Objective
The purpose of this study was to investigate alterations in brain metabolism in fetuses with intrauterine growth restriction (IUGR) and evidence of cerebral redistribution of blood flow.
Study Design
Biometry and Doppler assessment of blood flow was assessed with ultrasound in 28 fetuses with IUGR and cerebral redistribution and in 41 appropriately grown control subjects. Proton magnetic resonance spectroscopy of the fetal brain was then performed to determine the presence of choline (Cho), creatine (Cr), N-acetylaspartate (NAA), and lactate and to generate ratios for NAA:Cho, NAA:Cr, and Cho:Cr.
Results
Sixty-five percent of spectra were interpretable: N-acetylaspartate, choline, and creatine peaks were identified in all these spectra; lactate was present in 5 IUGR fetuses and in 3 appropriately grown fetuses. NAA:Cr and NAA:Cho ratios were significantly lower in IUGR fetuses with cerebral redistribution.
Conclusion
Cerebral redistribution is associated with altered brain metabolism that is evidenced by a reduction in NAA:Cho and NAA:Cr ratios.
Intrauterine growth restriction (IUGR) is a major clinical concern that complicates up to 10% of all pregnancies. It is associated with an increased risk of intrauterine death, neonatal death, and preterm delivery and significant postnatal morbidity. Neonates who do survive are at increased risk of neurodevelopmental sequelae in childhood and the metabolic syndrome in later life.
Previous studies that have assessed IUGR fetuses lack uniformity in diagnostic criteria and often do not distinguish between the terms small for gestational age and IUGR . The small-for-gestational-age fetus may be defined as having an estimated fetal weight and/or abdominal circumference of <10th percentile ; however, this definition will encompass constitutionally small but healthy fetuses. True fetal growth restriction that results from placental insufficiency is a pathologic condition that is associated with abnormal Doppler velocimetry in the umbilical artery, such as an increased pulsatility index and absent or reversed end diastolic blood flow.
True fetal growth restriction is also associated with the phenomenon known as brain sparing , which is defined as an increase in blood flow to the heart and brain at the expense of other organs, such as the kidneys and gastrointestinal tract. However, this supposedly neuroprotective mechanism is unable to compensate fully, which is evidenced by the higher incidence of cerebral palsy and neurodevelopmental sequelae in children who were growth restricted in utero.
Animal models have demonstrated alterations in energy metabolism in IUGR fetuses; it has been proposed that these metabolic disturbances could underlie the increased susceptibility of IUGR fetuses to hypoxic ischemic brain injury. In the fetal brain, energy is derived primarily from glucose metabolism through the Krebs cycle and oxidative phosphorylation within mitochondria. Animal models of IUGR have documented impairments of mitochondrial function in skeletal muscle, liver, and cerebral tissue. N-acetylaspartate (NAA) is a neuronal marker that also is localized to both mature and immature oligodendrocytes, is synthesized in neuronal mitochondria, is considered to reflect mitochondrial function. This study aims to investigate differences in brain levels of NAA between appropriately grown and growth-restricted human fetuses with the use of proton magnetic resonance spectroscopy ( H-MRS), which is a technique with which the chemical composition of in vivo tissue can be assessed noninvasively.
We propose that reduced cerebral levels of NAA that are evidenced by a reduction in NAA:choline (Cho) and NAA:creatine (Cr) ratios will be present in IUGR fetuses, which may be indicative of abnormal mitochondrial metabolism and may predispose these infants to the increased incidence of perinatal brain injury and later neurodevelopmental impairments.
Materials and Methods
Pregnant women were recruited between June 2007 and January 2010 from Queen Charlotte’s and Chelsea Hospital London; the study included appropriately grown fetuses after a normal anatomy ultrasound scan and with no associated pregnancy complications and IUGR fetuses from the Centre for Fetal Care. All fetuses were dated with a first trimester dating scan. Inclusion criteria for IUGR fetuses were an estimated fetal weight <10th percentile, abnormal Doppler velocimetry in the umbilical artery that was categorized in terms of disease severity (an increased pulsatility index of >95th percentile or absent/reversed end diastolic blood flow), and evidence of cerebral redistribution (a middle cerebral pulsatility index of <5th percentile). Amniocentesis, to exclude fetal karyotype abnormalities, was offered to all women with IUGR fetuses; this procedure was accepted by only 5 women. No karyotype abnormalities were reported after delivery. Fetuses with evidence of infection or with structural brain anomalies that were detected on ultrasound scanning and/or magnetic resonance imaging were excluded. Maternal demographic data were recorded for all women. Ethical approval was granted by Hammersmith Hospital Research Ethics Committee, and informed consent was obtained from all pregnant women for both magnetic resonance imaging and ultrasound scans (REC 07/H0707/105 and 07/Q0406/16).
All fetuses underwent ultrasound assessment of biometry and Doppler velocimetry of blood flow in the umbilical artery and middle cerebral artery with a 6-MHz curvilinear 6C2 transducer on a Siemens ultrasound system (Siemens Acuson Sequoia 512; Siemens AG, Munich, Germany). Fetuses with evidence of IUGR had disease severity recorded in terms of ultrasound Doppler findings in the umbilical artery.
A 1.5-T Philips Magnetic Resonance System (Philips Achieva; Philips Medical Systems, Best, The Netherlands) with a SENSE wrap around a 5-channel cardiac coil that is positioned as close as possible to the fetal head was used for magnetic resonance imaging. Women were positioned with a left-lateral tilt. No maternal sedation was used. The total length of the magnetic resonance examination did not exceed 1 hour, and spectral acquisition was completed in <7 minutes.
Conventional magnetic resonance imaging was first obtained using T-2 weighted single shot turbo spin echo sequences (echo time: 98 msec; repetition time: 1000 msec; number of signals averaged: 2; matrix: 400 × 178 encoding steps; field of vision: 36-40 cm; 4-mm slice thickness: 4 mm; slice gap: 0.4 mm; acquisition time: 20 seconds) in the axial, sagittal, and coronal planes. These images were analyzed visually by an experienced radiologist to confirm normal anatomy and exclude disease. One IUGR fetus was excluded because of the presence an intracerebral cyst.
MRS was obtained by positioning a volume of interest, 20 × 20 × 20 mm 3 , over central brain tissue that encompassed both central grey and white matter that had been selected from a survey sequence ( Figure 1 ) . A point resolved spectroscopy sequence of echo time of 136 msec and repetition time of 1500 msec was used to obtain the spectra. Data were acquired in blocks of 4 averages and saved independently, which allowed the elimination of individual spectra, if motion or contamination was suspected to have occurred in the postprocessing stage. An echo time of 136 msec was used, because this proved to be the most effective at detecting lactate, because the lactate peak is inverted at this echo time but positive at 270 msec when it can become contaminated by lipid and other macromolecule peaks. Water suppression was performed by selective excitation. Imaging was repeated after spectral acquisition to ensure significant fetal motion had not occurred. Data were discarded, if there was evidence of significant fetal motion.
Datasets were postprocessed with the jMRUI software package and metabolite ratios calculated for NAA:Cho, NAA:Cr, and Cho:Cr. Individual spectra were summed, and those with an aberrant residual water peak were eliminated. All spectra were postprocessed by the same operator who was blinded with regards to the presence or absence of IUGR and gestation at 1H-MRS. Good reproducibility of spectral acquisition was confirmed previously.
Statistical analysis was performed with the SPSS software package (version 17; SPSS Inc, Chicago, IL). Linear regression analysis was used to adjust for altered metabolite ratios with advancing gestational age. Analysis of covariance was used to assess the effects of metabolite ratios in IUGR when corrected for gestational age.
Delivery and neonatal outcome parameters were recorded. Customized birthweight percentiles were calculated based on delivery and maternal parameters using gestation network United Kingdom edition ( http://www.gestation.net/ ).
Results
Twenty-eight fetuses with IUGR (gestational age, 20 +3 –35 +0 weeks; median, 27 +6 weeks) and 41 appropriately grown fetuses (gestational age, 22 +0 –38 +6 weeks; median, 29 +5 weeks) were included. Two appropriately grown fetuses and 2 IUGR fetuses were scanned twice during gestation. One fetus with an intracerebral cyst that was detected on magnetic resonance imaging was excluded from the study. No other structural abnormalities were identified in any fetus.
At the time of imaging, all growth-restricted fetuses with interpretable spectra had an estimated fetal weight of <10th percentile. Seven fetuses had a raised pulsatility index and cerebral redistribution; 12 fetuses had absent end diastolic blood flow; and 2 fetuses had reversed end diastolic blood flow on umbilical artery Doppler images.
Sixty-five percent of the spectra were interpretable: 28 were obtained from appropriately grown control fetuses (22 +0 –37 +5 weeks’ gestation; median, 29 +5 weeks) and 21 from IUGR fetuses (20 +3 –35 +0 weeks gestation; median, 28 +0 weeks). One appropriately grown and 2 growth-restricted fetuses had 2 interpretable spectra at different gestations. Demographics of women with interpretable fetal spectra are given in Table 1 . There were no statistical differences between the maternal demographics or gestation of fetuses with interpretable and uninterpretable spectra.
| Maternal characteristic | Intrauterine growth-restricted pregnancy | Control pregnancy |
|---|---|---|
| Age, y | ||
| Range | 24-44 | 17-44 |
| Median | 32 | 32.5 |
| Body mass index, kg/m 2 | ||
| Range | 18.5-36 | 17-32 |
| Median | 24.5 | 22.3 |
| Parity, % | ||
| 0 | 53 | 64 |
| 1 | 26 | 25 |
| 2 | 16 | 10 |
| 3 | 5 | 0 |
| Ethnicity, % | ||
| White | 53 | 60 |
| Asian | 26 | 19 |
| Southeast Asian | 5 | 7 |
| Afro-Caribbean | 5 | 7 |
| African | 0 | 0 |
| Other | 11 | 7 |
Peaks from NAA, choline, and creatine were identified in all fetuses with interpretable spectra. An example of a fetal spectrum that was obtained at an echo time of 136 msec can be seen in Figure 2 . Lactate was detected in 5 growth-restricted fetuses and 3 appropriately grown fetuses.
In the appropriately grown fetuses, linear regression analysis demonstrated that the NAA:Cho ratio increased with gestation ( P = .003) and that the Cho:Cr ratio decreased with gestation ( P = .001). No difference in the NAA:Cr ratio ( P = .867) was found ( Figure 3 ) .
The NAA:Cho and NAA:Cr ratios that were corrected for gestation were significantly lower ( P = .001 and P = .003, respectively); the Cho:Cr ratios were unchanged in IUGR fetuses compared with appropriately grown fetuses ( P = .188; Figure 3 ).
The details of delivery parameters and perinatal outcomes of fetuses with interpretable spectra are given in Tables 2 and 3 . There were no statistically significant differences between fetuses with interpretable and uninterpretable spectra. None of the fetuses that were categorized as appropriately grown demonstrated evidence of growth restriction on subsequent ultrasound scans, and all had birthweights between the 10th and 90th percentile. Fetuses that were categorized as IUGR all had birthweights <10th percentile, which supports the antenatal diagnosis. Details of the 8 fetuses with cerebral lactate present are given in Table 4 .
