DOI:10.2214/AJR.04.1536
AJR 2005; 185:1328-1334
© American Roentgen Ray Society
MRI Calculation of Lung Volumes to Predict Outcome in Fetuses with Genitourinary Abnormalities
Michael Zaretsky1,
Ronald Ramus1,
Donald McIntire1,
Kevin Magee1 and
Diane M. Twickler1
1 All authors: Department of Radiology, University of Texas Southwestern Medical
Center, 5323 Harry Hines Blvd., Dallas, TX 75390-8896.
Received September 29, 2004;
accepted after revision December 6, 2004.
Address correspondence to D. M. Twickler
(Diane.Twickler{at}UTSouthwestern.edu).
Abstract
OBJECTIVE. The purpose of our study was to evaluate MRI total lung
volumes (TLV) for predicting outcome in fetuses with genitourinary
abnormalities and to compare lung volumes with the presence or absence of
oligohydramnios.
MATERIALS AND METHODS. Fetuses with genitourinary abnormalities
underwent blinded retrospective calculation of TLV. Distribution of the
TLV–gestational age ratios for survivors and nonsurvivors were compared
using the Wilcoxon's rank sum test. Lung volume calculation was compared with
the presence or absence of oligohydramnios.
RESULTS. There were 21 survivor and 24 nonsurvivor outcomes based on
neonatal discharge. TLV–gestational age ratios were significantly
different between the survivor and nonsurvivor groups (p = 0.0001).
No apparent difference was seen until after 26 weeks of gestation.
TLV–gestational age ratios were equal to the presence or absence of
oligohydramnios in predicting outcome after 26 weeks of gestation.
CONCLUSION. After 26 weeks' gestation, the prediction of outcome in
fetal genitourinary abnormalities using the MRI TLV–gestational age
ratio is comparable to the presence or absence of oligohydramnios.
Introduction
Fetal MRI has become a useful adjunct to sonography in the setting of
suspected central nervous system, chest, and genitourinary abnormalities
[1–7].
Fast-acquisition MRI, such as single-shot fast spin echo and other
vender-specific protocols, can produce images with excellent resolution of
fetal anatomy. These images can also be used to calculate fetal organ volumes
[8]. Fetal lung volumes have
been calculated, and nomograms have been published by previous investigators
for normal fetuses throughout gestation using MR volumetry
[9–12].
Lung volumes can be calculated via 3D sonography
[13–15]
with good correlation with MRI values. However, 3D sonography has the same
limitations that 2D sonography has, including near-field attenuation at bony
interfaces, difficulty of visualization with maternal obesity, anhydramnios,
and instances when fetal position limits orthogonal image acquisition of the
region of interest. Pulmonary hypoplasia can be caused by a number of factors,
including prolonged oligohydramnios, skeletal and neuromuscular disorders,
chest masses, and rare cases of aplasia. Lung volumes have been studied in the
settings of thoracic masses and congenital diaphragmatic hernias with the
promise of predicting lethal cases of pulmonary hypoplasia
[3,
16,
17]. Previous investigators
have had mixed results.
Pulmonary hypoplasia may be a complication of fetal genitourinary
abnormalities. Lung volumes have been rarely reported in this setting. In
cases in which there is not a clear-cut lethal genitourinary abnormality, it
is important to discuss the potential for pulmonary hypoplasia. In the
evaluation of these abnormalities it is often difficult to provide patients
with prognostic information. An example is a posterior urethral valve with
incomplete obstruction showing low or normal amnionic fluid. Fetal lung
volumes may have the potential to predict outcome in this setting and to
confirm suspected pulmonary hypoplasia in the setting of oligohydramnios. This
study was undertaken to evaluate the value of fetal MRI calculation of lung
volumes in predicting outcome in fetuses with genitourinary abnormalities.
Materials and Methods
Between July 2000 and July 2003, 70 cases of suspected genitourinary
abnormalities were referred for fetal MRI second opinion, of which 45 had
adequate information on outcome. Before these images were acquired, each
maternal patient was counseled about fetal safety issues, and written informed
consent was obtained as part of an indicated study. The remainder of the 70
patients were lost to follow-up. The gestational age ranged between 17 and 38
weeks. Outcome was based on review of medical records, including pathology
reports in cases of termination of pregnancy. Of these cases, 21 survived the
neonatal period and 24 either were terminated or were perinatal deaths. In
those women who terminated their pregnancy, the diagnosis of a potentially
lethal fetal outcome, such as renal agenesis, was suspected from the pathology
findings in all cases. Retrospective review of MRI studies was performed in
these fetuses with the diagnosis of genitourinary abnormalities based on
suspicious findings on sonography and MRI findings. All genitourinary
abnormalities were included in this analysis regardless of their potential for
pulmonary hypoplasia. Those cases with normal lung volumes were used as
controls. Retrospective calculation of total lung volume (TLV) was performed
by outlining regions of interest on consecutive axial MR images. Area lung
measurements were obtained by manual tracings of right and left lobes from
sequential 3- to 5-mm axial slices from a single sequence. The workstation
computer automatically calculated the area of the region of interest (Figs.
1A,
1B,
1C,
1D,
2A,
2B,
2C, and
2D). The sum of the areas from
each image was then multiplied by slice thickness to obtain volume (mL). The
average number of slices varied among cases and was determined by both
gestational age (size of the region of interest) and slice thickness. The
average postprocessing time to calculate TLV was approximately 5–10 min.
The investigator obtaining lung volumes was blinded to outcomes.
View larger version (146K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —MR images of 20-week fetus with bilateral multicystic
dysplastic kidneys. Selected coronal images from single 90-sec acquisition.
Total lung volume (TLV) was 6.48 mL and TLV–gestational age ratio was
0.32. This fetus did not survive and had significant pulmonary hypoplasia at
pathologic examination. Regions of interest are the areas of right and left (1
and 2) lung fields in A–D.
|
|
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —MR images of 20-week fetus with bilateral multicystic
dysplastic kidneys. Selected coronal images from single 90-sec acquisition.
Total lung volume (TLV) was 6.48 mL and TLV–gestational age ratio was
0.32. This fetus did not survive and had significant pulmonary hypoplasia at
pathologic examination. Regions of interest are the areas of right and left (1
and 2) lung fields in A–D.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —MR images of 20-week fetus with bilateral multicystic
dysplastic kidneys. Selected coronal images from single 90-sec acquisition.
Total lung volume (TLV) was 6.48 mL and TLV–gestational age ratio was
0.32. This fetus did not survive and had significant pulmonary hypoplasia at
pathologic examination. Regions of interest are the areas of right and left (1
and 2) lung fields in A–D.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —MR images of 20-week fetus with bilateral multicystic
dysplastic kidneys. Selected coronal images from single 90-sec acquisition.
Total lung volume (TLV) was 6.48 mL and TLV–gestational age ratio was
0.32. This fetus did not survive and had significant pulmonary hypoplasia at
pathologic examination. Regions of interest are the areas of right and left (1
and 2) lung fields in A–D.
|
|
View larger version (107K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —MR images of 31-week fetus with bladder outlet obstruction.
Selected axial images from single 90-sec acquisition. Total lung volume (TLV)
was 40.7 mL and TLV–gestational age ratio was 1.31. This fetus survived.
Regions of interest are the areas of right and left (1 and 2) lung fields in
A–D.
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —MR images of 31-week fetus with bladder outlet obstruction.
Selected axial images from single 90-sec acquisition. Total lung volume (TLV)
was 40.7 mL and TLV–gestational age ratio was 1.31. This fetus survived.
Regions of interest are the areas of right and left (1 and 2) lung fields in
A–D.
|
|
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —MR images of 31-week fetus with bladder outlet obstruction.
Selected axial images from single 90-sec acquisition. Total lung volume (TLV)
was 40.7 mL and TLV–gestational age ratio was 1.31. This fetus survived.
Regions of interest are the areas of right and left (1 and 2) lung fields in
A–D.
|
|
View larger version (109K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2D —MR images of 31-week fetus with bladder outlet obstruction.
Selected axial images from single 90-sec acquisition. Total lung volume (TLV)
was 40.7 mL and TLV–gestational age ratio was 1.31. This fetus survived.
Regions of interest are the areas of right and left (1 and 2) lung fields in
A–D.
|
|
The following parameters were used to obtain single-shot fast spin-echo
images: TR/effective TE range, 50–100; field of view, 12–36 cm;
matrix, 256 x 128 or 512 x 256; bandwidth, 31.2 or 62.5 Hz; 0.5
average (number of excitations, 0.5); and slice thickness varying between 3
and 7 mm without gaps. Average sequence acquisition time was 90 sec; the
images from these sequences are relatively T2 weighted. A 1.5-T Signa magnet
was used (GE Healthcare).
To normalize the values for age, a TLV–gestational age ratio was used
by dividing the calculated lung volume by the gestational age at the time of
the procedure, as described by Walsh et al.
[3]. Gestational age was
determined by sonographic or clinical criteria or both.
Distributions of the TLV–gestational age ratios were compared using
the Wilcoxon's rank sum test, and the slopes of the TLV by gestational age
were compared using analysis of covariance for the nonsurvivor and survivor
outcomes. A receiver operating characteristic (ROC) curve was developed to
determine the TLV–gestational age cutoff value that has the highest
sensitivity and specificity for predicting nonsurvivor outcomes.
The cases presented here were all referred for second-opinion MRI
evaluation, and oligohydramnios was determined both on sonography (amniotic
fluid index [AFI] < 5 cm) before referral and qualitatively from MR images
for each outcome. The investigators reviewing the MR images had extensive
experience with qualitative evaluation of amniotic fluid levels on both
sonography and MRI (> 600 clinical MRI studies). The sensitivity and
specificity of oligohydramnios for the nonsurvivor genitourinary outcome were
determined and then compared with the TLV–gestational age ratio. A
p value of less than 0.05 was considered statistically
significant.
Results
The diagnoses of the survivor and nonsurvivor outcomes are listed in Tables
1 and
2 along with gestational age,
TLV, TLV–gestational age ratio, and the presence or absence of
oligohydramnios. The TLV–gestational age ratio mean and SE for survivor
and nonsurvivor outcomes are 1.57 ± 0.13 and 0.58 ± 0.14,
respectively. The values are plotted graphically in
Figure 3 and quartiles are
listed in Table 3. There is a
statistically significant difference between TLV–gestational age ratios
between the two groups (p = 0.0001). The slopes of the TLV by
gestational age are plotted with the 95% confidence intervals in
Figure 4. The slopes of
survivor and nonsurvivor outcomes are significantly different (p =
0.005). The gestational age at which the 95% confidence intervals no longer
overlap is 26 weeks. Using the ROC curve, a TLV–gestational age ratio
cutoff value of less than 0.90 was determined; this value has a sensitivity
and specificity of 85.7% and 83.3%, respectively, for predicting nonsurvivor
outcomes. The ratio cutoff value of 0.90 has a sensitivity and specificity of
77.8% and 95%, respectively, after 26 weeks.
View larger version (5K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Graph shows slopes of total lung volume by gestational age
for survivor (upper slope, •) and nonsurvivor (lower slope, )
outcomes. Dotted slopes indicate 95% confidence intervals. Hypothesis of equal
slopes was rejected (p = 0.0054, analysis of covariance).
|
|
Oligohydramnios was present in 20 of 21 nonsurvivors versus four of 24
survivors. The sensitivity and specificity of oligohydramnios for predicting
nonsurvivor outcomes were 95.2% and 83.3%, respectively. No statistical
difference was seen between the sensitivities (p = 0.32) and
specificities (p = 1.00) of a TLV–gestational age ratio of less
than 0.90 and the presence of oligohydramnios. After 26 weeks, the sensitivity
and specificity for oligohydramnios were 100% and 95%, respectively. Again, no
statistically significant difference was seen between the sensitivities
(p = 0.16) and specificities (p = 1.00) of the
TLV–gestational age ratio of less than 0.90 and the presence of
oligohydramnios after 26 weeks.
Either oligohydramnios or a TLV–gestational age ratio of less than
0.90 was present in every nonsurvivor. In one case, amnionic fluid was normal
but the TLV–gestational age ratio was less than 0.90 (lymphangioma at 22
weeks). Three cases of oligohydramnios had a TLV–gestational age ratio
of more than 0.90 (two posterior urethral valve [one at 30 weeks and the other
at 35 weeks] and one renal agenesis at 25 weeks).
In the survivors, there were two cases with normal amnionic fluid and a
TLV–gestational age ratio of less than 0.90 (two prune-belly syndrome,
one at 22 weeks and one at 29 weeks), two cases in which there were
oligohydramnios and normal lung volumes (one megaureter at 23 weeks and one
unilateral renal agenesis at 28 weeks), and two cases in which there were both
oligohydramnios and a TLV–gestational age ratio of less than 0.90 (one
normal fetus at 19 weeks and one megaureter at 22 weeks).
Discussion
The purpose of this study was to determine the value of fetal TLV
calculations from MRI in predicting outcomes in cases of genitourinary
abnormalities. We found a significant difference for both the
TLV–gestational age ratios and slopes of the TLV by gestational age
between the survivors and nonsurvivors. However, despite these significant
findings, the clinical usefulness of lung volumes before 26 weeks' gestation
remains doubtful because of the overlap of the 95% confidence intervals. This
period is critical because it is a time when decisions such as whether to
terminate pregnancy are considered. When the TLV calculation is compared with
the presence or absence of oligohydramnios by MRI evaluation, no difference
was seen between their sensitivities and specificities for nonsurvivor
outcomes. The presence or absence of oligohydramnios before 26 weeks'
gestation is clearly a better prognostic indicator because of the significant
overlapping of lung volumes between the two groups.
It is unfortunate that our series did not have enough cases of
oligohydramnios with normal lung volumes or the converse of normal amnionic
fluid level with abnormal lung volumes and a TLV–gestational age ratio
of less than 0.90. It may be that in these cases in which there is discordance
between lung volumes and amnionic fluid levels, a true benefit arises from
calculation of lung volumes. It would be important to know, for example,
whether a TLV–gestational age ratio greater than 0.90 in the setting of
oligohydramnios is a reassuring finding, and likewise whether normal amnionic
fluid but TLV–gestational age ratio less than 0.90 is a predictor of
poor prognosis.
Our study found four discordant cases in the nonsurvivors. One was a
22-week fetus that had a large lymphangioma predominantly in the pelvis that
suffered a subsequent intrauterine death. The amnionic fluid was normal and
the TLV–gestational age ratio was less than 0.90. Pulmonary hypoplasia
was not likely responsible for the fetal death. There were two cases of
posterior urethral valves (30 and 35 weeks' gestation) and one case of renal
agenesis (25 weeks' gestation), both with oligohydramnios, that had
TLV–gestational age ratios of more than 0.90. The cases after 26 weeks'
gestation indicate false-negative results unless renal complications rather
than pulmonary hypoplasia were the cause of death. The case of renal agenesis
at 25 weeks had a TLV–gestational age ratio of 0.91. This is a
false-negative result and emphasizes the overlapping of ratios before 26
weeks' gestation and the superiority of the presence or absence of
oligohydramnios at this time.
The survivor group included two cases of prune-belly syndrome with
TLV–gestational age ratios of less than 0.90. One fetus was at 22 weeks
and the other at 29 weeks, and both had normal amnionic fluid levels. The
29-week fetus showed a false-positive TLV–gestational age ratio; the
presence or absence of oligohydramnios is most predictive at 22 weeks.
Two fetuses in the survivor group had oligohydramnios and normal lung
volumes. One had a megaureter at 32 weeks with preterm premature rupture of
membranes, and the other had unilateral renal agenesis with unexplained
oligohydramnios at 28 weeks. Both cases show the potential benefit of lung
volume calculations.
It is reasonable to continue further investigation of lung volumes in
fetuses referred for second-opinion MRI for genitourinary abnormalities after
26 weeks of gestation.
Kuwashima et al. [18] took
a novel approach to pulmonary hypoplasia by evaluating the MR signal intensity
of the lung in comparison with the signal intensity of the liver. That study
had 23 patients, nine of whom had genitourinary abnormalities. Fetuses without
pulmonary hypoplasia after 25 weeks' gestation showed a higher signal
intensity ratio than those with pulmonary hypoplasia (p < 0.01).
In the future, it might be possible to evaluate this technique in our series
of patients.
Other investigators have studied the usefulness of fetal lung volumes in
other settings. Walsh et al.
[3] calculated fetal lung
volumes in 41 fetuses with congenital diaphragmatic hernia (CDH) and found
that neither the right lung volume, the left lung volume, nor the total lung
volume was predictive of outcome. Devine et al.
[19] evaluated 16 patients and
Twickler et al. [17] evaluated
eight patients with CDH, and their conclusions were the same. Coakley et al.
[11] correlated lung volumes
with the lung–head ratio on sonography (r = 0.50, p
< 0.05) in 18 fetuses with CDH, but an outcome analysis was not performed
to determine the predictive value for lethal pulmonary hypoplasia.
One limitation of our study is that we did not include a group of healthy
controls in our series. Two patients in the survivor group gave birth to
healthy infants. Both of those infants had sonographic findings suspicious for
renal abnormalities but were found to be normal on MRI and neonatal follow-up.
If we compare the slope of the TLV by gestational age in the survivors in our
study with those of previous studies in healthy fetuses, the nomograms appear
nearly identical
[9–12].
Another limitation is the retrospective nature of our study. A prospective
analysis would be superior. One significant bias as a result of this
retrospective analysis is the fact that most of the nonsurvivors were
evaluated before 26 weeks, and most of the survivors were evaluated after 26
weeks.
At this juncture, the value of fetal MRI-calculated TLVs in the setting of
genitourinary abnormalities is no better than the evaluation of amniotic fluid
after 26 weeks' gestation. Further investigation should be focused on those
fetuses in which there is discordance between amnionic fluid levels and
calculated lung volumes after 26 weeks of gestation.
References
- Twickler DM, Magee KP, Caire JT, Zaretsky M, Ramus RM. Second
opinion magnetic resonance imaging for suspected fetal central nervous system
abnormalities. Am J Obstet Gynecol 2003;188
: 491-496
- Levine D, Barnes PD, Madsen JR, Abbott J, Mehta T, Edelman RR.
Central nervous system abnormalities assessed with prenatal magnetic resonance
imaging. Obstet Gynecol 1999;94
: 1011-1019[Abstract/Free Full Text]
- Walsh DS, Hubbard AM, Olutoye OO, et al. Assessment of fetal lung
volumes and liver herniation with magnetic resonance imaging in congenital
diaphragmatic hernia. Am J Obstet Gynecol2000; 183:1067
-1069[CrossRef][Medline]
- Caire JT, Ramus RM, Magee KP, Fullington BK, Ewalt DH, Twickler DM.
MRI of fetal genitourinary anomalies. AJR2003; 181:1381
-1385[Abstract/Free Full Text]
- Levine D, Barnewolt CE, Mehta TS, Trop I, Estroff J, Wong G. Fetal
thoracic abnormalities: MR imaging. Radiology2003; 228:379
-388[Abstract/Free Full Text]
- Hubbard AM, Adzick NS, Crombleholme TM, et al. Congenital chest
lesions: diagnosis and characterization with prenatal MR imaging.
Radiology 1999;212
: 43-48[Abstract/Free Full Text]
- Poutamo J, Vanninen R, Partanen K, Kirkinen P. Diagnosing fetal
urinary tract abnormalities: benefits of MRI compared to ultrasonography.
Acta Obstet Gynecol Scand 2000;79
: 65-71[CrossRef][Medline]
- Garden AS, Roberts N. Fetal and fetal organ volume estimations with
magnetic resonance imaging. Am J Obstet Gynecol1996; 175:442
-448[CrossRef][Medline]
- Duncan KR, Gowland PA, Moore RJ, Baker PN, Johnson IR. Assessment
of fetal lung growth in utero with echo-planar MR imaging.
Radiology 1999;210
: 197-200[Abstract/Free Full Text]
- Rypens F, Metens T, Rocourt N, et al. Fetal lung volume: estimation
at MR imaging—initial results. Radiology2001; 219:236
-241[Abstract/Free Full Text]
- Coakley FV, Lopoo JB, Lu Y, et al. Normal and hypoplastic fetal
lungs: volumetric assessment with prenatal single-shot rapid acquisition with
relaxation enhancement MR imaging. Radiology2000; 216:107
-111[Abstract/Free Full Text]
- Baker DM, Johnson IR, Gowland PA, Freeman A, Adams V, Mansfield P.
Estimation of fetal lung volume using echo-planar magnetic resonance imaging.
Obstet Gynecol 1994;83
: 951-954[Abstract]
- Ruano R, Joubin L, Sonigo P, et al. Fetal lung volume estimated by
3-dimensional ultrasonography and magnetic resonance imaging in cases with
isolated congenital diaphragmatic hernia. J Ultrasound
Med 2004; 23:353
-358[Abstract/Free Full Text]
- Sabogal JC, Becker E, Bega G, et al. Reproducibility of fetal lung
measurements with 3-dimensional ultrasonography. J Ultrasound
Med 2004; 23:347
-352[Abstract/Free Full Text]
- Kalache KD, Espinoza J, Chaiworspongsa T, et al. Three-dimensional
ultrasound fetal lung volume measurement: a systematic study comparing the
multiplanar method with the rotational (VOCAL) technique.
Ultrasound Obstet Gynecol 2003;21
: 111-118[Medline]
- Paek BW, Coakley FV, Lu Y, et al. Congenital diaphragmatic hernia:
prenatal evaluation with MR lung volumetry—preliminary experience.
Radiology 2001;220
: 63-67[Abstract/Free Full Text]
- Twickler DM, Caire J, McIntire DD, Magee KP, Ramus RM. Feasibility
of magnetic resonance lung volumes in fetuses at risk for pulmonary
hypoplasia. Am J Obstet Gynecol 2001;185
: S246[CrossRef]
- Kuwashima S, Nishimuro G, Imura F, Kohno T, Kohno A, Fujoka M. Low
intensity fetal lungs on MRI may suggest the diagnosis of pulmonary
hypoplasia. Pediatr Radiol 2001;31
: 669-672[CrossRef][Medline]
- Devine P, Budonick N, Alvarado G, Smok D, Graham G, D'Alton M. The
role of magnetic resonance imaging in fetuses with congenital diaphragmatic
hernia. Am J Obstet Gynecol 2002;187
: S209
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
J. F. Kazan-Tannus, V. Dialani, M. L. Kataoka, G. Chiang, H. A. Feldman, J. S. Brown, and D. Levine
MR Volumetry of Brain and CSF in Fetuses Referred for Ventriculomegaly
Am. J. Roentgenol.,
July 1, 2007;
189(1):
145 - 151.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
I. Klaassen, T. J. Neuhaus, D. E. Mueller-Wiefel, and M. J. Kemper
Antenatal oligohydramnios of renal origin: long-term outcome
Nephrol. Dial. Transplant.,
February 1, 2007;
22(2):
432 - 439.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
