| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Obstetric Imaging |
1 From the Departments of Radiology (M.C., F.D.K., G.M., S.D.) and Obstetrics and Gynaecology (J.C.J., R.D., D.V.S., I.W., J.A.D.), University Hospital Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium. Received July 22, 2005; revision requested September 26; revision received November 9; accepted December 2; final version accepted February 17, 2006. Address correspondence to S.D. (e-mail: Steven.Dymarkowski{at}uz.kuleuven.ac.be).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Materials and Methods: Oral informed consent was obtained for the prospective component of this ethics committee–approved study. MR imaging lung volumetry was performed in 36 fetuses with normally developed lungs between 18 and 39 weeks gestational age by using T2-weighted single-shot fast spin-echo imaging in fetal transverse and sagittal planes. Findings were then correlated with biometric variables and gestational age. The best-performing algorithm was applied to 37 fetuses (between 18 and 29 weeks gestational age) at risk for pulmonary hypoplasia to determine observed-expected lung volume ratio. This group was stratified according to pregnancy management, and observed-expected ratios were correlated with outcome. In fetuses with isolated congenital diaphragmatic hernia (CDH) (n = 19), observed-expected ratio was correlated with lung-head ratio, neonatal survival in pregnancies managed expectantly (n = 13), and/or lung–body weight ratio at necropsy (n = 9). For that purpose, linear regression correlation was used with the Pearson correlation coefficient; P < .05 was considered to indicate a significant difference.
Results: Total fetal lung volume correlated best with total FBV (r = 0.96, P < .05). Observed-expected ratio based on FBV correlated with lung-head ratio in patients with CDH (r = 0.71, P < .001) and with lung–body weight ratio at necropsy (r = 0.68, P < .05) and could be used to help predict neonatal survival.
Conclusion: FBV measured with MR imaging can be used as a single parameter in an algorithm and showed closest correlation with normal total fetal lung volume. In the transition from second to third trimester, this algorithm enabled calculation of the observed-expected ratio and prediction of outcome in fetuses at risk for pulmonary hypoplasia.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Results of a number of previous studies (3,4) on the use of fetal MR imaging in fetuses with normally developed lungs are available. In earlier studies, algorithms that used a single parameter to predict total fetal lung volume were based on gestational age. Intuitively, it seems more logical to rely on an algorithm based on biometric variables, discounting effects of variation in fetal growth patterns, as well as uncertain gestational age. Coakley et al (5) used fetal liver volume and other biometric variables measured with US, as well as gestational age, to improve prediction. As an alternative to such a complex process, a single biometric variable that can be obtained through only one examination would enable greater simplicity. Fetal body volume (FBV) seems an ideal candidate: it can be measured at fetal MR imaging, which has been shown to be superior to US in predicting fetal weight at term (6). FBV can be measured accurately with MR imaging in all three sectional planes (7).
Thus, the purpose of our study was to retrospectively determine an algorithm based on FBV determined with MR imaging for calculating relative lung volume in fetuses with normally developed lungs and prospectively assess the use of this algorithm in predicting pulmonary hypoplasia in the late second and early third trimesters for fetuses at risk for pulmonary hypoplasia.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Study Subjects and Design
One author (M.C.) first conducted a retrospective chart review of all fetuses that underwent fetal MR imaging between January 2002 and December 2004. Twenty-six of the 62 available examinations were excluded because of thoracic or other structural abnormalities that could potentially affect body volume, liver size, or lung volume; twin pregnancies; inappropriate images; or incompleteness of data. This left 36 fetuses (between 18 and 39 weeks gestational age) with central nervous system anomalies such as isolated mild hydrocephalus (n = 8), agenesis of the corpus callosum (n = 3), cerebral hemorrhage (n = 3), myelomeningocele (n = 2), and cytomegalovirus seroconversion (n = 12); annular pancreas (n = 1); mesoblastic nephroma (n = 1); renal cortical cyst (n = 3); or adrenal hemorrhage (n = 2); as well as a pregnancy in which placenta accreta was suspected (n = 1).
A second prospective study was initiated in January 2004; in this study, 37 fetuses at risk for pulmonary hypoplasia that could be assessed prior to 29 weeks gestational age were included (between 18 and 29 weeks gestational age) by December 2004. During fetal MR imaging, the attending radiologist ensured that the entire fetus was imaged, enabling later determination of biometric variables. Twenty-three fetuses were imaged for congenital diaphragmatic hernia (CDH), three for congenital cystic adenomatoid malformation, one for Jeune syndrome (ie, asphyxiating thoracic dystrophy), eight for renal dysplasia, and two for early (
21 weeks) premature prelabor rupture of membranes. For consistency of measurements, structural US was performed by a single operator (J.C.J., who had 9 years of experience) with high-frequency transducer US equipment (Sequoia; Acuson, Mountain View, Calif). This was done between 4 days before and 5 days after fetal MR imaging. For fetuses with isolated left-sided CDH, measurement of the right lung–head ratio was performed at the level of the four-chamber view and at 24–28 weeks of gestation, as described elsewhere (8).
MR Imaging Examination
MR imaging was performed with a clinical 1.5-T whole-body unit (Magnetom Sonata; Siemens, Erlangen, Germany) that had gradient switching capabilities of 25 mT/m in 300 microseconds. For maternal sedation, 0.5 mg of flunitrazepam (Rohypnol; Roche, Basel, Switzerland) was administered orally 30 minutes before MR imaging to reduce fetal movements and related motion artifacts. Patients were positioned in a left-lateral position to prevent supine hypotension syndrome, and a combined two-channel phased-array body and spine coil was positioned over the lower pelvic area. The MR imaging protocol consisted only of T2-weighted images. Geometric parameters of T2-weighted images were as follows: 38 adjacent sections; section thickness, 4 mm; intersection gap, 0 mm; field of view, 380 x 380 mm; matrix, 173 x 256; repetition time msec/echo time msec, 1000/88; partial Fourier factor, 5/8; resulting pixel resolution, 1.8 x 1.5 x 6.0 mm3; and bandwidth, 475 Hz/pixel. T2-weighted imaging was performed with a single-shot fast spin-echo sequence in orthogonal transverse, coronal, and sagittal planes according to fetal orientation. No breath hold was requested.
The radiologist (M.C., with 3 years of experience) adjusted the field of view, number of sections, and image orientation for each fetus as required for optimal acquisition of images and measurements. Sequences that were degraded by fetal motion were repeated. Mean examination time for the prospective study was 20 minutes ± 4 (standard deviation).
MR Imaging Planimetry
Planimetric measurements of lung and liver volumes were performed by a single trained operator (M.C.). Planimetric measurements of total FBV were performed by two independent operators (M.C. and J.C.J. [the latter with 1 year of experience]) to determine interobserver agreement. The latter operator had limited experience with fetal MR imaging. Lung and liver volumes were calculated on the basis of T2-weighted single-shot fast spin-echo sequences in the transverse plane by using sequences that allowed complete imaging of both lungs and liver volume without motion-induced artifacts.
As for the plane in which delineations were performed to obtain FBV, we used any of the transverse, coronal, or sagittal planes that yielded the best images, according to the methods of an earlier study (7). For estimations of liver and lung volumes, delineations were made in the transverse plane (9).
Lung, liver, and total fetal body areas were determined on each section by placing free-form regions of interest at a picture archiving and communication system (Impax; Agfa-Gevaert, Mortsel, Belgium). Measured areas were added and multiplied by section thickness to determine the entire volume of the right and left lungs and liver and total FBVs. Volumes of the right and left lungs were added to obtain total lung volume for each fetus (Fig 1).
|
|
|
|
Data and Statistical Analysis
Interobserver variability (M.C. and J.C.J.; one delineation each) was assessed for total FBV in 10 randomly chosen examinations in fetuses with normal lungs by using the intraclass correlation coefficient. Intraobserver variability (M.C., two delineations with at least 24 hours in between) was assessed for total fetal lung volume, liver volume, and total FBV in 10 randomly chosen fetuses at risk for pulmonary hypoplasia.
In the group of fetuses that served as control subjects, three independent linear regression analyses that involved least-squares optimization were performed between the following variables: total fetal lung volume and liver volume, total fetal lung volume and FBV, and total fetal lung volume and gestational age. In the prospective study of fetuses at risk for pulmonary hypoplasia, the algorithm with the highest correlation coefficient was applied to calculate expected total fetal lung volume. Subsequently, the observed-expected total fetal lung volume ratio (observed-expected ratio) was calculated by expressing the observed total fetal lung volume as a percentage of the expected total fetal lung volume. The mean observed-expected ratio in these fetuses (n = 37) was compared with that in normal fetuses (n = 36) by using the Mann-Whitney U test. The observed-expected ratio was also correlated with the lung-head ratio, neonatal outcome, and the lung–body weight ratio in post hoc–selected subgroups depending on outcome or management of pregnancy.
From among the fetuses at risk for pulmonary hypoplasia, three homogeneous subgroups were selected. One subgroup included 19 fetuses evaluated for left-sided CDH at 26–29 weeks gestational age (11 of these fetuses underwent fetal surgery and eight were treated after birth) in whom the observed-expected ratio was correlated (linear regression) with the right lung–head ratio. The second group consisted of fetuses at risk for pulmonary hypoplasia (n = 13), who were expectantly treated during pregnancy (Table 1). For these fetuses, the observed-expected ratio was correlated with neonatal survival at discharge. The third group comprised patients who opted for termination—irrespective of indication—and for which the observed-expected ratio was correlated with lung–body weight ratio at necropsy when available (n = 9).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
At the onset of this study, we speculated that it would be better to use a biometric variable in predicting lung volume rather than, for instance, gestational age. Biometric variables can lessen the problems that may arise with incorrect dating of the pregnancy or with substantial alterations in fetal growth, provided that measurements are obtained in an organ not affected by the disease. For practical reasons, it would also be helpful if the proposed biometric variable could easily be measured at the time of the examination involved (eg, during fetal MR imaging). Several biometric variables are eligible, including head or abdominal circumference and femur length; three-dimensional variables such as liver volume may also be used. To our knowledge, FBV has not previously been used in an algorithm to predict pulmonary hypoplasia. This is surprising because FBV determined with fetal MR imaging has previously been shown to correlate well with birth weight (7), which is the denominator of the lung–body weight ratio used to define pulmonary hypoplasia. It therefore seemed logical to investigate the use of lung and body volumes as a proxy for lung and body weight and hence their proportion (lung–body weight ratio).
We demonstrated that from the second trimester onward, FBV correlated better than liver volume with total fetal lung volume, with high intra- and interobserver agreement. Measurement of the FBV took approximately double the time needed for liver measurement, but delineation of the region of interest in control fetuses with normal amniotic fluid was straightforward and could be done as well by a less experienced operator. Moreover, the contrast difference at the interface between the fetal body and amniotic fluid may ultimately make volume delineation a semiautomated or even completely automated process.
As a second step, we applied this method to the population of interest (fetuses at risk for lung hypoplasia). We wanted to test the algorithm early in the third trimester at the latest, at which time fetal intervention might still be contemplated.
Thus, we correlated the observed-expected ratio with the outcomes of fetuses grouped by either clinical condition or pregnancy management. There was a group of nine fetuses in which pregnancy was terminated; as a result, actual lung and body weights, and therefore lung–body weight ratio, became available. Pulmonary hypoplasia could thus be established with certainty. Although this finding needs to be verified in larger groups, current data indicate that MR imaging measurements and lung–body weight ratio correlated well. In the other fetuses, no such direct confirmation could be obtained because the pregnancies continued or the fetuses were treated in utero. Among the latter was a group of fetuses with isolated left CDH evaluated at our institution with a view toward fetal therapy. In isolated left CDH, the lung-head ratio in combination with liver herniation has been validated as a means to help predict fatal pulmonary hypoplasia in the early third trimester (11). Also in this group, MR imaging–based lung measurements correlated well with the lung-head ratio.
The third group was more heterogeneous, and pulmonary outcome was judged on the basis of neonatal survival or the occurrence of respiratory death. We extrapolated a cutoff value for lung hypoplasia from earlier insights in fetuses with CDH; such fetuses usually do not survive when the lung-head ratio is less than 1.0 (11). This probably could not be simply extrapolated because the fetuses in our study had different ranges of gestational age as well as different conditions; nonetheless, a lung-head ratio of less than 1.0 would correlate with an observed-expected ratio of less than 35%. This indeed predicted most neonatal deaths. Moreover, earlier observations that the lung-head ratio should not be used as a single predictor for neonatal death were confirmed in our study. Two fetuses with CDH who had an observed-expected ratio of less than 35% survived, but they did not have liver herniation into the thorax. Indeed, it has previously been demonstrated that the combination of lung volume and liver herniation better predicts poor outcome (11). In one patient, the observed-expected ratio before viability did not predict neonatal death. This fetus had renal dysplasia and apparently had progressively worsening pulmonary development, as reflected by a later MR imaging–based observed-expected ratio of 33% and death after respiratory failure. This case observation confirms the well-known heterogeneity of lung hypoplasia in different conditions and individual variations (12), which ultimately may limit individual prediction.
We also compared a gestational age–based algorithm previously described by Mahieu-Caputo et al (4). With a similar cutoff of 35.73% (corresponding to a lung-head ratio of 1.0), that algorithm would have predicted outcome incorrectly in three fetuses. Of note, in one fetus that died, survival was predicted by an observed-expected ratio that was based on gestational age but not on FBV. That fetus had a relatively high body weight (in the 90th percentile), an effect that is discounted for in the observed-expected ratio that is based on FBV. The other two incorrect predictions were two fetuses with CDH in whom the liver remained in the abdomen; for these fetuses, our algorithm wrongly predicted neonatal death. Also of interest, our algorithms predicted comparable (low) lung volumes, which can be expected for fetuses with a weight in the 50th percentile.
We acknowledge that our study was limited by the number of fetuses. However, it seems logical to base the observed-expected ratio on a parameter that can be measured in the individual fetus of interest (ie, FBV) rather than on gestational age. Along the same lines, the equally low numbers and heterogeneity of the populations in the subgroups of the prospective study were a limiting factor. Larger data sets will also allow for analysis of covariance between given (biometric) variables or other features that predict outcome, such as presence of liver herniation or early polyhydramnios in fetuses with CDH (13). In addition, our study focused on a particular time period in gestation, but pulmonary hypoplasia is probably a dynamic phenomenon. The fetus with renal dysplasia illustrates this; therefore, the evolution of the observed-expected ratio as pregnancy proceeds may be an even more powerful predictor.
In conclusion, we have demonstrated that lung volume can be predicted by a single biometric variable (ie, FBV) that can be measured during the same MR imaging examination; in a group at high risk for pulmonary hypoplasia evaluated before 29 weeks gestational age, it was a good predictor of outcome.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: CDH = congenital diaphragmatic hernia • FBV = fetal body volume
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, M.C., J.C.J., S.D., J.A.D.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, M.C., J.C.J., J.A.D.; clinical studies, M.C., J.C.J., D.V.S., J.A.D.; experimental studies, M.C., J.C.J., S.D., J.A.D.; statistical analysis, M.C., J.C.J., F.D.K.; and manuscript editing, M.C., J.C.J., F.D.K., I.W., G.M., S.D., J.A.D.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  M. M. Cannie, J. C. Jani, F. De Keyzer, K. Allegaert, S. Dymarkowski, and J. Deprest Evidence and Patterns in Lung Response after Fetal Tracheal Occlusion: Clinical Controlled Study Radiology, June 9, 2009; (2009) 2522081955. [Abstract] [Full Text]
|
|
|
|
|
|
M. M. Cannie, J. C. Jani, F. Van Kerkhove, J. Meerschaert, F. De Keyzer, L. Lewi, J. A. Deprest, and S. Dymarkowski Fetal Body Volume at MR Imaging to Quantify Total Fetal Lung Volume: Normal Ranges Radiology, April 1, 2008; 247(1): 197 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Gorincour, D. Eurin, F. E. Avni, For the GRRIF Group (Groupe Radiopediatrique de Re, M. Cannie, J. C. Jani, F. De Keyzer, S. Dymarkowski, and J. A. Deprest Prenatal Prediction of Pulmonary Hypoplasia: US and MR Imaging Working Together Radiology, November 1, 2007; 245(2): 608 - 609. [Full Text] [PDF]
|
|
|
|
|
|
J. C. Jani, M. Cannie, C. F. A. Peralta, J. A. Deprest, K. H. Nicolaides, and S. Dymarkowski Lung Volumes in Fetuses with Congenital Diaphragmatic Hernia: Comparison of 3D US and MR Imaging Assessments Radiology, August 1, 2007; 244(2): 575 - 582. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
