- Email: [email protected]
Magnetic resonance imaging in neonates with total asphyxia
Brain & Development 35 (2013) 53–60 www.elsevier.com/locate/braindev
Original article
Magnetic resonance imaging in neonates with total asphyxia Hiroshi Sugiura a, Masanori Kouwaki b,⇑, Tohru Kato c, Tsutomu Ogata d, Rie Sakamoto e, Atsushi Ieshima f, Kenji Yokochi g a
Department of Neonatology, Seirei-Hamamatsu General Hospital, Hamamatsu, Shizuoka, Japan b Department of Neonatology, Toyohashi Municipal Hospital, Toyohashi, Aichi, Japan c Department of Pediatrics, Okazaki City Hospital, Okazaki, Aichi, Japan d Department of Pediatrics, Nagasaki Medical Center, Ohmura, Nagasaki, Japan e Department of Perinatogy, Funahashi Central Hospital, Funahashi, Chiba, Japan f Department of Pediatrics and Neuropediatrics, Ibaraki Welfare and Medical Center for Disabled Children, Mito, Ibaraki, Japan g Department of Pediatric Neurology, Seirei-Mikatahara General Hospital, Hamamatsu, Shizuoka, Japan Received 11 December 2011; received in revised form 13 April 2012; accepted 18 April 2012
Abstract Magnetic resonance (MR) findings in cases of total asphyxia, whose lesions are mainly in the brainstem and deep nuclei, have not been clarified. In this study, we investigated MR images in neonates with total asphyxia. MR images of six infants (three males and three females; gestational age, 35–39 weeks; birth weights, 1880–3572 g) with total asphyxia were examined. In all subjects, neonatal cortical MR lesions were limited to the hippocampus with highlighting on T1-weighted imaging (T1-WI). The neonatal MR lesions of the cerebral white matter were limited to the white matter between the insula and putamen in four infants, and were diffusely involved in two infants. The ventral lateral nucleus of the thalamus was hyperintense on T1-WI in all of the subjects. Other nuclei in the thalamus, the globus pallidus and the putamen were involved in neonatal MR images of all subjects. High intensity areas on T2- weighted imaging were observed at the dorsal areas in the midbrain, pons and medulla oblongata in all or most of the subjects at the neonatal period. Also, high intensity areas on T1-WI were observed in the tegmentum of the pons and the midbrain in five cases. Neonates with total asphyxia had lesions mainly in the tegmentem of the brainstem, thalamus, putamen and globus palludus. Some of the infants had extensive lesions of the white matter. Ó 2012 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. Keywords: Total asphyxia; Hypoxic ischemic encephalopathy; Tegmentum; Term infant
1. Introduction Myers [1,2] examined perinatal brain damage after total and partial asphyxia in fullterm fetal monkeys. Total asphyxia led to specific patterns of injury in the brainstem of the monkeys. Children with brainstem lesions resulting from perinatal insults at term have been reported to show supratentorial lesions [3–13]. Therefore, ⇑ Corresponding author. Address: Toyohashi Municipal Hospital, 50 Aza Hachiken Nishi, Aotake–Cho, Toyohashi, Aichi 441-8570, Japan. Tel.: +81-532-33-6111; fax: +81-532-33-6177. E-mail address: [email protected] (M. Kouwaki).
it is unclear whether total asphyxia results in isolated brain lesions in humans. Furthermore, insults during the fetal period have been suggested to result in brainstem dysgenesis [14,15]. Partial asphyxia has been associated with parasagittal cerebral injury [16]. A mixture of total and partial asphyxia has been suggested to cause selective neuronal necrosis. Volpe [16] classified this neuronal necrosis into three patterns: diffuse, cerebral cortex-deep nuclear and deep nuclear-brainstem patterns resulting from very severe and prolonged insults, moderate to severe and prolonged insults, and severe and abrupt insults,
0387-7604/$ - see front matter Ó 2012 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.braindev.2012.04.002
54
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
respectively. In the deep nuclear-brainstem pattern, clinical correlates include ptosis, oculomotor disturbances, facial diparesis, ventilatory disturbances, and impaired sucking and swallowing [16]. We prefer the term “total asphyxia” for this pattern, because it reveals a causative mechanism [9]. Although magnetic resonance (MR) findings of brainstem lesions in fullterm infants with hypoxic ischemic encephalopathy (HIE) have been reported [13,17,18], MR findings characterizing total asphyxia have not been clarified. In this study, we investigated MR images in neonates with total asphyxia. 2. Subjects and methods Six infants (three males and three females, born in 2006–2009) were studied. They were selected from patients attending the Seirei-Mikatahara General Hospital for rehabilitation, and from those reported in the Japanese conference of fetal and neonatal neurology or the Zao mailing conference. Inclusion criteria of total asphyxia were as follows. First, acute, brief and severe insults that occurred within 2 weeks before delivery, during delivery or within 24 h after delivery were assumed to result in brain damage. This assumption
Table 1 Clinical profiles of the subjects and methods of MR scanning. Case No. (sex) Perinatal events (gestational age, birth weight, Apgar score (1 min/5 min))
was justified by a loss of fetal heart-rate variability, Apgar score, neurological assessment or laboratory findings such as acidosis. Second, the characteristic neurological abnormalities mentioned below persisted for >1 month after birth. Spontaneous body movement was scant, and was limited to bending backward within a small range. Muscles of facial expression and the jaw were almost immobile. Swallowing of saliva was not observed, and tube-feeding was required. The eyes and eye-lids moved only within a small range. Third, MR imaging was performed within 4 weeks after birth, and a lesion in the brainstem was revealed by MR imaging. However, no lesion forming cyst could be present in the brain, except for a periventicular cyst persisting from prenatal periods. Finally, patients diagnosed with metabolic disease or chromosomal abnormality were excluded. Clinical profiles of the subjects are shown in Table 1. In three infants (Cases 1, 2, 3), an acute insult was assumed to have occurred during delivery. In two infants (Cases 4, 5) and one (Case 6), the insult was assumed to occur before delivery and after delivery, respectively. Cases 2, 4 and 5 underwent MR scanning three times after birth, and Cases 1, 3 and 6 underwent MR scanning 1, 1 and 2 times, respectively (Table 1).
Outcome (age at last examination)
MR scanning (age at performing, machine, slice thickness,* SE(TR/TE ms)*)
1 (F)
Hysterorrhexis Cesarean section 39 weeks 3 days, 2586 g, 0/2
2 years 3 months Almost immobile Tube-feeding Laryngo-tracheal diversion
21 days, 2 years 3 months 1.5T GE, 4 mm, T1: 600/9, T2: 3500/89
2 (M)
Fetal distress Cesarean section 40 weeks 5 days, 3572 g, 1/4
5 years 0 months Almost immobile Tube-feeding
8, 16 and 37 days 1.5T GE, 6 mm, T1: 500/13–14, T2: 3000/98–107
3 (F)
Hysterorrhexis Cesarean section 39 weeks 2 days, 3096 g, 1/4
2 years 1 months Almost immobile Artificial ventilation Tube-feeding
20 days 1.5T Siemens, 4 mm, T1 : 650/15, T2 : 4500/120
4 (F)
Disappearance of fetal movements at 2 weeks before delivery Fetal distress Cesarean section 35 weeks 4 days, 1880 g, 3/5
1 years 10 months
9 days, 44 days, 21 weeks
Almost immobile Artificial ventilation Tube-feeding
1.5T GE, 5 mm, T1: 480/14, T2: 2860/81
2 years4 months
8 days, 41 days, 20 weeks
Almost immobile Artificial ventilation Tube-feeding
1.5T Philips, 5 mm, T1: 450/11, T2: 2721/90
1 years 11 months Almost immobile Tube-feeding
21 days, 1 year 11 months 1.5T GE, 4 mm, T1: 600/9, T2: 3500/90
5 (M)
6 (M)
*
Disappearance of fetal movements at 3 days before delivery Fetal distress Cesarean section 36 weeks 0 days, 2356 g, 2/3 39 weeks 3 days, 2550 g, 8/9 Cardiopulmonary arrest at 16 h after birth
These values are shown when MR scanning is performed within 28 days after birth.
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Axial T1-weighted and T2-weighted images were analyzed, and their spine-echo sequences are shown in Table 1. 3. Results The representative MR images are shown in Figs. 1–6. In all subjects, neonatal cortical MR lesions were limited to the hippocampus with highlighting on T1weighted imaging (T1-WI). As for neonatal MR lesions of the cerebral white matter, lesions were limited to the white matter between the insula and putamen in four (Cases 1, 2, 3, 6) of six infants, and the periventricular areas or watershed areas of the cerebral arteries were not involved. In two infants (Cases 4, 5), the white matter was diffusely involved. This lesion was connected to
55
impairment of the myelination process and atrophy of the white matter in early infancy. Myelination in the posterior limb of the internal capsule was not observed on T1-WI at the neonatal periods. In Cases 4 and 5, high intensity areas on T1-WI in the posterior limb of the internal capsule was not observed at 21 and 20 weeks old, respectively. Although low intensity areas on T2WI, which indicated myelination, were observed in the posterior limb of the internal capsule in two infants (Cases 1, 6) undergoing MR scanning after 23 months old, they were not observed in the anterior limb of it. In Case 1, it was accompanied by the frontal white matter lesions. For thalamic MR lesions, areas corresponding to the ventral lateral nucleus [19] were hyperintense on T1-WI at the neonatal periods in all subjects. Also, areas at the
Fig. 1. Case 1. MR images at 21 days old and 2 years and 4 months old. T2-weighted and T1-weighted images estimated as normal in the brainstem during the neonatal period at the same hospital are shown in the middle row for comparison. At the dorsal part in the lower medulla oblongata, curved belt-shaped areas with high intensity on T2-WI are seen (an up white arrowhead). At the dorsal paracentral part in the upper medulla oblongata, high intensity areas on T2-WI are observed bilaterally (an up arrowhead). At the middle of the tegmentum in the pons, high intensity areas on T2-WI are observed bilaterally (up arrows). At the middle areas in the tegmentum of the midbrain, round high intensity areas on T2-WI are observed bilaterally (down arrows). The tegmentum of the pons and the midbrain is faintly hyperintense on T1-WI (right arrowheads). The hippocampus is bilaterally hyperintense on T1-WI (a white arrow); the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2-WI and hypointense on T1-WI (left white arrowheads). The lateral part of the thalamus, where the ventral lateral nucleus is located [19], is bilaterally hyperintense on T1-WI and on T2-WI (left arrows). The central part of the thalamus is bilaterally hyperintense on T2-WI and faintly hyperintense on T1-WI (up long arrows). The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI (curved arrows). On T2-WI, its central part is bilaterally hypointense and its peripheral part is hyperintense. The posterior and lateral parts of the putamen are bilaterally hyperintense on T1-WI and isointense on T2-WI (down long arrows); its surrounding areas are bilaterally hyperintense on T2-WI and isointense on T1-WI. No high intensity areas on T1-WI, which indicates myelination, are observed in the posterior limb of the internal capsule. At 2 years and 4 months old, the dorsal part in the pons is hyperintense on T2-WI (up arrowheads). Almost the entire midbrain is hyperintense on T2-WI (left arrowheads). The third and lateral ventricle dilated; the inferior horn (*) is markedly dilated. The cortical sulci are dilated predominantly at the frontal areas. The cerebral white matter with high intensity areas on T2-WI is atrophic predominantly at the frontal areas. The thalamic areas at the ventral lateral nucleus, ventral posterolateral nucleus and other nuclei located inferiorly are hyperintense on T2-WI. The inferior and posterior parts of the globus pallidus and the posterior part of the putamen are hyperintense on T2-WI (left arrows). No high intensity areas on T2-WI, which indicates myelination, are observed in the anterior limb of the internal capsule.
56
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Fig. 2. Case 2. MR images at 8, 16 and 37 days old. At all three times examined, faintly high intensity areas on T2-WI are observed bilaterally at the dorsal part in the lower medulla oblongata (left arrowheads), and at the dorsal paracentral parts in the upper medulla oblongata (right arrowheads). At the tegmentum of the midbrain, high intensity areas on T2-WI are observed bilaterally at 8 and 16 days old (right arrows). The tegmentum of the pons and the dorsal part of the midbrain are hypertense on T1-WI at 37 days old (left arrows). At all times examined, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hypointense on T1-WI and slightly hyperintense on T2-WI at 16 and 37 days old (left white arrowheads). On T1WI, the thalamic lesions in the ventral lateral nucleus (an up white arrow) and those in the ventral posterolateral nucleus [19] (an up arrow) are bilaterally hyperintense at 37 days old. Then, the same areas seem hyperintense on T2-WI. At 16 days old, the areas with high intensity on T1-WI are larger and the intensity is less intense than at 37 days old. Then, almost the entire thalamus is hyperintense on T2-WI with the peripheral isointense areas; the hyperintense areas on T1-WI cannot be discriminated from the T2-weighted images. At 8 days old, the hyperintense areas on T1-WI are least clearly exhibited; the lateral and posterior borders of the ventral lateral nucleus are most hyperintense on T1-WI (an up arrowhead). Then, the more extensive thalamic areas are hyperintense on T2-WI than at 16 days old. At all three times examined, hyperintense areas on T1-WI or on T2-WI are intermingled in the globus palllidus (down arrows) and in the putamen. No high intensity areas on T1-WI are observed in posterior limb of the internal capsule.
Fig. 3. Case 3. MR images at 20 days old. At the dorsal part in the lower medulla oblongata, high intensity areas on T2-WI are observed (a left arrowhead). At the dorsal paracentral parts in the upper medulla oblongata, faintly high intensity areas on T2-WI are observed bilaterally (right arrowheads). At the middle of the tegmentum in the pons, faintly high intensity areas on T2-WI are seen bilaterally (up arrowheads). The tegmentum in the pons is faintly hyperintense on T1-WI (right white arrowheads). The dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). The hippocampus is bilaterally hyperintense on T1-WI (a white arrow); the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2-WI and hypointense on T1-WI (left white arrowheads). The ventral lateral nucleus of the thalamus is bilaterally hyperintense on T1-WI and hypointense on T2-WI (left arrows). Its surrounding areas are hyperintense on T2-WI and isointense or faintly hyperintense on T1-WI; the latter hyperintense areas on T1-WI are located in the ventral posterolateral nucleus [19] (an up arrow). The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI and hypointense on T2-WI (not shown in the Figure). The posterior and lateral parts of the putamen are bilaterally hyperintense on T1-WI and hypointense on T2-WI (a down arrow); its surrounding areas are bilaterally hyperintense on T2-WI and isointense on T1-WI. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
57
Fig. 4. Case 4. MR images at 9 days old, 44 days old and 21 weeks old. At 9 days old, faintly high intensity areas on T2-WI are observed bilaterally at the middle part in the lower medulla oblongata (a right arrowhead), and at the dorsal paracentral parts in the upper medulla oblongata (a left arrowhead). The dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). Then, the dorsal parts of the upper medulla oblongata, pons and midbrain are hypertense on T1-WI (up white arrowheads). At 44 days old, they are less intense than at 9 days old. At 9 days old, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. Then, the cerebral white matter is diffusely hyperintense on T2-WI and hypointense on T1-WI; the subependymal cyst is observed bilaterally (down arrowheads). The lateral ventricle and third ventricle become dilated as the age progress. At 21 weeks old, no myelination in the cerebral white matter is considered to emerge on T1-WI or on T2-WI, and the volume of the cerebral white matter is reduced. At 21 weeks old, the entire thalamus exhibits faintly high intensity on T1-WI, and no abnormal intensity on T2-WI; the thalamus becomes atrophied. At 9 days old, the peripheral areas of the thalamus are hyperintense on T1-WI, and its central areas are hyperintense on T2-WI. At 21 weeks old, the whole globus palllidus and putamen exhibits no abnormal intensity on T2-WI or T1-WI; the two become atrophied. At 9 days old, hyperintense areas on T1-WI or on T2-WI are intermingled in the globus palllidus and in the putamen. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
ventral posterolateral nucleus [19] were involved in 5 subjects (Cases 1, 2, 3, 5, 6). In 2 infants (Cases 1, 6), high intensity areas at the ventral lateral nucleus on T1-WI at the neonatal period were shown to develop to high intensity areas on T2-WI after 23 months old. In contrast, the central part of the thalamus was hyperintense on T2-WI in most infants. In all three infants undergoing successive MR scanning (Cases 2, 4, 5), this high intensity area was most intense at the first scanning, and it disappeared at the last scanning. In the globus pallidus, high intensity areas on T1-WI were observed in all subjects at the neonatal period. However, in three cases (Cases 1, 3, 6), these areas were mostly limited to the inferior and posterior parts, with accompanying low intensity areas on T2-WI. In the putamen, high intensity areas on T2-WI were mixed with those on T1-WI in all subjects at the neonatal period. High intensity areas on T2-WI were observed at the dorsal midbrain in all subjects at the neonatal period. While round and hyperintense areas at the upper midbrain were seen clearly at 21 days old in Case 1, almost the entire midbrain was hyperintense on T2-WI at 2 years 4 months old. In the pons, high intensity areas on T2-WI were observed in the middle of the tegmentum in three cases (Cases 1, 3, 6). Also, high intensity areas on T1-WI were observed in the entire tegmentum of
the pons and midbrain in five cases (Cases 1, 2, 3, 4, 5). In the medulla oblongata, faint high intensity areas on T2-WI were observed in five subjects (Cases 1, 2, 3, 4, 6). They were observed at the dorsal paracentral parts in the upper medulla oblongata and at the dorsal or central part in the lower medulla oblongata. Cerebellar MR lesions were not observed in the subjects. 4. Discussion Nervous system damage from total asphyxia in monkeys was reported to affect nuclei in the brainstem tegmentum including the central nucleus of the inferior colliculus, posterior and lateral ventral nuclei of the thalamus, and the intermediolateral gray column of the spinal cord [1,2]. MR imaging of brainstem nuclei in the human neonates after total asphyxia showed injuries similar to those demonstrated in monkeys after total asphyxia. In monkeys with the total asphyxia [1,2], the cerebral lesions were limited to the thalamus. In human subjects with total asphyxia, not only the thalamus, but also the basal ganglia were involved. In addition, the cerebral cortex and white matter were involved, despite being mild. This is a crucial difference between monkeys with total asphyxia and humans with total asphyxia. Thalamic lesions were predominant in the ventral lateral nucleus, and lesions of the putamen were posterior
58
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Fig. 5. Case 5. MR images at 8 days old, 41 days old and 20 weeks old. At 8 days old, the dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). Then, the dorsal parts of the upper medulla oblongata, pons and midbrain are hypertense on T1-WI (up white arrowheads). At 41 days old, they become less intense than at 8 days old. At 8 days old, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. Then, the cerebral white matter is diffusely hyperintense on T2-WI and hypointense on T1WI. The lateral ventricle and third ventricle become dilated as the age progress. At 20 weeks old, no myelination in the cerebral white matter is considered to emerge on T1-WI or on T2-WI, and the volume of the cerebral white matter is reduced. At 20 weeks old, the ventral lateral nucleus in the thalamus is hyperintense on T1-WI, and it is slightly hypointense on T2-WI. At 8 days old, the ventral lateral nucleus of the thalamus is bilaterally hyperintense on T1-WI and hypointense on T2-WI (left arrows). The surrounding areas are hyperintense on T2-WI, and isointense or faintly hyperintense on T1-WI; the latter hyperintense areas on T1-WI are located in the ventral posterolateral nucleus [19] (an up arrow). At 20 weeks old, the inferior and posterior part of the globus palllidus and the posterior part of the putamen are hyperintense on T2-WI and isointense on T1-WI. At 8 days old, hyperintense areas on T1-WI or on T2-WI are intermingled in the inferior and posterior parts of the globus palllidus and the posterior part of the putamen. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
predominant. This is similar to lesions of the basal ganglia and thalamus with cerebral-deep nuclear neuronal injury [16] or profound asphyxia [20]. Different from these two categories, cortical highlighting other than in the hippocampus was not observed, and lesions of the white matter were limited to the white matter between the insula and putamen or were diffusely distributed in the subjects with total asphyxia. However, diffusionweighted imaging performed at an earlier time after birth may detect the other lesions in the cortex or white matter. Low intensity on T1-weighted imaging in the posterior limb of the internal capsule may be interpreted as a defect in myelination, but it also may reflect damage to upper neurons or neuronal pathways. The MR thalamic lesions were combined with high intensity areas on T1-WI and those on T2-WI. In the subjects performing successive MR scanning, the lesions with high intensity on T2-WI were shown to emerge earlier than those on T1-WI and to become unclear. This may indicate that the lesions exhibited by T2-WI are mild or have no accompanying sequela. In other words, dam-
age of the tissue with high water content may be mild or minimum. In contrast, lesions with high intensity areas on T1-WI emerged later than those on T2-WI, and were replaced with those on T2-WI in the future, as shown in Cases 1 and 6. The lesions exhibited by T1-WI certainly connect with neurological sequela. Suggested possible causes of T1 shortening include hemorrhage, calcification, lipid release from myelination breakdown, and the paramagnetic effects of free radicals [21]. However, they remain undetermined. In two subjects (Cases 3, 5), the ventral lateral nucleus was shown as high intensity on T1-WI and as low intensity on T2-WI. This is considered to result from hemorrhage associated from anoxic damage. Also, the MR lesions in the globus pallidus and in the putamen are interpreted, as well as the thalamus. Topography of the affected tegmental nuclei corresponds to watershed areas of the brainstem [15]. They are considered to be most vulnerable to abrupt, severe anoxia. However, the watershed areas of the cerebrum are damaged by prolonged circulatory failure in term
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
59
Fig. 6. Case 6. MR images at 18 days old and 1 year and 11 months old. At 18 days old, high intensity areas on T2-WI are bilaterally seen at the dorsal part in the lower medulla oblongata (a right arrowhead), and at the dorsal paracentral parts in the upper medulla oblongata (left arrowheads). At the middle of the tegmentum in the pons, faintly high intensity areas on T2-WI are bilaterally seen (an up arrowhead). The dorsal part of the midbrain is hyperintense on T2-WI (right arrows). The hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2WI and hypointense on T1-WI (left white arrowheads). Areas in the ventral lateral nucleus of the thalamus connected with part of the ventral posterolateral nucleus [19] are bilaterally hyperintense on T1-WI, and they are isointense or hyperintense on T2-WI (left arrows). The central areas in the thalamus are hyperintense on T2-WI and isointense on T1-WI. The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI; on T2-WI, its central part is bilaterally hypointense and its peripheral part is hyperintense (not seen in the Figure). No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule. At 1 year and 11 months old, high intensity areas on T2-WI are seen at the posterior part of the ventral lateral nucleus and at the posterior and the lateral parts of the putamen. No high intensity areas on T2-WI are observed in the anterior limb of the internal capsule. The midbrain seems vaguely hyperintense on T2-WI.
neonates [16]. This difference can be explained as follows. In experimental studies of acute asphyxia in near-term fetal sheep [22], acute hypoxia has been shown to result in an increase in blood flow to the brainstem, contrary to the decrease in the cerebrum. The fetal brainstem may be protected by the dynamic changes of the fetal circulation responding to hypoxemia and asphyxia. This compensatory mechanism may be difficult to operate fully at the watershed areas of the brainstem. Therefore, the tegmental nuclei in the watershed areas of the brainstem are considered to be vulnerable to abrupt, severe anoxia. Structures in the dorsal midbrain were found to be affected in this study, and lesions in this region were more prominent than lesions in the pons and medulla oblongata. The dorsal midbrain includes the central nucleus of the inferior colliculus, oculomotor nucleus, trochlear nucleus, mesencephalic nucleus of trigeminal nerve, locus ceruleus, reticular formation, and red nucleus. These lesions result in oculomotor disturbances and ptosis, which are characteristic of total asphyxia. In the dorsal pons, MR lesions on T2-WI were observed in only some of the subjects. In the subjects demonstrating no MR lesions by this method, diffusion-weighted imaging performed at an earlier time after birth may have detected lesions [17,18]. Distinct high intensity areas in the dorsal pons on T1-WI were observed in Cases 4 and 5, in whom the dorsal pons was not considered as hyperintense on T2-WI. Thus, high-intensity areas in the dorsal pons on T1-WI may emerge after a decline or disappearance of high-intensity
areas on T2-WI. Although the histological differences are unclear, this phenomenon may be similar to that in the thalamus. The damaged structures exhibited by MR imaging in the dorsal pons include the facial nucleus, vestibular nucleus, abducens nucleus, motor nucleus, spinal nucleus, main sensory nucleus and mesencephalic nucleus of trigeminal nerve, and reticular formation. Lesions in the facial nucleus are responsible for the lack of facial expression as a characterizing symptom of total asphyxia. Of the tegmenta of the brainstem, the tegmentum of the medulla oblongata is considered to be especially vulnerable to HIE [13]. In this region, MR lesions were identified in all of the subjects, although their findings were faint. The damaged structures exhibited by MR imaging in the tegmentum of the medulla oblongata include the hypoglossal nucleus, dorsal nucleus of the vagus nerve, nucleus ambiguus, gracile nucleus of Goll, cuneate nucleus of Burdach, spinal nucleus of the trigeminal nerve, and reticular formation. These lesions cause impaired sucking and swallowing, ventilatory disturbances and gastro-esophageal reflux as characteristic symptoms of total asphyxia. Although the subjects with total asphyxia showed profound motor and intellectual disabilities, their cerebral MR lesions were considered to be insufficient for the corresponding to the severest cerebral symptom. For example, the final T2-weighted image in Case 6 is similar to that of a child with athetotic cerebral palsy [23]. It is difficult to identify a lesion responsible for the most severe disabilities. It is possible that a primary
60
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
nervous center replaced by the cerebrum exists in the brainstem tegmentum and that damage to its center interferes with this replacement. This study has limitations in clarifying the brain lesions with total asphyxia, because MR scanning was performed only at late neonatal periods. However, diffusion-weighted imaging performed within a week after birth is a sensitive method to detect hypoxic or ischemic lesions [17,18]. References [1] Myers RE. Two patterns of perinatal brain damage and their conditions of occurrence. Am J Obstet Gynecol 1972;112:246–76. [2] Myers RE. Four patterns of perinatal brain damage and their conditions of occurrence in primates. Adv Neurol 1975;10:223–34. [3] Schneider H, Ballowitz L, Schachinger H, Hanefeld F, Dro¨szus JU. Anoxic encephalopathy with predominant involvement of basal ganglia, brain stem and spinal cord in the perinatal period. Report on seven newborns. Acta Neuropathol 1975;32:287–98. [4] Leech RW, Alvord Jr EC. Anoxic-ischemic encephalopathy in the human neonatal period. The significance of brain stem involvement. Arch Neurol 1977;34:109–13. [5] Dambska M, Laure-Kamionowska M, Liebhart M. Brainstem lesions in the course of chronic fetal asphyxia. Clin Neuropathol 1987;6:110–5. [6] Leech RW, Brumback RA. Massive brain stem necrosis in the human neonate: presentation of three cases with review of the literature. J Child Neurol 1988;3:258–62. [7] Roland EH, Hill A, Norman MG, Flodmark O, MacNab AJ. Selective brainstem injury in an asphyxiated newborn. Ann Neurol 1988;23:89–92. [8] Pasternak JF, Predey TA, Mikhael MA. Neonatal asphyxia: vulnerability of basal ganglia, thalamus, and brainstem. Pediatr Neurol 1991;7:147–9. [9] Natsume J, Watanabe K, Kuno K, Hayakawa F, Hashizume Y. Clinical, neurophysiologic, and neuropathological features of an infant with brain damage of total asphyxia type (Myers). Pediatr Neurol 1995;13:61–4. [10] Pasternak JF, Gorey MT. The syndrome of acute near-total intrauterine asphyxia in the term infant. Pediatr Neurol 1998;18: 391–8.
[11] Sugama S, Eto Y. Brainstem lesions in children with perinatal brain injury. Pediatr Neurol 2003;28:212–5. [12] Saito Y, Kawashima Y, Kondo A, Chikumaru Y, Matsui A, Nagata I, et al. Dysphagia-gastroesophageal reflux complex: complications due to dysfunction of solitary tract nucleusmediated vago-vagal reflex. Neuropediatrics 2006;37:115–20. [13] Quattrocchi CC, Longo D, Delfino LN, Cilio MR, Piersigilli F, Capua MD, et al. Dorsal brain stem syndrome: MR imaging location of brain stem tegmental lesions in neonates with oral motor dysfunction. Am J Neuroradiol 2010;31:1438–42. [14] Roig M, Grataco`s M, Vazquez E, Del Toro M, Foguet A, Ferrer I, et al. Brainstem dysgenesis: report of five patients with congenital hypotonia, multiple cranial nerve involvement, and ocular motor apraxia. Dev Med Child Neurol 2003;45:489–93. [15] Sarnat HB. Watershed infarcts in the fetal and neonatal brainstem. An aetiology of central hypoventilation, dysphagia, Mo¨ibius syndrome and micrognathia. Eur J Paediatr Neurol 2004;8:71–87. [16] Volpe JJ. Neurology of the newborn. 5th ed. Philadelphia: Saunders; 2008. [17] Barkovich AJ, Miller SP, Bartha A, Newton N, Hamrick SE, Mukherjee P, et al. MR imaging, MR spectroscopy, and diffusion tensor imaging of sequential studies in neonates with encephalopathy. Am J Neuroradiol 2006;27:533–47. [18] Foran A, Cinnante C, Groves A, Azzopardi DV, Rutherford MA, Cowan FM. Patterns of brain injury and outcome in term neonates presenting with postnatal collapse. Arch Dis Child Fetal Neonatal Ed 2009;94:F168–77. [19] Morel A, Magnin M, Jeanmonod D. Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387:588–630. [20] Barkovich AJ. MR and CT evaluation of profound neonatal and infantile asphyxia. Am J Neuroradiol 1992;13:959–72. [21] Huang BY, Castillo M. Hypoxic-ischemic brain injury: imaging findings from birth to adulthood. Radiographics 2008;28:417–39. [22] Jensen A, Garnier Y, Berger R. Dynamics of fetal circulatory responses to hypoxia and asphyxia. Eur J Obstet Gynecol Reprod Biol 1999;84:155–72. [23] Yokochi K, Aiba K, Kodama M, Fujimoto S. Magnetic resonance imaging in athetotic cerebral palsied children. Acta Paediatr Scand 1991;80:818–23.
Original article
Magnetic resonance imaging in neonates with total asphyxia Hiroshi Sugiura a, Masanori Kouwaki b,⇑, Tohru Kato c, Tsutomu Ogata d, Rie Sakamoto e, Atsushi Ieshima f, Kenji Yokochi g a
Department of Neonatology, Seirei-Hamamatsu General Hospital, Hamamatsu, Shizuoka, Japan b Department of Neonatology, Toyohashi Municipal Hospital, Toyohashi, Aichi, Japan c Department of Pediatrics, Okazaki City Hospital, Okazaki, Aichi, Japan d Department of Pediatrics, Nagasaki Medical Center, Ohmura, Nagasaki, Japan e Department of Perinatogy, Funahashi Central Hospital, Funahashi, Chiba, Japan f Department of Pediatrics and Neuropediatrics, Ibaraki Welfare and Medical Center for Disabled Children, Mito, Ibaraki, Japan g Department of Pediatric Neurology, Seirei-Mikatahara General Hospital, Hamamatsu, Shizuoka, Japan Received 11 December 2011; received in revised form 13 April 2012; accepted 18 April 2012
Abstract Magnetic resonance (MR) findings in cases of total asphyxia, whose lesions are mainly in the brainstem and deep nuclei, have not been clarified. In this study, we investigated MR images in neonates with total asphyxia. MR images of six infants (three males and three females; gestational age, 35–39 weeks; birth weights, 1880–3572 g) with total asphyxia were examined. In all subjects, neonatal cortical MR lesions were limited to the hippocampus with highlighting on T1-weighted imaging (T1-WI). The neonatal MR lesions of the cerebral white matter were limited to the white matter between the insula and putamen in four infants, and were diffusely involved in two infants. The ventral lateral nucleus of the thalamus was hyperintense on T1-WI in all of the subjects. Other nuclei in the thalamus, the globus pallidus and the putamen were involved in neonatal MR images of all subjects. High intensity areas on T2- weighted imaging were observed at the dorsal areas in the midbrain, pons and medulla oblongata in all or most of the subjects at the neonatal period. Also, high intensity areas on T1-WI were observed in the tegmentum of the pons and the midbrain in five cases. Neonates with total asphyxia had lesions mainly in the tegmentem of the brainstem, thalamus, putamen and globus palludus. Some of the infants had extensive lesions of the white matter. Ó 2012 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. Keywords: Total asphyxia; Hypoxic ischemic encephalopathy; Tegmentum; Term infant
1. Introduction Myers [1,2] examined perinatal brain damage after total and partial asphyxia in fullterm fetal monkeys. Total asphyxia led to specific patterns of injury in the brainstem of the monkeys. Children with brainstem lesions resulting from perinatal insults at term have been reported to show supratentorial lesions [3–13]. Therefore, ⇑ Corresponding author. Address: Toyohashi Municipal Hospital, 50 Aza Hachiken Nishi, Aotake–Cho, Toyohashi, Aichi 441-8570, Japan. Tel.: +81-532-33-6111; fax: +81-532-33-6177. E-mail address: [email protected] (M. Kouwaki).
it is unclear whether total asphyxia results in isolated brain lesions in humans. Furthermore, insults during the fetal period have been suggested to result in brainstem dysgenesis [14,15]. Partial asphyxia has been associated with parasagittal cerebral injury [16]. A mixture of total and partial asphyxia has been suggested to cause selective neuronal necrosis. Volpe [16] classified this neuronal necrosis into three patterns: diffuse, cerebral cortex-deep nuclear and deep nuclear-brainstem patterns resulting from very severe and prolonged insults, moderate to severe and prolonged insults, and severe and abrupt insults,
0387-7604/$ - see front matter Ó 2012 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.braindev.2012.04.002
54
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
respectively. In the deep nuclear-brainstem pattern, clinical correlates include ptosis, oculomotor disturbances, facial diparesis, ventilatory disturbances, and impaired sucking and swallowing [16]. We prefer the term “total asphyxia” for this pattern, because it reveals a causative mechanism [9]. Although magnetic resonance (MR) findings of brainstem lesions in fullterm infants with hypoxic ischemic encephalopathy (HIE) have been reported [13,17,18], MR findings characterizing total asphyxia have not been clarified. In this study, we investigated MR images in neonates with total asphyxia. 2. Subjects and methods Six infants (three males and three females, born in 2006–2009) were studied. They were selected from patients attending the Seirei-Mikatahara General Hospital for rehabilitation, and from those reported in the Japanese conference of fetal and neonatal neurology or the Zao mailing conference. Inclusion criteria of total asphyxia were as follows. First, acute, brief and severe insults that occurred within 2 weeks before delivery, during delivery or within 24 h after delivery were assumed to result in brain damage. This assumption
Table 1 Clinical profiles of the subjects and methods of MR scanning. Case No. (sex) Perinatal events (gestational age, birth weight, Apgar score (1 min/5 min))
was justified by a loss of fetal heart-rate variability, Apgar score, neurological assessment or laboratory findings such as acidosis. Second, the characteristic neurological abnormalities mentioned below persisted for >1 month after birth. Spontaneous body movement was scant, and was limited to bending backward within a small range. Muscles of facial expression and the jaw were almost immobile. Swallowing of saliva was not observed, and tube-feeding was required. The eyes and eye-lids moved only within a small range. Third, MR imaging was performed within 4 weeks after birth, and a lesion in the brainstem was revealed by MR imaging. However, no lesion forming cyst could be present in the brain, except for a periventicular cyst persisting from prenatal periods. Finally, patients diagnosed with metabolic disease or chromosomal abnormality were excluded. Clinical profiles of the subjects are shown in Table 1. In three infants (Cases 1, 2, 3), an acute insult was assumed to have occurred during delivery. In two infants (Cases 4, 5) and one (Case 6), the insult was assumed to occur before delivery and after delivery, respectively. Cases 2, 4 and 5 underwent MR scanning three times after birth, and Cases 1, 3 and 6 underwent MR scanning 1, 1 and 2 times, respectively (Table 1).
Outcome (age at last examination)
MR scanning (age at performing, machine, slice thickness,* SE(TR/TE ms)*)
1 (F)
Hysterorrhexis Cesarean section 39 weeks 3 days, 2586 g, 0/2
2 years 3 months Almost immobile Tube-feeding Laryngo-tracheal diversion
21 days, 2 years 3 months 1.5T GE, 4 mm, T1: 600/9, T2: 3500/89
2 (M)
Fetal distress Cesarean section 40 weeks 5 days, 3572 g, 1/4
5 years 0 months Almost immobile Tube-feeding
8, 16 and 37 days 1.5T GE, 6 mm, T1: 500/13–14, T2: 3000/98–107
3 (F)
Hysterorrhexis Cesarean section 39 weeks 2 days, 3096 g, 1/4
2 years 1 months Almost immobile Artificial ventilation Tube-feeding
20 days 1.5T Siemens, 4 mm, T1 : 650/15, T2 : 4500/120
4 (F)
Disappearance of fetal movements at 2 weeks before delivery Fetal distress Cesarean section 35 weeks 4 days, 1880 g, 3/5
1 years 10 months
9 days, 44 days, 21 weeks
Almost immobile Artificial ventilation Tube-feeding
1.5T GE, 5 mm, T1: 480/14, T2: 2860/81
2 years4 months
8 days, 41 days, 20 weeks
Almost immobile Artificial ventilation Tube-feeding
1.5T Philips, 5 mm, T1: 450/11, T2: 2721/90
1 years 11 months Almost immobile Tube-feeding
21 days, 1 year 11 months 1.5T GE, 4 mm, T1: 600/9, T2: 3500/90
5 (M)
6 (M)
*
Disappearance of fetal movements at 3 days before delivery Fetal distress Cesarean section 36 weeks 0 days, 2356 g, 2/3 39 weeks 3 days, 2550 g, 8/9 Cardiopulmonary arrest at 16 h after birth
These values are shown when MR scanning is performed within 28 days after birth.
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Axial T1-weighted and T2-weighted images were analyzed, and their spine-echo sequences are shown in Table 1. 3. Results The representative MR images are shown in Figs. 1–6. In all subjects, neonatal cortical MR lesions were limited to the hippocampus with highlighting on T1weighted imaging (T1-WI). As for neonatal MR lesions of the cerebral white matter, lesions were limited to the white matter between the insula and putamen in four (Cases 1, 2, 3, 6) of six infants, and the periventricular areas or watershed areas of the cerebral arteries were not involved. In two infants (Cases 4, 5), the white matter was diffusely involved. This lesion was connected to
55
impairment of the myelination process and atrophy of the white matter in early infancy. Myelination in the posterior limb of the internal capsule was not observed on T1-WI at the neonatal periods. In Cases 4 and 5, high intensity areas on T1-WI in the posterior limb of the internal capsule was not observed at 21 and 20 weeks old, respectively. Although low intensity areas on T2WI, which indicated myelination, were observed in the posterior limb of the internal capsule in two infants (Cases 1, 6) undergoing MR scanning after 23 months old, they were not observed in the anterior limb of it. In Case 1, it was accompanied by the frontal white matter lesions. For thalamic MR lesions, areas corresponding to the ventral lateral nucleus [19] were hyperintense on T1-WI at the neonatal periods in all subjects. Also, areas at the
Fig. 1. Case 1. MR images at 21 days old and 2 years and 4 months old. T2-weighted and T1-weighted images estimated as normal in the brainstem during the neonatal period at the same hospital are shown in the middle row for comparison. At the dorsal part in the lower medulla oblongata, curved belt-shaped areas with high intensity on T2-WI are seen (an up white arrowhead). At the dorsal paracentral part in the upper medulla oblongata, high intensity areas on T2-WI are observed bilaterally (an up arrowhead). At the middle of the tegmentum in the pons, high intensity areas on T2-WI are observed bilaterally (up arrows). At the middle areas in the tegmentum of the midbrain, round high intensity areas on T2-WI are observed bilaterally (down arrows). The tegmentum of the pons and the midbrain is faintly hyperintense on T1-WI (right arrowheads). The hippocampus is bilaterally hyperintense on T1-WI (a white arrow); the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2-WI and hypointense on T1-WI (left white arrowheads). The lateral part of the thalamus, where the ventral lateral nucleus is located [19], is bilaterally hyperintense on T1-WI and on T2-WI (left arrows). The central part of the thalamus is bilaterally hyperintense on T2-WI and faintly hyperintense on T1-WI (up long arrows). The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI (curved arrows). On T2-WI, its central part is bilaterally hypointense and its peripheral part is hyperintense. The posterior and lateral parts of the putamen are bilaterally hyperintense on T1-WI and isointense on T2-WI (down long arrows); its surrounding areas are bilaterally hyperintense on T2-WI and isointense on T1-WI. No high intensity areas on T1-WI, which indicates myelination, are observed in the posterior limb of the internal capsule. At 2 years and 4 months old, the dorsal part in the pons is hyperintense on T2-WI (up arrowheads). Almost the entire midbrain is hyperintense on T2-WI (left arrowheads). The third and lateral ventricle dilated; the inferior horn (*) is markedly dilated. The cortical sulci are dilated predominantly at the frontal areas. The cerebral white matter with high intensity areas on T2-WI is atrophic predominantly at the frontal areas. The thalamic areas at the ventral lateral nucleus, ventral posterolateral nucleus and other nuclei located inferiorly are hyperintense on T2-WI. The inferior and posterior parts of the globus pallidus and the posterior part of the putamen are hyperintense on T2-WI (left arrows). No high intensity areas on T2-WI, which indicates myelination, are observed in the anterior limb of the internal capsule.
56
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Fig. 2. Case 2. MR images at 8, 16 and 37 days old. At all three times examined, faintly high intensity areas on T2-WI are observed bilaterally at the dorsal part in the lower medulla oblongata (left arrowheads), and at the dorsal paracentral parts in the upper medulla oblongata (right arrowheads). At the tegmentum of the midbrain, high intensity areas on T2-WI are observed bilaterally at 8 and 16 days old (right arrows). The tegmentum of the pons and the dorsal part of the midbrain are hypertense on T1-WI at 37 days old (left arrows). At all times examined, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hypointense on T1-WI and slightly hyperintense on T2-WI at 16 and 37 days old (left white arrowheads). On T1WI, the thalamic lesions in the ventral lateral nucleus (an up white arrow) and those in the ventral posterolateral nucleus [19] (an up arrow) are bilaterally hyperintense at 37 days old. Then, the same areas seem hyperintense on T2-WI. At 16 days old, the areas with high intensity on T1-WI are larger and the intensity is less intense than at 37 days old. Then, almost the entire thalamus is hyperintense on T2-WI with the peripheral isointense areas; the hyperintense areas on T1-WI cannot be discriminated from the T2-weighted images. At 8 days old, the hyperintense areas on T1-WI are least clearly exhibited; the lateral and posterior borders of the ventral lateral nucleus are most hyperintense on T1-WI (an up arrowhead). Then, the more extensive thalamic areas are hyperintense on T2-WI than at 16 days old. At all three times examined, hyperintense areas on T1-WI or on T2-WI are intermingled in the globus palllidus (down arrows) and in the putamen. No high intensity areas on T1-WI are observed in posterior limb of the internal capsule.
Fig. 3. Case 3. MR images at 20 days old. At the dorsal part in the lower medulla oblongata, high intensity areas on T2-WI are observed (a left arrowhead). At the dorsal paracentral parts in the upper medulla oblongata, faintly high intensity areas on T2-WI are observed bilaterally (right arrowheads). At the middle of the tegmentum in the pons, faintly high intensity areas on T2-WI are seen bilaterally (up arrowheads). The tegmentum in the pons is faintly hyperintense on T1-WI (right white arrowheads). The dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). The hippocampus is bilaterally hyperintense on T1-WI (a white arrow); the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2-WI and hypointense on T1-WI (left white arrowheads). The ventral lateral nucleus of the thalamus is bilaterally hyperintense on T1-WI and hypointense on T2-WI (left arrows). Its surrounding areas are hyperintense on T2-WI and isointense or faintly hyperintense on T1-WI; the latter hyperintense areas on T1-WI are located in the ventral posterolateral nucleus [19] (an up arrow). The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI and hypointense on T2-WI (not shown in the Figure). The posterior and lateral parts of the putamen are bilaterally hyperintense on T1-WI and hypointense on T2-WI (a down arrow); its surrounding areas are bilaterally hyperintense on T2-WI and isointense on T1-WI. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
57
Fig. 4. Case 4. MR images at 9 days old, 44 days old and 21 weeks old. At 9 days old, faintly high intensity areas on T2-WI are observed bilaterally at the middle part in the lower medulla oblongata (a right arrowhead), and at the dorsal paracentral parts in the upper medulla oblongata (a left arrowhead). The dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). Then, the dorsal parts of the upper medulla oblongata, pons and midbrain are hypertense on T1-WI (up white arrowheads). At 44 days old, they are less intense than at 9 days old. At 9 days old, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. Then, the cerebral white matter is diffusely hyperintense on T2-WI and hypointense on T1-WI; the subependymal cyst is observed bilaterally (down arrowheads). The lateral ventricle and third ventricle become dilated as the age progress. At 21 weeks old, no myelination in the cerebral white matter is considered to emerge on T1-WI or on T2-WI, and the volume of the cerebral white matter is reduced. At 21 weeks old, the entire thalamus exhibits faintly high intensity on T1-WI, and no abnormal intensity on T2-WI; the thalamus becomes atrophied. At 9 days old, the peripheral areas of the thalamus are hyperintense on T1-WI, and its central areas are hyperintense on T2-WI. At 21 weeks old, the whole globus palllidus and putamen exhibits no abnormal intensity on T2-WI or T1-WI; the two become atrophied. At 9 days old, hyperintense areas on T1-WI or on T2-WI are intermingled in the globus palllidus and in the putamen. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
ventral posterolateral nucleus [19] were involved in 5 subjects (Cases 1, 2, 3, 5, 6). In 2 infants (Cases 1, 6), high intensity areas at the ventral lateral nucleus on T1-WI at the neonatal period were shown to develop to high intensity areas on T2-WI after 23 months old. In contrast, the central part of the thalamus was hyperintense on T2-WI in most infants. In all three infants undergoing successive MR scanning (Cases 2, 4, 5), this high intensity area was most intense at the first scanning, and it disappeared at the last scanning. In the globus pallidus, high intensity areas on T1-WI were observed in all subjects at the neonatal period. However, in three cases (Cases 1, 3, 6), these areas were mostly limited to the inferior and posterior parts, with accompanying low intensity areas on T2-WI. In the putamen, high intensity areas on T2-WI were mixed with those on T1-WI in all subjects at the neonatal period. High intensity areas on T2-WI were observed at the dorsal midbrain in all subjects at the neonatal period. While round and hyperintense areas at the upper midbrain were seen clearly at 21 days old in Case 1, almost the entire midbrain was hyperintense on T2-WI at 2 years 4 months old. In the pons, high intensity areas on T2-WI were observed in the middle of the tegmentum in three cases (Cases 1, 3, 6). Also, high intensity areas on T1-WI were observed in the entire tegmentum of
the pons and midbrain in five cases (Cases 1, 2, 3, 4, 5). In the medulla oblongata, faint high intensity areas on T2-WI were observed in five subjects (Cases 1, 2, 3, 4, 6). They were observed at the dorsal paracentral parts in the upper medulla oblongata and at the dorsal or central part in the lower medulla oblongata. Cerebellar MR lesions were not observed in the subjects. 4. Discussion Nervous system damage from total asphyxia in monkeys was reported to affect nuclei in the brainstem tegmentum including the central nucleus of the inferior colliculus, posterior and lateral ventral nuclei of the thalamus, and the intermediolateral gray column of the spinal cord [1,2]. MR imaging of brainstem nuclei in the human neonates after total asphyxia showed injuries similar to those demonstrated in monkeys after total asphyxia. In monkeys with the total asphyxia [1,2], the cerebral lesions were limited to the thalamus. In human subjects with total asphyxia, not only the thalamus, but also the basal ganglia were involved. In addition, the cerebral cortex and white matter were involved, despite being mild. This is a crucial difference between monkeys with total asphyxia and humans with total asphyxia. Thalamic lesions were predominant in the ventral lateral nucleus, and lesions of the putamen were posterior
58
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
Fig. 5. Case 5. MR images at 8 days old, 41 days old and 20 weeks old. At 8 days old, the dorsal part of the midbrain is hyperintense on T2-WI (a right arrow). Then, the dorsal parts of the upper medulla oblongata, pons and midbrain are hypertense on T1-WI (up white arrowheads). At 41 days old, they become less intense than at 8 days old. At 8 days old, the hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. Then, the cerebral white matter is diffusely hyperintense on T2-WI and hypointense on T1WI. The lateral ventricle and third ventricle become dilated as the age progress. At 20 weeks old, no myelination in the cerebral white matter is considered to emerge on T1-WI or on T2-WI, and the volume of the cerebral white matter is reduced. At 20 weeks old, the ventral lateral nucleus in the thalamus is hyperintense on T1-WI, and it is slightly hypointense on T2-WI. At 8 days old, the ventral lateral nucleus of the thalamus is bilaterally hyperintense on T1-WI and hypointense on T2-WI (left arrows). The surrounding areas are hyperintense on T2-WI, and isointense or faintly hyperintense on T1-WI; the latter hyperintense areas on T1-WI are located in the ventral posterolateral nucleus [19] (an up arrow). At 20 weeks old, the inferior and posterior part of the globus palllidus and the posterior part of the putamen are hyperintense on T2-WI and isointense on T1-WI. At 8 days old, hyperintense areas on T1-WI or on T2-WI are intermingled in the inferior and posterior parts of the globus palllidus and the posterior part of the putamen. No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule.
predominant. This is similar to lesions of the basal ganglia and thalamus with cerebral-deep nuclear neuronal injury [16] or profound asphyxia [20]. Different from these two categories, cortical highlighting other than in the hippocampus was not observed, and lesions of the white matter were limited to the white matter between the insula and putamen or were diffusely distributed in the subjects with total asphyxia. However, diffusionweighted imaging performed at an earlier time after birth may detect the other lesions in the cortex or white matter. Low intensity on T1-weighted imaging in the posterior limb of the internal capsule may be interpreted as a defect in myelination, but it also may reflect damage to upper neurons or neuronal pathways. The MR thalamic lesions were combined with high intensity areas on T1-WI and those on T2-WI. In the subjects performing successive MR scanning, the lesions with high intensity on T2-WI were shown to emerge earlier than those on T1-WI and to become unclear. This may indicate that the lesions exhibited by T2-WI are mild or have no accompanying sequela. In other words, dam-
age of the tissue with high water content may be mild or minimum. In contrast, lesions with high intensity areas on T1-WI emerged later than those on T2-WI, and were replaced with those on T2-WI in the future, as shown in Cases 1 and 6. The lesions exhibited by T1-WI certainly connect with neurological sequela. Suggested possible causes of T1 shortening include hemorrhage, calcification, lipid release from myelination breakdown, and the paramagnetic effects of free radicals [21]. However, they remain undetermined. In two subjects (Cases 3, 5), the ventral lateral nucleus was shown as high intensity on T1-WI and as low intensity on T2-WI. This is considered to result from hemorrhage associated from anoxic damage. Also, the MR lesions in the globus pallidus and in the putamen are interpreted, as well as the thalamus. Topography of the affected tegmental nuclei corresponds to watershed areas of the brainstem [15]. They are considered to be most vulnerable to abrupt, severe anoxia. However, the watershed areas of the cerebrum are damaged by prolonged circulatory failure in term
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
59
Fig. 6. Case 6. MR images at 18 days old and 1 year and 11 months old. At 18 days old, high intensity areas on T2-WI are bilaterally seen at the dorsal part in the lower medulla oblongata (a right arrowhead), and at the dorsal paracentral parts in the upper medulla oblongata (left arrowheads). At the middle of the tegmentum in the pons, faintly high intensity areas on T2-WI are bilaterally seen (an up arrowhead). The dorsal part of the midbrain is hyperintense on T2-WI (right arrows). The hippocampus is bilaterally hyperintense on T1-WI; the cerebral cortex except for the hippocampus does not exhibit highlighting on T1-WI. The cerebral white matter between the insula and putamen is bilaterally hyperintense on T2WI and hypointense on T1-WI (left white arrowheads). Areas in the ventral lateral nucleus of the thalamus connected with part of the ventral posterolateral nucleus [19] are bilaterally hyperintense on T1-WI, and they are isointense or hyperintense on T2-WI (left arrows). The central areas in the thalamus are hyperintense on T2-WI and isointense on T1-WI. The inferior and posterior parts of the globus palllidus are bilaterally hyperintense on T1-WI; on T2-WI, its central part is bilaterally hypointense and its peripheral part is hyperintense (not seen in the Figure). No high intensity areas on T1-WI are observed in the posterior limb of the internal capsule. At 1 year and 11 months old, high intensity areas on T2-WI are seen at the posterior part of the ventral lateral nucleus and at the posterior and the lateral parts of the putamen. No high intensity areas on T2-WI are observed in the anterior limb of the internal capsule. The midbrain seems vaguely hyperintense on T2-WI.
neonates [16]. This difference can be explained as follows. In experimental studies of acute asphyxia in near-term fetal sheep [22], acute hypoxia has been shown to result in an increase in blood flow to the brainstem, contrary to the decrease in the cerebrum. The fetal brainstem may be protected by the dynamic changes of the fetal circulation responding to hypoxemia and asphyxia. This compensatory mechanism may be difficult to operate fully at the watershed areas of the brainstem. Therefore, the tegmental nuclei in the watershed areas of the brainstem are considered to be vulnerable to abrupt, severe anoxia. Structures in the dorsal midbrain were found to be affected in this study, and lesions in this region were more prominent than lesions in the pons and medulla oblongata. The dorsal midbrain includes the central nucleus of the inferior colliculus, oculomotor nucleus, trochlear nucleus, mesencephalic nucleus of trigeminal nerve, locus ceruleus, reticular formation, and red nucleus. These lesions result in oculomotor disturbances and ptosis, which are characteristic of total asphyxia. In the dorsal pons, MR lesions on T2-WI were observed in only some of the subjects. In the subjects demonstrating no MR lesions by this method, diffusion-weighted imaging performed at an earlier time after birth may have detected lesions [17,18]. Distinct high intensity areas in the dorsal pons on T1-WI were observed in Cases 4 and 5, in whom the dorsal pons was not considered as hyperintense on T2-WI. Thus, high-intensity areas in the dorsal pons on T1-WI may emerge after a decline or disappearance of high-intensity
areas on T2-WI. Although the histological differences are unclear, this phenomenon may be similar to that in the thalamus. The damaged structures exhibited by MR imaging in the dorsal pons include the facial nucleus, vestibular nucleus, abducens nucleus, motor nucleus, spinal nucleus, main sensory nucleus and mesencephalic nucleus of trigeminal nerve, and reticular formation. Lesions in the facial nucleus are responsible for the lack of facial expression as a characterizing symptom of total asphyxia. Of the tegmenta of the brainstem, the tegmentum of the medulla oblongata is considered to be especially vulnerable to HIE [13]. In this region, MR lesions were identified in all of the subjects, although their findings were faint. The damaged structures exhibited by MR imaging in the tegmentum of the medulla oblongata include the hypoglossal nucleus, dorsal nucleus of the vagus nerve, nucleus ambiguus, gracile nucleus of Goll, cuneate nucleus of Burdach, spinal nucleus of the trigeminal nerve, and reticular formation. These lesions cause impaired sucking and swallowing, ventilatory disturbances and gastro-esophageal reflux as characteristic symptoms of total asphyxia. Although the subjects with total asphyxia showed profound motor and intellectual disabilities, their cerebral MR lesions were considered to be insufficient for the corresponding to the severest cerebral symptom. For example, the final T2-weighted image in Case 6 is similar to that of a child with athetotic cerebral palsy [23]. It is difficult to identify a lesion responsible for the most severe disabilities. It is possible that a primary
60
H. Sugiura et al. / Brain & Development 35 (2013) 53–60
nervous center replaced by the cerebrum exists in the brainstem tegmentum and that damage to its center interferes with this replacement. This study has limitations in clarifying the brain lesions with total asphyxia, because MR scanning was performed only at late neonatal periods. However, diffusion-weighted imaging performed within a week after birth is a sensitive method to detect hypoxic or ischemic lesions [17,18]. References [1] Myers RE. Two patterns of perinatal brain damage and their conditions of occurrence. Am J Obstet Gynecol 1972;112:246–76. [2] Myers RE. Four patterns of perinatal brain damage and their conditions of occurrence in primates. Adv Neurol 1975;10:223–34. [3] Schneider H, Ballowitz L, Schachinger H, Hanefeld F, Dro¨szus JU. Anoxic encephalopathy with predominant involvement of basal ganglia, brain stem and spinal cord in the perinatal period. Report on seven newborns. Acta Neuropathol 1975;32:287–98. [4] Leech RW, Alvord Jr EC. Anoxic-ischemic encephalopathy in the human neonatal period. The significance of brain stem involvement. Arch Neurol 1977;34:109–13. [5] Dambska M, Laure-Kamionowska M, Liebhart M. Brainstem lesions in the course of chronic fetal asphyxia. Clin Neuropathol 1987;6:110–5. [6] Leech RW, Brumback RA. Massive brain stem necrosis in the human neonate: presentation of three cases with review of the literature. J Child Neurol 1988;3:258–62. [7] Roland EH, Hill A, Norman MG, Flodmark O, MacNab AJ. Selective brainstem injury in an asphyxiated newborn. Ann Neurol 1988;23:89–92. [8] Pasternak JF, Predey TA, Mikhael MA. Neonatal asphyxia: vulnerability of basal ganglia, thalamus, and brainstem. Pediatr Neurol 1991;7:147–9. [9] Natsume J, Watanabe K, Kuno K, Hayakawa F, Hashizume Y. Clinical, neurophysiologic, and neuropathological features of an infant with brain damage of total asphyxia type (Myers). Pediatr Neurol 1995;13:61–4. [10] Pasternak JF, Gorey MT. The syndrome of acute near-total intrauterine asphyxia in the term infant. Pediatr Neurol 1998;18: 391–8.
[11] Sugama S, Eto Y. Brainstem lesions in children with perinatal brain injury. Pediatr Neurol 2003;28:212–5. [12] Saito Y, Kawashima Y, Kondo A, Chikumaru Y, Matsui A, Nagata I, et al. Dysphagia-gastroesophageal reflux complex: complications due to dysfunction of solitary tract nucleusmediated vago-vagal reflex. Neuropediatrics 2006;37:115–20. [13] Quattrocchi CC, Longo D, Delfino LN, Cilio MR, Piersigilli F, Capua MD, et al. Dorsal brain stem syndrome: MR imaging location of brain stem tegmental lesions in neonates with oral motor dysfunction. Am J Neuroradiol 2010;31:1438–42. [14] Roig M, Grataco`s M, Vazquez E, Del Toro M, Foguet A, Ferrer I, et al. Brainstem dysgenesis: report of five patients with congenital hypotonia, multiple cranial nerve involvement, and ocular motor apraxia. Dev Med Child Neurol 2003;45:489–93. [15] Sarnat HB. Watershed infarcts in the fetal and neonatal brainstem. An aetiology of central hypoventilation, dysphagia, Mo¨ibius syndrome and micrognathia. Eur J Paediatr Neurol 2004;8:71–87. [16] Volpe JJ. Neurology of the newborn. 5th ed. Philadelphia: Saunders; 2008. [17] Barkovich AJ, Miller SP, Bartha A, Newton N, Hamrick SE, Mukherjee P, et al. MR imaging, MR spectroscopy, and diffusion tensor imaging of sequential studies in neonates with encephalopathy. Am J Neuroradiol 2006;27:533–47. [18] Foran A, Cinnante C, Groves A, Azzopardi DV, Rutherford MA, Cowan FM. Patterns of brain injury and outcome in term neonates presenting with postnatal collapse. Arch Dis Child Fetal Neonatal Ed 2009;94:F168–77. [19] Morel A, Magnin M, Jeanmonod D. Multiarchitectonic and stereotactic atlas of the human thalamus. J Comp Neurol 1997;387:588–630. [20] Barkovich AJ. MR and CT evaluation of profound neonatal and infantile asphyxia. Am J Neuroradiol 1992;13:959–72. [21] Huang BY, Castillo M. Hypoxic-ischemic brain injury: imaging findings from birth to adulthood. Radiographics 2008;28:417–39. [22] Jensen A, Garnier Y, Berger R. Dynamics of fetal circulatory responses to hypoxia and asphyxia. Eur J Obstet Gynecol Reprod Biol 1999;84:155–72. [23] Yokochi K, Aiba K, Kodama M, Fujimoto S. Magnetic resonance imaging in athetotic cerebral palsied children. Acta Paediatr Scand 1991;80:818–23.