Keywords antenatal corticosteroids - premature birth - timing - time interval
Introduction
The administration of antenatal corticosteroids (ACS) in premature births delivered
prior to gestational week (GW) 34 leads to a significant reduction in perinatal morbidity
and mortality
[1 ]
[2 ]. Their effectiveness depends crucially on optimal timing [3 ]. As shown in a meta-analysis from 2006, there is no effect on respiratory distress
syndrome (RDS) < 24 hours after the first administration of betamethasone (RR 0.87,
95% CI 0.66–1.15); this effect only becomes apparent at < 48 hours (RR 0.63, 95% CI
0.43–0.92). A significant effect can therefore be expected > 24 hours after the first
administration
of ACS. After a period of seven days following corticosteroid administration, no further
reduction in RDS can be detected (RR 0.82, 95% CI 0.53–1.28) [4 ]. However, a prospective cohort study by the German Neonatal Network describes positive
effects on the rate of intraventricular brain hemorrhage (OR 0.43, 95% CI 0.25–0.72)
and on the need for mechanical ventilation (OR 0.43, 95% CI 0.27–0.71) even seven
days after administration of ACS [5 ]. Nevertheless, there is also evidence that extremely premature infants born before
GW 28 who are delivered more than 10 days after the first corticosteroid administration
have a more than twofold higher rate of brain hemorrhage (17% vs. 7%; aOR 4.16, 95%
CI 1.59–10.87)[6 ], although the retrospective design of this study of course does not allow a causal
conclusion to be drawn. In addition, children born full-term after antenatal
administration of corticosteroids may have more mental and behavioral impairments
compared to children born full-term who did not receive corticosteroids during pregnancy
(HR 1.47, 95% CI
1.36–1.69). This effect persisted even after taking socioeconomic influences into
account (HR 1.38, 95% CI 1.21–1.58) [7 ]. The same applies to the psychological and neurosensory development of these children
[8 ]. A further prospective cohort study confirms these results for dexamethasone [9 ].
However, as numerous studies have shown, only 25–40% of patients are delivered within
the optimal timeframe after administration of ACS [10 ]
[11 ]
[12 ]. The optimal timing depends crucially on the indication for administration of ACS
[10 ]. For example, this treatment is clearly very successful in cases of severe pre-eclampsia
or preterm premature rupture of membranes (PPROM), but appears to be less
successful in patients with asymptomatic cervical insufficiency ([Fig. 1 ]) [10 ]. In this review, we explain the reasons for this and present ways for further optimization.
Fig. 1
Interval between administration of antenatal corticosteroids (ACS) and delivery according
to the indication (data from [10 ]) Patients as % with an interval between ACS administration and delivery of < 24
hours (blue), 24 hours to 7 days (red), and over 7 days (green). Asymptomatic =
positive fibronectin test, shortened cervical length, asymptomatic cervical opening;
Fetal = intrauterine growth restriction, oligohydramnios; HTN = hypertensive diseases
of pregnancy;
Maternal = maternal diseases other than pregnancy-induced hypertension; PPROM = preterm
premature rupture of membranes.
Literature Search
A selective literature search up to May 2023 was conducted in PubMed for the keywords
“corticosteroid”, “timing”, “preterm birth”, “preterm delivery”, “pregnancy prolongation”,
and “delivery
delay”. Prospective randomized trials, reviews, and meta-analyses relevant to the
topic were selected. Cross-references to other important studies have been taken into
account.
PPROM
More than half of all patients who undergo PPROM deliver within a week. The median
duration of pregnancy in a cohort of 239 patients who were negative for B streptococci
was 6.1 days. The
cumulative delivery rate was 27% after 48 hours, 56% after 7 days, 76% after 14 days,
and 86% after 21 days [13 ]. The latency period prior to birth is inversely correlated to gestational age at
the time of PPROM [14 ]; the greater the remaining volume of amniotic fluid, the longer the latency [15 ]. Spontaneous rupture of membranes is very rare, unless it is the result of an amniocentesis
[16 ].
In view of the high probability of delivery within one week of PPROM, the administration
of ACS is indicated in these cases. However, the question arises as to what clinical
management should
look like for the approx. 50% of women who have not yet given birth seven days after
PPROM. The guideline “Prevention and treatment of premature birth” from the Association
of Scientific
Medical Societies in Germany (AWMF) recommends the following: Women treated more than
7 days previously with steroids for threatened preterm birth before week 29+0 of gestation,
may receive a
further single dose of steroids after re-evaluation if they have an increasing risk
of immediately threatened preterm birth [17 ].
ACS Booster
Due to the 50 percent likelihood that women who are still pregnant one week after
PPROM will give birth in the following week, a repeat administration of ACS appears
to make perfect sense.
However, a recent prospective randomized study shows that this view is controversial
[18 ]. In this study, 192 patients with PPROM occurring between GW 24+0 and GW 31+6 who
had already received one dose of ACS and were still pregnant after seven days were
randomized to receive either a second dose of ACS (booster) or placebo. The primary
study endpoint was combined neonatal morbidity or neonatal death. No significant difference
was found
between the groups for either the primary or secondary study endpoints. Moreover,
this observation was independent of the time interval between ACS booster and delivery,
as well as the
gestational age at birth ([Table 1 ]) [18 ]. Two further studies based on a secondary analysis of a prospective randomized study
on neuroprotection with magnesium were also unable to demonstrate any effect of an
ACS booster on the RDS rate [19 ]
[20 ].
Table 1
ACS booster for preterm premature rupture of membranes. 192 patients with preterm
premature rupture of membranes occurring between GW 24+0 and GW
31+6 who had already received antenatal corticosteroids (ACS) and were still pregnant
after seven days were randomized to receive either a second administration of ACS
(booster) or
placebo. The primary study endpoint was combined neonatal morbidity or neonatal death
(data from [18 ]).
ACS Booster
n = 94 (%)
Placebo
n = 98 (%)
P value
Primary study endpoint: neonatal morbidity and/or mortality
60/94 (64)
63/98 (64)
0.54
Individual components of the primary study endpoint
Respiratory distress syndrome
57/94 (61)
63/98 (64)
0.44
Bronchopulmonary dysplasia
13/94 (14)
11/98 (11)
0.67
Intraventricular brain hemorrhage, grade 3 and 4
4/94 (4)
3/98 (3)
0.69
Periventricular leukomalacia
0/94 (0)
1/98 (1)
0.31
Sepsis detected by culture
5/94 (5)
3/98 (3)
0.44
Necrotizing enterocolitis
4/94 (4)
3/98 (3)
0.70
Neonatal death
2/94 (2)
4/98 (4)
0.38
Primary study endpoint according to the time interval between ACS booster and delivery
< 48 hours
7/9 (78)
5/12 (50)
0.47
24 hours to < 7 days
25/35 (71)
15/23 (65)
0.88
48 hours to < 7 days
22/32 (67)
13/18 (72)
0.58
7 to < 14 days
8/14 (57)
20/31 (65)
0.83
≥ 14 days
20/35 (57)
26/35 (74)
0.19
Primary study endpoint according to gestational age at delivery
Delivery < GW 29
33/36 (92)
23/28 (82)
0.26
Delivery ≥ GW 29
27/58 (47)
42/70 (60)
0.16
It is known that an infection can impair the effectiveness of glucocorticoids. Webster
et al. showed that TNF-α, an inflammatory mediator, stimulates synthesis of the glucocorticoid
β-receptor and thus induces glucocorticoid resistance [21 ]. Endotoxins also modulate glucocorticoid receptor expression and the associated
signaling mechanism [22 ]. Ascending inflammatory processes may have impaired the efficacy of the second administration
of ACS in the above-mentioned study.
However, the data on the administration of an ACS booster with an intact amniotic
sac is also mixed. While Garite et al. were able to observe a reduction in RDS after
a second administration
[23 ], this effect has not been demonstrated in other studies [24 ]
[25 ]. However, RDS was defined differently in these studies. Taking this into account,
it is very likely that the administration of an ACS booster only reduces the incidence
of mild RDS, whereas it has no effect on the incidence of severe RDS or other parameters
of neonatal morbidity [24 ]. In general, caution is advised when using an ACS booster, as it can lead to a significant
increase in small for gestational age (SGA) infants (4.9% vs 10.6%; aOR, 1.63;
95% CI, 1.07–2.47) [24 ]. This fact is taken into account in the AWMF guideline “Prevention and treatment
of premature birth”. In this guideline, a booster is only recommended before GW 29
and
only if there is a very high risk of a premature birth occurring within seven days
[17 ].
Hypertensive Disorders of Pregnancy
Hypertensive Disorders of Pregnancy
The sFlt/PlGF ratio can be helpful in identifying patients likely to develop pre-eclampsia
during the course of their pregnancy [26 ]. For example, the Pregnancy Outcome Prediction Study showed in a non-selected patient
cohort that the positive predictive value (PPV) of an sFlt/PlGF ratio > 38
measured at GW 28 identified 32% of patients who went on to suffer a premature birth
due to pre-eclampsia [27 ]. A value of > 38 between GW 30 and GW 37 detects 79% of all patients who have to
be delivered within a week due to pre-eclampsia, with a false-positive rate of 4.5%
[28 ].
The INSPIRE trial showed that an sFlt/PlGF ratio > 85 had a PPV of 71.7% for development
of pre-eclampsia within the next four weeks [29 ]. Similarly, in the Rule Out Pre-Eclampsia Study, an sFlt/PlGF ratio > 85 was found
to have a PPV of 74% for the development of severe pre-eclampsia within two weeks
in
patients prior to GW 34 [30 ]. Women with an extremely high sFlt/PlGF ratio > 655 have a significantly shorter
time interval before delivery [31 ]. Changes in the sFlt/PlGF ratio over time also appear to be significant. Patients
who develop pre-eclampsia have a greater increase in sFlt/PlGF ratio within two weeks
than those who do not develop pre-eclampsia (31.22 vs. 1.45) [32 ].
Although the sFlt/PlGF ratio can be helpful in identifying patients with an increased
risk of pre-eclampsia, its predictive value is not sufficient to determine the optimal
timing for
administration of ACS. There is currently no sufficiently reliable way of predicting
an imminent delivery within a period of seven days in patients prior to GW 34. The
decision on the
administration of ACS must therefore be based on the patient’s clinical symptoms;
however, making an accurate predication and thus determining the correct timing of
administration is also
difficult in this context. The effect of esomeprazole on a possible prolongation of
pregnancy in patients with early, severe pre-eclampsia between GW 26+0 and GW 31+6
has been investigated in
a prospective randomized study. The average systolic blood pressure of the women in
the control group at the time of randomization was 168 ± 16.4 mmHg, and the diastolic
pressure was 103 ±
11.4 mmHg. The average 24-hour protein urine value was 1.06 (0.57–16.86) g/24 hours
(median value and interquartile range). In the control group, a median pregnancy prolongation
of 8.3 days
(interquartile range: 3.8–19.6 days) was achieved [33 ]. In other words, administration of ACS following diagnosis would have fallen within
the seven-day timeframe in only just on 50% of cases.
Very similar values have also been described by other study groups [34 ]
[35 ]
[36 ]
[37 ]. In a prospective observational study, Haddad et al. investigated a group of 239
women with severe pre-eclampsia occurring between GW 24 and GW 33. In the context
of a
watchful waiting approach, they reported a median prolongation of the duration of
pregnancy of 6 days before GW 29 (range: 2–35 days), 4 days between GW 29 and GW 32
(range: 2–32 days), and 4
days after GW 32 (range: 2–12 days) [34 ]. Chammas et al. also described a pregnancy prolongation of 6 days in patients with
severe pre-eclampsia occurring before GW 34. If there was also fetal growth
restriction, the prolongation was only 3 days [35 ].
These figures show that in early, severe pre-eclampsia, optimal timing of ACS administration
can be achieved with high success in many cases. In the aforementioned study, Levin
et al. found
that optimal timing (interval from administration to delivery: 24 hours to 7 days)
in hypertensive pregnancy was achieved in as many as 62% of patients [10 ].
Clinical prediction models for pre-eclampsia are now also available with fullPIERS
and PREP to predict the occurrence of maternal and fetal complications within seven
days. One example of
this is the prospective cohort study by Dadelszen et al. in which 106 out of 2023
women who were hospitalized for pre-eclampsia developed life-threatening complications
within 48 hours. The
fullPIERS model, which includes gestational age, breast pain, dyspnea, oxygen saturation,
platelet count, serum creatinine, and transaminases, showed an AUC of 0.88 (95% CI
0.84–0.92) for the
occurrence of these complications [38 ]. These models could certainly be developed further in order to predict the optimal
timing of ACS administration.
Fetal Growth Restriction
Two prospective randomized studies have investigated the watchful waiting approach
to managing intrauterine growth restriction (IUGR) [39 ]
[40 ]. The GRIT trial recruited 588 patients with IUGR between GW 24 and GW 36. In 77%
of cases, there was pathological end-diastolic flow in the umbilical artery. The doctors
in charge were unsure whether or not they should deliver these patients immediately.
After administration of one cycle of betamethasone, the women were randomized to undergo
either immediate
delivery or watchful waiting. The primary study endpoint of death or severe disability
at two years of age was 19% in the first group and 16% in the second group (OR 1.1,
95% CI 0.7–1.8).
However, the rate of disability in children born at ≤ GW 30 was 13% with immediate
delivery and only 5% with a watchful waiting approach. The corresponding median interval
between
randomization and delivery was 0.9 days (interquartile range: 0.4–1.3) and 4.9 days
(interquartile range: 2.0–11.0 days); for a gestational age ≤ 30 weeks, the respective
values were 0.8 days
(0.3–1.1) versus 3.2 days (1.5–8.0) [39 ].
The TRUFFLE trial recruited 542 patients between GW 26 and GW 32 who had early-onset
fetal growth restriction (abdominal circumference < 10th percentile) and a pathological
pulsatility
index (PI) in the umbilical artery (> 95th percentile). The women were randomized
into three groups characterized by different strategies used to decide when to deliver.
In all groups, the
patients were monitored using Oxford CTG and Doppler ultrasound of the umbilical artery;
monitoring of the ductus venosus was only planned in two groups. In the first group
the decision to
deliver was based on the short-term variability of the Oxford CTG (cut-off: 2.6 ms
< GW 29 or 3.0 ms at GW 29 to GW 32), in the second group the decision was made based
on early changes in
the ductus venosus (pulsatility index > 95th percentile), and in the third group it
was based on later changes in the ductus venosus (A-wave absent or negative). Of the
children who
survived without neurological defects, significantly more belonged to group three
than to group one (95% [95% CI 90–98] vs. 85% [95% CI 78–90]; p = 0.005). The respective
median interval from
randomization to delivery for the three groups was seven days (interquartile range:
0.5–61), seven days (interquartile range: 0.5–56]), and nine days (interquartile range:
0.5–88) [40 ].
Approximately 40% of patients in the TRUFFLE study had absent or reverse flow in the
umbilical artery at the time of randomization [40 ]. The time interval until manifestation of fetal distress is five or two days respectively,
while the probability of delivery within the next seven days is significantly
lower in the case of early changes in the PI in the umbilical artery [41 ]. Absent or reverse flow in the umbilical artery is therefore an indication for the
administration of ACS, while restraint should be exercised in the case of early changes
in the PI. If ACS is administered immediately in these cases, it usually does not
fall within the optimal timeframe.
Taking into account the pathophysiological development of end-diastolic umbilical
blood flow on Doppler ultrasound in IUGR fetuses, optimal timing of ACS administration
should generally also
be possible in this indication, similar to early pre-eclampsia. It should be borne
in mind that in IUGR fetuses, an improvement in end-diastolic flow in the umbilical
artery is often observed
following administration of ACS. However, this is often an expression of an increased
cardiac output, rather than reduced placental resistance [42 ].
There are currently no prospective randomized studies that have investigated the effect
of ACS administration on neonatal morbidity in IUGR fetuses. There is concern that
glucocorticoids may
exacerbate the cardiovascular and endocrinological alterations associated with intrauterine
growth restriction. However, in a 2001 prospective cohort study, Schaap et al. showed
that IUGR
fetuses delivered by caesarean section 24 hours to seven days after administration
of ACS had a higher probability of survival without disability at the age of two years
(OR 3.2, 95% CI
1.1–11.2) [43 ]. A meta-analysis from 2017, which included five retrospective or prospective studies,
found no reduction in neonatal morbidity after administration of ACS in IUGR
infants; however, it did show a clear trend towards a reduction in the rate of brain
hemorrhage. However, it is unclear whether the optimal timeframe of 24 hours to seven
days after ACS
administration was considered in this analysis [44 ].
Similarly for fetuses with late-onset intrauterine growth restriction, in another
prospective cohort study, no reduction in neonatal morbidity or perinatal mortality
was observed following
ACS administration between GW 32+0 and GW 36+6. However, the neonatal morbidity parameters
investigated in this study (e.g., pH value in the umbilical artery < 7.00, grade III–IV
brain
hemorrhage, grade II–III periventricular leukomalacia, respiratory support for more
than seven days, mechanical ventilation, etc.) were not suitable for detecting a potentially
small benefit
from ACS administration [45 ]. After GW 32, the benefit from administering ACS is hardly demonstrable in any case,
as will be explained in the section “Benefit of Antenatal Corticosteroids According
to Gestational Age”.
Ex Utero Bleeding
In the event of ex utero bleeding, it is difficult to predict how long the interval
will be until delivery becomes necessary. In the case of vasa previa, placental abruption,
or placenta
previa with heavy bleeding, an emergency caesarean section is indicated. If a watchful
waiting approach is possible in the case of placenta previa with moderate bleeding
or placental edge
bleeding, ACS administration should always be considered. As no information can be
found in the literature on the time until delivery in these situations, the decision
as to whether ACS
administration is indicated must be made based on a subjective assessment in each
case. We do not administer ACS if the bleeding is very light, but we recommend it
if the bleeding is as severe
as during menstruation.
Asymptomatic Cervical Insufficiency
Asymptomatic Cervical Insufficiency
As Levin et al. were able to show in their retrospective study, optimal timing is
not adequately achieved in the group of patients with asymptomatic cervical insufficiency.
Only 12% of the
women were delivered within the timeframe between 24 hours and seven days after administration
of ACS [10 ]. This phenomenon is due to the inadequate predictive power of the diagnostic methods
available to us in this context.
Esplin et al. showed that for cervical length measured by vaginal ultrasound between
GW 22 and GW 30 in asymptomatic patients, the AUC (95% CI) for premature birth before
GW 37 was only 0.67
(0.64–0.70) [46 ]. A further prospective cohort study investigated the significance of cervical length
measured by vaginal ultrasound between GW 31 and GW 34 for predicting the occurrence
of a premature birth between GW 32 and GW 36. For asymptomatic patients, the AUC (95%
CI) was only 0.700 (0.627–0.773) [47 ].
In addition to ultrasound measurement of cervical length, we have various tests at
our disposal that allow us to predict premature birth within seven days by measuring
proteins in the
cervical secretions (PAMG-1, fibronectin, IGFBP-1). However, as Esplin et al. also
showed, with a cervical length > 15 mm, the sensitivity and positive predictive value
of an fFN ≥ 50 ng/mL
measured between GW 22 and GW 30 for the occurrence of a premature birth before GW
32 were only 32.1% and 3.1% respectively. The use of higher or lower threshold values
did not improve the
test quality either. The fibronectin test did not increase the predictive value of
cervical length measured by vaginal ultrasound for the occurrence of premature birth
before GW 37 (AUC for
cervical length: 0.67; AUC for fibronectin: 0.59; AUC for cervical length + fibronectin:
0.67) [46 ].
In general, the risk of women with asymptomatic cervical insufficiency giving birth
within seven days is very low. For example, in a retrospective study that included
126 asymptomatic
patients with a cervical length ≤ 25 mm between GW 23 and GW 28, no patients were
delivered within seven days, and only one patient was delivered within 14 days. The
length of this patient’s
cervix was less than 10 mm [48 ]. These data are supported by a further retrospective study of 367 largely asymptomatic
women – vaginal spotting and pressure or pulling in the lower abdomen were not
considered exclusion criteria – with a cervical length of less than 25 mm between
GW 24 and GW 34. Only two of these patients gave birth within seven days [49 ].
A retrospective analysis investigated the negative predictive value of the fibronectin
test in asymptomatic patients between GW 22 and GW 32 whose cervical length was less
than 10 mm. This
value was 100% for a birth within seven or 14 days [50 ]. Another retrospective study shows almost identical results [51 ].
Based on these results, the following recommendation was made in the AWMF guideline:
If asymptomatic patients with a cervical length of 5–15 mm who have tested negative
for fibronectin,
phIGFBP-1 or PAMG-1 have no additional risk factors for preterm birth, they should
not be administered antenatal steroids because of the very low probability (< 1%)
that they will give
birth within 7 days. Nevertheless, the patient should continue to be monitored closely
with regards to her risk of preterm birth [17 ].
In order to determine when ACS is indicated in this patient group, we need diagnostic
methods that have a better positive predictive value. However, despite numerous innovative
approaches,
there are currently no viable solutions on the horizon [52 ]
[53 ].
Premature Labor
Premature labor alone has a < 50% predictive value for the occurrence of a premature
birth; e.g., premature labor stops spontaneously in 30% of cases, 50–70% of pregnant
women treated with
placebo give birth close to term [54 ], and only 12–17% give birth within one week [55 ]. The fibronectin test and ultrasound measurement of cervical length can help to
enable better assessment of the risk of these patients giving birth within seven days.
In
a prospective cohort study of 655 patients experiencing preterm labor, van Baaren
et al. observed the rate of delivery within seven days to be 12% [56 ]. If the cervical length was < 15 mm, the delivery rate was 47% regardless of the
fibronectin test. If the cervical length was between 15 and 30 mm, 2.7% (4/149) of
children were born within seven days if the fibronectin test was negative, and 14.1%
(21/148) if the test was positive. With a cervical length > 30 mm, only 0.7% of women
gave birth within
seven days, regardless of the fibronectin test.
Within the group of women experiencing preterm labor, in addition to those with a
cervical length < 15 mm, there is another group of patients who are highly likely
to undergo a premature
birth within seven days. These are women with an intra-amniotic infection/microbial
invasion of the amniotic cavity (IAI/MIAC) [57 ]
[58 ]
[59 ]. In a cohort of 358 women experiencing preterm labor, Cobo et al. found the condition
known as MIAC to be present in 68 patients, diagnosed by means of amniocentesis. In
these women, the gestational age at delivery was significantly lower (GW 26.9 [25.2–31.1]
vs. GW 35.0 [29.7–38.3]; p = 0.001) (median value and interquartile range). The time
until delivery
was also significantly shorter (1 day [0–3] vs. 31 days [6–62]; p = 0.001) [60 ]. This observation is supported by other studies [57 ]
[58 ]. Studies are currently planned to investigate the extent to which the neonatal outcome
can be improved by the administration of antibiotics in cases of proven
intra-amniotic infection [61 ].
There is an urgent need for research to optimize the identification of patients with
intra-amniotic infection. It is precisely in this small patient group, representing
approx. 10% of all
women who experience premature labor [62 ], that the administration of ACS is indicated. Strict restraint should be exercised
in the remaining cohort, as evidenced by the long median latency period of 31 days
observed in the study by Cobo et al. [60 ]. Since an intra-amniotic infection can only be diagnosed through amniocentesis,
better identification of affected patients can only be achieved through increased
use of
this method in the clinical setting. From a technical perspective, amniocentesis should
be easy to perform in this situation for the vast majority of patients.
Combined Indications
Of course, in everyday clinical practice there can often be a number of obstetric
clinical pictures that do not individually constitute an indication for ACS administration,
but may do so in
combination. As there is a general absence of information in the literature on the
interval between administration and delivery in situations of this kind, it is up
to the team providing care
to estimate the remaining time until birth on a case-by-case basis.
Timing of the Indication
ACS administration is not performed as an emergency measure, usually not even before
transfer to a perinatal center. Therefore, whenever possible, the indication for ACS
administration should
be determined by a clinician who is highly experienced in perinatology. In particular,
patients who experience preterm labor or have asymptomatic cervical insufficiency
with a closed cervix
have a low probability of giving birth within the next seven days. The same applies
to IUGR fetuses with an incipient reduction in end-diastolic flow in the umbilical
artery. The timing of ACS
administration can probably be further optimized through appropriate organizational
management.
Quality Assurance
Unfortunately, the previous quality parameter from IQTIG QI 330, defined as “Antenatal
corticosteroid therapy in premature births with a prepartum inpatient stay of at least
two calendar
days”, led to a strong disincentive. With a required cut-off > 95%, the optimal timeframe
of 24 hours to seven days was largely disregarded when administering ACS. This has
now been
corrected following evaluation of the 2021 cohort. That quality parameter has been
removed. Instead, the number of mothers who had a premature birth before GW 34 and
for whom the
administration of ACS did not occur within the optimal timeframe of 24–168 hours is
now stated (72.9%; n = 11873/16278), as is the number of women who received ACS and
then did not give birth
until after GW 34 (41.2%; n = 6715/16278) [63 ].
The Society for Maternal Fetal Medicine is thinking along very similar lines and has
defined two quality parameters. The first is a ratio which has as its denominator
the total number of
mothers with premature infants born between GW 24+0 and GW 33+6, excluding stillbirths,
and as its numerator the total number of mothers who received a complete or partial
ACS administration
or a first ACS booster 6–168 hours before giving birth. The second quality parameter
is a ratio consisting of the total number of women who gave birth at full term over
the total number of
women who received one or more doses of ACS [64 ].
Quality parameters of this kind can help to objectively assess the management of ACS
timing, providing an impetus for internal evaluation processes that ideally lead to
an improvement in
neonatal morbidity and mortality.
Benefit of Antenatal Corticosteroids According to Gestational Age
Benefit of Antenatal Corticosteroids According to Gestational Age
As numerous prospective randomized studies have shown, ACS undoubtedly contribute
to a reduction in perinatal morbidity and mortality. However, almost all of these
studies were carried out
more than 20 years ago and in no way represent the current standard in perinatology.
This can be seen from the fact that only around 100 children born before GW 28 are
included in these
studies; moreover, this group of children probably did not receive any surfactant
treatment or neuroprotection with magnesium [1 ]
[4 ]. Since randomized studies on ACS administration prior to GW 34 are now considered
unethical, no further information on this topic will be available in the foreseeable
future.
As an alternative, large prospective cohort studies have been conducted in an attempt
to obtain information on the effect of ACS on perinatal mortality and morbidity in
relation to
gestational age. Carlo et al. studied a prospective cohort of 10541 children born
between GW 22 and GW 25. The primary study endpoint of death or neurological impairment
was significantly
reduced following ACS administration between GW 23 and GW 25 (GW 23: aOR, 0.58 [95%
CI, 0.42–0.80]; GW 24: aOR, 0.62 [95% CI, 0.49–0.78]; GW 25: aOR, 0.61 [95% CI, 0.50–0.74]),
but not at GW
22 (aOR, 0.80 [95% CI, 0.29–2.21]). The results for severe grade 3/4 brain hemorrhage
were almost identical (GW 22: aOR, 0.94 [95% CI, 0.20–4.49]; GW 23: aOR, 0.59 [95%
CI, 0.40–0.87]; GW 24:
aOR, 0.81 [95% CI, 0.61–1.08]; GW 25: aOR, 0.56 [95% CI, 0.44–0.72]) [65 ]. In a prospective cohort study of almost 118000 women, Travers et al. were able
to show that in children born between GW 23 and GW 34, mortality before discharge
following ACS administration was significantly lower for almost every gestational
age within this range (aOR 0.47–0.32). However, the number needed to treat at GW 23
was 6, while at GW 34 it
was 798. ACS was observed to have a significant impact on the rate of severe cerebral
hemorrhage up to GW 30, and on survival without severe disability up to GW 27 ([Table 2 ]) [66 ]. The EPICE cohort study of children born between GW 24 and GW 31 also showed a 50%
reduction in perinatal mortality after administration of ACS [67 ].
Table 2
Benefit of ACS in children born between GW 23 and GW 34. Adjusted odds ratio (aOR)
and 95% confidence interval (CI); antenatal corticosteroids
(ACS) (data from [66 ]).
GW 23
n (%)
GW 24
n (%)
GW 25
n (%)
GW 26
n (%)
GW 27
n (%)
GW 28
n (%)
GW 29
n (%)
GW 30
n (%)
GW 31
n (%)
GW 32
n (%)
GW 33
n (%)
GW 34
n (%)
Death before discharge
With ACS
439/754 (58.2)
642/1781 (36.0)
432/2161 (20.0)
302/2602 (11.6)
213/3315 (6.4)
141/4237 (3.3)
90/5019 (1.8)
75/6466 (1.2)
45/8556 (0.5)
44/13203 (0.3)
22/16810 (0.1)
9/16928 (0.1)
Without ACS
331/447 (74.0)
182/352 (51.7)
114/381 (29.9)
77/444 (17.3)
51/468 (109)
46/685 (6.7)
30/813 (3.7)
22/1172 (1.9)
21/1591 (1.3)
18/3070 (0.6)
18/5954 (03)
37/20732 (0.2)
aOR (95% CI)
0.47 (0.36–0.62)
0.51 (0.40–0.64)
0.52 (0.41–0.67)
0.55 (0.42–0.74)
0.50 (0.36–0.71)
0.48 (0.34–0.69)
0.44 (0.29–0.68)
0.66 (0.41–1.12)
0.42 (0.25–0.74)
0.61 (0.36–1.08)
0.43 (0.23–0.80)
0.32 (0.14–0.63)
Survival without severe disability
With ACS
51/754 (6.8)
270/1781 (15.2)
594/2161 (27.5)
1127/2602 (43.3)
1988/3315 (60.0)
3068/4237 (72.4)
4125/5019 (82.2)
5676/6466 (87.8)
7817/8556 (91.4)
12409/13203 (94.0)
16177/16810 (96.2)
16507/16928 (97.5)
Without ACS
8/447 (1.8)
38/352 (10.8)
96/381 (25.2)
175/444 (39.4)
265/468 (56.6)
494/685 (72.1)
653/813 (80.3)
1008/1172 (86.0)
1443/1591 (90.7)
2876/3070 (93.7)
5728/5954 (9.2)
20220/20732 (9.5)
aOR (95% CI)
4.2 (2.1–9.7)
1.64 (1.15–2.41)
1.26 (0.98–1.64)
1.27 (1.03–1.58)
1.23 (1.00–1.50)
1.09 (0.90–1.31)
1.19 (0.98–1.44)
1.18 (0.98–1.42)
1.08 (0.89–1.30)
1.08 (0.91–1.27)
1.04 (0.88–1.21)
0.99 (0.87–1.13)
Severe brain hemorrhage
With ACS
186/745 (24.7)
314/1781 (17.6)
267/2161 (12.4)
204/2602 (7.8)
146/3315 (4.4)
126/4237 (3.0)
76/5019 (1.5)
63/6466 (1.0)
50/8556 (0.6)
34/13203 (0.3)
25/16810 (0.1)
12/16928 (0.1)
Without ACS
133/447 (29.8)
92/352 (26.1)
71/381 (18.6)
78/444 (17.6)
47/468 (10.0)
36/685 (5.3)
25/813 (3.1)
26/1172 (2.2)
15/1591 (0.9)
12/3070 (0.4)
10/5954 (0.2)
13/20732 (0.1)
aOR (95% CI)
0.75 (0.57–0.98)
0.61 (0.46–0.80)
0.60 (0.45–0.81)
0.40 (0.30–0.54)
0.40 (0.29–0.58)
0.56 (0.38–0.83)
0.50 (0.32–0.81)
0.44 (0.28–0.71)
0.61 (0.15–1.12)
0.65 (0.35–1.32)
0.87 (0.43–1.90)
1.08 (0.48–2.40)
However, when considering these cohort studies it must always be borne in mind that
nowadays approx. 90% of children born before GW 34+0 have received ACS. The mothers
who were not given ACS
before delivery exhibit differing patient characteristics; these can only be taken
into account to a limited extent, even using multiple regression analysis. This is
also evident, for example,
from the EPICE cohort study. Of the group of patients who had received ACS, only 18%
were delivered on the day of admission, compared to 67% of women who did not receive
ACS [67 ]. It is very likely that ACS administration is a significant indicator of orderly
and structured patient care, while the lack of ACS administration indicates emergency
situations. Due to the increased risk profile of patients who do not receive ACS,
the benefit of corticosteroids for neonatal morbidity and mortality is likely to be
overestimated in this
study design.
Nevertheless, the benefit of administering ACS before GW 30 is obvious. Beyond GW
32, the number needed to treat in order to reduce mortality before discharge is extremely
high. An influence
on the rate of brain hemorrhage is hardly to be expected at this point, as this is
now only in the per-thousands range [66 ]. Surfactant is now available for the treatment of RDS. In light of this, serious
questions may be asked as to whether it still makes sense to administer ACS at GW
≥ 32.
Unfortunately, we have no data from prospective randomized studies that would enable
a precise assessment.
Conclusion
ACS should ideally be administered 24 hours to seven days prior to delivery, as corticosteroids
only reduce the rate of RDS within this timeframe. Their impact on neonatal morbidity
and
mortality is much higher in extremely premature births than it is after GW 32. In
addition, children born full-term after ACS administration are significantly more
likely to have mental and
psychological development disorders. However, delivery within the optimal timeframe
only occurs in approx. 25–40% of cases. This means that the indication for ACS administration
needs to be
much stricter than it has previously been. ACS is always indicated in PPROM, severe
early pre-eclampsia, fetal IUGR with absent or reverse flow in the umbilical artery,
placenta previa with
bleeding, and patients experiencing preterm labor with a cervical length < 15 mm.
The risk of women with asymptomatic cervical insufficiency giving birth within seven
days is very low. In
such cases, ACS should not be administered even in patients with a cervical length
less than 15 mm, provided that the cervix is closed and there are no other risk factors
for premature birth
([Table 3 ]). In order to optimize the assessment of when ACS administration is indicated in
this patient population, we need diagnostic methods that have a better positive
predictive value. Caution is also indicated for women experiencing premature labor
who have a cervical length ≥ 15 mm. Further studies using amniocentesis are needed
in order to identify the
patient population with MIAC/IAI and to define threshold values for determining when
delivery is indicated. Whenever possible, the indication for ACS administration should
be determined by a
clinician who is highly experienced in perinatology. Emergency administration of ACS
prior to transfer to a perinatal center should be avoided if possible, except in clear
cases (see above).
In addition, quality parameters can help to objectively assess how the timing of ACS
administration is managed, providing impetus for internal evaluation processes that
will ideally lead to an
improvement in neonatal morbidity and mortality.
Table 3
Indication for the administration of antenatal corticosteroids (ACS).
Indication for the administration of antenatal corticosteroids
Indicated
Not indicated
Early premature rupture of membranes
+
Severe early pre-eclampsia
+
Fetal intrauterine growth restriction with absent or reverse flow in the umbilical
artery
+
Placenta previa with bleeding or premature placental abruption, if a watchful waiting
approach is possible
+
Asymptomatic cervical insufficiency with a cervical length < 15 mm, with a closed
cervix and no other risk factors for premature birth
+
Premature labor with cervical length < 15 mm
+
