Prognostic value of a hernia sac in congenital diaphragmatic hernia

Authors

  • E. Spaggiari,

    1. Department of Obstetrics and Maternal–Fetal Medicine, GHU Necker Enfants-Malades, AP-HP and Université Paris Descartes, Paris, France
    2. Department of Developmental Biology, Hopital Robert Debré, AP-HP and Université Paris Diderot and Paris Sorbonne-Cité, Paris, France
    Search for more papers by this author
  • J. Stirnemann,

    Corresponding author
    • Department of Obstetrics and Maternal–Fetal Medicine, GHU Necker Enfants-Malades, AP-HP and Université Paris Descartes, Paris, France
    Search for more papers by this author
  • J.-P. Bernard,

    1. Department of Obstetrics and Maternal–Fetal Medicine, GHU Necker Enfants-Malades, AP-HP and Université Paris Descartes, Paris, France
    Search for more papers by this author
  • L. De Saint Blanquat,

    1. Department of Neonatal Intensive Care and Pediatric Reanimation, GHU Necker Enfants Malades, AP-HP and Université Paris Descartes, Paris, France
    Search for more papers by this author
  • S. Beaudoin,

    1. Department of Pediatric Surgery, GHU Necker Enfants Malades, AP-HP and Université Paris Descartes, Paris, France
    Search for more papers by this author
  • Y. Ville

    1. Department of Obstetrics and Maternal–Fetal Medicine, GHU Necker Enfants-Malades, AP-HP and Université Paris Descartes, Paris, France
    Search for more papers by this author

Correspondence to: Dr J. Stirnemann, Maternité et Medicine Materno-foetale, GHU Necker-Enfants Malades, 149 rue de Sèvres, 75015, Paris, France (e-mail: j.stirnemann@gmail.com)

ABSTRACT

Objective

To investigate the prognostic value of a hernia sac in isolated congenital diaphragmatic hernia (CDH).

Methods

Our database was searched to identify all consecutive cases of CDH referred to our fetal medicine unit between January 2004 and August 2011. Presence or absence of a hernia sac was assessed in liveborn cases using surgery or postnatal autopsy reports. We studied the correlation between the presence of a hernia sac and prenatal findings and perinatal morbidity and mortality.

Results

Over the study period, there were 70 cases with isolated CDH born alive in which either a surgery or autopsy report was available. Neonatal death, either preoperative or postoperative, occurred in 1/18 (5.6%) infants with a hernia sac and in 17/52 (32.7%) cases without a hernia sac (P = 0.03). Patients with a hernia sac had a significantly higher observed to expected pulmonary volume on prenatal magnetic resonance imaging (51.9 vs 39.3%, P = 0.01). Neonatal morbidity in surviving infants was lower in the group with a hernia sac, although not significantly.

Conclusion

The presence of a hernia sac is associated with a higher pulmonary volume and a better overall prognosis for CDH. Copyright © 2012 ISUOG. Published by John Wiley & Sons, Ltd.

INTRODUCTION

Congenital diaphragmatic hernia (CDH) is a severe fetal malformation, occurring in about one in 3000 live births, that carries high mortality and morbidity rates[1]. Associations with genetic, chromosomal or structural anomalies occur in about 30% of cases and heavily impact on the overall prognosis[2]. Despite advances in neonatal intensive care, the overall mortality rate still ranges from 30 to 50%[3-5]. Pulmonary hypoplasia and pulmonary hypertension are the leading causes of postnatal morbidity and mortality[6-8]. The prenatal prognostic assessment generally relies upon liver position, lung-to-head ratio (LHR) and magnetic resonance imaging (MRI) measurement of pulmonary volume[9-16]. Other prenatal prognostic factors such as stomach location or the presence of polyhydramnios have also been evaluated, but these factors remain controversial[13, 17, 18]. Fetal pulmonary vasculature also appears promising as a prognostic factor[19-21].

In this study, we investigated the potential prognostic value of the presence of a hernia sac in isolated CDH. The most common forms of CDH consist of a direct communication between the abdominal and thoracic cavities. Anatomically, a hernia sac is a pleuroperitoneal sac covering and enclosing the herniated organs in the thorax[22]. It is different from a diaphragmatic eventration, which is characterized by an upward displacement of an intact thin muscular diaphragm resulting from insufficient muscularization[23]. In contrast to CDH with a sac, a congenital diaphragmatic eventration is the failure of muscularization of the diaphragm, not a failure of fusion of its parts. Whereas diaphragmatic eventration has been shown to carry a better prognosis[24], the prognostic value of hernia sacs has not been assessed.

METHODS

We conducted a retrospective single-center study in which we searched the database to identify all consecutives cases of CDH referred to the fetal medicine center of Necker Hospital in Paris between January 2004 and August 2011. Prenatal prognostic variables included: side of the hernia, position of the liver, position of the stomach, amniotic fluid volume and lung size as assessed using both the LHR on ultrasound and measurement of lung volume on fetal MRI. Observed to expected (O/E) LHR was calculated using a formula derived from Peralta et al.[25]. Similarly, we considered O/E ratios for MRI volumes using the formula described by Mahieu-Caputo et al.[26]. In our protocol, the LHR is usually measured between 22 and 28 weeks' gestation, and MRI is performed between 28 and 34 weeks. However, the O/E-LHR was also evaluated in patients referred to our center after 28 weeks. Other potential prognostic factors such as the degree of pulmonary vascularization were not routinely analyzed and were therefore not available for this population.

Only liveborn infants with an isolated CDH and with either a surgery or autopsy report available were selected for final analysis (Figure 1). We excluded all fetuses with chromosomal abnormalities or malformations, as well as medical termination of pregnancy and intrauterine fetal death. We also excluded one case with diaphragmatic eventration and cases that underwent fetal endoluminal tracheal occlusion (FETO). Prenatal karyotype was obtained from amniotic fluid testing in all cases. Following delivery, all infants were treated following a standard protocol described by Datin-Dorriere et al.[9]. In brief, neonates were intubated nasally in the delivery room and immediately transferred to the neonatal intensive care unit (NICU). The first ventilator strategy used was high-frequency oscillation. Pulmonary artery hypertension (PAH) was evaluated by pre- and post-ductal saturation or by echocardiography and was managed by administering a high fraction of inspired oxygen and inhaled nitric oxide. Surgical repair was performed only after respiratory and hemodynamic stabilization had been achieved. The treatment strategy did not include administration of exogenous surfactant or extracorporeal membrane oxygenation. Survival was defined by discharge from the NICU. The presence of a hernia sac was recorded at the time of surgery or during autopsy. In surviving infants, postnatal morbidity was defined as the need for prosthetic patch repair, length of assisted ventilation, need for supplemental oxygen at 28 days of age, time in NICU or occurrence of a chylothorax.

Figure 1.

Flow chart showing outcomes of pregnancies with prenatal diagnosis of congenital diaphragmatic hernia (CDH), according to presence or absence of a hernia sac. Liveborn infants with isolated CDH for whom either a surgery or an autopsy report was available (n = 70) were included for analysis in the present study. FETO, fetal endoluminal tracheal occlusion; IUFD, intrauterine fetal death; TOP, termination of pregnancy.

Median values were compared using the Mann–Whitney test, and proportions were compared using Fisher's exact test. Two-sided P ≤ 0.05 was considered to be statistically significant. Statistical analysis was conducted using R software, version 2.0.0 (www.r-project.org, Foundation for Statistical Computing, Vienna, Austria).

This study was exempt from ethical review board approval because of its retrospective nature and because at the time of prenatal diagnosis LHR measurement and MRI were part of the routine diagnostic work-up.

RESULTS

Between January 2004 and August 2011, 143 patients with a prenatal diagnosis of CDH were referred to our center for perinatal care. Forty-five cases that did not meet the inclusion criteria were excluded: 17 terminations of pregnancy, six cases with associated malformations discovered after birth, one intrauterine fetal death, one eventration diagnosed after birth and 20 FETO procedures. Figure 1 summarizes management from prenatal diagnosis onwards. There were 98 liveborn infants with an isolated CDH. The overall postnatal survival rate as defined by discharge from the NICU was 55/98 (56.1%). All patients underwent high-frequency oscillatory ventilation. Thirty-three (33/98, 33.7%) patients died before surgery. Among these 33 cases, postmortem examination was performed in only eight cases owing to parental objections. Surgery was performed in 65/98 (66.3%) cases. In three cases, however, a surgical report was not available. Thus, 70 patients' records were kept for analysis, comprising 62 cases with a well documented surgical procedure and eight cases with a documented autopsy. Median gestational age at birth was 39 (range, 35–41) weeks and median birth weight was 3070 (range, 1500–4220) g. Among these 70 patients, 39 were male (55.7%).

Within this population of liveborn neonates with an isolated CDH, the neonatal mortality rate was 18/70 (25.7%). Forty infants (57.1%) subsequently developed pulmonary hypertension. A hernia sac was found in 18/70 (25.7%) cases on surgery or autopsy. Neonatal death occurred in 1/18 (5.6%) cases with a hernia sac and in 17/52 (32.7%) cases without a hernia sac (P = 0.03). The rate of PAH was lower in cases with a hernia sac, although not significantly. None of the eight children who died before surgery had a hernia sac. In one case, the hernia sac was prenatally suspected and confirmed postnatally; in this case, diaphragmatic hernia presented as a displacement of the left kidney into the thorax.

The relationship between outcome, especially neonatal morbidity, in surviving infants and the presence of a hernia sac is presented in Table 1. Although none of the neonatal morbidity variables reached statistical significance, patients without a hernia sac showed a higher rate of prosthetic patch repair, longer hospitalization, the need for prolonged assisted ventilation and supplemental oxygen and a higher occurrence of chylothorax.

Table 1. Outcomes in 70 liveborn patients with isolated congenital diaphragmatic hernia (CDH) and neonatal measures of morbidity in the 52 surviving patients, according to presence or absence of a hernia sac
CharacteristicCDH with sac (n = 18)CDH without sac (n = 52)P
  1. Data are given as n (%) or median (interquartile range). PAH, pulmonary artery hypertension.

PAH7 (38.9)33 (63.5)0.098
Survival at discharge17 (94.4)35 (67.3)0.03
  Need for patch repair0 (0)4 (11.4)0.29
  Need for supplementaloxygen at 28 days2 (11.8)10 (28.6)0.29
  Chylothorax3 (17.6)12 (34.3)0.33
  Duration of hospitalization(days)12 (6–28)19 (10–37)0.14
  Duration of assistedventilation (days)8 (2.5–20)11 (6–21)0.42

The relationship between prenatal characteristics and the presence of a hernia sac is presented in Table 2. The side of the hernia, amniotic fluid volume, stomach position, position of the liver, LHR and O/E-LHR were not significantly associated with the presence of a hernia sac. Gestational age at diagnosis was the same in both groups. The O/E pulmonary volume as measured by MRI was significantly higher in cases with a hernia sac: 51.9% (interquartile range (IQR), 40–58.7%) and 39.3% (IQR, 32.5–46.1%) in cases with and without a hernia sac, respectively (P = 0.01). The overall median gestational age at evaluation of O/E-LHR was 27 (IQR, 24–32) weeks, and median gestational age at MRI was 31 (IQR, 28.5–33) weeks.

Table 2. Prenatal characteristics of 70 patients with isolated congenital diaphragmatic hernia (CDH) according to presence or absence of a hernia sac
CharacteristicCDH with sac (n = 18)CDH without sac (n = 52)P
  1. Data are given as n (%) or median (interquartile range).

  2. GA, gestational age at diagnosis; LHR, lung-to-head ratio; O/E, observed to expected.

Right-sided hernia1 (5.6)8 (15.4)0.43
Intrathoracic liver4 (22.2)11 (21.2)1
Intrathoracic stomach14 (77.8)30 (57.7)0.16
Polyhydramnios4 (22.2)10 (19.2)0.74
GA (weeks)23 (22–32)23.5 (22–32.5)0.45
LHR2.07 (1.42–1.97)1.75 (1.41–2.27)0.45
O/E LHR (%)48.8 (38–57.9)51.9 (31.6–69.7)0.91
O/E total fetal lung volume (%)51.9 (40–58.7)39.3 (32.5–46.1)0.01

Among the 17 patients who opted for a termination of pregnancy, an autopsy was performed in 10 cases and no hernia sacs were found. Among those that underwent FETO, data on the presence of a hernia sac were available in 6/20 cases; one child had a hernia sac and survived after discharge from the NICU.

DISCUSSION

We have shown that the presence of a hernia sac significantly improves the prognosis in isolated CDH. A hernia sac is formed of parietal peritoneum and lung pleura and has been reported in approximately 20% of CDH cases[27], consistent with the prevalence of 26% in our population. Of the few studies that have addressed the impact of hernia sacs, most featured cases with a late postnatal diagnosis of CDH[28, 29]. In such cases of late presentation of CDH, a hernia sac is found in about 35%[28, 29]. However, to our knowledge this study is the first one addressing the prognostic value of a hernia sac.

One working hypothesis is that the sac prevents the upward displacement of the abdominal organs, resulting in less compression of the lungs, thus reducing the risk and degree of pulmonary hypoplasia. However, the incidence of intrathoracic liver and intrathoracic stomach was not significantly different in the two groups in our study, suggesting that the mechanism is probably more complex. In its most common presentation, CDH originates from the failure of closure of the pleuroperitoneal canal by the pleuroperitoneal membranes, resulting in a complete defect without peritoneum, pleura or muscle. Conversely, a hernia sac may result from a timely closure of the pleuroperitoneal canal without the appropriate subsequent muscularization of the defect. Therefore, the presence of a hernia sac could be the visible evidence of a late herniation during the embryological period[22, 30]. The timing of the defect could explain the observed improvement in prognosis rather than mechanical retention of organs in the abdominal cavity. This hypothesis is supported by the observed increase in fetal lung volume as assessed by MRI in cases with a hernia sac. Indeed, we showed that O/E ratios on MRI were significantly higher in cases of presence of a sac, resulting in an improvement in outcome. Conversely, ultrasound LHR indices were related to neither the presence of a hernia sac nor to postnatal outcome in our population. However LHR is more prone to high inter- and intraoperator variability and is therefore probably less reliable than MRI in assessing pulmonary hypoplasia. Indeed many factors may impair the reliability of ultrasound measurements of LHR, such as an accurate reference plane, visualization of the limits of the lungs, rib ultrasound shadowing and placement of calipers.

An autopsy was performed in only eight of the 33 children who died before surgery, which could cause a bias. However, parental refusal of an autopsy was the only reason for missing information and should therefore not have influenced the results.

Although not statistically significant, our results suggest lower morbidity in cases of CDH with a hernia sac. Although it appears to be a strong prognostic factor, prenatal identification of a hernia sac using a combination of ultrasound and MRI remains challenging. In our study, diagnosis of a hernia sac was suspected prenatally and confirmed in only one case. However, this was a retrospective study and prenatal imaging in the cases included did not specifically focus on the detection of hernia sacs. Advances in imaging techniques and awareness of its potential prognostic value should help in improving the prenatal assessment of the presence or absence of a hernia sac. Efforts should focus on refining the prenatal ultrasound and MRI findings and their reproducibility, with two goals in mind: first, a better understanding of the pathophysiology of CDH, particularly the relationship between the timing of the defect in cases with a hernia sac and its relationship with pulmonary hypoplasia; and second, improvement in prenatal counseling, especially in the era of prenatal therapy. Indeed, fetal therapy currently plays a significant role in the management of CDH, and the choice of its use is essentially based on prenatal predictor factors[31-33].

In conclusion, CDH with a hernia sac has a better prognosis in relation to increased pulmonary volume. The presence of a hernia sac could be associated with lesser morbidity, although this requires confirmation.

Ancillary