Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print April 26, 2006, 10.1148/radiol.2393050027
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2393050027v1
239/3/839    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Boichot, C.
Right arrow Articles by Brunotte, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Boichot, C.
Right arrow Articles by Brunotte, F.
(Radiology 2006;239:839-848.)
© RSNA, 2006

Pediatric Imaging

Term Neonate Prognoses after Perinatal Asphyxia: Contributions of MR Imaging, MR Spectroscopy, Relaxation Times, and Apparent Diffusion Coefficients1

Christophe Boichot, MD, Paul M. Walker, PhD, Christine Durand, MD, Marianne Grimaldi, MD, Séverine Chapuis, MD, Jean B. Gouyon, MD and François Brunotte, MD

1 From the Departments of Magnetic Resonance Spectroscopy (C.B., P.M.W., F.B.), Radiology (C.D., S.C.), and Neonatology (M.G., J.B.G.), Hopital d'Enfants, University Hospital of Dijon, Dijon, France. Received January 7, 2005; revision requested March 11; revision received June 9; accepted July 11; final version accepted September 1. Address correspondence to P.M.W., Laboratoire de Physiopathologie et Pharmacologie Cardiovasculaires Expérimentales, Faculte de Medecine, Blvd Jeanne d'Arc, 21000 Dijon, France (e-mail: pwalker{at}u-bourgogne.fr).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 
Purpose: To retrospectively evaluate magnetic resonance (MR) imaging, hydrogen 1 (1H) MR spectroscopy, apparent diffusion coefficient (ADC), T1, and T2 measurements for prediction of late neurologic outcome in term neonates after severe perinatal asphyxia.

Materials and Methods: This study was approved by the local ethics committee. Informed consent from parents was not required. Thirty term neonates (12 boys, 18 girls; age range, 2–12 days) with severe hypoxic-ischemic encephalopathy were examined during the first 12 days of life with conventional and diffusion-weighted cerebral MR imaging, 1H MR spectroscopy with absolute quantification, and T1 and T2 measurements. Quantitative 1H MR spectroscopy, T1, and T2 data were acquired on one 10-mm slab positioned at the level of the basal ganglia. The neonates were assigned to one of two groups according to their late (>12-month follow-up) neurologic outcome: those with an unfavorable outcome—that is, death or severe disability—and those with a favorable outcome. Clinical data, MR signal intensity abnormalities, ADCs, 1H MR spectroscopy findings, and relaxation times were compared by using {chi}2 testing and analysis of variance to individualize the prognostic indicators.

Results: The unfavorable (n = 16) and favorable (n = 14) outcome groups were similar in terms of clinical data (ie, Apgar scores, visceral hypoxic injuries), visualization of brain edema on MR images, and T1 and T2 relaxation times. Late unfavorable neurologic outcome was associated with a mixed pattern of cortical and basal ganglia signal intensity abnormalities on MR images (13 babies with unfavorable vs three babies with favorable outcomes, P = .001) and with decreased absolute N-acetylaspartate (NAA) and choline concentrations in all brain structures, especially the basal ganglia (mean NAA concentration: 2.72 mmol/L in unfavorable outcome group vs 4.66 mmol/L in favorable outcome group, P < 5 x 10–9), as measured with MR spectroscopy. In the basal ganglia, an NAA concentration lower than 4 mmol/L indicated an unfavorable individual prognosis with 94% sensitivity and 93% specificity. Significantly reduced ADCs also were noted in the unfavorable outcome group, but only during the first 6 days of life.

Conclusion: Conventional MR imaging findings, spectroscopically measured absolute NAA and choline concentrations, and ADCs are complementary tools for predicting the individual outcomes of severely asphyxiated term neonates.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 
Perinatal asphyxia is an acute cerebral hypoxia-ischemia due to a critical impairment of the intrapartum gas exchange (1). Perinatal asphyxia remains a major cause of pediatric mortality and morbidity, with possible long-term neurologic sequelae such as cerebral palsy, mental retardation, or epilepsy (2). Because the prognosis for any given baby is uncertain, reliable prognostic indicators are needed. Multiple clinical parameters have been suggested, but few have been used successfully (3).

For more than a decade, conventional T1- and T2-weighted magnetic resonance (MR) imaging has been considered the best modality for imaging the neonatal brain (4–6). MR imaging enables early visualization of brain lesions in babies with hypoxic-ischemic encephalopathy (7–12). However, interobserver reproducibility appears low (13), and the value of MR imaging in predicting the outcome associated with this abnormality appears to be limited (14,15). Many attempts to improve the prognostic value of MR imaging have been made and include the use of contrast enhancement (16) and scoring systems based on visual analysis (17–19). More recently, MR spectroscopy and diffusion-weighted imaging of the brain have revealed brain ischemic injuries earlier than T1- or T2-weighted MR imaging and have been shown to be of potential prognostic value (20–32). Thus, the purpose of our study was to retrospectively evaluate MR imaging, hydrogen 1 (1H) MR spectroscopy, apparent diffusion coefficient (ADC), T1, and T2 measurements for prediction of late neurologic outcome in term neonates after severe perinatal asphyxia.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 APPENDIX
 References
 
Patient Selection and Clinical Data Collection
From review of our hospital database, we retrospectively identified neonates who had been examined with brain MR imaging between January 1997 and June 2003. Babies were selected if they were term neonates (gestational age, >37 weeks) who had had severe perinatal asphyxia and been examined with MR imaging during the first 12 days of life. Severe perinatal asphyxia was defined on the basis of the criteria established by the Neonatal Encephalopathy Committee Opinion in 2003 (33) and thus was diagnosed when the neonate had encephalopathy with seizures (corresponding to stages 2 and 3 of the Sarnat-Sarnat classification [34]) in association with at least two of the following three criteria: (a) intrapartum fetal distress, fetal heart rate abnormalities, or meconium-stained amniotic fluid; (b) neonatal distress, as indicated by a low 5-minute Apgar score (<5), an umbilical artery pH level of less than 7.10, or the need for immediate resuscitation; and (c) organ dysfunction indicating asphyxia (35). Exclusion criteria were consanguinity, intrauterine infection or trauma (ie, skull fracture or blunt trauma of the maternal abdomen), cardiac or central nervous system malformation, chromosomal abnormality, and/or inborn metabolic error.

The data on 30 term neonates (12 boys, 18 girls) aged 2–12 days (mean, 6 days) were included in this study, which was approved by the local ethics committee. Informed consent from the parents was not required. In this study, no MR imaging data were lost owing to technical problems, thanks to the use of a rigorous shimming procedure and MR-dedicated equipment for mechanical ventilation and monitoring. To limit head motion, the neonate's head was immobilized with cushions.

Outcome Assessments
The neurologic examinations were routinely performed by the pediatricians in charge of the neonates. Information regarding the outcomes of the children—such as motor and tone evaluation results, seizures, sensory impairment, and psychodevelopmental milestone achievements—was collected from the medical records by a pediatrician (M.G.) who specialized in neurology and was blinded to the MR imaging and MR spectroscopy data. To correctly evaluate the outcomes of the babies, at least 12 months of follow-up were judged to be necessary. Neurodevelopment was categorized, according to World Health Organization criteria, into four groups: category 1, no disability; category 2, mild to moderate disability; category 3, severe disability; and category 4, death. In the present study, we simplified this classification so that only two groups were considered: favorable outcome, which encompassed categories 1 and 2, and unfavorable outcome, which encompassed categories 3 and 4.

MR Imaging Examinations
MR imaging examination was indicated for medical reasons, and MR spectroscopy was considered a part of the MR examination. The neonates were examined as soon as possible after their birth (within 2–12 days of life; mean, 6 days), during natural sleep. The MR imaging and MR spectroscopy examinations were performed with a 1.5-T unit (Magnetom Vision; Siemens, Erlangen, Germany). The neonates were monitored with electrocardiography and pulse oximetry, and a pediatrician was present throughout the examination. If necessary, an MR-compatible ventilator was used. The average total examination time was 60–70 minutes.

MR imaging methods.—The MR imaging sequences involved the acquisition of 5-mm transverse and sagittal T1-weighted spin-echo images (500/12 [repetition time msec/echo time msec]), 5-mm transverse T2-weighted dual-spin-echo images (3000/17–119), 4-mm transverse inversion-recovery images (7000/60/400 [repetition time msec/echo time msec/inversion time msec]), and diffusion-weighted multisection images. The field of view used—generally 140–160 x 160 mm, with a 154–224 x 256 matrix—was adapted to the given neonate's head size. Diffusion-weighted MR imaging was performed by using a multisection echo-planar sequence with 4000/100, a 96 x 128 matrix, and a 210-mm field of view. The diffusion-weighted MR sequence involved the use of three b values (from 0 to 1000 sec/mm2) in each of the three orthogonal directions. ADCs were measured directly on the ADC maps generated by the MR unit software (Siemens).

A multiecho T2-weighted spin-echo sequence (with 6000-msec repetition time and 16 echoes from 50 to 800 msec) was used to quantify the T2 relaxation times of water in tissue. To quantify the T1 relaxation times of water in tissue, a series of 14 T1-weighted turbo fast low-angle shot images (11/4.2/100–5000) also were acquired at the same section position.

MR spectroscopy methods.—Spectra were acquired by using a point-resolved spatially localized spectroscopy–chemical shift imaging sequence. Spectroscopy was performed in the transverse plane on one 10-mm slab positioned at the level of the basal ganglia. From this slab, we obtained data on the basal ganglia and cortical gray matter and the frontal and parieto-occipital white matter. Sixteen by 16 partitions were acquired and yielded voxel dimensions of 10 x 10 x 10 mm (1 mL). Water suppression was achieved by applying chemical shift–selective saturation pulses. For absolute metabolite quantification, the neonates were examined at an echo time of 270 msec with water suppression and at an echo time of 80 msec without water suppression. A repetition time of 1500 msec was used with and without water suppression.

Analysis of MR Imaging Data
Conventional MR imaging.—The MR images obtained in all neonates (ie, T1- and T2-weighted, inversion-recovery, and diffusion-weighted images, but not ADC maps) were reviewed independently by two experienced radiologists (S.C. and C.D., with 6 and 12 years of experience, respectively, in neonatal brain MR imaging) who were blinded to the clinical outcomes and MR spectroscopy information. Disagreements regarding image findings were resolved by means of discussion and mutual agreement. The presence or absence of edema—defined as sulcal and ventricular effacement—and intra- or extraparenchymal hematoma was determined. The following predefined structures were analyzed: basal ganglia regions (thalami, caudate and lentiform nuclei, anterior and posterior limbs of the internal capsule), cerebral cortex, periventricular and subcortical white matter, corpus callosum, brainstem, and cerebellum. For each structure and substructure, any MR signal intensity abnormality that was not attributable to edema or hematoma was documented as a low- or high-signal-intensity lesion.

MR spectroscopy and other quantitative measurements.—The spectra, T1 and T2 quantitative data, and diffusion data were analyzed by an MR scientist (P.M.W.) who had 15 years of experience in brain MR spectroscopy and was blinded to the clinical outcomes and the MR image interpretations. The chemical shift imaging data were processed by using the MRUI (Magnetic Resonance User Interface), version 99.1b (www.mrui.uab.es/mrui), spectroscopic analysis package. The resonances of four metabolites were quantified: the N-acetylaspartate (NAA) peak at 2.02 ppm, the creatine (Cr) and phosphocreatine peak at 3.02 ppm, the choline (Cho) peak at 3.20 ppm, and the lactate doublet at 1.33 ppm. Details about the quantification procedures are given in the Appendix. From the parametric (ie, ADC, T1, and T2) maps, small regions of interest encompassing the different structures were used; that is, the entire basal ganglia—not individual elements of this structure, such as the caudate nuclei, pallidum, and lentiform nuclei—was analyzed, and the frontal and parieto-occipital white matter was clearly isolated from the surrounding gray matter (P.M.W.). All of these regions were systematically analyzed, irrespective of the abnormality present. However, with the relatively thick (10-mm) sections, partial volume effects could not be avoided entirely.

Statistical Analyses
Results are presented as means ± standard deviations. The statistical tests were performed by using the SYSTAT 7.0 for Windows (SPSS, Evanston, Ill) statistical package. P < .05 indicated statistical significance. An analysis of variance with post hoc Bonferroni testing was used to compare the quantitative data (ie, clinical data [gestational ages, birth weights, Apgar scores, follow-up durations], metabolite peak area ratios, and absolute metabolite concentrations) of the neonates with an unfavorable outcome with those of the neonates with a favorable outcome. The {chi}2 test was used to assess qualitative data (clinical: sex ratios, fetal heart rate and amniotic fluid abnormalities, need for resuscitation or adrenaline use, organ dysfunction; MR imaging: presence or absence of edema, gray and white matter abnormalities).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 APPENDIX
 References
 
Clinical Data
Of the 30 term neonates, 16 had an unfavorable outcome and 14 had a favorable outcome. Clinical data (Table 1) indicated no significant differences between the two groups (favorable vs unfavorable outcome) in terms of follow-up durations (12–82 months); sex ratios; medical conditions at birth; delayed MR examinations after birth; and neurologic, hematologic, cardiac, and hepatic neonatal injuries. Only oliguria—defined as a urinary flow rate of less than 1 mL/kg per hour—was associated with an unfavorable outcome.


View this table:
[in this window]
[in a new window]

  		Table 1. Clinical Data for Favorable Outcome Group versus Unfavorable Outcome Group

 
MR Imaging Data
The two groups had a similar high frequency of cerebral edema, as indicated by sulcal and ventricular effacement (higher than 80% in each group). Globally, basal ganglia changes were frequent in the two groups but predominated in the unfavorable outcome group (94% vs 57%; P = .018). However, when each basal ganglia substructure was considered individually, no significant differences were noted (Table 2). In terms of cortical injuries, unfavorable outcome was related to signal intensity abnormalities (81% vs 36%, P = .011), especially in the frontal and perirolandic territories. In the basal ganglia and cortex, the characteristic patterns of abnormalities, which were visible as early as the second day of life, consisted of high-signal-intensity areas on T1-weighted images and low-signal-intensity areas on T2-weighted images. These changes were assumed to be due to petechial hemorrhage and not calcifications, although the latter have been described in preterm neonates. This assumption was also retrospectively confirmed in 10 patients in our study who underwent a second MR examination 14–28 days later: Bright T1 and dark T2 areas either were reduced or had disappeared at the second examination and therefore were incompatible with calcifications. On the other hand, the white matter signal intensity was little affected by hypoxia-ischemia, irrespective of the location and the prognosis.


View this table:
[in this window]
[in a new window]

  		Table 2. MR Signal Intensity Abnormalities for Favorable Outcome Group versus Unfavorable Outcome Group

 
The patterns of gray matter signal intensity changes in the two neonate groups (Table 3) indicated no or only isolated cortical abnormalities in the favorable outcome group, whereas mixed involvement (cortical and basal ganglia injuries) was found predominantly in the unfavorable outcome group. Isolated basal ganglia involvement was present in the two groups but tended to be more frequent in the favorable outcome group.


View this table:
[in this window]
[in a new window]

  		Table 3. Comparisons of Cortical and Basal Ganglia MR Signal Intensity Abnormality Patterns in the Two Outcome Groups

 
MR Spectroscopy
In the asphyxiated neonates with unfavorable outcomes, the spectroscopic data indicated dramatic decreases in absolute NAA concentrations accompanied by decreased absolute Cho and Cr concentrations. Likewise, NAA ratios were reduced, while lactose ratios increased (Fig 1). This pattern of metabolic changes was observed in all structures but was predominant in the gray matter and mainly at the basal ganglia level (Tables 4 and 5). Among these metabolite parameters, the absolute NAA concentration—particularly that in the basal ganglia—showed the most significant variation between the two groups. A cutoff value of about 4 mmol/L (Fig 2) for absolute NAA concentration was identified, and this concentration threshold enabled correct determination of the outcome of the neonates with 94% (15 of 16 neonates) sensitivity and 93% (13 of 14 neonates) specificity.


Figure 1
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: MR image and spectrum obtained in 3-day-old neonate with unfavorable outcome. (a) Transverse T1-weighted MR image (500/12) shows bilateral high-signal-intensity lenticular nuclei and lateral aspect of thalami. Arrowheads point to right-sided lentiform nucleus (top) and posterior limb of internal capsule (bottom). The basal ganglia is outlined by the box. (b) 1H MR spectrum shows decreased NAA concentration with elevated lactate (Lac) doublet concentration in the basal ganglia voxel (measured in the region of interest).

 

Figure 1
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: MR image and spectrum obtained in 3-day-old neonate with unfavorable outcome. (a) Transverse T1-weighted MR image (500/12) shows bilateral high-signal-intensity lenticular nuclei and lateral aspect of thalami. Arrowheads point to right-sided lentiform nucleus (top) and posterior limb of internal capsule (bottom). The basal ganglia is outlined by the box. (b) 1H MR spectrum shows decreased NAA concentration with elevated lactate (Lac) doublet concentration in the basal ganglia voxel (measured in the region of interest).

 

View this table:
[in this window]
[in a new window]

  		Table 4. 1H Spectroscopy Data from the Gray Matter Structures in the Two Outcome Groups

 

View this table:
[in this window]
[in a new window]

  		Table 5. 1H Spectroscopy Data from the White Matter Structures in the Two Outcome Groups

 

Figure 2
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Graph shows absolute NAA concentrations in the basal ganglia of neonates in unfavorable outcome and favorable outcome groups.

 
T1, T2, and ADC Values
Compared with T1 and T2 relaxation times in the favorable outcome group, T1 and T2 relaxation times in the unfavorable outcome group tended to be increased in all studied structures but were only slightly significant for outcome prediction in the basal ganglia. In the unfavorable outcome group, ADCs were reduced in all studied structures but were significant for outcome prediction in only the occipital gray matter and white matter. However, when only those neonates who were examined during the first 6 days of life (n = 19) were taken into account, ADCs were dramatically decreased in the unfavorable outcome group in every location (Tables 4 and 5), especially the basal ganglia (79[10–5 x mm2]/sec ± 25 vs 120[10–5 x mm2]/sec ± 8, P = .00002)


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
APPENDIX
 References
 
Our study data suggest that MR techniques are useful for examining term neonates early after severe perinatal asphyxia and can help to determine their individual late neurologic outcome. MR spectroscopy with absolute quantification of brain metabolites is particularly interesting and enables discrimination between gray and white matter involvement.

MR Imaging and Outcome
Our study results indicate a high frequency of signal intensity abnormalities in the basal ganglia: This finding may suggest, as shown with computed tomography (36), that abnormal signal intensity of the basal ganglia could be an indicator of the severity of the profound perinatal asphyxia (11). On the other hand, the signal intensity of the cortex was abnormal mainly in the unfavorable outcome group in our study. All of the neonates with cortical abnormalities also had basal ganglia abnormalities; thus, this mixed pattern (basal ganglia and cortical injuries) seems to correspond to the most severe cases of perinatal asphyxia and is consistent with findings in previous studies (15,17–19).

T1 and T2 Values and Outcome
In the present study, no useful prognostic information was extracted from the T1 and T2 relaxation times. This result is comparable to and strengthens the work of Coskun et al (37), who used the ratio of the signal intensity of the given structure to the signal intensity of the vitreous of the ocular globe for the indirect evaluation of the relaxation times. Because the MR signal intensity is mainly related to water content, the edema should have increased the T1 and T2 values in the two groups in the same way and thereby decreased the effect of hypoxia on relaxation changes. This hypothesis is difficult to verify because relatively few studies with healthy neonates have been performed to determine normal relaxation times and the results of these investigations are discordant (38–42). T1 values of 1700–2300 msec in white matter and 1140–1500 msec in gray matter have been reported (38,39). The ranges of T2 values are greater, ranging between 130 and 394 msec in white matter and between 100 and 206 msec in the basal ganglia (38,40–42).

ADCs and Outcome
Diffusion-weighted MR images enable the detection of brain lesions in babies within the first hours after a hypoxic-ischemic insult because they can depict a water mobility impairment before any signal intensity changes on T1- or T2-weighted images occur (28). In a study involving 26 patients, Johnson et al (24) found diffusion abnormalities to be predictive of adverse outcomes, with a positive predictive value of 83% (10 of 12 patients) and a negative predictive value of 86% (12 of 14 patients). In 13 term babies suspected of having hypoxic-ischemic encephalopathy, Wolf et al (29) observed decreased ADCs in the basal ganglia and the frontal and parietal white matter, but they did not assess the relationship between quantitative ADC measurements and outcome.

However, despite the positive contribution of diffusion-weighted imaging to outcome determinations, the present study revealed some limitations in predicting outcomes by using quantified ADCs. In agreement with the findings of Zarifi et al (30), the ADC measured during the first 10 days of life was not associated with a late prognosis. However, when we limited the analysis to that of data collected during the first 6 days of life, we observed highly significant differences. In agreement with other authors (28,32), we assume that this result may be explained by the temporal evolution of the ADC after a hypoxic-ischemic brain injury: With adult stroke (43) and experimental hypoxic-ischemic injury in animals (44), a maximal decrease in the ADC has been observed about 1–2 days after the insult and followed by a progressive increase, with pseudonormalized values by day 8 and elevated values after day 10. Nevertheless, a larger study is needed to validate the hypothesis that a similar pattern of ADC behavior occurs in neonates after perinatal asphyxia.

1H MR Spectroscopy and Outcome
Compared with single-voxel MR spectroscopy, where spectra are obtained from one or two voxels 5 cm3 or larger, multivoxel MR spectroscopy allows spectra to be obtained simultaneously from many smaller (~1 cm3) voxels. With this technique, the studied voxels encompass most of the pertinent cerebral structures in the transverse plane of the basal ganglia, so prognostic data can be extracted simultaneously from the cortex, basal ganglia, and white matter. The multivoxel approach does have disadvantages, however, such as longer acquisition times, nonnegligible intervoxel contamination, and difficulties when short-echo-time sequences are used. Nevertheless, the present study findings emphasize the prognostic value of 1H MR spectroscopy for assessment of perinatal asphyxia: Metabolic impairment involving every brain structure—even the frontal and occipital white matter, where MR images showed few signal intensity abnormalities—was identified in the unfavorable outcome group.

Although our work mirrors previous studies (28,31) with results suggesting that metabolite ratios are interesting predictors of outcome in neonates, metabolite ratios remain an indirect method of evaluating metabolite concentration: Variations in the numerator metabolite value can be assessed accurately only when the metabolite value used as the denominator is stable under abnormal conditions. The present study findings partially support the proposal that the Cr concentration could be such an ideal denominator in cases of hypoxia-ischemia (45): The absolute Cr concentration in half of the brain locations studied was not significantly modified in the unfavorable outcome group. The lactate-Cho ratio, which has often been an important prognostic factor in previous studies (30), also had an echo—albeit a somewhat weaker one—in the present study. The presence of lactate in spectra is transitory, and the amplitude of this metabolite will depend on the severity of the ischemic insult. Moreover, the delay to spectroscopy becomes crucial. In our study, the dispersion of the lactate ratio was important and thereby reduced the statistical power of this metabolite measurement.

If metabolite ratios are deemed inappropriate, the absolute quantification of cerebral metabolites can also be performed by using external or internal references, such as tissue water. Among the proposed methods, the use of an external reference generally yields much larger standard deviations than does the use of tissue water as an internal reference (46). Ideally, extrapolation of the internal water calibration method to compare absolute metabolite concentrations between two abnormality groups would require that the following two assumptions be met: first, that the content of internal water is known and is the same in the two groups (39) and second, that the T1 and T2 relaxation times of the water and the studied metabolites also are known or are unchanged by the abnormality. According to the first hypothesis, we can suggest that in the present study, because the brain tissue relaxation times were similar in the two groups, the water content, 95% of which is detectable with use of the MR approach (5% of the tissue water remains invisible to the technique) (41), may have been similar in the two groups. According to the second hypothesis, the behavior of T1 and T2 relaxation times during an hypoxic-ischemic insult has not been well characterized: Cady et al described increased NAA, Cho, and Cr T2 values in term neonates (45) and newborn piglets (47), while decreased NAA T2 values were observed with adult stroke by Walker et al (43) and with rat brain global ischemia by Kettunen et al (48). To avoid problems with tissue water concentrations, the T1 and T2 values in tissue water should be measured, whereas metabolite relaxation time measurements would not be feasible in the clinical context. Whatever the potential errors were, the absolute NAA concentration in the basal ganglia was revealed as a powerful prognostic indicator of individual neurologic outcome. Furthermore, quantification is particularly useful for comparing metabolite levels in serial examinations of the same neonate.

As in most studies, in the present work there were a number of technical limitations. The use of a single 10-mm slab for MR spectroscopy meant that only a limited portion of the entire brain structure was sampled. Recently developed multisection MR spectroscopy techniques are now available for routine applications. However, such options were not available at our clinical site during the study period. Likewise, the use of a 10-mm slab in the newborn brain evidently led to some volume averaging, and small structures such as the basal ganglia cannot be sampled without some degree of contamination from neighboring structures.

It must also be recognized that the strict inclusion criteria used for this study, which enabled us to identify a homogeneous population, were the main drawback: The determined prognostic indicators are applicable only under the conditions established in this study—that is, in the setting of retrospectively confirmed severe perinatal asphyxia. These results should be extrapolated to ongoing evaluations of neonatal encephalopathy with caution—that is, with the prior elimination of MR imaging findings of nonhypoxic encephalopathy (such as hemorrhage, malformation, or lesions suggesting metabolic disorders or infection).


   APPENDIX
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
References
 
The residual water resonance was removed by using Hankle single-value deconvolution filtering (49). Peak detection and quantification of selected resonances were performed in the time domain by using a VARPRO (variable projection)-like algorithm called advanced method for accurate, robust, and efficient spectral fitting, or AMARES (50), which enables the inclusion of a large amount of prior knowledge. The resonances of four metabolites were quantified: the NAA peak at 2.02 ppm, the Cr and phosphocreatine peak at 3.02 ppm, the Cho peak at 3.20 ppm, and the lactate doublet at 1.33 ppm. Peak integrals were quantified by means of fitting to a Gaussian line shape.

The absolute concentrations of the proton metabolites in the brain can be estimated by using the following equation:

Formula
where |M| and |H2O| are the absolute concentrations of the given metabolite and the tissue water, respectively, in millimoles per liter; Sm and Sw are the peak integrals for the metabolite and water, respectively; T1m and T1w are the T1 relaxation times of the metabolite and water, respectively; T2m and T2w are the T2 relaxation times of the metabolite and water, respectively; TE is the echo time; TR is the repetition time; {Delta} is the difference between the receiver gain used for the metabolite acquisitions and the receiver gain used for the unsuppressed water acquisitions, in decibels; and n is the number of protons contributing to the metabolite resonance: n = 3 for NAA, n = 9 for Cho, and n = 3 for Cr and phosphocreatine. Any contribution from the cerebral spinal fluid was taken into account in the parameter absolute tissue water concentration.

For spectroscopic voxel quantification, the T2 relaxation data were processed by using the PV-WAVE 6.10 (Visual Numerics, Boulder, Colo) image-processing program. A grid defined by the 16 x 16 chemical shift imaging acquisition matrix was applied to the T2-weighted images to subdivide them into 10 x 10-mm voxels. Thus, the tissue water T2 data could be directly compared with both the metabolite T2 data and the water T2 data collected at MR spectroscopy. The spin-spin relaxation times of the tissue water were calculated by using a biexponential regression model based on a modified Levenberg-Marquardt algorithm. This two-component model took into account the presence of varying amounts of cerebral spinal fluid within the voxel. The T1 relaxation data were processed in a similar fashion but were calculated by using a monoexponential type regression model.


   FOOTNOTES
 

Abbreviations: ADC = apparent diffusion coefficient • Cho = choline • Cr = creatine • NAA = N-acetylaspartate

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, P.M.W., F.B.; 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, C.B., P.M.W.; clinical studies, all authors; statistical analysis, C.B., P.M.W.; and manuscript editing, C.B., P.M.W., C.D., F.B.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 APPENDIX
 References
 

  1. Low JA. Intrapartum fetal asphyxia: definition, diagnosis, and classification. Am J Obstet Gynecol 1997;176:957–959.[CrossRef][Medline]
  2. Dilenge ME, Majnemer A, Shevell MI. Long-term developmental outcome of asphyxiated term neonates. J Child Neurol 2001;16:781–792.[Medline]
  3. Patel J, Edwards AD. Prediction of outcome after perinatal asphyxia. Curr Opin Pediatr 1997;9:128–132.[CrossRef][Medline]
  4. McArdle CB, Richardson CJ, Nicholas DA, Mirfakhraee M, Hayden CK, Amparo EG. Developmental features of the neonatal brain: MR imaging. I. Gray-white matter differentiation and myelination. Radiology 1987;162:223–229.
  5. Barkovich AJ, Truwit CL. Brain damage from perinatal asphyxia: correlation of MR findings with gestational age. AJNR Am J Neuroradiol 1990;11:1087–1096.[Abstract]
  6. Barkovich AJ. The encephalopathic neonate: choosing the proper imaging technique. AJNR Am J Neuroradiol 1997;18:1816–1820.[Medline]
  7. McArdle CB, Richardson CJ, Hayden CK, Nicholas DA, Amparo EG. Abnormalities of the neonatal brain: MR imaging. II. Hypoxic-ischemic brain injury. Radiology 1987;163:395–403.
  8. Keeney SE, Adcock EW, McArdle CB. Prospective observations of 100 high-risk neonates by high-field (1.5 tesla) magnetic resonance imaging of the central nervous system. II. Lesions associated with hypoxic-ischemic encephalopathy. Pediatrics 1991;87:431–438.
  9. Barkovich AJ. MR and CT evaluation of profound neonatal and infantile asphyxia. AJNR Am J Neuroradiol 1992;13:959–972.[Abstract]
  10. Baenziger O, Martin E, Steinlin M, et al. Early pattern recognition in severe perinatal asphyxia: a prospective MRI study. Neuroradiology 1993;35:437–442.[CrossRef][Medline]
  11. Barkovich AJ, Westmark K, Partridge C, Sola A, Ferriero DM. Perinatal asphyxia: MR findings in the first 10 days. AJNR Am J Neuroradiol 1995;16:427–438.[Abstract]
  12. Sie LT, van der Knaap MS, Oosting J, de Vries LS, Lafeber HN, Valk J. MR patterns of hypoxic-ischemic brain damage after prenatal, perinatal or postnatal asphyxia. Neuropediatrics 2000;31:128–136.[CrossRef][Medline]
  13. Jouvet P, Cowan FM, Cox P, et al. Reproducibility and accuracy of MR imaging of the brain after severe birth asphyxia. AJNR Am J Neuroradiol 1999;20:1343–1348.[Abstract/Free Full Text]
  14. Rutherford MA, Pennock JM, Counsell SJ, et al. Abnormal magnetic resonance signal in the internal capsule predicts poor neurodevelopmental outcome in infants with hypoxic-ischemic encephalopathy. Pediatrics 1998;102:323–328.[Abstract/Free Full Text]
  15. Aida N, Nishimura G, Hachiya Y, Matsui K, Takeuchi M, Itani Y. MR imaging of perinatal brain damage: comparison of clinical outcome with initial and follow-up MR findings. AJNR Am J Neuroradiol 1998;19:1909–1921.[Abstract]
  16. Westmark KD, Barkovich AJ, Sola A, Ferriero D, Partridge JC. Patterns and implications of MR contrast enhancement in perinatal asphyxia: a preliminary report. AJNR Am J Neuroradiol 1995;16:685–692.[Abstract]
  17. Kuenzle C, Baenziger O, Martin E, et al. Prognostic value of early MR imaging in term infants with severe perinatal asphyxia. Neuropediatrics 1994;25:191–200.[Medline]
  18. Rutherford MA, Pennock JM, Schwieso JE, Cowan FM, Dubowitz LM. Hypoxic ischaemic encephalopathy: early magnetic resonance imaging findings and their evolution. Neuropediatrics 1995;26:183–191.[Medline]
  19. Barkovich AJ, Hajnal BL, Vigneron D, et al. Prediction of neuromotor outcome in perinatal asphyxia: evaluation of MR scoring systems. AJNR Am J Neuroradiol 1998;19:143–149.[Abstract]
  20. Peden CJ, Rutherford MA, Sargentoni J, Cox IJ, Bryant DJ, Dubowitz LM. Proton spectroscopy of the neonatal brain following hypoxic-ischaemic injury. Dev Med Child Neurol 1993;35:502–510.[Medline]
  21. Groenendaal F, Veenhoven RH, van der Grond J, Jansen GH, Witkamp TD, de Vries LS. Cerebral lactate and N-acetyl-aspartate/choline ratios in asphyxiated full-term neonates demonstrated in vivo using proton magnetic resonance spectroscopy. Pediatr Res 1994;35:148–151.[Medline]
  22. Leth H, Toft PB, Peitersen B, Lou HC, Henriksen O. Use of brain lactate levels to predict outcome after perinatal asphyxia. Acta Paediatr 1996;85:859–864.[Medline]
  23. Penrice J, Cady EB, Lorek A, et al. Proton magnetic resonance spectroscopy of the brain in normal preterm and term infants, and early changes after perinatal hypoxia-ischemia. Pediatr Res 1996;40:6–14.[Medline]
  24. Johnson AJ, Lee BC, Lin W. Echoplanar diffusion-weighted imaging in neonates and infants with suspected hypoxic-ischemic injury: correlation with patient outcome. AJR Am J Roentgenol 1999;172:219–226.[Abstract/Free Full Text]
  25. Hanrahan JD, Cox IJ, Azzopardi D, et al. Relation between proton magnetic resonance spectroscopy within 18 hours of birth asphyxia and neurodevelopment at 1 year of age. Dev Med Child Neurol 1999;41:76–82.[CrossRef][Medline]
  26. Barkovich AJ, Baranski K, Vigneron D, et al. Proton MR spectroscopy for the evaluation of brain injury in asphyxiated, term neonates. AJNR Am J Neuroradiol 1999;20:1399–1405.[Abstract/Free Full Text]
  27. Amess PN, Penrice J, Wylezinska M, et al. Early brain proton magnetic resonance spectroscopy and neonatal neurology related to neurodevelopmental outcome at 1 year in term infants after presumed hypoxic-ischaemic brain injury. Dev Med Child Neurol 1999;41:436–445.[CrossRef][Medline]
  28. Barkovich AJ, Westmark KD, Bedi HS, Partridge JC, Ferriero DM, Vigneron DB. Proton spectroscopy and diffusion imaging on the first day of life after perinatal asphyxia: preliminary report. AJNR Am J Neuroradiol 2001;22:1786–1794.[Abstract/Free Full Text]
  29. Wolf RL, Zimmerman RA, Clancy R, Haselgrove JH. Quantitative apparent diffusion coefficient measurements in term neonates for early detection of hypoxic-ischemic brain injury: initial experience. Radiology 2001;218:825–833.[Abstract/Free Full Text]
  30. Zarifi MK, Astrakas LG, Poussaint TY, Plessis AA, Zurakowski D, Tzika AA. Prediction of adverse outcome with cerebral lactate level and apparent diffusion coefficient in infants with perinatal asphyxia. Radiology 2002;225:859–870.[Abstract/Free Full Text]
  31. Malik GK, Pandey M, Kumar R, Chawla S, Rathi B, Gupta RK. MR imaging and in vivo proton spectroscopy of the brain in neonates with hypoxic ischemic encephalopathy. Eur J Radiol 2002;43:6–13.[CrossRef][Medline]
  32. Forbes KP, Pipe JG, Bird R. Neonatal hypoxic-ischemic encephalopathy: detection with diffusion-weighted MR imaging. AJNR Am J Neuroradiol 2000;21:1490–1496.[Abstract/Free Full Text]
  33. Phelan JP, Martin GI, Korst LM. Birth asphyxia and cerebral palsy. Clin Perinatol 2005;32:61–76.[CrossRef][Medline]
  34. Sarnat HB, Sarnat MS. Neonatal encephalopathy following fetal distress: a clinical and electroencephalographic study. Arch Neurol 1976;33:696–705.[Abstract]
  35. Carter BS, Haverkamp AD, Merenstein GB. The definition of acute perinatal asphyxia. Clin Perinatol 1993;20:287–304.[Medline]
  36. Roland EH, Poskitt K, Rodriguez E, Lupton BA, Hill A. Perinatal hypoxic-ischemic thalamic injury: clinical features and neuroimaging. Ann Neurol 1998;44:161–166.[CrossRef][Medline]
  37. Coskun A, Lequin M, Segal M, Vigneron DB, Ferriero DM, Barkovich AJ. Quantitative analysis of MR images in asphyxiated neonates: correlation with neurodevelopmental outcome. AJNR Am J Neuroradiol 2001;22:400–405.[Abstract/Free Full Text]
  38. Jones RA, Palasis S, Grattan-Smith JD. MRI of the neonatal brain: optimization of spin-echo parameters. AJR Am J Roentgenol 2004;182:367–372.[Abstract/Free Full Text]
  39. Steen RG, Hunte M, Traipe E, et al. Brain T1 in young children with sickle cell disease: evidence of early abnormalities in brain development. Magn Reson Imaging 2004;22:299–306.[CrossRef][Medline]
  40. Engelbrecht V, Rassek M, Preiss S, Wald C, Modder U. Age-dependent changes in magnetization transfer contrast of white matter in the pediatric brain. AJNR Am J Neuroradiol 1998;19:1923–1929.[Abstract]
  41. Ferrie JC, Barantin L, Saliba E, et al. MR assessment of the brain maturation during the perinatal period: quantitative T2 MR study in premature newborns. Magn Reson Imaging 1999;17:1275–1288.[CrossRef][Medline]
  42. Thornton JS, Amess PN, Penrice J, Chong WK, Wyatt JS, Ordidge RJ. Cerebral tissue water spin-spin relaxation times in human neonates at 2.4 tesla: methodology and the effects of maturation. Magn Reson Imaging 1999;17:1289–1295.[CrossRef][Medline]
  43. Walker PM, Ben Salem D, Lalande A, Giroud M, Brunotte F. Time course of NAA T2 and ADC(w) in ischaemic stroke patients: 1H MRS imaging and diffusion-weighted MRI. J Neurol Sci 2004;220:23–28.[CrossRef][Medline]
  44. Thornton JS, Ordidge RJ, Penrice J, et al. Temporal and anatomical variations of brain water apparent diffusion coefficient in perinatal cerebral hypoxic-ischemic injury: relationships to cerebral energy metabolism. Magn Reson Med 1998;39:920–927.[Medline]
  45. Cady EB. Metabolite concentrations and relaxation in perinatal cerebral hypoxic-ischemic injury. Neurochem Res 1996;21:1043–1052.[Medline]
  46. Longo R, Bampo A, Vidimari R, Magnaldi S, Giorgini A. Absolute quantitation of brain 1H nuclear magnetic resonance spectra: comparison of different approaches. Invest Radiol 1995;30:199–203.[CrossRef][Medline]
  47. Cady EB, Lorek A, Penrice J, et al. Brain-metabolite transverse relaxation times in magnetic resonance spectroscopy increase as adenosine triphosphate depletes during secondary energy failure following acute hypoxia-ischaemia in the newborn piglet. Neurosci Lett 1994;182:201–204.[CrossRef][Medline]
  48. Kettunen MI, Grohn OH, Kauppinen RA. Quantitative T1rho NMR spectroscopy of rat cerebral metabolites in vivo: effects of global ischemia. Magn Reson Med 2004;51:875–880.[CrossRef][Medline]
  49. de Beer R, van den Boogaart A, van Ormondt D, et al. Application of time-domain fitting in the quantification of in vivo 1H spectroscopic imaging data sets. NMR Biomed 1992;5:171–178.[Medline]
  50. Vanhamme L, van den Boogaart A, van Huffel S. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson 1997;129:35–43.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 PediatricsHome page
E. W. Y. Tam, E. Widjaja, S. I. Blaser, D. L. MacGregor, P. Satodia, and A. M. Moore
Occipital Lobe Injury and Cortical Visual Outcomes After Neonatal Hypoglycemia
Pediatrics, September 1, 2008; 122(3): 507 - 512.
[Abstract] [Full Text] [PDF]

Home page
 J Child NeurolHome page
C. Hoffmann, B. Ben-Zeev, Y. Anikster, A. Nissenkorn, N. Brand, J. Kuint, and T. Kushnir
Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy in Isolated Sulfite Oxidase Deficiency
J Child Neurol, October 1, 2007; 22(10): 1214 - 1221.
[Abstract] [PDF]

Home page
 RadioGraphicsHome page
C. P. Chao, C. G. Zaleski, and A. C. Patton
Neonatal Hypoxic-Ischemic Encephalopathy: Multimodality Imaging Findings
RadioGraphics, October 1, 2006; 26(suppl_1): S159 - S172.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2393050027v1
239/3/839    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Boichot, C.
Right arrow Articles by Brunotte, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Boichot, C.
Right arrow Articles by Brunotte, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE