Reliability of prenatal sonographic lung biometry in the diagnosis of pulmonary hypoplasia

Magnetic Resonance Imaging Assessment of Pulmonary Vascularity in Infants with Congenital Diaphragmatic Hernia: A Novel Tool for Direct Assessment of Severity of Pulmonary Hypertension and Hypoplasia

OBJECTIVES

To assess the feasibility of magnetic resonance imaging (MRI) for postnatal assessment of pulmonary vascularity in infants with congenital diaphragmatic hernia (CDH).

STUDY DESIGN

Infants with prenatally diagnosed CDH (n = 24) received postnatal pulmonary MRI. Infants with nonpulmonary birth defects served as controls (n = 5). Semiautomatic segmentation was performed to obtain total vascular volume using time of flight images to assess vascularity.

RESULTS

Average vascular density (vascular volume/lung volume) in control infants was 0.23 ± 0.06 mm³/mm³ compared with 0.18 ± 0.06 mm³/mm³ in infants with CDH is (P = .09). When stratified further based on CDH severity, the difference between control infants and moderate CDH group was statistically significant. (0.23 mm³/mm³ vs 0.15 mm³/mm³, P = .01). Ipsilateral vascular density on MRI in infants with CDH significantly correlated with the prenatal pulmonary hypertensive index (P = .0004, Spearman R = +0.87) and with number of days on mechanical ventilation (P = .04, Spearman R = -0.44), total days on inhaled nitric oxide (P = .02, Spearman R = -0.47), use of epoprostenol for acute pulmonary hypertension (PH) (0.14 mm³/mm³ vs 0.20 mm³/mm³, P = .005), and use of sildenafil for chronic PH (0.15 mm³/mm³ vs 0.19 mm³/mm³, P = .03).

CONCLUSIONS

Our results suggest that postnatal pulmonary vascularity assessed by MRI strongly correlates with prenatal and postnatal markers of PH severity and that pulmonary vascularity may serve as a direct measure of pulmonary vascular hypoplasia in infants with CDH.

Percent predicted lung volume changes on fetal magnetic resonance imaging throughout gestation in congenital diaphragmatic hernia

PURPOSE

Percent predicted lung volume (PPLV)<15% on fetal MRI predicts high-risk CDH. Potential changes in PPLV throughout gestation and impact on risk stratification are unknown. We reviewed CDH patients with serial fetal MRIs to follow PPLV and determine correlation with postnatal outcomes.

METHODS

CDH patients with serial fetal MRIs from 2005 to 2015 were included. We recorded prenatal MRI gestational age (GA) and PPLV, postnatal ECMO use, and survival. Data were analyzed by logistic regression and Fisher's exact test.

RESULTS

57 patients had 127 fetal MRI studies. PPLV decreased from mean 25.4% to 19.6% between GA 22.1 and 32.6weeks. A steeper decline in PPLV, regardless of final PPLV, was independently predictive of higher ECMO use (p=0.046) and death (p=0.045). All patients with first PPLV<15% remained high-risk with poor outcomes. Of those with first PPLV>15%, 31% dropped below 15%, having similar ECMO use as the high-risk cohort, but trending toward greater survival (p=0.09). Those with first and final PPLV>15% had significantly less ECMO use (p=0.015) and greater survival (p<0.001) than the high-risk cohort.

CONCLUSIONS

On average, PPLV decreases throughout gestation in fetuses with CDH. Serial MRI is recommended for those with initial PPLV>15%, as clinical outcomes tend to mirror the lowest PPLV.

TYPE OF STUDY

Treatment study LEVEL OF EVIDENCE: III.

Prediction of lethal pulmonary hypoplasia by means fetal lung volume in skeletal dysplasias: a three-dimensional ultrasound assessment

OBJECTIVE

The aim of this study was to assess the capacity of three-dimensional ultrasound (3DUS) for predicting lethality in fetuses with skeletal dysplasia.

METHODS

Twenty-four fetuses between 20 and 32 weeks of gestation were assessed. Bilateral lung volume scans were performed three times in each fetus during one ultrasound session. The virtual organ computer-aided analysis method was used to obtain a sequence of six sections of each lung around a fixed axis, and a rotation angle of 30° was adopted. Fetal lung volume measurements were analyzed according to the reference range. After birth, lung hypoplasia was diagnosed considering clinical and radiological criteria.

RESULTS

Of all cases of skeletal dysplasia, 18 (75%) were lethal. Among the lethal cases, after postnatal diagnosis, four were osteogenesis imperfecta type II, three were thanatophoric dysplasia and two were campomelic dysplasia. The remaining nine cases remained without a definitive diagnosis. The accuracy of 3DUS in predicting lethality in fetuses with skeletal dysplasia was high, with a sensitivity of 83.3%, specificity of 100%, positive predictive value of 100% and negative predictive value of 66.7%. The kappa index of 0.174 showed a good agreement between the possibility of lethality when the 3DUS volume measurement was altered and real lethality after birth (p < 0.001).

CONCLUSION

This study suggests that the 3DUS lung volume measurement is a good predictor of lethal pulmonary hypoplasia in fetuses with skeletal dysplasia, with high accuracy.

Magnetic resonance fetal right lung volumetry and its efficacy in predicting postnatal short-term outcomes of congenital left-sided diaphragmatic hernia

AIM

We aimed to investigate whether the ratio of magnetic resonance imaging (MRI)-measured right lung volume (RLV) to ultrasonography-estimated bodyweight (RLV/BW) and observed-to-expected (o/e) RLV are of diagnostic value in predicting postnatal outcomes of left congenital diaphragmatic hernia (CDH).

MATERIAL AND METHODS

We included 32 CDH patients and 34 control subjects. Manually outlined fetal right lung areas on MRI were multiplied by the slice thickness and added to determine the entire volume. The association between RLV and RLV/BW with gestational age in the controls was examined using regression analysis. RLV/BW and o/e RLV were compared between surviving and non-surviving neonates with CDH.

RESULTS

The expected fetal RLV was derived using the formula RLV (mm(3)) = 1.717 × (gestational weeks)(2.82). In the controls, RLV/BW was nearly constant during the third trimester. The 27 survivors with CDH had a median RLV/BW of 10.7 and a median o/e RLV of 60.0, whereas the five non-surviving neonates had a median RLV/BW of 4.3 and a median o/e RLV of 22.6; the differences were statistically significant.

CONCLUSION

Assessment of fetal lungs by MRI volumetry is reliable for clinical use. RLV/BW and o/e RLV are potential predictors of postnatal outcomes of left CDH.

New anatomical landmarks to study the relationship between fetal lung area and thoracic circumference by three-dimensional ultrasonography

OBJECTIVE

To evaluate the relation between total lung area (TLA) and thoracic circumference (TC) ratio by three-dimensional (3D) ultrasonography applying new anatomical landmarks as the fetal aorta and inferior angle of the scapula.

METHODS

A longitudinal prospective study was conducted with 56 uncomplicated pregnancies between 24 and 32 weeks of gestation. Polynomial regressions were used to evaluate the correlation between TC and gestational age (GA) as well as TC and estimated fetal weight (EFW). A simple linear regression was used to evaluate the correlation between TLA and Total thoracic area (TTA) and GA. The intraclass correlation coefficient (ICC) was used to assess the intra and interobserver variability.

RESULTS

127 examinations were performed. TC values ranged from 150 to 174 mm (mean 166 mm) at 24 weeks and 215-248 mm (mean 231 mm) at 32 weeks. The TLA/TC ratio ranged from a mean of 0.64 at 24 weeks (range 0.56-0.70) to 0.90 at 32 weeks gestation (range 0.79-1.01). The intraobserver variability using the ICC was of 0.919 for TC; 0.916 for TTA; 0.860 for right lung area (RLA) and 0.910 for left lung area (LLA). Interobserver reproducibility was with an ICC of 0.970 for TC; 0.984 for RLA and 0.910 for LLA.

CONCLUSIONS

Measurement of fetal TC and the relationship between TLA and TC by 3D-ultrasonography applying new anatomical landmarks shows good reproducibility and allows a new assessment of thoracic and lung growth.

Three-dimensional ultrasound fetal lung volumes and infant respiratory outcome: a prospective observational study

OBJECTIVE

To determine if fetal lung volumes (FLVs), determined by three-dimensional rotational ultrasound and virtual organ computer-aided analysis software (vocal), correlated with neonatal respiratory outcomes in surviving infants who had a high risk [fetuses with congenital diaphragmatic hernia (CDH)], lower risk [fetuses with anterior wall defects (AWDs)] and no risk (controls) of abnormal antenatal lung growth.

DESIGN

Prospective observational study.

SETTING

Tertiary fetal medicine and neonatal intensive care units.

POPULATION

Sixty fetuses (25 with CDH, 25 with AWDs and ten controls).

METHODS

FLVs were measured and expressed as the percentage of the observed compared with the expected for gestational age.

MAIN OUTCOME MEASURES

Neonatal respiratory outcome was determined by the duration of supplemental oxygen, mechanical ventilation and dependencies, and assessment of lung volume using a gas dilution technique to measure functional residual capacity (FRC).

RESULTS

The infants with CDH had lower FLV results than both the infants with AWDs (P=0.05) and the controls (P<0.05). The infants with CDH had longer durations of mechanical ventilation (P<0.001) and supplementary oxygen (P<0.001) dependence, compared with infants with AWDs. The infants with CDH had a lower median FRC than both the infants with AWDs (P<0.001) and the controls (P<0.001). FLV results correlated significantly with the durations of dependency on ventilation (r= -0.744, P<0.01) and oxygen (r= -0.788, P<0.001), and with FRC results (r=0.429, P=0.001).

CONCLUSIONS

These results suggest that FLVs obtained using three-dimensional rotational ultrasound might be useful in predicting neonatal respiratory outcome in surviving infants who had varying risks of abnormal lung growth. Larger and more comprehensive studies are needed to clarify the role that lung volume measurements have in assessing lung function and growth.

Reference range of fetal lung volume by 3D-ultrasonography using the rotational method (VOCAL)

AIM

To determine reference ranges for fetal lung volume by 3-dimensional ultrasonography using the VOCAL (Virtual Organ Computed-aided Analysis) method.

METHODS

A longitudinal prospective study was conducted with 61 uncomplicated pregnancies between 24 and 32 weeks of gestation. A separate measurement of both lungs volume was conducted by VOCAL with 30 degrees rotation angle. For each gestational age, an average, standard deviation and maximal and minimal values were established in addition to the percentiles 5, 10, 25, 50, 75 and 90 for right and left lung volume. To evaluate the correlation between lung volume and gestational age, and estimated fetal weight, a polynomial regression with determination coefficient adjustment (R(2)) was used. The intra-observer reproducibility was evaluated by the intraclass correlation coefficient (ICC), whereas the inter-observer reproducibility was evaluated by Cronbach alpha statistic test.

RESULTS

The average right lung volume varied from 12.5+/-0.7 cm(3) at the 24(th) week to 31.8 cm(3)+/-1.8 cm(3) at the 32(nd) week. The average left lung volume varied from 9.2+/-0.9 cm(3) at the 24(th) week to 22.0 cm(3)+/-1.6 cm(3) at the 32(nd) week. We observed a strong correlation between right lung volume and gestational age (R(2)=0.975) and estimated fetal weight (R(2)=0.905), as well as between the left lung volume with gestational age (R(2)=0.970) and estimated fetal weight (R(2)=0.908). We observed a good intra-observer reproducibility for the right lung volume (ICC=0.990) and for the left lung volume (ICC=0.986). Similarly, we observed good inter-observer reproducibility for right lung volume (0.975) and left lung volume (0.962).

CONCLUSION

Reference range of fetal lung volume by 3D-ultrasonography using the VOCAL method was determined.

[Fetal lung imaging]

Exponential improvements in imaging techniques over the last ten years, through patients' and physicians' wishes for less invasive fetal work-up, now allow us to better explore and understand fetal lung physiology during pregnancy. Diagnostic and prognostic consequences at stake are huge, especially for fetuses at risk of pulmonary hypoplasia. We will decline in three parts (normal lung, malformative lung and pulmonary hypoplasia), through a review of the literature and at the light of our experience, the potentialities and limitations of all imaging modalities (Ultrasound, Doppler, 3D, MRI). Then, we will dwell on future leads and the need for large-scale collaborative studies.

Comparison of two- and three-dimensional ultrasonography in lung volume measurement of normal fetuses

AIM

The purpose of this study was to compare the two- and three-dimensional methods for measuring fetal lung volume of normal fetuses.

METHODS

A cross-sectional study was performed with 51 normal pregnant women between 20 and 35 weeks. The ellipsoid formula (X*Y*Z*0.52) was used for volume calculation with the two-dimensional (2D) method. With the VOCAL (Virtual Organ Computer-aided Analysis) method, a rotation angle of 30 degrees was used. Intraclass correlation coefficient (ICC), paired Student's t-test and Bland-Altman plots were used to compare the techniques. To calculate the intraobserver variability we used the ICC and compared the means between the two measures using the paired Student's p-test.

RESULTS

VOCAL and 2D methods were highly correlated (ICC=0.919 and 0.873 for the right and left lungs, respectively), however, there was a disagreement. The fetal lung volume means calculated by the 2D method were always overestimated in relation to the means obtained by the VOCAL, for the right lung (24.02 mL x 19.15 mL; P<0.001), as well as for the left (16.03 mL x 13.77 mL; P=0.002). As for the intraobserver variability, a good reproducibility was observed for the volume measurement of the left lung by the 2D technique (mean=0.40 mL; P=0.57) and by the VOCAL (mean=-0.22 mL; P=0.63). The 2D method, however, presented low reproducibility for the right lung (mean=1.73 mL; P=0.31).

CONCLUSION

The two-dimensional method had low agreement and low reproducibility in relation to the three-dimensional method for measurement of fetal lung volume in normal fetuses.

Percent predicted lung volumes as measured on fetal magnetic resonance imaging: a useful biometric parameter for risk stratification in congenital diaphragmatic hernia

PURPOSE

This study was aimed at determining whether a new method of analyzing lung volumes on fetal magnetic resonance (MR) imaging could be used to predict the degree of pulmonary compromise in congenital diaphragmatic hernia (CDH).

METHODS

Seventeen fetuses with CDH were prospectively evaluated by MR. Lung volumes were measured using an established technique and expressed as a percentage of the predicted lung volume (PPLV). Predicted lung volume was determined by subtracting measured mediastinal volume from total measured thoracic volume. The PPLV was correlated with postnatal outcomes. Statistical analyses were performed using the Mann-Whitney, Spearman correlation, or Fisher exact tests (P < .05).

RESULTS

Of the 14 liveborn patients, the PPLV was 20.3+/-10.4 (gestational age at MR, 22.3 +/- 5.7 weeks). The PPLV was significantly associated with extracorporeal membrane oxygenation (ECMO) use, hospital length of stay, and survival. All patients with a PPLV of less than 15 required prolonged ECMO support and had a 40% survival rate. In contrast, only 11% of patients with a PPLV of greater than 15 required ECMO, and survival was 100%.

CONCLUSION

The PPLV as measured by fetal MR imaging can accurately predict disease severity in CDH. A value of less than 15 is associated with a significantly higher risk for prolonged support and/or death, despite aggressive postnatal management.

Fetal MRI in experimental tracheal occlusion

Congenital diaphragmatic hernia (CDH) is associated with a high mortality, which is mainly due to pulmonary hypoplasia and secondary pulmonary hypertension. In severely affected fetuses, tracheal occlusion (TO) is performed prenatally to reverse pulmonary hypoplasia, because TO leads to accelerated lung growth. Prenatal imaging is important to identify fetuses with pulmonary hypoplasia, to diagnose high-risk fetuses who would benefit from TO, and to monitor the effect of TO after surgery. In fetal imaging, ultrasound (US) is the method of choice, because it is widely available, less expensive, and less time-consuming to perform than magnetic resonance imaging (MRI). However, there are some limitations for US in the evaluation of CDH fetuses. In those cases, MRI is helpful because of a better tissue contrast between liver and lung, which enables evaluation of liver herniation for the diagnosis of a high-risk fetus. MRI provides the ability to determine absolute lung volumes to detect lung hypoplasia. In fetal sheep with normal and hyperplastic lungs after TO, lung growth was assessed on the basis of cross-sectional US measurements, after initial lung volume determination by MRI. To monitor fetal lung growth after prenatal TO, both MRI and US seem to be useful methods.

Three- and 4-dimensional ultrasound in obstetric practice: does it help?

OBJECTIVE

The purpose of this article was to review the published literature on 3-dimensional ultrasound (3DUS) and 4-dimensional ultrasound (4DUS) in obstetrics and determine whether 3DUS adds diagnostic information to what is currently provided by 2-dimensional ultrasound (2DUS) and, if so, in what areas.

METHODS

A PubMed search was conducted for articles reporting on the use of 3DUS or 4DUS in obstetrics. Seven-hundred six articles were identified, and among those, 525 were actually related to the subject of this review. Articles describing technical developments, clinical studies, reviews, editorials, and studies on fetal behavior or maternal-fetal bonding were reviewed.

RESULTS

Three-dimensional ultrasound provides additional diagnostic information for the diagnosis of facial anomalies, especially facial clefts. There is also evidence that 3DUS provides additional diagnostic information in neural tube defects and skeletal malformations. Large studies comparing 2DUS and 3DUS for the diagnosis of congenital anomalies have not provided conclusive results. Preliminary evidence suggests that sonographic tomography may decrease the examination time of the obstetric ultrasound examination, with minimal impact on the visualization rates of anatomic structures.

CONCLUSIONS

Three-dimensional ultrasound provides additional diagnostic information for the diagnosis of facial anomalies, evaluation of neural tube defects, and skeletal malformations. Additional research is needed to determine the clinical role of 3DUS and 4DUS for the diagnosis of congenital heart disease and central nervous system anomalies. Future studies should determine whether the information contained in the volume data set, by itself, is sufficient to evaluate fetal biometric measurements and diagnose congenital anomalies.

Reliability of the lung-to-head ratio in predicting outcome and neonatal ventilation parameters in fetuses with congenital diaphragmatic hernia

OBJECTIVES

The ratio of the lung area (on the contralateral side of the hernia) to the head circumference, the lung-to-head ratio (LHR), has been proposed as a reliable tool in the assessment of the prognosis of congenital diaphragmatic hernia (CDH). An LHR < 0.6 has been associated with poor outcome whereas one > 1.4 has been associated with survival. We aimed to analyze the role of LHR in predicting fetal outcome and ventilation parameters in cases of isolated CDH in our center.

METHODS

During the 40-month study period, 22 fetuses with an isolated CDH were delivered alive under optimized conditions at our perinatal center. LHR was measured at the time of diagnosis (median, 27 weeks' gestation). In addition to survival, postnatal ventilation parameters including pCO2, pO2, inspiratory O2 partial pressure, inspiratory pressure and oxygenation index were determined, as was the occurrence of pulmonary hypertension.

RESULTS

The overall survival rate was 59%. The LHR was not able to predict fetal outcome reliably. There was no correlation between the liver herniation, LHR and fetal outcome. The prenatally determined lung size reflected in the LHR did not show any significant association with individual ventilation parameters. Eleven of 17 infants examined had signs of pulmonary hypertension and the LHR did not predict this condition.

CONCLUSION

Our study cannot support the optimistic results reported by other groups on the use of LHR as a reliable predictor of outcome in fetuses with CDH. The LHR, as a reflection of lung size, correlates neither with survival patterns nor with various postnatal ventilation parameters.

Fetal Sheep with Tracheal Occlusion: Monitoring Lung Development with MR Imaging and B-Mode US

PURPOSE

To assess the accuracy of magnetic resonance (MR) imaging in determining fetal lung volume (FLV) and to observe fetal lung development with B-mode ultrasonography (US) and MR imaging.

MATERIALS AND METHODS

Seven sheep fetuses between 92 and 141 gestational days (term, 145 days) with and without tracheal occlusion (controls) underwent serial MR imaging and US. FLV at MR imaging was measured with true fast imaging with steady-state precession in coronal and transverse planes. The combined cross-sectional left- and right-lung area was measured with US at three transverse levels. FLV was measured at autopsy. Statistical evaluations included linear regression analysis and calculation of the mean and 95% CI.

RESULTS

No differences in FLV were observed on coronal or transverse MR images (r2 = 0.98; slope = 0.91; 95% CI: 0.82, 1.01). FLV at MR imaging at termination of the experiment was significantly related to FLV at autopsy (r2 = 0.96; slope = 1.27; 95% CI: 0.97, 1.57; n = 6). FLV at MR imaging increased more rapidly with gestational age in fetuses with tracheal occlusion (21.0 mL/d; 95% CI: 10.7, 31.3) than in controls (4.7 mL/d; 95% CI: 1.7, 7.7). Increase in left- and right-lung area at US was accelerated in fetuses with tracheal occlusion (1.60 cm2/d; 95% CI: 1.3, 1.9) compared with controls (0.38 cm2/d; 95% CI: 0.23, 0.53). Left- and right-lung area at US and FLV at MR imaging were significantly correlated (r2 = 0.82).

CONCLUSION

FLV can be measured with moderate accuracy at MR imaging on both coronal and transverse images. MR imaging and B-mode US are useful tools for monitoring and quantifying tracheal occlusion-stimulated fetal lung growth in sheep fetuses.

Three-dimensional ultrasound fetal lung volume measurement: a systematic study comparing the multiplanar method with the rotational (VOCAL) technique

OBJECTIVES

This study was designed to compare a conventional multiplanar technique for three-dimensional (3D) ultrasound measurement of fetal lung volume with a rotational method using VOCAL trade mark (Virtual Organ Computer-aided AnaLysis).

METHODS

Thirty-two fetuses with a variety of conditions at risk for pulmonary hypoplasia were studied. 3D volume data sets of the fetal lungs were acquired using a commercially available ultrasound system. The right and left lung volumes were calculated separately using VOCAL and the multiplanar technique. The level of agreement between two independent observers in categorizing the 3D volume data set as measurable or non-measurable was determined. The interobserver and intermethod variabilities were also evaluated for both methods.

RESULTS

The intermethod variability was excellent (correlation r = 0.93 and r = 0.96 for the left and right lung, respectively), and there was substantial agreement between the results of both approaches (limits of agreement - 4.4 to 8.9 and - 3.4 to 4.8 mL for the right and left lung, respectively). Fetal lung estimation with VOCAL had a significantly higher interobserver variability than the multiplanar technique. Interobserver agreement in categorizing lung volume data sets as measurable or non-measurable was lower when VOCAL was used.

CONCLUSION

Fetal lung volume measurements can be undertaken interchangeably using the multiplanar technique or the rotational method with VOCAL. However, the latter was less reproducible (lower degree of agreement and significantly higher interobserver variability) than the former.

Prenatal evaluation for fetal surgery

The selection of fetuses that may benefit with in utero surgery is being developed. Noninvasive and invasive prenatal diagnosis techniques are utilized to try and gain as much knowledge about the fetus so that the appropriate counseling of parents can be undertaken. The most common invasive techniques are amniocentesis and chorionic villus sampling for fetal karyotyping and genetic diagnosis. Noninvasive techniques include ultrasound (2D and 3D), fetal echocardiography and magnetic resonance imaging. Additional techniques such as specific Doppler evaluation of vascular components, new techniques to look at fetal electrocardiograms and the use of computer tomography are also considered. The most common conditions for which in utero fetal surgery is also being considered are twin to twin transfusion syndrome, myelomeningocele, sacrococcygeal teratoma, cystic adenomatoid malformation of the lung with fetal hydrops and other monochorionic twin abnormalities (severe discordant birth defects or twin reversal arterial perfusion sequences). Ongoing evaluation of the sensitivity, specificity, positive and negative predicted values of these evaluation tools is required so that appropriate selection of fetuses for the surgery can be made.