Abstract
Objective. The most common finding on magnetic resonance imaging (MRI) of the brain in preterm infants at term-equivalent age is diffuse excessive high signal intensity (DEHSI) in the white matter. It is unclear whether DEHSI represents a biological abnormality. This study used diffusion-weighted imaging (DWI) to compare apparent diffusion coefficient (ADC) values in DEHSI with infants with normal imaging and those with overt brain damage to determine whether DEHSI shows the diffusion characteristics of normal or abnormal tissue.
Methods. MRI, using conventional and diffusion-weighted imaging (DWI), was performed in 50 preterm infants at term-equivalent age using a 1.5 Tesla MR scanner. The infants were divided into 3 groups on the basis of their MRI results: 1) normal white matter, 2) DEHSI, or 3) overt white matter pathology. ADC values were measured in the frontal, central, and posterior white matter at the level of the centrum semiovale. ADC values in the 3 groups of preterm infants were compared using a 1-way analysis of variance with a Bonferroni test for multiple comparisons.
Results. ADC values were significantly higher in infants with DEHSI and infants with overt white matter pathology than in infants with normal white matter. There was no significant difference between ADC values in infants with DEHSI and those with overt white matter pathology.
Conclusions. This study provides objective evidence that DEHSI represents diffuse white matter abnormality.
Improvements in neonatal intensive care have resulted in a decline in the incidence of periventricular leukomalacia (PVL), periventricular hemorrhagic infarction (PHI), and major intraventricular hemorrhage (IVH).1,2 However, magnetic resonance imaging (MRI) has demonstrated diffuse excessive high signal intensity (DEHSI) in the cerebral white matter on T2-weighted imaging in 75% of preterm infants at term-equivalent age.3 In the absence of postmortem correlation, it has been unclear whether DEHSI is an imaging correlate of diffuse white matter disease. The assessment of DEHSI by visual analysis is difficult, as the appearances are influenced by the windowing used before image processing. Therefore, an objective method of assessing the cerebral white matter in preterm infants is required to determine whether DEHSI reflects a true white matter abnormality.
Diffusion-weighted imaging (DWI) is a MR technique that demonstrates the molecular motion of water in tissue,4 and previous studies have shown that DWI is able to reveal abnormalities in the cerebral white matter of the preterm brain that are not demonstrated on conventional MRI.5–7 Apparent diffusion coefficient (ADC) values (quantitative measures of water motion) can be calculated from DWI and may be useful to assess objectively the cerebral white matter in DEHSI. The aim of this study was to quantify ADC values in the cerebral white matter in preterm infants to determine whether DEHSI has the same diffusion characteristics as normal-appearing white matter or those of overt white matter pathology.
METHODS
Ethical permission for this study was granted by the Hammersmith Hospital Research Ethics Committee, and informed parental consent was obtained for each infant. We studied 50 preterm infants between October 2000 and June 2002 at term-equivalent age. The infants were either referred for MRI because of a preexisting abnormality on cranial ultrasound or recruited at random from the neonatal intensive care unit as part of an ongoing MRI project to examine brain development in preterm infants. They represented 14% of the eligible population. The median (range) gestational age (GA) of the infants at birth was 29 weeks (25–34 weeks), and the median birth weight was 1086 g (502–2120 g). The median postmenstrual age (PMA) at scanning was 41 weeks (37–44 weeks). No infants required mechanical ventilation at the time of the MRI examination.
MRI was performed on a 1.5 Tesla Eclipse scanner (Philips Medical Systems, Cleveland, OH) using a dedicated pediatric head coil. The infants were sedated for imaging with oral chloral hydrate (20–30 mg/kg), and pulse oximetry and electrocardiograph were monitored throughout the procedure. Ear protection was used for each infant (Natus MiniMuffs, Natus Medical Inc, San Carlos, CA). An experienced neonatologist, trained in MRI procedures, was in attendance throughout the MRI examination. Transverse T1-weighted conventional spin echo (CSE; TR 500/TE 15 ms) and T2-weighted fast spin echo (FSE; TR 4500/TEeff 210 ms) images were obtained before the DWI. Single-shot echo planar DWI was obtained using the following pulse sequence parameters; TR ∞, TE 100 ms, 100 × 100 matrix, FOV 24 cm, slice thickness 5 mm. A reference image was obtained with a b value of 0 (nominal value), and DWIs were obtained with a b value of 1000 s/mm2 in the read, phase, and slice directions. The total scanning time of the DWI sequence was 37 seconds. Accuracy and reproducibility of the DWI sequence was assessed using a distilled water phantom at 20°C.
Circular regions of interest (ROIs) were positioned in the frontal, central, and posterior white matter bilaterally on the reference image (b = 0) and on the phase, read, and slice DWIs at the level of the centrum semiovale (on the transverse slice above the level of the lateral ventricles, where the central sulcus was at its maximum depth; Fig 1). Care was taken to position the ROIs so as to avoid partial volume averaging from cerebrospinal fluid or cortical gray matter. The size of the ROIs depended on the area of white matter at this level (ROI diameter: 4.2–5.4 mm). Consistency of positioning was ensured by having all ROIs positioned by a single investigator. The ADC value for the ROIs in each direction of sensitization was calculated as follows, and the directionally averaged ADC for each ROI was calculated:
where S is signal in the DW image, S0 is the signal in the reference image, and b is given by the following equation:
where γ is gyromagnetic ratio for protons, G is amplitude of the pulsed gradient, δ is duration of the pulsed gradient, and Δ is time interval between the leading edges of the 2 pulsed gradients.
Positioning of ROIs in the white matter at the level of the centrum semiovale on the read DWI.
For infants with overt white matter pathology, the ROIs were positioned as consistently as possible but avoiding regions of obvious focal pathology.
The conventional MR images were reviewed separately by 2 investigators who were unaware of the DWI results (M.A.R. and J.M.A.), and the infants were divided into 3 groups on the basis of their MR images: 1) normal white matter, 2) DEHSI, or 3) overt white matter lesions. Interobserver and intraobserver variability for the assessment of DEHSI were analyzed by calculating the κ statistic, and the following ranges for agreement were used: 0.00, poor; 0.00 to 0.20, slight; 0.21 to 0.40, fair; 0.41 to 0.6, moderate; 0.61 to 0.8, substantial; and 0.81 to 1.0, almost perfect.8 ADC analysis was performed by another investigator (S.J.C.) who was unaware of the MRI findings.
The data were tested for normality using a Shapiro Wilks test and found to be compatible with a normal distribution. Statistical analysis of the ADC values between groups and regional variation within groups was performed using a 1-way analysis of variance with a Bonferroni test for multiple comparisons. In addition, linear regression analysis was performed to examine the relationship between PMA at scanning and ADC values for each group. Repeated measures reliability testing was evaluated by calculating coefficients of reliability for each ROI (√ [2 × standard deviation of the differences between the 2 measurements2]).9
RESULTS
Phantom Studies
The mean ADC value of the distilled water was 2.01 ± 0.05 × 10−3 mm2/s, which is comparable to published values,10–12 and the reproducibility of the DWI sequence was <2.5%.
MRI
There were no complications during or immediately after the MRI studies. Thirteen infants had normal white matter (Fig 2), 23 had DEHSI (Fig 3), and 11 had overt white matter lesions. There was no significant difference in the GA at birth (P = .61), birth weight (P = .41), or PMA at scanning (P = .50) of the infants in the 3 groups. Table 1 summarizes the characteristics of the infants in the 3 groups. Of the infants with normal white matter, 1 had evidence of previous IVH, 1 had unilateral minimal ventricular dilation, and 1 had bilateral minimal ventricular dilation (Fig 4), but none had any parenchymal abnormality. Of the infants with DEHSI, 2 had evidence of previous unilateral germinal layer hemorrhage, 2 had evidence of previous bilateral germinal layer hemorrhage, and 2 had bilateral minimal ventricular dilation. None of the infants in this group had focal parenchymal lesions.
Transverse T2-weighted FSE image at the level of the centrum semiovale of an infant at 42 weeks’ PMA, who was born at 31 weeks’ GA, demonstrating normal signal intensity in the cerebral white matter.
Transverse T2-weighted FSE image at the level of the centrum semiovale of an infant at 42 weeks’ PMA, who was born at 28 weeks’ GA, demonstrating DEHSI within the cerebral white matter (arrows).
T2-weighted FSE image at the level of the basal ganglia of an infant at 40 weeks’ PMA, who was born at 30 weeks’ GA, demonstrating minimal bilateral ventricular dilation and normal-appearing white matter.
Characteristics of the Infants in Each Group
The overt white matter lesions included bilateral periventricular leukomalacia (n = 5; Fig 5), a unilateral cystic lesion in the white matter lateral to the right lentiform nucleus, consistent with an area of infarction (n = 1), unilateral PHI (n = 2), and bilateral multiple punctate lesions (n = 3; Fig 6). Two infants with PVL, all 3 of the infants with multiple punctate lesions, both infants with PHI, and the infant with the cystic lesion in the white matter lateral to the right lentiform nucleus also had areas of long T2 within the “unaffected” white matter. The 3 remaining infants with PVL demonstrated white matter atrophy and bilateral ventricular dilation. Although visual analysis of DEHSI is subjective, the κ statistic for interobserver and intraobserver variability of differentiation between normal white matter and DEHSI was high (κ = 0.68 and κ = 0.72, respectively), representing substantial agreement.8
A, T1-weighted CSE image at the midventricular level of an infant at 39 weeks’ PMA, who was born at 27 weeks’ GA, demonstrating cystic PVL in the white matter posterior and anterior to the lateral ventricles (arrows). B, T2-weighted FSE image of the same infant demonstrating the cystic lesions as high signal intensity (arrows).
T1-weighted CSE image at the midventricular level of an infant at 41 weeks’ PMA, who was born at 32 weeks’ GA, demonstrating multiple punctate high signal intensity lesions in the cerebral white matter (arrows).
ADC Values
Unpaired t tests showed no significant difference between the right and left hemispheres (P > .05), so the mean was calculated from the bilateral measurements to give ADC values for each region. The mean ADC values obtained in the frontal, central, and posterior white matter are shown in Table 2. ADC values were significantly higher in infants with DEHSI (frontal white matter, P < .0001; central white matter, P = .007; posterior white matter, P < .0001) and in infants with overt white matter lesions (frontal white matter, P < .0001; central white matter, P = .001; posterior white matter, P < .0001) than in those with normal white matter. There was no significant difference between infants with DEHSI and those with overt white matter lesions (frontal white matter, P = 1.0; central white matter, P = .42; posterior white matter, P = 1.0). Figure 7 demonstrates the ADC values obtained in the posterior white matter in the 3 groups of infants.
Graph demonstrating ADC values in the posterior white matter of the centrum semiovale in infants with normal white matter, infants with DEHSI, and infants with overt white matter pathology.
Mean Adc Values for the 3 Groups of Infants
Analysis of regional variation within each infant group showed no significant difference between ADC values in the frontal, central, and posterior white matter regions for infants with normal-appearing white matter (frontal vs central white matter, P = 1.0; central versus posterior white matter, P = 1.0; frontal vs posterior white matter, P = 1.0) and those with overt white matter lesions (frontal vs central white matter, P = 1.0; central vs posterior white matter, P = 1.0; frontal vs posterior white matter, P = 1.0). For infants with DEHSI, there was no significant difference in ADC values between the frontal and posterior white matter (P = 1.0) or between the central and posterior white matter (P = .16). However, ADC values in the frontal white matter were significantly higher than in the central white matter (P = .01) in this group of infants.
A significant negative correlation between PMA and ADC values was demonstrated in the frontal white matter for infants with normal-appearing white matter (P = .04). However, there was no significant correlation between PMA at scanning and ADC values for central (P = .59) and posterior (P = .81) white matter in infants with normal-appearing white matter, in infants with DEHSI (frontal white matter, P = 0. 54; central white matter, P = .20; posterior white matter, P = .35), or infants with overt white matter lesions (frontal white matter, P = .57; central white matter, P = .89; posterior white matter, P = .71). The coefficient of reliability for the measurement of ADC values in each region was as follows: frontal white matter, 4.4%; central white matter, 4.6%; posterior white matter, 3.5%.
DISCUSSION
This study demonstrates that ADC values in the cerebral white matter are higher in infants with DEHSI and overt white matter pathology than in infants with normal white matter on MRI and that there is no significant difference in ADC values between infants with DEHSI and those with overt white matter pathology. Therefore, DEHSI has the same diffusion characteristics as the white matter of infants with overt white matter pathology. DEHSI, the phenomenon of excessive high signal intensity on T2-weighted MRI and corresponding low signal on T1-weighted imaging, is usually most marked in the periventricular white matter but often evident throughout the white matter.3 The signal intensity associated with DEHSI exceeds the high signal intensity of normal unmyelinated white matter in the neonatal brain. Although visual analysis of this signal is subjective, interobserver agreement was good in this study.
Using line scan diffusion tensor imaging (DTI), ADC values of 1.4 ± 0.2 × 10−3 mm2/s5 and 1.5 ± 0.2 × 10−3 mm2/s13 have been found in the central white matter of preterm infants at term, which are slightly higher than our values of 1.29 ± 0.12 × 10−3 mm2/s in this region. Normative ADC values in full-term infants have been reported as 1.2 ± 0.10 × 10−3 mm2/s using line scan DTI5 and 1.43 ± 0.14 × 10−3 mm2/s using echo planar DTI14 in the central white matter and in the frontal white matter, 1.62 ± 0.16 × 10−3 mm2/s using echo planar DWI.10 Although previous studies have found a significant reduction in ADC values in normal white matter with increasing PMA from ≤32 weeks’ GA to term-equivalent age,5,7,11,14,15 with the exception of the frontal region in infants with normal-appearing white matter, we found no significant decrease in ADC with increasing PMA in this study. This is likely to be attributable to the narrow range of PMAs (37–44 weeks) studied here.
Elevations in ADC values may be caused by an increase in water content and a decrease in restriction to water motion. The reduction in ADC values in the cerebral white matter with increasing PMA in the preterm brain is probably attributable to a reduction in the cerebral water content and a decreased extracellular space, which reduces separation of structures such as cell membranes and so impedes diffusion of water.14 As other studies have demonstrated a reduction in ADC values in the cerebral white matter from the preterm period to term-equivalent age, it is possible that the elevated ADC values in DEHSI represent delayed white matter maturation in these infants.
A previous study that examined ADC values in the white matter at the level of the centrum semiovale found no regional variation between frontal, central, and posterior white matter areas.14 However, another study, which examined a greater number of white matter regions, reported varying ADC values across white matter regions in both infants with normal white matter and those with white matter injury.7 Although we found no regional variation in ADC values at the level of the centrum semiovale in infants with normal-appearing white matter or those with overt white matter lesions, ADC values in the frontal region were higher than in the central white matter in DEHSI. It is possible that elevated ADC values in the frontal white matter represent an area of more severe damage or delayed maturation compared with the central white matter at this level. It is interesting that Miller et al7 reported an absence of the normal maturational increase in anisotropy (the directional dependence of water diffusion in tissue) in this region in infants with minimal white matter injury. These results suggest an increase in susceptibility to injury in the frontal white matter.
The elevation in ADC values in infants with overt pathology reported here suggests diffuse involvement of the white matter, beyond the visually obvious lesions. These findings concur with a recent DTI study, which reported a significant increase in ADC values with increasing PMA in the frontal white matter and visual association areas in moderate white matter injury.7 Indeed, 8 of the infants with overt lesions in our study also demonstrated the signal characteristics of DEHSI throughout the white matter on visual analysis. In addition, elevated ADC values have been reported in a preterm infant with PVL at term-equivalent age in the cerebral white matter surrounding the cystic lesions.6 The elevation of ADC values in the white matter distant from the focal lesions in PVL probably reflects the diffuse white matter damage that has been identified on histopathological studies.16 In PHI, positron emission tomography studies have shown that cerebral blood flow is impaired in the ipsilateral hemisphere distant from the focal lesion, suggesting that the injury extends beyond the obvious focal lesion.17 This may account for the elevated ADC in the ipsilateral hemisphere in PHI. ADC values were elevated in both hemispheres in the 2 infants with PHI in this study, suggesting bilateral and diffuse white matter involvement. Visual analysis of the images showed DEHSI in the contralateral hemisphere in PHI.
A line scan DTI study of preterm infants at term-equivalent age found that relative anisotropy was reduced in the central white matter in infants with focal white matter pathology.13 However, they found no difference in ADC values between infants with pathology and those with no evidence of pathology on MRI. The reasons for the differing results between this line scan study and our work cannot be explained by the different techniques used. However, it is possible that the line scan DTI study13 included infants with DEHSI in the “no focal pathology” group. If this were the case, then ADC values in this group would be elevated, and so there may be no difference in ADC values between the 2 groups of infants.
It has been shown that ADC values are higher and anisotropy is lower in preterm infants with no evidence of abnormality on MRI at term-equivalent age compared with infants who are born at term,5 suggesting differences in the development of cerebral white matter in preterm infants compared with infants who are born at term. The reasons for this are not clear, but the white matter in preterm infants is more susceptible to injury. This is attributed to the low blood flow to the cerebral white matter in preterm infants18 and the susceptibility of immature oligodendrocytes to injury from free radicals,19 certain cytokines,20,21 and glutamate toxicity.22,23 In addition, steroid exposure has been shown to affect the development of the preterm brain.24 Although the relationship between lesions such as PVL and cerebral palsy are established, the neuropathological correlates for the less severe neurodevelopmental impairments in preterm infants are incompletely defined. As focal parenchymal lesions are relatively uncommon, DEHSI may provide the radiologic substrate for the high incidence of neurocognitive deficits seen in preterm infants. DEHSI may represent delayed white matter maturation or diffuse pathology in the white matter. However, in view of the fact that DEHSI is clearly distinct from normal-appearing white matter, we propose that DEHSI should be regarded as a diffuse white matter abnormality.
CONCLUSIONS
DWI is able to provide a noninvasive, objective assessment of the cerebral white matter in preterm infants. Elevated ADC values in DEHSI and infants with overt white matter pathology may be caused by an increase in water content and a decrease in restriction to water motion in the cerebral white matter and suggest diffuse white matter abnormality in these infants. Follow-up studies will determine whether DEHSI is related to poor neurodevelopmental outcome.
Acknowledgments
We are grateful for the support of the Medical Research Council, Philips Medical Systems, and the Garfield Weston Foundation.
Footnotes
- Received September 11, 2002.
- Accepted December 20, 2002.
- Reprint requests to (M.A.R.) Robert Steiner MR Unit, Imaging Sciences Department, MRC Clinical Sciences Centre, Imperial College, Hammersmith Campus, DuCane Rd, London W12 0HS, United Kingdom. E-mail: m.rutherford{at}imperial.ac.uk
REFERENCES
- Copyright © 2003 by the American Academy of Pediatrics
In this issue
Jump to section
Related Articles
Cited By...
- Diffuse excessive high signal intensity on term equivalent MRI does not predict disability: a systematic review and meta-analysis
- The Integrated Stress Response Is Not a Target for Diffuse White Matter Injury in Premature Infants
- Apparent Diffusion Coefficient Value Changes and Clinical Correlation in 90 Cases of Cytomegalovirus-Infected Fetuses with Unremarkable Fetal MRI Results
- Brain Development in Fetuses of Mothers with Diabetes: A Case-Control MR Imaging Study
- Microstructure of the Default Mode Network in Preterm Infants
- Optimal Timing of Cerebral MRI in Preterm Infants to Predict Long-Term Neurodevelopmental Outcome: A Systematic Review
- High Signal Intensity on T2-Weighted MR Imaging at Term-Equivalent Age in Preterm Infants Does Not Predict 2-Year Neurodevelopmental Outcomes
- Cellular Changes Underlying Hyperoxia-Induced Delay of White Matter Development
- Periventricular Leukomalacia
- Early versus late MRI in asphyxiated newborns treated with hypothermia
- The Anatomic Variations of the Circle of Willis in Preterm-at-Term and Term-Born Infants: An MR Angiography Study at 3T
- The Association of Lung Disease With Cerebral White Matter Abnormalities in Preterm Infants
- Abnormal White Matter Signal on MR Imaging Is Related to Abnormal Tissue Microstructure
- Predicting Motor Development in Very Preterm Infants at 12 Months' Corrected Age: The Role of Qualitative Magnetic Resonance Imaging and General Movements Assessments
- Functional Magnetic Resonance Imaging of the Sensorimotor System in Preterm Infants
- Characterization of White Matter Injury in a Hypoxic-Ischemic Neonatal Rat Model by Diffusion Tensor MRI
- Cortical vision, MRI and developmental outcome in preterm infants
- Quality of General Movements Is Related to White Matter Pathology in Very Preterm Infants
- Prematurely Born Children Demonstrate White Matter Microstructural Differences at 12 Years of Age, Relative to Term Control Subjects: An Investigation of Group and Gender Effects
- Relationship Between White Matter Apparent Diffusion Coefficients in Preterm Infants at Term-Equivalent Age and Developmental Outcome at 2 Years
- Quantification of Deep Gray Matter in Preterm Infants at Term-Equivalent Age Using Manual Volumetry of 3-Tesla Magnetic Resonance Images
- Natural History of Brain Lesions in Extremely Preterm Infants Studied With Serial Magnetic Resonance Imaging From Birth and Neurodevelopmental Assessment
- Pathophysiology of Perinatal Hypoxia-Ischemia and the Prospects for Repair from Endogenous and Exogenous Stem Cells
- Structural, Immunocytochemical, and MR Imaging Properties of Periventricular Crossroads of Growing Cortical Pathways in Preterm Infants
- Neonatal cranial ultrasound versus MRI and neurodevelopmental outcome at school age in children born preterm
- Encephalopathy of Prematurity Includes Neuronal Abnormalities
- Abnormal Cerebral Structure Is Present at Term in Premature Infants
- Cerebral White Matter Injury of the Premature Infant--More Common Than You Think
More in this TOC Section
Similar Articles
Subjects
Keywords
- magnetic resonance imaging
- diffusion weighted imaging
- DEHSI
- preterm
- brain
- PVL, periventricular leukomalacia
- PHI, periventricular hemorrhagic infarction
- IVH, intraventricular hemorrhage
- MRI, magnetic resonance imaging
- DEHSI, diffuse excessive high signal intensity
- DWI, diffusion-weighted imaging
- ADC, apparent diffusion coefficient
- GA, gestational age
- PMA, postmenstrual age
- CSE, conventional spin echo
- FSE, fast spin echo
- ROI, region of interest
- DTI, diffusion tensor imaging