions and other causes of unexplained

origin (in about 50 % of patients) – is approximately identical after three, four or

five abortions without treatment. ASRM modified the definition of RM in 2008

(two or more abortions instead of three or more) because the occurrence rates of

the causal factors investigated until now are the same after the second and the

third or further abortions. Therefore, there is no point waiting until the third

abortion to perform medical examinations.

This observation also means that the chance of losing the following pregnancies is almost the same whether these causes are present or not. If the prevalence

of miscarriage was higher with any of these causal factors after, for example, two

abortions, then this cause would have to occur more frequently during the following pregnancies compared to the other factors and so on. However, based on

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

1 Most Investigated Causes and Risk Factors of Recurrent Miscarriage

Tải bản đầy đủ ngay(0 tr)