Abstract
Objective
Head ultrasonography (HUS) is a reliable and easy to perform bedside imaging technique that can give valuable information about degree of brain injury/edema after perinatal asphyxia in term neonates. The goals of our study were to determine whether semiquantitative markers such as standardized white matter/gray matter (WM/GM) echogenicity ratio and resistive index (RI) value measured by HUS differs between asphyxiated term neonates and healthy controls.
Study Design
Thirty-one carefully selected term neonates who suffered from perinatal hypoxic-ischemic encephalopathy (HIE) were included in the study. The ratio of the WM/GM echogenicity of the cingulate gyrus was calculated. In addition, the RI value was measured in the anterior cerebral artery. US scalars were compared with 11 healthy neonates.
Result
WM/GM ratio is significantly increased and RI value significantly decreased in asphyxiated term neonates compared with healthy subjects.
Conclusion
WM/GM ratio and RI value allows discriminating between asphyxiated neonates and healthy subjects. These US scalars may serve as valuable, easy to acquire semiquantitative bedside markers of brain HIE, when magnetic resonance imaging is unavailable or cannot be performed in the acute setting.
Keywords: white–gray matter echogenicity ratio, resistive index, head ultrasonography, hypoxic–ischemic encephalopathy, hypothermia
Introduction
Among term infants, hypoxic–ischemic encephalopathy (HIE) due to acute perinatal asphyxia is an important cause of neurological morbidity and mortality.1 With advances in diagnosis and treatment options, the incidence of HIE has significantly decreased, with a current estimated incidence of 2 to 5 per 1000 live births.1,2 The results of the Cool-Cap3 and the National Institute of Child Health and Human Development (NICHD) whole-body cooling4 trials demonstrated the beneficial effects of hypothermia in reducing death and adverse neurodevelopmental deficits in HIE.5,6 Currently, therapeutic hypothermia represents the only intervention, which has been demonstrated to favorably alter the outcome of term infants who suffered from perinatal asphyxia.5
An early and reliable diagnostic test is essential to guide and monitor treatment. Head ultrasonography (HUS) and magnetic resonance imaging (MRI) are frequently used in the acute/subacute phase of HIE. MRI has proven to be the most reliable imaging modality to evaluate the extent, quality and degree of brain injury.7–9 However, MRI examinations of the acutely asphyxiated neonate may be difficult to perform in the immediate postnatal setting. However, a rapid evaluation of the neonatal brain is critical because the decision to cool the neonate has to be made as early as possible, preferably within hours of birth. In addition, transportation of the neonate to the MRI suite may be critical, image quality may be degraded by motion because of the long acquisition times, and vital access to the critically sick neonate is limited during the MRI examination and finally maintaining therapeutic cooling of the neonate within the MRI is limited. Part of the MRI limitations may be solved by having an MRI unit within the neonatal intensive care unit; however, only very few hospitals have this setup available. There is consequently a need for an alternative, fast, reliable and easy-to-perform bedside imaging technique that can give valuable, information about degree and quality of brain injury (brain edema versus focal infarction or hemorrhage). HUS fulfills most of these criteria if performed on high-end US scanners using meticulous examination techniques.10,11 HUS can be performed at bedside, is readily available, does not require neonatal sedation, is free of ionizing radiation, can be repeated as often as necessary and gives both anatomical (gray scale imaging) and functional (color Doppler sonography) information. Moreover, HUS may be performed in the immediate period after birth while the neonate is being stabilized and/or prepared for cooling in the neonatal intensive care unit.
Qualitative sonographic assessment of the neonatal brain has proven to be reliable for diagnosing focal intracerebral lesions such as germinal matrix or intraventricular hemorrhage, infarction and hydrocephalus. Diffuse processes that may result from perinatal asphyxia such as WM edema are however more difficult to recognize and appears to be more dependent on the skills and experience of the radiologist. In HIE, the echogenicity of the cerebral WM increases relatively in relation to the hypoechogenic cortical gray matter (GM).8–10 The WM/GM differentiation increases accordingly (Figures 1a and b). Depending on the used US technique and acquisition settings, US transducer, size of the fontanelle and the selected window/level settings of the US images, the degree of WM edema may be under- or overestimated if the US images are reviewed only `visually'. Even experienced pediatric neuroradiologists may misinterpret the degree of WM echogenicity. Previous studies12,13 proposed various more `objective' techniques to quantify neonatal WM lesions. Simaeys et al.12 used a software-based method to compensate for the variable US acquisition factors by comparing the echogenicity of intracerebral lesions with that of the choroid plexus whose echogenicity is considered to be relatively constant. Padilla et al.14 showed that calculation of the relative echogenicity (ratio of the mean pixel brightness measured in brain regions and the bone at the same depth of sampling) may offer a semiquantitative method to evaluate various anatomical regions of the neonatal brain on US.
Figure 1.
(a, b) Coronal and sagittal images of a healthy control (a) and a neonate with hypoxic–ischemic encephalopathy (HIE) (b). In the control subject the echogenicity of the white matter (WM) appears nearly isoechogenic to the cortical gray matter (GM), whereas in the asphyxiated neonate the WM is significantly more echogenic compared with the overlying cortical GM. This difference in echogenicity is especially well seen in the cingulate gyrus and increases the WM/GM echogenicity ratio.
In addition, the resistive index (RI), which is calculated as the ratio between the end diastolic and peak systolic velocity, measured within the anterior or middle cerebral artery by duplex sonography, may give objective functional data about brain hemodynamics/perfusion. Previous studies have shown that the RI value may be lowered in asphyxiated neonates.15 In addition, Archer et al.17 showed that a RI value below 0.55 (measured in the anterior cerebral artery) correlates with a poor neurodevelopment prognosis,16 especially if measured in the first 24 to 72 h after HIE.
The goals of our study were to compare two easy-to-collect semiquantitative US markers including a calculated WM/GM echogenicity ratio and simultaneously measured RI value in a cohort of asphyxiated term neonates versus a group of healthy controls.
Methods
Study design and clinical features
Forty consecutive term neonates with clinically confirmed HIE who had been selected for cooling treatment were included in our study (January 2008 to August 2009). Patients were selected from our `hypothermia' database. Patients in this database had been carefully selected for cooling treatment based on their clinical history and metabolic findings by an expert team of neurologists/neonatologists. Apart from the clinical evidence of encephalopathy, at least two of the following findings had to be present: (1) APGAR score of ⩽5 at 10 min, (2) need of mechanical ventilation at 10 min and (3) metabolic acidosis (umbilical cord pH<7.00).
A second group of 11 age-matched control subjects were selected from our radiological database for comparison. All patients in this control group had normal HUS studies and had a normal initial and follow-up neurological examination.
All patients were treated with a standard protocol including whole-body cooling to a rectal temperature of 33 to 34 °C for the duration of 72 h.
Institutional Review Board's approval was obtained for our study.
Imaging acquisition
All neonates were examined in the neonatal intensive care unit. No sedation was performed for purpose of the US studies. HUS was performed by experienced pediatric sonographers, using state-of-the-art US equipment (Zonare, Medical Systems, Mountain View, CA, USA). A standardized set of images were obtained through the anterior fontanel in coronal and sagittal planes using curved and linear array transducers (8 to 17 MHz) (Figures 1a and b). The angle of the transducer was varied when needed to evaluate the periphery of the brain. Penetration depth, number of US foci, US gain and tissue brightness were optimized for each individual examination.
Color-coded duplex technique was used to evaluate the intracranial arteries and veins. Spectral Doppler tracings were obtained from the A1 or A2 segment of the anterior cerebral artery in a sagittal plane. The Doppler angle was corrected for the course of the vessel. The RI was recorded (RI = peak systolic velocity—end diastolic velocity/peak systolic velocity). In all neonates, the patency of the superior sagittal sinus was also studied per standard protocol. Both, patients and controls were studied with exactly the same technique. Images were transferred to the patient archive and communications system for offline analysis. All neonates who suffered from perinatal asphyxia were studied by US within 48 h of birth. The controls were studied within the first 6 days of life.
Imaging analysis
Of the 40 patients, nine neonates were excluded because HUS was performed later than 48 h after birth. All images were evaluated for diagnostic quality, performed technique and completeness of data. The WM/GM echogenicity was measured using a region of interest manually positioned in the WM/GM of the cingulate gyrus on a paramedian sagittal image (Figure 2). Care was given to prevent partial volume effects. The region of interest within the WM was typically ovoid shaped, whereas the region of interest in the GM was specifically adapted to the particular anatomy of the cortex. All measurements were repeated twice and the measurement with the smallest standard deviation was selected for further statistical analysis. The pediatric neuroradiologist outlining the region of interest was blinded for the clinical findings on both occasions.
Figure 2.
Sample `regions of interest' positioned within the cortical gray matter (GM) (left image) and white matter (WM) (right image) of the cingulate gyrus in head ultrasound sagittal images of a hypoxic–ischemic encephalopathy (HIE) patient. White matter echogenicity average (61.9587) is significantly higher than in the GM (36.6328), increasing the WM/GM ratio.
The measured values of echogenicity (mean pixel brightness) of WM/GM including the standard deviation and size of the cross-sectional areas were transferred into our database for further analysis. WM/GM ratio were consequently calculated and recorded.
The measured RI values were extracted from the available imaging data sets and transferred into the study data set.
The clinical charts of the controls were studied to exclude systemic or focal diseases affecting the central nervous system.
Statistical analysis
A two-tailed, unpaired t-test was used to check for differences in the WM/GM ratio as well as RI values between patients and controls. All statistical analysis was carried out in Prism and STATA v11 (STATA Corporation, College Station, TX, USA, 2009) software.
A P-value of less than 0.05 was considered statistically significant.
Results
HUS images of 31 patients (15 boys; 16 girls) older than 36 weeks of gestation with clinical diagnosis of HIE were included in the study and proved to be of high diagnostic quality. The main demographic and clinical characteristics of the patient population are shown in Table 1. The mean age when HUS was performed in the asphyxiated neonates was 1.70 days (range: 1 to 2 days).
Table 1.
Clinical and demographical characteristics of the patients
| Gestational age (n = 31) (in years) | 39+2 (36+0–42+0) |
| Umbilical cord, pH (n = 24) | 6.93 (6.70–7.25) |
| APGAE 1′ (n = 24) | 2.27 (0–8) |
| APGAE 5′ (n = 24) | 4.1 (0–9) |
| APGAE 10′ (n = 24) | 5.25 (2–9) |
| LOHS (n = 24) | 17.24 (5–54) |
Abbreviation: LOHS, length of hospital stay.
Eleven children were chosen as controls. The reasons for HUS were the following: large head circumference (n = 2), cardiac abnormalities (n = 2), congenital adenomatoid lung malformation, transient neonatal melanosis, maternal HIV and cytomegalovirus infection during pregnancy, bulging fontanel, subgaleal hematoma, rhesus iso-immunization and mild dysmorphic features in one patient each. The mean age when HUS was performed in the controls was 2.63 days. Ultrasound results were negative in all control subjects; neurological follow-up examinations excluded neurological disease in all control subjects.
Within the HIE group, WM/GM ratio was measured/calculated in all patients. RI values were available for 16 neonates. In the control group, WM/GM ratios and RI values were available in all neonates.
Mean WM/GM ratio for the HIE patients was 1.436 (s.d. = 0.222), and for the controls 1.234 (s.d. = 0.159) (Figure 3). The difference was statistically significant (P = 0.008).
Figure 3.
White matter/gray matter (WM/GM) ratio is significantly higher in hypoxic–ischemic neonates when compared with normal controls (P = 0.008).
Mean RI value for the HIE patients was 0.612 (s.d. = 0.061), while for the controls, it was 0.683 (s.d. = 0.060) (Figure 4). This difference was statistically significant (P = 0.006).
Figure 4.
Resistive index (RI) values are significantly lower in hypoxic–ischemic neonates when compared with normal controls (P = 0.006).
Discussion
Our study showed that the semiquantitative WM/GM ratio as measured by HUS is statistically significantly increased in neonates who suffered from perinatal asphyxia compared with healthy controls. The relative increase in echogenicity of the cerebral WM in relation to the cortical GM is believed to result from a combination of HIE-related acute global edema and possibly cerebral tissue necrosis.10 This semiquantitative US scalar may allow estimating the degree of tissue injury/edema, independent of the US settings, in a more objective approach. In addition, the WM/GM ratio may possibly be used to monitor treatment and predict outcome and need for adjuvant therapies in children who suffered from perinatal asphyxia. Future prospective studies, evaluating the temporal evolution of the WM/GM ratio during treatment and correlation with other biomarkers of brain injury (imaging, neurophysiology and so on) and neurodevelopmental follow-up data are however mandatory to better determine the significance of the WM/GM ratio.
In addition, our analysis showed that the RI values are statistically significantly decreased in neonates who suffered from HIE compared with healthy subjects. These results match previous studies, which showed that the RI values decrease in asphyxiated neonates.16,18,19 The lowered RI values result from higher end diastolic flow velocities in relation to the peak systolic velocities measured in major branches of the circle of Willis. The exact pathophysiology behind the changes in cerebral hemodynamics is a focus of ongoing research. A loss of the normal autoregulation of brain perfusion due to the HIE is debated.20
WM/GM ratio and RI values should be correlated with high-end functional MRI scalars such as concentration of various neurometabolites as measured by 1H-MR spectroscopy (1H MRS), apparent diffusion coefficient, fractional anisotropy values as measured by diffusion tensor imaging and cerebral blood volume, cerebral blood flow, mean transit time as measured by perfusion-weighted imaging. In addition, new functional sequences such as pH-weighted MRI can give imaging data/scalars that may better correlate with the measured US scalars. Prospective comparative functional MRI studies may further validate the US findings.
We appreciate several limitations in our study. The study is retrospective in design and evaluates a relatively small number of neonates. The severity of perinatal depression and asphyxia was not prospectively evaluated by a quantitative clinical scoring system. The patient population was however a well controlled and homogeneous group of neonates because of the fact that they had to fulfill strict criteria to be included in the hypothermia treatment/protocol. Other limitations include the lack of RI values or clinical markers in some HIE neonates, the lack of information about long-term outcome and the unknown early influence of the hypothermia treatment on the US scalars. Hypothermia is recognized to limit degree of HIE in infants with moderate degrees of perinatal asphyxia, but its influence on brain edema is still lacking in the literature. Regarding the imaging acquisition and analysis, there are also some limitations. First, the WM/GM ratio was measured on only one time point and in only one brain region (cingulate gyrus). Our objective was however to get an easy-to-apply, reproducible and reliable bedside semiquantitative scalar for evaluating brain injury/edema. We have chosen the cingulate gyrus because this region is close to the US transducer, which guarantees a good image quality, and IT is easy to locate on both imaging planes. Certainly, additional brain regions may also be affected by perinatal asphyxia, such as basal ganglia and thalami. However, as the goal of our pilot study was to evaluate a semiquantitative scalar of WM injury/edema, we focused on a region close to the US transducer that incorporated GM and WM simultaneously. This semiquantitative WM/GM ratio appears especially important, because most experienced pediatric neuroradiologists use the WM echogenicity as one of the markers of HIE-related brain injury. Future studies should however evaluate multiple GM and WM regions. This is particularly important given the well-recognized selective vulnerability of the various brain regions for HIE. Similarly, the RI values were also measured on one time point after perinatal asphyxia and in only one intracranial artery (anterior cerebral artery). It is well known that a loss of autoregulation may result in an increased variability of the RI values and that the various intracranial arteries may show different hemodynamic responses to hypoxia. Future studies should evaluate the temporal evolution and variability of the RI values. In addition, the RI values should be sampled for multiple intracranial arteries. The variability of the RI value during the time course of a study and in between studies may harbor additional valuable information. Finally, MRI of the brain was not available as gold standard in the acute phase because the decision to cool the neonates was made based upon the clinical findings and history. Collecting MRI data immediately after birth would unnecessarily delay the start of cooling treatment. In addition, follow-up MRI could not be used as a late gold standard because the impact of the cooling treatment on the MRI findings is not yet completely understood. Currently, it seems impossible to simultaneously collect acute, pre-cooling US and MRI data, unless an MRI unit is available within the neonatal intensive care unit that allows ongoing cooling. Evaluating neonates who do not require cooling may allow early, time-matched US and MRI studies, but this would inevitably result in a selection bias, because this would focus on neonates with less severe perinatal asphyxia.
Conclusion
In conclusion, the measured US scalars (WM/GM ratio and RI values) are significantly different in brains of neonates who suffered from perinatal HIE compared with healthy controls. These US markers, especially the WM/GM ratio may serve as an easy to acquire, bedside marker for brain edema/injury and could help to semiquantitatively evaluate brain edema/injury, and possibly help to monitor treatment and predict neurofunctional outcome or help select patients in need of additional adjuvant therapies. Future prospective studies with larger patient groups that evaluate the temporal evolution of the WM/GM ratio at multiple early and late time points and correlation with sensitive clinical markers and outcome scores are mandatory.
Footnotes
Conflict of interest The authors declare no conflict of interest.
References
- 1.Wu YW, Backstrand KH, Zhao S, Fullerton HJ, Johnston SC. Declining diagnosis of birth asphyxia in california: 1991–2000. Pediatrics. 2004;114(6):1584–1590. doi: 10.1542/peds.2004-0708. [DOI] [PubMed] [Google Scholar]
- 2.Smith J, Wells L, Dodd K. The continuing fall in incidence of hypoxic-ischemic encephalopathy in term infants. BJOG. 2000;107(4):461–466. doi: 10.1111/j.1471-0528.2000.tb13262.x. [DOI] [PubMed] [Google Scholar]
- 3.Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards AD, Ferriero DM, et al. Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. Lancet. 2005;365(9460):663–670. doi: 10.1016/S0140-6736(05)17946-X. [DOI] [PubMed] [Google Scholar]
- 4.Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDonald SA, Donovan EF, et al. Whole-body hypothermia for neonates with hypoxic-ischemic encephalopathy. N Engl J Med. 2005;353(15):1574–1584. doi: 10.1056/NEJMcps050929. [DOI] [PubMed] [Google Scholar]
- 5.Laptook AR. Use of therapeutic hypothermia for term infants with hypoxic-ischemic encephalopathy. Pediatr Clin North Am. 2009;56(3):601–661. doi: 10.1016/j.pcl.2009.03.007. [DOI] [PubMed] [Google Scholar]
- 6.Shah PS. Hypothermia: a systematic review and meta-analysis of clinical trials. Semin Fetal Neonatal Med. 2010;15(5):238–246. doi: 10.1016/j.siny.2010.02.003. [DOI] [PubMed] [Google Scholar]
- 7.Blankenberg FG, Loh NN, Bracci P, D'Arceuil HE, Rhine WD, Norbash AM, et al. Sonography, CT, and MR imaging: a prospective comparison of neonates with suspected intracranial ischemia and hemorrhage. AJNR Am J Neuroradiol. 2000;21(1):213–218. [PMC free article] [PubMed] [Google Scholar]
- 8.Childs AM, Cornette L, Ramenghi LA, Tanner SF, Arthur RJ, Martinez D, et al. Magnetic resonance and cranial ultrasound characteristics of periventricular white matter abnormalities in newborn infants. Clin Radiol. 2001;56(8):647–655. doi: 10.1053/crad.2001.0754. [DOI] [PubMed] [Google Scholar]
- 9.Sie LT, van der Knaap MS, van Wezel-Meijler G, Taets van Amerongen AH, Lafeber HN, Valk J. Early mr features of hypoxic-ischemic brain injury in neonates with periventricular densities on sonograms. AJNR Am J Neuroradiol. 2000;21(5):852–861. [PMC free article] [PubMed] [Google Scholar]
- 10.Daneman A, Epelman M, Blaser S, Jarrin JR. Imaging of the brain in full-term neonates: does sonography still play a role? Pediatr Radiol. 2006;36(7):636–646. doi: 10.1007/s00247-006-0201-7. [DOI] [PubMed] [Google Scholar]
- 11.Epelman M, Daneman A, Kellenberger CJ, Aziz A, Konen O, Moineddin R, et al. Neonatal encephalopathy: a prospective comparison of head US and MRI. Pediatr Radiol. 2010;40(10):1640–1650. doi: 10.1007/s00247-010-1634-6. [DOI] [PubMed] [Google Scholar]
- 12.Kuban K, Adler I, Allred EN, Batton D, Bezinque S, Betz BW, et al. Observer variability assessing us scans of the preterm brain: The Elgan study. Pediatr Radiol. 2007;37(12):1201–1208. doi: 10.1007/s00247-007-0605-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Simaeys B, Philips W, Lemahieu I, Govaert P. Quantitative analysis of the neonatal brain by ultrasound. Comput Med Imaging Graph. 2000;24(1):11–18. doi: 10.1016/s0895-6111(99)00041-5. [DOI] [PubMed] [Google Scholar]
- 14.Padilla NF, Enriquez G, Jansson T, Gratacos E, Hernandez-Andrade E. Quantitative tissue echogenicity of the neonatal brain assessed by ultrasound imaging. Ultrasound Med Biol. 2009;35(9):1421–1426. doi: 10.1016/j.ultrasmedbio.2009.04.014. [DOI] [PubMed] [Google Scholar]
- 15.Chao CP, Zaleski CG, Patton AC. Neonatal hypoxic-ischemic encephalopathy: multimodality imaging findings. Radiographics. 2006;26(Suppl 1):S159–S172. doi: 10.1148/rg.26si065504. [DOI] [PubMed] [Google Scholar]
- 16.Archer LN, Levene MI, Evans DH. Cerebral artery doppler ultrasonography for prediction of outcome after perinatal asphyxia. Lancet. 1986;2(8516):1116–1118. doi: 10.1016/s0140-6736(86)90528-3. [DOI] [PubMed] [Google Scholar]
- 17.Ilves P, Lintrop M, Talvik I, Muug K, Maipuu L, Metsvaht T. Low cerebral blood flow velocity and head circumference in infants with severe hypoxic ischemic encephalopathy and poor outcome. Acta Paediatr. 2009;98(3):459–465. doi: 10.1111/j.1651-2227.2008.01119.x. [DOI] [PubMed] [Google Scholar]
- 18.Levene MI, Fenton AC, Evans DH, Archer LN, Shortland DB, Gibson NA. Severe birth asphyxia and abnormal cerebral blood-flow velocity. Dev Med Child Neurol. 1989;31(4):427–434. doi: 10.1111/j.1469-8749.1989.tb04020.x. [DOI] [PubMed] [Google Scholar]
- 19.Stark JE, Seibert JJ. Cerebral artery doppler ultrasonography for prediction of outcome after perinatal asphyxia. J Ultrasound Med. 1994;13(8):595–600. doi: 10.7863/jum.1994.13.8.595. [DOI] [PubMed] [Google Scholar]
- 20.Brady KM, Lee JK, Kibler KK, Easley RB, Koehler RC, Shaffner DH. Continuous measurement of autoregulation by spontaneous fluctuations in cerebral perfusion pressure: comparison of 3 methods. Stroke. 2008;39(9):2531–2537. doi: 10.1161/STROKEAHA.108.514877. [DOI] [PMC free article] [PubMed] [Google Scholar]