Middle cerebral artery Doppler pulsatility index as a predictor of intrapartum meconium release in prolonged pregnancies

1 Introduction

Prolongation of pregnancy beyond the 41st week of gestation has been associated with increased neonatal and maternal morbidity [1–5]. Therefore, there is no general consensus on what is the ideal protocol for the management of these pregnancies.

In general terms, once 41 weeks are reached, there are two options: planned induction of labor or expectant management with intermittent fetal monitoring. Although some studies have shown no differences in maternal and fetal outcomes when induction of labor is performed [6, 7], the opinions of experts vary concerning this topic [8, 9]. In addition, there is a non-negligible number of pregnant women who prefer to wait for the spontaneous onset of labor and thus avoid obstetric intervention. However, there is no agreement on which prenatal test to perform and on the optimal frequency of the evaluation if expectant management is performed [10].

Morbidity and mortality in post-term pregnancies are thought to be secondary to placental insufficiency that leads to fetal hypoxia [11]. One of the most important adaptive fetal response to hypoxia is redistribution of blood flow towards the fetal brain, known as the ‘brain sparing effect’, which can be detected and quantified by Doppler. Normally, middle cerebral artery (MCA) resistance decreases at the end of pregnancy [12]. However, some authors have related that fetal redistribution is associated with an increased risk of adverse perinatal event, regardless of fetal size [13]. Thus fetal redistribution may not be a benign protective mechanism in all cases.

The aim of this study was to asses the correlation of middle cerebral artery pulsatility index (MCA-PI) and perinatal outcome in prolonged pregnancies.

2 Subjects and methods

This was a prospective observational study conducted in all consecutive pregnant women with a single uncomplicated pregnancy reaching 41 or more weeks’ gestation and referred for prolonged pregnancy surveillance, from February 2016 to February 2018. Gestational age was established by first trimester ultrasonographic measurement of crown-rump length. The setting was the Ultrasound Unit of the Department of Obstetrics and Gynecology of a non-tertiary community hospital. The study was approved by the Ethics and Research Committee of our hospital.

Our management protocol for prolonged pregnancies consists of ultrasound control of the amniotic fluid index (AFI) and non stress test (NST) twice a week, from 41 + 0 to 41 + 6 weeks. Exclusion criteria are fetal chromosomal or structural malformations and obstetric complications such as: intrauterine growth restriction, hypertensive disorders, insulin-dependent diabetes or other maternal pathology.

Under prior informed consent, fetal ultrasound assessment was performed by four experienced examiners using GE Voluson E6 (3.5-MHz convex probe with spatial peak temporal average intensities below 50 mW/cm2 and the high-pass filter set at 50–100 Hz). MCA-PI was measured at the straight portion of the artery at the end of the greater wing of the sphenoid bone, without head compression by the transducer nor fetal movements. A minimum of three consecutive waveforms were used for Doppler assessment, and the mean value of them was recorded. As a cut-off, the MCA-PI was considered abnormal below the 5th percentile (<p5), following reference ranges in prolonged pregnancies published by Palacio [14], which were used due to the similarity of their demographic characteristics with our population.

Oligohydramnios was defined by a maximum vertical pocket of amniotic fluid less than 2 cm [15]. Induction of labor was recommended under the presence of non reassuring or pathological cardiotocography, oligohydramnios or 41 + 6 weeks’ gestation.

After delivery, the occurrence of any of the following events: cesarean delivery for presumed fetal distress, 5-min Apgar score <7, arterial cord pH < 7.15, presence of thick meconium, need for admission to neonatal intensive care unit (NICU) and/or neonatal death were recorded. Cesarean delivery for suspicion of loss of fetal wellbeing was performed when persistent late decelerations were recorded or when fetal scalp blood sample showed pH below <7.10. Meconium was qualified as “thin” when the fluid was just tinted yellowish or slightly greenish, and “thick” when it was green and dense.

Statistical analysis was conducted using the SPSS 15.0 software package (SPSS, Chicago, IL, USA). Diagnostic accuracy was evaluated calculating the true- and false-positive rates. Categorical variables were analyzed by chi-square test or Fisher’s exact test, as appropriate, considering a p-value < 0.05 as statistically significant.

3 Results

Three hundred and forty-nine pregnancies met initially the inclusion criteria. Four patients refused consent to the recruitment. Immediate induction of labor was undertaken in 30 cases: 2 patients because of pathological cardiotocography, and 28 because oligohydramnios detected by ultrasound. These cases were not included in the analysis.

Fourteen cases were finally excluded: 3 patients delivered at another center, 10 newborns because cord gas analysis was not performed, and another one because shoulder dystocia occurred at the time of birth. Thus, 301 expectantly managed prolonged pregnancies were included for the final study. The median maternal age was 33 years (range 19–43); 19.2% were not Caucasian and 45.2% were nulliparous women. The median gestational age at inclusion was 41 + 0 weeks’ gestation (287 days; range 287 to 293). Of them, 31 (10.3%) fetuses had an MCA-PI < p5.

Three hundred babies were born alive, and one fetus had an antepartum death at 41 + 3 weeks’ gestation. The median gestational age at delivery was 290 days (range 287–294; SD 2), and the mean birth weight was 3559 g (SD 418). There were 16 (5.3%) non prenatally detected fetuses small for gestational age (SGA) (birth weight <10th percentile), but none of them was lower than the 3rd percentile. One hundred and seventy-three (57.5%) fetuses had spontaneous onset of labor and 128 (42.5%) required labor induction: 22 cases for rupture of membranes, 8 cases for non-reassuring cardiotocography, 2 cases for oligohydramnios, one because onset of clinically symptomatic cholestasis, and the remainder 95 in reaching 41.6 week’s, following our management protocol. Sixty (19.9%) pregnancies were delivered by cesarean section. Of them, 22 (7.3%) were emergent for fetal distress.

In our study population, thick meconium was observed in 18 (5.9%) fetuses: 8 (44.4%) of them required emergency delivery by cesarean section. The fetus who had antepartum death showed abundant and thick meconium at delivery, which was shown to be present in the oropharynx and bronchi of the fetus at necropsy. Two fetuses had a 5-min Apgar score <7 (0.7%), and 14 (4.7%) showed arterial cord pH < 7.15. Only one baby (0.3%) required admission to NICU.

MCA-PI < p5 was not statiscally related to the occurrence of: cesarean delivery for fetal distress (p = 0.262), 5-min Apgar score <7 (p = 0.804), cord arterial pH < 7.15 (p = 0.165), need for admission to the NICU (p = 0.926) nor neonatal death (p = 0.897). However, a statistical significance was demonstrated between MCA-PI < p5 and the presence of thick meconium at birth (p < 0.001). Clinical characteristics and perinatal outcomes of the study population are summarized in Table 1.

The main finding of our study is that an abnormal MCA-PI in prolonged pregnancies shows a strong relation with the presence of thick meconium atdelivery.

Post-term pregnancy has been related to increased risk of both intrauterine and postnatal death, adverse perinatal outcomes such as cesarean section rate (especially the emergency cesarean rate), neonatal acidosis and admissions to the NICU [16]. In the same way, it is known that the amniotic fluid volume decreases physiologically from about 37 weeks, and during the post-dates period, it is estimated that there is a decrease in amniotic fluid volume of about 33% per week [17]. Oligohydramnios, combined with the increased incidence of meconium staining of the amniotic fluid in post-term pregnancies, results in an increased risk of meconium aspiration syndrome [18]. Therefore, antenatal fetal surveillance in the form of semiweekly NST [19] and measurement of the largest fluid pocket is recommended for these high risk pregnancies [16]. According to these recommendations, we offered induction of labour in 30 prolonged pregnancies, because of either a pathological cardiotocography or oligohydramnios.

Late placental insufficiency and fetal hypoxemia have been suggested to explain the increased risk of adverse perinatal outcome in prolonged pregnancies, thus fetal Doppler has been proposed as a strategy in evaluating and monitoring these high-risk pregnancies. Doppler assessment of the cerebroplacental ratio (CPR) and MCA-PI are widely used to assess small fetuses with growth restriction [20–22]. However, very few data are available in the literature about its usefulness in evaluating prolonged gestations have yielded inconsistent results [23–30]. Studies on MCA Doppler in the post-term population have shown a continuous decrease in vascular flow resistance at the end of pregnancy [23–25]. This might be explained by a physiological change associated with an increase in cerebral metabolic requirements [24], or be secondary to a mild placental insufficiency with fetal hypoxemia, which may occur at this stage of gestation. However, other studies have not verified any decrease in MCA-PI [26] and suggest that the brain-sparing effect occurs only in situations with acute or chronic hypoxia. Devine and colleagues [27] demonstrated that a low MCA to UA resistance ratio, also called the cerebroplacental ratio (CPR), is associated with an increased risk of fetal distress. A recently published meta-analysis on prognostic accuracy of CPR and MCA for adverse perinatal outcome in prolonged pregnancies concluded that CPR seems to relate better than MCA for all adverse perinatal outcomes [22]. However, too many different thresholds (different indices, different reference ranges of normality curves) were used in these studies, which renders to be cautious about its application in routine practice [22]. In addition, other studies found poor relation of CPR performance and adverse perinatal outcome both in low risk [31], and in prolonged pregnancies [28, 29].

Our study population involves a large number of uncomplicated prolonged pregnancies, and Doppler findings were a closed reflection of fetal status prior to delivery. We have used reference ranges obtained by Palacio [14] in which gestational age is considered a continuous variable in complete days, since we agree it is methodologically more correct. The use of a single cutoff point [33] facilitates interpretation of Doppler parameters, though the use of centiles may be more accurate to identify fetuses with cerebral vasodilation, thus at risk for adverse perinatal outcome. In addition, the characteristics of Palacio’s population are very similar to ours, as a good justification to be used as our reference curves.

The portion of the MCA employed in the flow measurement is another factor of variability in the studies published in this regard. Doppler measurements of MCA flow velocities can be carried out in the proximal, the middle or the distal part of the vessel. But, flow in the latter segment has been related to be more influenced by behavioral states [34], and therefore less reproducible. Also, the PI tends to be lower in the proximal part than in the two distal parts [25, 34]. The clinical importance of different sample sites was addressed by Figueras and colleagues [29], who concluded that proximal MCA-PI better predicts umbilical cord pO2 and distal MCA-PI best predicts cord pH. We have followed this methodology, and our results show association between MCA-PI < 5p measured in the proximal part of the vessel and the release of thick meconium at birth.

This finding, showing that MCA-PI is better than AFI or umbilical artery PI Doppler to predict the risk of thick meconium-stained liquor, has been previously reported [30]. Passage of meconium in amniotic fluid is associated with increase of neonatal mortality and morbidity [35]. This complication occurs in about 12% of deliveries at term, and is more widespread in post-term neonates [35]. The underlying mechanisms are still not completely understood and are debatable, and different theories have been suggested. Passage of meconium could be due to mesenteric vasoconstriction, causing hyperperistalsis and anal sphincter relaxation [36]. Other authors referred passage of meconium when the umbilical venous oxygen saturation decreased to below 30% [37], and when vagal stimulation secondary to cord compression occurred as well [38]. Also, association with chorioamnionitis has been reported [39], and suggest that in such cases meconium aspiration syndrome (MAS) may occur more frequently. This condition, MAS, is accompanied by an increased incidence of low Apgar scores and high morbidity and mortality [40], though not all studies have confirmed these relationship [41]. Therefore, the presence of meconium stained fluid in an otherwise normal labor without fetal heart rate abnormalities, independently of gestational age, does not indicate by itself a risk for an adverse fetal outcome and can be managed by just close monitoring [42]. But, we want to highlight that in our study, almost half of the cases in which thick meconium was observed required emergency delivery by cesarean section. In addition, necropsy of the only fetus of the series that resulted in intrauterine death showed abundant and thick meconium in the oropharynx and bronchi, though neither oligohydramnios nor MCA-PI < 5p had been detected three days previously. However, in the subsequent maternal study it was detected that the mother was a carrier of a previously unknown thrombophilia (protein S deficiency), which could explain an antepartum death of acute cause, without giving time enough to significantly modify the Doppler parameters.

In spite of this, we recognize that the distinction between thin and thick meconium may be subjective.

In conclusion, abnormal MCA-PI < 5p in prolonged pregnancies is related to meconium emission at the time of delivery, but if that is properly managed it is not associated with an increased risk of adverse perinatal outcome.

Therefore, we advocate to further explore the role of Doppler MCA-PI in differentiating which pregnancies are not safe candidates to expectant management in prolonged pregnancy in the context of randomized trials, to determine the best indices and reference ranges to apply to general population, in the same way ongoing trials are currently evaluating the role of CPR in low risk fetuses [43].

Conflict of interest

All the authors declared no conflict of interest.