Premature monochorionic monoamniotic twins have lower lung compliance at birth than matched dichorionic diamniotic twins

Abstract

BACKGROUND:

Premature infants are born with immature lungs that demonstrate abnormal pulmonary function with differences in passive respiratory system compliance and resistance, and functional residual capacity. To our knowledge, no studies have evaluated differences in neonatal pulmonary function based on the type of twin gestation, or chorionicity. Given the effect of chorionicity on outcomes, we aimed to study the effect of twin type, monochorionic monoamniotic (MCMA) vs dichorionic diamniotic (DCDA), on neonatal early pulmonary function tests.

METHODS:

In this prospective cohort study, 5 sets of DCDA twins were matched to 5 sets of MCMA twins on gestational age at delivery, latency from antenatal corticosteroid exposure, birthweight, race and gender. Mean values were compared for passive respiratory system compliance and resistance, functional residual capacity, and tidal volume.

RESULTS:

MCMA infants had a significantly lower compliance (0.64 vs 1.25mL/cm H₂O /kg; p = 0.0001) and significantly higher resistance (0.130 vs 0.087 cm H₂O /mL/sec; p = 0.0003) than DCDA infants. Functional residual capacity was lower for MCMA than DCDA infants (17.5 vs 23.4mL/kg; p = 0.17). Further, 80% of MCMA infants required intubation for surfactant administration compared to 20% of DCDA infants, indicating the clinical significance of these objective measures.

CONCLUSIONS:

Due to the matched case-control design, causality cannot be established. However, we speculate that these differences in lung function may derive from differential exposure to preterm labor and endogenous maternal corticosteroid exposure. Further study is necessary to establish the true causal relationship.

1. Background

Premature infants are born with immature lungs that demonstrate abnormal function with differences in passive respiratory compliance and resistance, and functional residual capacity that are evident in the neonatal period [1–3]. Though some studies have evaluated early lung function in premature neonates, few have included twins. Twin gestations occur at a frequency of 34 per 1000 live births. Of these, approximately 50% are delivered prematurely including approximately 10% who are delivered before 32 weeks’ gestation [4, 5]. Among twin pregnancies, the type of twinning appears to play a role in outcomes. Monochronic monoamniotic (MCMA) twins in which the fetuses share a common chorion and amnion, are a rare occurrence accounting for only 1% of twin gestations. Compared to dichorionic diamniotic (DCDA) gestations, in which each fetus has a separate chorion and amnion, MCMA gestations carry a significantly higher risk from complications including intrauterine growth restriction, twin-to-twin transfusion syndrome, and cord entanglement [6,7]. For this reason, MCMA gestations are often delivered electively by Cesarean section and often prior to the onset of labor, thus without labor-related endogenous steroids. Because these high-risk pregnancies frequently require urgent or emergent deliveries, optimal timing of ACS administration is not always possible. The current recommendation from the Society for Maternal-Fetal Medicine (SMFM) is ACS treatment within 7 days before anticipated delivery and planned cesarean delivery at 32–34 weeks [8].

Capitalizing on the opportunity to distinguish the effects of shared genetics and shared environment from individual characteristics, several studies have examined pulmonary function tests (PFTs) in twins. On the whole, these studies have shown that growth restriction plays a significant role in trajectories of lung function among twins. For example, in the case of twins with discordant growth, the growth restricted twin demonstrates significantly worse lung function than the normally grown co-twin [9]. Additionally, given the risk of prematurity in twin gestations and the known decrease in lung function among premature neonates, the use of ACS to improve neonatal lung function is the standard of care in the setting of threatened preterm delivery [10]. Pharmacologic studies have also indicated that the dose of ACS (betamethasone or dexamethasone) typically used for singleton gestations is adequate in twin gestations as well [11]. Given the effect of chorionicity on outcomes, in this study, we aimed to evaluate the association of the type of twin gestation, MCMA vs DCDA, on neonatal pulmonary function tests. To our knowledge, no studies of neonatal PFTs among premature twins have been evaluated by chorionicity in this manner.

2. Methods

This prospective study was performed in the Doernbecher Neonatal Intensive Care Unit at Oregon Health & Science University as part of an ongoing study examining the association between the latency from ACS treatment to preterm delivery and subsequent neonatal PFTs. The study was approved by the Institutional Review Board of the hospital and informed consent was obtained from the parents. Inclusion criteria for the study group included: 1) MCMA twins born at < 34 weeks’ gestation; 2) maternal treatment with 2 doses of betamethasone ( 12 mg intramuscular injection/dose, Celestone Soluspan; Merck and Co, Inc, Whitehouse Station, New Jersey) given 24 hours apart, 3) PFT performed with 72 hours of delivery and before surfactant, if required. Infants were excluded from the study for the following: 1) exposure to more than one course of ACS; 2) exposure to < 6 hours of ACS prior to delivery; 3) insulin dependent maternal diabetes; 4) fetal congenital malformations or chromosomal abnormalities; 5) multiple gestations of triplets or greater. The comparison group was made up of DCDA twins who were matched as closely as possible to the MCMA twins on gestational age at delivery, latency from ACS exposure, birthweight, race, and gender, and also required the completion of a PFT as outlined for the MCMA twins. All infants were born remotely (>14 days) from an initial ACS course and none received rescue ACS treatment. Surfactant was given as rescue therapy within the first 24 hours of age if the patient required > 0.40 fractional inspired oxygen concentration (FiO₂) to maintain oxygen saturation > 90%, had a pH < 7.25 in the setting of respiratory acidosis, and/or had increased work of breathing despite adequate continuous positive airway pressure. All infants requiring surfactant were intubated and placed on mechanical ventilation, but were extubated as soon as clinically appropriate. Repeat surfactant doses were given if the infant continued to meet the above criteria.

PFTs including measurements of passive respiratory system compliance (Crs), passive respiratory system resistance (Rrs), functional residual capacity (FRC), and tidal volumes were performed in the NICU per American Thoracic Society (ATS)/European Respiratory Society (ERS) standards as previously described [12]. Briefly, infant PFTs were obtained using the SensorMedics 2600 infant pulmonary function cart (SensorMedics, Inc., Yorba Linda, CA). These measurements can be done in both non-intubated and intubated infants. Crs and Rrs were obtained using the single-breath occlusion technique after inducing the Hering-Breuer reflex on unsedated infants in quiet sleep. FRC was obtained using the nitrogen washout technique. Tidal volumes were calculated from a minimum of 12 flow volume loops with less than 15% variation between the inspiratory and expiratory tidal volumes [13,14]. If the infant was intubated, the PFT was done prior to surfactant administration and performed on clinical ventilator settings with a PEEP of 4–5cm H₂O.

2.1. Statistics

Mean values and standard deviations were calculated for Crs, Crs/kg, Rrs, FRC, FRC/kg, tidal volume, and tidal volume/kg and were compared using Wilcoxon rank sum test. P values of < 0.05were deemed significant and were determined using linear mixed modelling to account for confounders and nonindependence of covariates among twins [15].All analyses were performed using Stata 14.2 (StataCorp, College Station, Texas).

3. Results

In this prospective cohort study, 5 sets of DCDA twins were matched to 5 sets of MCMA on gestational age at delivery, latency from ACS exposure, birthweight, race and gender. Despite matching, there was a 94 g difference in birthweight between the two groups, but this difference was not statistically significant. None of the infants was small for gestational age (i.e.<10th weight percentile for gestational age) or growth restricted. As expected, all of the five MCMA deliveries were by Cesarean section compared to only two of the five DCDA deliveries. Maternal and infant demographic data is shown in Table 1.

Table 1.

Maternal and delivery demographics and infant clinical outcomes

Maternal and Delivery Demographics	MCMA(n = 5 sets)	DCDA (n = 5 sets)
Maternal age, mean ± SD	23.4 ± 5.3	27.2 ± 5.1
Maternal age, median (25th, 75th percentile)	24.0(18.5,28)	27 (23, 31.5)
Maternal Caucasian, n (%)	5 (100)	5 (100)
Gravida, median (25th, 75th percentile)	2 (1, 4.5)	3 (1.5, 3.5)
Tocolytic exposure, n (%)	0 (0)	5 (100)
Antibiotic exposure, n (%)	1 (20)	2 (40)
Preeclampsia, n (%)	0 (0)	0(0)
Preterm labor, n (%)	1 (20)	5 (100)
Rupture of membranes more than 24 hours, n (%)	0 (0)	2 (40)
Maternal smoking, n (%)	1 (20)	0 (0)
Latency from ACS to delivery in days, mean ± SD	29.0 ± 7.0	33.4 ± 10.0
Cesarean delivery, n (%)	5 (100)	2 (40)
Gestational age at delivery in weeks, mean ± SD	31.1 ± 1.6	31.5 ± 2.1
Infant Outcomes	MCMA (n =10 infants)	DCDA (n = 10 infants)
Male, n (%)	4 (40)	4 (40)
Birthweight in grams, mean ± SD	1560±134	1654±367
1-minute Apgar score, median (25th, 75th percentile)	7.5 (7, 8)	7 (6.5, 7.3)
5-minute Apgar score, median (25th, 75th percentile)	9 (8, 9)	8 (7.5, 8)
Age at PFT in hours, median (25th, 75th percentile)	6.8(4,41)	42(18.8,63.8)
FiO2 at PFT, median % (25th, 75th percentile)	41 (21, 56)	21 (21,21)
Required surfactant, n (%)	8 (80)	2 (20)

All infants had PFTs within 72 hours of delivery and before the administration of any surfactant, if clinically indicated. MCMA infants were found to have a mean Crs (normalized per kilogram) after birth that was approximately half the mean Crs for DCDA twins, p < 0.001 (Table 2). MCMA infants were also found to have a significantly higher Rrs than DCDA infants, p < 0.001. Further, MCMA infants were found to have lower FRC than DCDA infants, though this difference was not statistically significant. Of note, FRC measurements were not technically acceptable on one infant in the DCDA group. There was no difference in tidal volumes between the two groups of infants. In addition, 80% of MCMA infants required intubation for the delivery of surfactant, given as rescue treatment, compared to 20% of DCDA infants. Three of the MCMA twins required two doses of surfactant. Finally, a robust correlation was found between Crs and surfactant administration such that all infants with Crs ≤ 0.83 mL/cm H₂O/kg required treatment with surfactant while no infants with Crs > 0.83 mL/cm H₂O/kg did.

Table 2.

Infant pulmonary function tests

PFT parameter	MCMA twins (n =10)	DCDA twins (n = 10)	p value†
Crs (mL/cmH2O/kg)	0.64 ± 0.32	1.25 ± 0.54	<0.001
Total Crs (mL/cmH2O)	0.95 ± 0.39	2.07 ± 0.87	<0.001
FRC (mL/kg)*	17.5 ± 4.7	23.4 ± 8.1	0.17
Total FRC (mL)*	27.1 ± 8.3	38.7 ± 15.2	<0.001
Rrs (cm H2O/mL/sec)	0.130 ± 0.08	0.087 ± 0.07	<0.001
Tidal volume (mL/kg)	5.2 ± 1.5	6.3 ± 1.0	0.07
Total tidal volume (mL)	8.2 ± 3.1	10.3 ± 2.4	<0.001

^*FRC obtained in 19 infants, 10 in MCMA and 9 in DCDA,

^†adjusted for the correlation between twins. Crs, passive respiratory system compliance; Rrs, passive respiratory system resistance; FRC, functional residual capacity.

4. Discussion

In this case-control study, we found that MCMA infants had a significantly lower Crs and higher Rrs as well as a trend toward lower FRC than matched DCDA infants. We also found that MCMA twins had a greater need for surfactant after delivery. The differences in neonatal PFTs between the MCMA and DCDA twins, and specifically the significantly lower Crs in the MCMA twins (about 50% lower than in the DCDA twins), translated into a higher need for surfactant treatment. For comparison, the values for the Crs and FRC in the MCMA twins were similar to values we previously published in preterm infants delivered without any ACS treatment [16]. The fact that the MCMA twins demonstrated a greater need for surfactant substantiates the correlation of the objective neonatal PFT measurements with clinically important neonatal pulmonary outcomes. It is notable that 8 of the 10 MCMA infants in this study had PFTs performed while on mechanical ventilation (but before surfactant administration) compared to 2 of the 10 DCDA infants. Non-intubated infants were supported with CPAP. However, in our experience with infant PFTs, mechanical ventilation tends to stabilize FRC, compared to non-intubated PFTs. Thus, in this study, the effect of mechanical ventilation itself on the PFTs, if any, would have been to dampen the differences we measured in Crs and FRC between MCMA and DCDA infants. It is possible that the increased Rrs measured in the MCMA group compared to the DCDA group is related to the higher incidence of PFTs obtained on intubated infants in the former group. However, these infants would have been too unstable prior to intubation for the performance of a PFT. To our knowledge, no study has compared early objectives measures of pulmonary function tests in premature infant twins with different chorionicity.

The type of chorionicity in twin pregnancies is known to impact neonatal outcomes. Though the reasons are likely multifactorial, our data suggests that chorionicity and its management, may be an important clinical predictor of neonatal respiratory distress after delivery requiring surfactant therapy. One potential mechanism by which the difference in neonatal PFTs for MCMA and DCDA twins may derive is via a differential response to ACS and the presence of intermittent spontaneous labor. As hypothesized by Ballabh and colleagues, the existence two fetoplacental units may influence the metabolism and clearance of corticosteroids. MCMA twins may also have decreased responsiveness or lose the effect of steroids more rapidly due to this larger, in effect, singular, fetal and placental mass [17]. The current clinical practice guideline for the management of MCMA twins from the SMFM recommends the administration of ACS if preterm delivery is anticipated within 7 days [8].However, though all infants in the study had been exposed to ACS (betamethasone) during the pregnancy, the exposures were all > 20 days prior to delivery. When latencies between ACS exposure and delivery are of > 14 days, the benefit of ACS on neonatal PFTs appears to wane significantly [2,18]. Due to the matched case-control design, no formal cause-effect relationship can be established in this study. While it is possible that the differences in chorionicity or response to ACS could play a causal role in the differences in lung function, we speculate that the differences between these two groups may derive from differences in the exposure to endogenous maternal corticosteroids in these two groups related to preterm labor. The differences observed in this study may also be explained by the presence or length of labor prior to delivery, a phenomenon that has been most well-studied in term neonates where the frequency of respiratory morbidity in term neonates is highest when Cesarean delivery is performed prior to the onset of labor [19–21]. Although DCDA twins are potentially allowed to proceed with vaginal delivery, MCMA twins are much less likely to experience labor. The length of labor prior to delivery, which was not addressed in this study, may have served as a confounding variable. Additionally, there may have been differing etiologies precipitating delivery leading to different levels of stress and hypothalamus-pituitary-adrenal system activation prior to delivery. All 5 DCDA (100%) pregnancies were complicated by preterm labor, compared to only 1 of the MCMA pregnancies (20%). This potential difference in endogenous maternal corticosteroid could account for the differences seen in lung function and the need for surfactant. While no studies have directly tested the effect of endogenous maternal corticosteroids on neonatal PFTs, studies have demonstrated that the levels of maternal glucocorticoids needed to prevent RDS in premature neonates is in the same range as is seen in physiologic stress states [22, 23]. Given the lower Crs and increased need for surfactant administration seen in MCMA infants, the immediate respiratory outcomes in these infants could potentially benefit from a repeat course / rescue ACS, in the proper clinical context.

The strengths of this study include the fact that all PFT measurements were obtained by research-dedicated respiratory therapists and were subject to strict standardization per ATS/ERS guidelines, thus generating high quality data. Further, all PFTs were obtained in the NICU, without sedation, within 72 hours of delivery, and before any administration of surfactant. This study is also strengthened by the internal consistency shown in the correlation between the objective lung function tests and the clinical need for surfactant administration. The MCMA infants, who had lower Crs than DCDA infants, also had a significantly greater respiratory distress requiring surfactant.

There are also some notable limitations to the study. Due to the relative rarity of MCMA pregnancies and difficulty in performing PFTs prior to clinically indicated surfactant delivery, this study has a relatively small sample size limiting generalizability. In this matched case-control design, efforts were made to match cases on gestational age at delivery, latency from ACS exposure, birthweight, race and gender. However, it is possible that additional unmeasured confounders might be present. Although infants of mothers with insulin-dependent diabetes mellitus were excluded, diet-controlled diabetes mellitus was not measured. Additionally, no objective biomarkers of maternal stress, such as cortisol, were obtained in this study. Finally, 8 of the 10 MCMA infants’ PFTs were performed while mechanically ventilated compared to only 2 of 10 of the DCDA infants. This could potentially affect the PFT measurements but would be expected to dampen, rather than augment, the differences Crs and FRC found in this study though it could increase Rrs.

5. Conclusion

In a matched case-control study of 5 sets of MCMA twins and 5 sets of DCDA twins (20 total premature twin infants), we found that MCMA infants had lower Crs and higher Rrs than DCDA infants even after matching for gestational age at delivery, latency from ACS exposure, birthweight, race and gender. Due to the matched case-control design, no formal cause-effect relationship can be established in this study. However, we speculate that these differences in lung function may derive from differential exposure to preterm labor and endogenous maternal corticosteroid exposure. Further study is necessary to formally establish the true causal relationship.