Periventricular Leukomalacia: Ultrasonic and NeuropathologicaI CorreIations
CASE REPORTS
J. Rodriguez D. Claw G. Verellen G. Lyon
Studies of correlations between ultrasonic changes and neuropathologica~ findings in periventricular leukomalacia (PVL) are relatively scarce, and frequently are based purely on macroscopic evidence or on fragmentary microscopic studies (Hill et al. 1982, Bowerman et al. 1984, Nwaesei et al. 1984, Fawer et al. 1985, Trounce et al. 1986). We report what we believe to be the first complete neuropathological study of a case of PVL, in which periventricular echo-free cavities discovered in a preterm neonate at 34 weeks postmenstrual age (PMA) were no longer visible at 42 weeks PMA, three weeks before the infant died.
Case report The child was evaluated and treated in the Neonatal Unit at St. Luc Hospital. Ultrasound scans were obtained with a MARK 100 (ATL) real-time multifrequency scan-head (3-5-7-SMHz). Using a transfontanelle approach, scans were taken in the coronal and sagittal planes.
Clinical findings This boy was born at 27 weeks PMA, weighing 930g, after an uneventful pregnancy. At birth he was asphyxiated. with severe acidosis. Apgar scores were I at one minute and 5 at five minutes. He was intubated and ventilated for 19 days because of hyaline membrane disease. There was no evidence of haemodynamic failure during this period. At six weeks of age (33 weeks PMA), prolonged apnoea required re-intubation and assisted ventilation,
which was complicated by bilateral pneumothorax and severe circulatory collapse. Intestinal resection for nccrotizing enterocolitis was performed one week Neurological development was much delayed. The infant was hypotonic, with intermittent hypertonic spasms. He died at 18 weeks of age from an intestinal infection (Bacillus cereus).
Sonographic findings Increased echogenicity in the left periventricular white-matter was observed at three weeks of age (30 weeks PMA). At seven weeks, a string of echo-free cavities appeared in the periventricular hyperechogenic area, the largest measuring 0.9cm at its widest, located at the level of the atrium (point of confluence of the body of the lateral ventricle with the temporal and occipital horns) (Fig. la, b). The echo-free cavities remained stable for at least one month, but were no longer visible at 15 weeks of age, when the periventricular hyperechogenicity had also disappeared. No ventricular dilatation was observed. A CT scan at that time also showed no periventricular cavity (Fig. Ic). Neuroparhological findings The postmortem was done in the Laboratory of Developmental Neuropathology at the Paediatric Neurology Centre. The brain was fixed in 20 per cent formaldehyde. Blocks of brain-tissue were embedded in paraffin and sectioned semi-serially; and the brainstem and cerebellum horizontally. The 10 to ISpm-[hick sections were stained with haematoxylin and eosin and cresyl violet (Nissl) according to Mallory's method. Frozen sections were stained with oil red 0. The brain had a normal weight of 4708. The Ieptomeninges, meningeal vessels and external surface of the brain looked normal. No macroscopic
n
m r-I
4
34 7
f
u
Fig. 2. Left hemisphere: coronal section through the parietal lobe shows slit-like cavity (arrow) in centre of thin glial band extending from corona radiata into the paracentral white-matter (arrowhead). (Cresyl violet x 10.)
Fig. 1. Ultrasound scan at seven weeks of age. (Aj Left sagittal view showing 9mm (greatest diameterj echo-free area (arrow). suggesting cavity within periventricular white-matter at level of atrium. /Bj Additional smaller cavities (arrowheadsj, rostrally disposed. (Cj At I5 weeks all cavities have disappeared.
348
changes were observed in the hemispheric whitematter on coronal section. The cerebellum and brainstem were normal. On microscopic examination, the left hemisphere had an elongated glial scar extending from the deep part of the centrum semi-ovale, near the angle of the lateral ventricle, into the paracentral white-matter (Figs. 2, 4). In some coronal sections it was seen to extend to the core of two adjacent gyri. In a sagittal plane, the glial band was seen from the mid-frontal to the posterior parietal region. Astrocytic nuclei and bundles of glial fibres were parallel to the axis of the gyrus. A few cell fragments were encrusted in calcium. The posterior part of the glial band, at the level of the pulvinar, was excavated by a slit-like cavity
containing a few lipid-laden oil red 0-positive macrophages (Fig. 3). A narrow zone of loose gliosis surrounded the dense glial core. The rest of the whitematter was normal and well developed ‘myelination oligo-dendrocytes’ (Roback and Scherer 1935), with their typical eosinophilic flame-like cytoplasm, were present. No changes were detected in the whitematter of the right hemisphere. In the thalamus bilaterally, in the right globus pallidus and in the left neocortex, small postischaemic foci of neuronal loss, dense gliosis. macrophages and cells encrusted in calcium were observed. The lesion seemed to be contemporary with the white-matter changes.
Discussion Sonographic findings in this neonate are consistent with PVL. A left-periventricular echogenic area became apparent at three weeks of age. In the posterior half of this zone, one large and several small echo-free cavities visualized at seven weeks were no longer detectable by ultrasonography at 15 weeks. The destructive white-matter lesions were more likely to be a result of ischaemia than of haemorrhage, because
u-l m u-l
4
P D
Fig. 3. Left hemisphere: same coronal section at higher magnuication shows excavation of linear scar. Cavity containing lipid-laden macrophages amid dense glial tissue. Glial nuclei and fibres run parallel to gyral axis. Moderate degree of astrocytic proli/eration can be seen in limited area of surrounding white-matter. (Cresyl violet x 300.)
in the latter case some trace of bleeding (for instance haemosiderin in macrophages) should have been found at autopsy. The prolonged hypotension at six weeks may have played an aggravating r81e in the evolution of white-matter lesions and a determinating r81e in the pathogenesis of the focal grey-matter infarcts. Disappearance of periventricular cysts on ultrasound imaging after three weeks of age has been well documented (Bowerman ef al. 1984, Dubowitz ef a/. 1985, Trounce ef al. 1986, de Vries ef al. 1987), but precise correlations of this phenomenon with neuropathological findings-other than macroscopic (Trounce ef al. 1986)-have not been reported to our knowledge. De Reuck ef a/. (1972) mention that the cavity may ultimately collapse, leaving nothing but a periventricular scar, but give no illustration of this occurrence.
Fig. 4. Left hemisphere: coronal section. Glial band, seen here at more rostra1 level, extends into periventricular white-matter (arrow). (Cresyl violet x 75.)
In our case no abnormality could be detected on the formalin-fixed brain, but on microscopic examination a linear glial scar was seen, extending in a frontal plane from the periventricular white-matter (or the centrum semi-ovale) well into the paracentral white-matter, and in the sagittal plane from the anterior to the posterior poles of the thalamus, approximately in the area of extension of the periventricular cavities. In the posterior part of the glial streak, at a point corresponding to the large echofree cyst on ultrasound scanning, there was a slit-like cavity containing a few macrophages. There was no ventricular enlargement, as there was in the cases of De Reuck ef a/. (1 972) and Trounce ef a/. (1986). It seems highly probable that this small cavity in a glial scar was what remained of the large periventricular cyst observed at seven weeks of age, after its collapse.
349
c v) )
i: 2 s u
Otherwise one would have to assume that the cyst seen on ultrasound scans had vanished entirely. The mechanism of disappearance or considerable reduction in size of periventricular cavities is not clear. However, it seems logical to postulate-at least in the present casethat it is the result of a change in the pressure gradient between the cavity and the surrounding brain substance. A decrease in intracavity pressure could be related to resorption of intracystic fluid (or blood) and to a progressive reduction of the mass of macrophages. Pericystic pressure, on the other hand, is bound to increase as a result of rapid brain-growth during late fetal and early postnatal life. This is mainly because of an enormous increase in neuronal volume and by glial proliferation, particularly the multiplication of oligodendrocytes, which precedes myelination in the central whitematter. Between 34 and 42 weeks of gestation, brain weight increases from 290g to 420g (Gruenwald and Minh 1960).
It is interesting that the cortical, thalamic and pallidal ischaemic foci, which might have been of clinical significance, were not detected by ultrasound scanning. This exemplifes the difficulty of making an exact estimation of brain damage in the neonate by ultrasonography . Another point is the significance of the pathological changes found in the left hemisphere. PVL frequently leads to a stellate scar: from a periventricular area of demyelination and gliosis, long linear extensions radiate into the white axis of one or more gyri, as first described by Marie et al. (1959). In our case, the glial scar is quite consistent with PVL, although a strictly
350
periventricular softening was inconspicuous in most areas. We have observed this pattern in brains in which there was typical periventricular softening in adjacent areas. The topography of the streak of isomorphic glial tissue in which astrocytic nuclei and glial fibres are parallel to the gyral axis is not readily accounted for by the usual explanation of border-zone ischaemic softening (Banken and Larroche 1962, Lindenberg 1963, De Reuck 1971, 1984; De Reuck et al. 1972, Armstrong and Norman 1974, Takashima and Tanaka 1978). Nor does it correspond to the relatively avascular subcortical triangle described by Takashima et al. (1978). The remarkable study by Kuban and Gilles (1985) on the development of telencephalic microvasculature casts much doubt on the pathophysiological explanations cited above, and on the reality of an avascular subcortical zone. The characteristic intra-axial linear glial scar, or the much more frequent stellate scar, are more readily explained by the experiments of Young et al. (1982) on newborn dogs with severe hypotension. Those authors found a selective diminution of blood-flow in the whitematter. Interestingly, hypoperfusion was more marked precisely in the periventricular and intra-axial white-matter, which are .the usual sites of brain damage. Accepted for publication 9th May 1989. Authors’ Appointments J. Rodriguez; D. Claus; G. Verellen; *G. Lyon; Service de Neurologie Pkdiatrique, Hapita1 Universitaire St. Luc, UCL, Avenue Hippocrate 10, B-1200 Bruxelles, Belgium. *Correspondence to last author.
SUMMARY Ultrasound scans of a preterm neonate (27 weeks gestation) at seven weeks of age showed periventricular echo-free cavities, but these were no longer visible at 15 weeks, three weeks before the infant died. At autopsy, a linear glial scar, extending from the periventricular white-matter into the white axis of the parasagittal gyrus. was found in the area occupied by the periventricular cysts. The larger cavity was reduced to a slit-like excavation in the midst of glial tissue. Unsuspected focal infarcts in the cerebral cortex were also found. This observation demonstrates that transient echofree cavities represent foci of cystic necrosis, which are subject to secondary collapse. In the authors’ experience, the linear extension of periventricular leukomalacia (PVL) into the core of parasagittal gyri is a frequent feature of PVL, and one which cannot easily be accounted for by the usual explanations of border-zone ischaemic softening.
n m n
RI~SUMJ~ Leucomalacie pdri-ventriculaire: corrdlation entre donndes kchographiques et neuropathologiques Des kchographies ont etk realistes chez un prkmature (27 semaines de gestation) A 1’8ge de sept semaines. Des cavitks pkri-ventriculaires sans k h o s ont etk dkcouvertes, mais n’ktaient plus visible A 15 semaines, trois semaines avant la mort. A l’autopsie, une cicatrice linkaire gliale s’ktendant de la substance blanche pkri-ventriculaire jusqu’ri I’axe blanc du gyrus a etk trouvke dans la region occupee par les cavitks peri-ventriculaires. La plus grande cavite etait rkduite A une fissure au milieu du tissu glial. Un infarcissement local insoupconne fut aussi trouvk dans le cortex ckrkbral. Cette observation dkmontre que les cavitks transitoires sans kchos representent des foyers de nkcroses cystiques sujets A un effondrement secondaire. L’extension linkaire de la leucomalacie pkri-ventriculaire au coeur des gyri para-sagitaux est dans notre experience une caractkristique frkquente de cette condition et elle ne peut repondre facilement ti l’explication habituelle de ramollissement ischkmique en bordure.
p’ I
z
N-
m
Q
ZUSAMMENFASSUNG Periventrikulare Leukomalazie: Korrelationen zwischen Ultraschall und Neuropathologie Bei einem Friihgeborenen (27. Gestationswoche) wurde im Alter von sieben Wochen eine Ultraschalluntersuchung durchgefiihrt. Es wurden periventrikulare echofreie Raume gefunden, die aber im Alter von 15 Wochen-drei Wochen vor dem Tod-nicht mehr nachweisbar waren. Bei der Autopsie fand sich in dem Bereich der periventrikularen Cysten ein lineares Gliaband, das sich von der periventrikularen weifien Substanz bis zur Achse des parasagittalen Gyrus erstreckte. Der grofiere Hohlraum war auf eine schlitzformige Aushohlung im Gliagewebe reduziert. AuRerdem fanden sich unvermutete Infarkte im cerebralen Cortex. Diese Beobachtung zeigt ,daR passagere echofreie Raume Herde cystischer Nekrosen darstellen, die spater kollabieren. Die lineare Ausdehnung der periventrikularen Leukomalazie bis in das Horn des parasagittalen Gyrus ist nach unserer Erfahrung ein typisches Merkmal bei diesem Befund und kann nicht einfach durch ischamische Randzonenerweichung erklart werden. RESUMEN Leucomalacia periventricular: correlaciones ultrasdnicas y neuropatologicas Se realizaron exAmenes ultrasbnicos en recikn nacidos pretkrmino (27 semanas de gestacibn) a las siete semanas de edad. Se decubrieron cavidades periventriculares libres de eco, que ya no eran visibles a las 15 semanas, tres semanas antes de la muerte. En la autopsia se hallb una cicatriz lineal glial. que se extendia desde la substancia blanca periventricular hacia adentro del eje blanco del giro parasagital, en el Area ocupada por el quiste periventricular. La gran cavidad se habia reducido a una excavacibn como una rendija en medio del tejido glial. Tambikn se hallaron infartos focales insospechados en el cbrtex cerebral. Esta observacibn demuestra que las cavidades libres de eco representan focos de necrosis quistica que posteriormente sufriran un colapso. La linea extensa de leucomalacia periventricular dentro de 10s giros parasagitales. es en nuestra experiencia una caracteristica frecuente de esta patologia, que no puede ser fAcilmente explicada por el argument0 usual de un reblandecimiento isquemico de 10s bordes. References Armstrong, D., Norman, M. G. (1974) ‘Periventricular leukomalacia in neonates: complications and sequelae.’ Archives of Disease in Childhood, 49, 361-375. Banken. B., Larroche, J. (1962) ‘Periventricular leucomalacia in infancy.’ Archives of Neurology, 7, 386-410. Bowerman, R. A., Donn, S. M., Di Pietro, M. A,, D’Amato, C. J., Hicks, S. P. (1984) ‘Periventricular leucomalacia in the preterm newborn infant: sonographic and clinical features.’ Radiology, 151, 383-388. De Reuck, J . (1971) ‘The human periventricular arterial blood supply and the anatomy of cerebral infarctions.’ European Neurology, 5, 321-334. - (1984) ‘Cerebral angioarchitecture and perinatal brain lesions in premature and full-term infants.’ Acto Neurologrca Scandinavica, 70, 391-395. - Chattha, A. S., Richardson, E. P. (1972)
‘Pathogenesis and evolution of periventricular leukomalacia in infancy.’ Archives of Neurology, 27, 229-236. De Vries, L. S., Connell, J. A., Dubowitz, L. M. S.. Oozeer. R. C., Dubowitz, V., Pennock, J. M. (1987) ‘Neurological, electrophysiological and MRI abnormalities in infants with extensive cystic leukomalacia.’ Neuropediafrics. 18, 61-66.
Dubowitz, L. M. S., Bydder, G . M., Mushin, J. (1985) ‘Developmentalsequence of periventricular leukomalacia.’ Archives of Disease in Childhood, 60, 349-355. Fawer, C-L.. Calame, A., Perentes, A., Anderegg. A. (1985) ‘Periventricular leukomalacia: a correlation study between real-time ultrasound and autopsy findings.’ Neuroradiology. 27, 292-300.
Gruenwald, P., Minh. H. N. (1960) ‘Evaluation of body and organ weights in perinatal pathology. I. Normal standards derived from autopsies.’ American Journal of Clinical Pathology. 34, 247-253.
Hill, A., Melson, G. L., Clark, H. B. (1982) ‘Hemorrhagic periventricular leukomalacia: diagnosis by real-time ultrasound and correlation with autopsy findings.’ Pediatrics, 69, 144-149. Kuban, K. C. K., Gilles, F. H. (1985) ‘Human Telencephalic angiogenesis.’ Annals of Neurology. 17, 539-548. Lindenberg. R. (1963) ‘Patterns of CNS vulnerability in acute, hypoxaemia including anaesthesia accidents. In SchadC, J. P., McMenemey, W . H. (Eds.) Selective Vulnerability of the Brain in Hypoxaemia. Philadelphia: F. A. Davis. Marie, J., Lyon. G . , Bargeton. E. (1959) ‘La
351
sclkrose cerebrate centro-lobaire. A propos de I’etude anatamo-clinique d’un cas de diplkgie spasmodique congenitale.’ Presse MPdicale. 60, 2286-2289. Nwaesei, C. G.. Pape, K. E., Martin, D. J., Becker, L. E., Fitz, C. R. (105) ‘Periventricular infarction diagnosed by ultrasound. A postmortem correlation.’ Journal of Pediatrics, 105, 106-1 10. Roback, H. N., Scherer, J. J. (1935) ‘Uber die feinere Morphologie des fruhkindlichen Hirnes unter besonderer Berucksichtigung des Gliaentwicklung.’ Virchow’s Archiv fur Puthologische Anatomie. 294, 365-413. Takashima, S., Tanaka, K. (1978) ‘Development of cerebrovascular architecture and fts relationship to periventricular leukomalacia. Archives of
Neurology, 35, 11-16. Armstrong, D. L., Becker, L. E. (1978) ‘Subcortical leukomalacia: relationship to development of, the cerebral sulcus and its vascular supply. Archives of Neurology, 35, 470-472. Trounce, J. Q.,Fagan, D., Levene, M. I. (1986) ‘Intraventricular haemorrhage and periventricular leukomalacia: ultrasound and autopsy correlation.’ Archives of Disease in Childhood, 61, 1203-1207. Young, R. S. K., Hernandez, M. J., Yagel, S. K. (1982) ‘Selective reduction of blood flow to white matter during hypotension in newborn dogs: a possible mechanism of periventricular leukomalacia.’ Annals of Neurology, 12, 445448.
An extended Family with a Domi nant ly Inherited Speech Disorder
family attended normal schools and had normal speech and language development. The unusual regular transmission of the speech disorder prompts us t o report this family, since it suggests that at least one type of dyspraxia is simply inherted.
J. A . Hurst
M. Baraitser E. Auger F. Graham S. Norell
352
Few speech disorders are inherited in a simple Mendelian way, and when familial clustering does occur, both genetic and environmental factors are involved. The genetic contribution in most cases involves more than one gene (polygenic inheritance), although a dominant gene with reduced penetrance is possible. In practice, it is unusual to find full penetrance of a speech disorder occurring in a regular manner over generations. The family reported here was referred to the genetics clinic from a speech and language unit, where seven affected members were pupils and other children in the same family had previously been pupils. The children had serious communication difficulties and their speech disorder was classified as a severe form of developmental verbal apraxia, since both speech and expressive language were involved. The unaffected members of the
-
Representative case histories CASE 1 R.N. (d.0.b. 27.2.72) is patient 1111 on the pedigree shown in Figure 1. This boy is the oldest of nine children born to unrelated parents. The pregnancy was normal and there were no neonatal problems. He spoke no words at 18 months, and by 2 years 5 months his mother was concerned about his speech delay. Other aspects of his development were normal. From the age of six years he attended a school for children with speech and language difficulties. A report by an educational psychologist at the age of 12 years 9 months concluded that he functioned over a wide range of levels, but most were clustered around his chronological age or slightly higher. At 16 years his height was on the 7Sth percentile, weight was just under the 50th percentile and occipitofrontal circumference (OFC) was 56cm (50th percentile). On examination, there was no facial weakness and there were full movements of the tongue and palate. He had no abnormal neurological signs. His speech was understandable, but not fluent because of searching for words. He used short, simple sentences, which were often telegrammatic. His comprehension of speech and his reading, spelling and mathematical abilities were adequate for most social occasions, but below those expected for his chronological age. Hearing and vision were normal. CASE 2 D.N. (d.0.b. 14.7.83) is patient 1119 on the pedigree (Fig. I). She is the youngest of nine children and was the second of dizygotic twins, with a birthweight of 36308. Apart from mild jaundice, there were no neonatal problems. She sucked and fed well. Leftsided plagiocephaly was noted and kyphoscoliosis developed. X-rays of the spine showed fusion of the 5th to 10th thoracic vertebrae. Early development
CASE REPORTS
J. Rodriguez D. Claw G. Verellen G. Lyon
Studies of correlations between ultrasonic changes and neuropathologica~ findings in periventricular leukomalacia (PVL) are relatively scarce, and frequently are based purely on macroscopic evidence or on fragmentary microscopic studies (Hill et al. 1982, Bowerman et al. 1984, Nwaesei et al. 1984, Fawer et al. 1985, Trounce et al. 1986). We report what we believe to be the first complete neuropathological study of a case of PVL, in which periventricular echo-free cavities discovered in a preterm neonate at 34 weeks postmenstrual age (PMA) were no longer visible at 42 weeks PMA, three weeks before the infant died.
Case report The child was evaluated and treated in the Neonatal Unit at St. Luc Hospital. Ultrasound scans were obtained with a MARK 100 (ATL) real-time multifrequency scan-head (3-5-7-SMHz). Using a transfontanelle approach, scans were taken in the coronal and sagittal planes.
Clinical findings This boy was born at 27 weeks PMA, weighing 930g, after an uneventful pregnancy. At birth he was asphyxiated. with severe acidosis. Apgar scores were I at one minute and 5 at five minutes. He was intubated and ventilated for 19 days because of hyaline membrane disease. There was no evidence of haemodynamic failure during this period. At six weeks of age (33 weeks PMA), prolonged apnoea required re-intubation and assisted ventilation,
which was complicated by bilateral pneumothorax and severe circulatory collapse. Intestinal resection for nccrotizing enterocolitis was performed one week Neurological development was much delayed. The infant was hypotonic, with intermittent hypertonic spasms. He died at 18 weeks of age from an intestinal infection (Bacillus cereus).
Sonographic findings Increased echogenicity in the left periventricular white-matter was observed at three weeks of age (30 weeks PMA). At seven weeks, a string of echo-free cavities appeared in the periventricular hyperechogenic area, the largest measuring 0.9cm at its widest, located at the level of the atrium (point of confluence of the body of the lateral ventricle with the temporal and occipital horns) (Fig. la, b). The echo-free cavities remained stable for at least one month, but were no longer visible at 15 weeks of age, when the periventricular hyperechogenicity had also disappeared. No ventricular dilatation was observed. A CT scan at that time also showed no periventricular cavity (Fig. Ic). Neuroparhological findings The postmortem was done in the Laboratory of Developmental Neuropathology at the Paediatric Neurology Centre. The brain was fixed in 20 per cent formaldehyde. Blocks of brain-tissue were embedded in paraffin and sectioned semi-serially; and the brainstem and cerebellum horizontally. The 10 to ISpm-[hick sections were stained with haematoxylin and eosin and cresyl violet (Nissl) according to Mallory's method. Frozen sections were stained with oil red 0. The brain had a normal weight of 4708. The Ieptomeninges, meningeal vessels and external surface of the brain looked normal. No macroscopic
n
m r-I
4
34 7
f
u
Fig. 2. Left hemisphere: coronal section through the parietal lobe shows slit-like cavity (arrow) in centre of thin glial band extending from corona radiata into the paracentral white-matter (arrowhead). (Cresyl violet x 10.)
Fig. 1. Ultrasound scan at seven weeks of age. (Aj Left sagittal view showing 9mm (greatest diameterj echo-free area (arrow). suggesting cavity within periventricular white-matter at level of atrium. /Bj Additional smaller cavities (arrowheadsj, rostrally disposed. (Cj At I5 weeks all cavities have disappeared.
348
changes were observed in the hemispheric whitematter on coronal section. The cerebellum and brainstem were normal. On microscopic examination, the left hemisphere had an elongated glial scar extending from the deep part of the centrum semi-ovale, near the angle of the lateral ventricle, into the paracentral white-matter (Figs. 2, 4). In some coronal sections it was seen to extend to the core of two adjacent gyri. In a sagittal plane, the glial band was seen from the mid-frontal to the posterior parietal region. Astrocytic nuclei and bundles of glial fibres were parallel to the axis of the gyrus. A few cell fragments were encrusted in calcium. The posterior part of the glial band, at the level of the pulvinar, was excavated by a slit-like cavity
containing a few lipid-laden oil red 0-positive macrophages (Fig. 3). A narrow zone of loose gliosis surrounded the dense glial core. The rest of the whitematter was normal and well developed ‘myelination oligo-dendrocytes’ (Roback and Scherer 1935), with their typical eosinophilic flame-like cytoplasm, were present. No changes were detected in the whitematter of the right hemisphere. In the thalamus bilaterally, in the right globus pallidus and in the left neocortex, small postischaemic foci of neuronal loss, dense gliosis. macrophages and cells encrusted in calcium were observed. The lesion seemed to be contemporary with the white-matter changes.
Discussion Sonographic findings in this neonate are consistent with PVL. A left-periventricular echogenic area became apparent at three weeks of age. In the posterior half of this zone, one large and several small echo-free cavities visualized at seven weeks were no longer detectable by ultrasonography at 15 weeks. The destructive white-matter lesions were more likely to be a result of ischaemia than of haemorrhage, because
u-l m u-l
4
P D
Fig. 3. Left hemisphere: same coronal section at higher magnuication shows excavation of linear scar. Cavity containing lipid-laden macrophages amid dense glial tissue. Glial nuclei and fibres run parallel to gyral axis. Moderate degree of astrocytic proli/eration can be seen in limited area of surrounding white-matter. (Cresyl violet x 300.)
in the latter case some trace of bleeding (for instance haemosiderin in macrophages) should have been found at autopsy. The prolonged hypotension at six weeks may have played an aggravating r81e in the evolution of white-matter lesions and a determinating r81e in the pathogenesis of the focal grey-matter infarcts. Disappearance of periventricular cysts on ultrasound imaging after three weeks of age has been well documented (Bowerman ef al. 1984, Dubowitz ef a/. 1985, Trounce ef al. 1986, de Vries ef al. 1987), but precise correlations of this phenomenon with neuropathological findings-other than macroscopic (Trounce ef al. 1986)-have not been reported to our knowledge. De Reuck ef a/. (1972) mention that the cavity may ultimately collapse, leaving nothing but a periventricular scar, but give no illustration of this occurrence.
Fig. 4. Left hemisphere: coronal section. Glial band, seen here at more rostra1 level, extends into periventricular white-matter (arrow). (Cresyl violet x 75.)
In our case no abnormality could be detected on the formalin-fixed brain, but on microscopic examination a linear glial scar was seen, extending in a frontal plane from the periventricular white-matter (or the centrum semi-ovale) well into the paracentral white-matter, and in the sagittal plane from the anterior to the posterior poles of the thalamus, approximately in the area of extension of the periventricular cavities. In the posterior part of the glial streak, at a point corresponding to the large echofree cyst on ultrasound scanning, there was a slit-like cavity containing a few macrophages. There was no ventricular enlargement, as there was in the cases of De Reuck ef a/. (1 972) and Trounce ef a/. (1986). It seems highly probable that this small cavity in a glial scar was what remained of the large periventricular cyst observed at seven weeks of age, after its collapse.
349
c v) )
i: 2 s u
Otherwise one would have to assume that the cyst seen on ultrasound scans had vanished entirely. The mechanism of disappearance or considerable reduction in size of periventricular cavities is not clear. However, it seems logical to postulate-at least in the present casethat it is the result of a change in the pressure gradient between the cavity and the surrounding brain substance. A decrease in intracavity pressure could be related to resorption of intracystic fluid (or blood) and to a progressive reduction of the mass of macrophages. Pericystic pressure, on the other hand, is bound to increase as a result of rapid brain-growth during late fetal and early postnatal life. This is mainly because of an enormous increase in neuronal volume and by glial proliferation, particularly the multiplication of oligodendrocytes, which precedes myelination in the central whitematter. Between 34 and 42 weeks of gestation, brain weight increases from 290g to 420g (Gruenwald and Minh 1960).
It is interesting that the cortical, thalamic and pallidal ischaemic foci, which might have been of clinical significance, were not detected by ultrasound scanning. This exemplifes the difficulty of making an exact estimation of brain damage in the neonate by ultrasonography . Another point is the significance of the pathological changes found in the left hemisphere. PVL frequently leads to a stellate scar: from a periventricular area of demyelination and gliosis, long linear extensions radiate into the white axis of one or more gyri, as first described by Marie et al. (1959). In our case, the glial scar is quite consistent with PVL, although a strictly
350
periventricular softening was inconspicuous in most areas. We have observed this pattern in brains in which there was typical periventricular softening in adjacent areas. The topography of the streak of isomorphic glial tissue in which astrocytic nuclei and glial fibres are parallel to the gyral axis is not readily accounted for by the usual explanation of border-zone ischaemic softening (Banken and Larroche 1962, Lindenberg 1963, De Reuck 1971, 1984; De Reuck et al. 1972, Armstrong and Norman 1974, Takashima and Tanaka 1978). Nor does it correspond to the relatively avascular subcortical triangle described by Takashima et al. (1978). The remarkable study by Kuban and Gilles (1985) on the development of telencephalic microvasculature casts much doubt on the pathophysiological explanations cited above, and on the reality of an avascular subcortical zone. The characteristic intra-axial linear glial scar, or the much more frequent stellate scar, are more readily explained by the experiments of Young et al. (1982) on newborn dogs with severe hypotension. Those authors found a selective diminution of blood-flow in the whitematter. Interestingly, hypoperfusion was more marked precisely in the periventricular and intra-axial white-matter, which are .the usual sites of brain damage. Accepted for publication 9th May 1989. Authors’ Appointments J. Rodriguez; D. Claus; G. Verellen; *G. Lyon; Service de Neurologie Pkdiatrique, Hapita1 Universitaire St. Luc, UCL, Avenue Hippocrate 10, B-1200 Bruxelles, Belgium. *Correspondence to last author.
SUMMARY Ultrasound scans of a preterm neonate (27 weeks gestation) at seven weeks of age showed periventricular echo-free cavities, but these were no longer visible at 15 weeks, three weeks before the infant died. At autopsy, a linear glial scar, extending from the periventricular white-matter into the white axis of the parasagittal gyrus. was found in the area occupied by the periventricular cysts. The larger cavity was reduced to a slit-like excavation in the midst of glial tissue. Unsuspected focal infarcts in the cerebral cortex were also found. This observation demonstrates that transient echofree cavities represent foci of cystic necrosis, which are subject to secondary collapse. In the authors’ experience, the linear extension of periventricular leukomalacia (PVL) into the core of parasagittal gyri is a frequent feature of PVL, and one which cannot easily be accounted for by the usual explanations of border-zone ischaemic softening.
n m n
RI~SUMJ~ Leucomalacie pdri-ventriculaire: corrdlation entre donndes kchographiques et neuropathologiques Des kchographies ont etk realistes chez un prkmature (27 semaines de gestation) A 1’8ge de sept semaines. Des cavitks pkri-ventriculaires sans k h o s ont etk dkcouvertes, mais n’ktaient plus visible A 15 semaines, trois semaines avant la mort. A l’autopsie, une cicatrice linkaire gliale s’ktendant de la substance blanche pkri-ventriculaire jusqu’ri I’axe blanc du gyrus a etk trouvke dans la region occupee par les cavitks peri-ventriculaires. La plus grande cavite etait rkduite A une fissure au milieu du tissu glial. Un infarcissement local insoupconne fut aussi trouvk dans le cortex ckrkbral. Cette observation dkmontre que les cavitks transitoires sans kchos representent des foyers de nkcroses cystiques sujets A un effondrement secondaire. L’extension linkaire de la leucomalacie pkri-ventriculaire au coeur des gyri para-sagitaux est dans notre experience une caractkristique frkquente de cette condition et elle ne peut repondre facilement ti l’explication habituelle de ramollissement ischkmique en bordure.
p’ I
z
N-
m
Q
ZUSAMMENFASSUNG Periventrikulare Leukomalazie: Korrelationen zwischen Ultraschall und Neuropathologie Bei einem Friihgeborenen (27. Gestationswoche) wurde im Alter von sieben Wochen eine Ultraschalluntersuchung durchgefiihrt. Es wurden periventrikulare echofreie Raume gefunden, die aber im Alter von 15 Wochen-drei Wochen vor dem Tod-nicht mehr nachweisbar waren. Bei der Autopsie fand sich in dem Bereich der periventrikularen Cysten ein lineares Gliaband, das sich von der periventrikularen weifien Substanz bis zur Achse des parasagittalen Gyrus erstreckte. Der grofiere Hohlraum war auf eine schlitzformige Aushohlung im Gliagewebe reduziert. AuRerdem fanden sich unvermutete Infarkte im cerebralen Cortex. Diese Beobachtung zeigt ,daR passagere echofreie Raume Herde cystischer Nekrosen darstellen, die spater kollabieren. Die lineare Ausdehnung der periventrikularen Leukomalazie bis in das Horn des parasagittalen Gyrus ist nach unserer Erfahrung ein typisches Merkmal bei diesem Befund und kann nicht einfach durch ischamische Randzonenerweichung erklart werden. RESUMEN Leucomalacia periventricular: correlaciones ultrasdnicas y neuropatologicas Se realizaron exAmenes ultrasbnicos en recikn nacidos pretkrmino (27 semanas de gestacibn) a las siete semanas de edad. Se decubrieron cavidades periventriculares libres de eco, que ya no eran visibles a las 15 semanas, tres semanas antes de la muerte. En la autopsia se hallb una cicatriz lineal glial. que se extendia desde la substancia blanca periventricular hacia adentro del eje blanco del giro parasagital, en el Area ocupada por el quiste periventricular. La gran cavidad se habia reducido a una excavacibn como una rendija en medio del tejido glial. Tambikn se hallaron infartos focales insospechados en el cbrtex cerebral. Esta observacibn demuestra que las cavidades libres de eco representan focos de necrosis quistica que posteriormente sufriran un colapso. La linea extensa de leucomalacia periventricular dentro de 10s giros parasagitales. es en nuestra experiencia una caracteristica frecuente de esta patologia, que no puede ser fAcilmente explicada por el argument0 usual de un reblandecimiento isquemico de 10s bordes. References Armstrong, D., Norman, M. G. (1974) ‘Periventricular leukomalacia in neonates: complications and sequelae.’ Archives of Disease in Childhood, 49, 361-375. Banken. B., Larroche, J. (1962) ‘Periventricular leucomalacia in infancy.’ Archives of Neurology, 7, 386-410. Bowerman, R. A., Donn, S. M., Di Pietro, M. A,, D’Amato, C. J., Hicks, S. P. (1984) ‘Periventricular leucomalacia in the preterm newborn infant: sonographic and clinical features.’ Radiology, 151, 383-388. De Reuck, J . (1971) ‘The human periventricular arterial blood supply and the anatomy of cerebral infarctions.’ European Neurology, 5, 321-334. - (1984) ‘Cerebral angioarchitecture and perinatal brain lesions in premature and full-term infants.’ Acto Neurologrca Scandinavica, 70, 391-395. - Chattha, A. S., Richardson, E. P. (1972)
‘Pathogenesis and evolution of periventricular leukomalacia in infancy.’ Archives of Neurology, 27, 229-236. De Vries, L. S., Connell, J. A., Dubowitz, L. M. S.. Oozeer. R. C., Dubowitz, V., Pennock, J. M. (1987) ‘Neurological, electrophysiological and MRI abnormalities in infants with extensive cystic leukomalacia.’ Neuropediafrics. 18, 61-66.
Dubowitz, L. M. S., Bydder, G . M., Mushin, J. (1985) ‘Developmentalsequence of periventricular leukomalacia.’ Archives of Disease in Childhood, 60, 349-355. Fawer, C-L.. Calame, A., Perentes, A., Anderegg. A. (1985) ‘Periventricular leukomalacia: a correlation study between real-time ultrasound and autopsy findings.’ Neuroradiology. 27, 292-300.
Gruenwald, P., Minh. H. N. (1960) ‘Evaluation of body and organ weights in perinatal pathology. I. Normal standards derived from autopsies.’ American Journal of Clinical Pathology. 34, 247-253.
Hill, A., Melson, G. L., Clark, H. B. (1982) ‘Hemorrhagic periventricular leukomalacia: diagnosis by real-time ultrasound and correlation with autopsy findings.’ Pediatrics, 69, 144-149. Kuban, K. C. K., Gilles, F. H. (1985) ‘Human Telencephalic angiogenesis.’ Annals of Neurology. 17, 539-548. Lindenberg. R. (1963) ‘Patterns of CNS vulnerability in acute, hypoxaemia including anaesthesia accidents. In SchadC, J. P., McMenemey, W . H. (Eds.) Selective Vulnerability of the Brain in Hypoxaemia. Philadelphia: F. A. Davis. Marie, J., Lyon. G . , Bargeton. E. (1959) ‘La
351
sclkrose cerebrate centro-lobaire. A propos de I’etude anatamo-clinique d’un cas de diplkgie spasmodique congenitale.’ Presse MPdicale. 60, 2286-2289. Nwaesei, C. G.. Pape, K. E., Martin, D. J., Becker, L. E., Fitz, C. R. (105) ‘Periventricular infarction diagnosed by ultrasound. A postmortem correlation.’ Journal of Pediatrics, 105, 106-1 10. Roback, H. N., Scherer, J. J. (1935) ‘Uber die feinere Morphologie des fruhkindlichen Hirnes unter besonderer Berucksichtigung des Gliaentwicklung.’ Virchow’s Archiv fur Puthologische Anatomie. 294, 365-413. Takashima, S., Tanaka, K. (1978) ‘Development of cerebrovascular architecture and fts relationship to periventricular leukomalacia. Archives of
Neurology, 35, 11-16. Armstrong, D. L., Becker, L. E. (1978) ‘Subcortical leukomalacia: relationship to development of, the cerebral sulcus and its vascular supply. Archives of Neurology, 35, 470-472. Trounce, J. Q.,Fagan, D., Levene, M. I. (1986) ‘Intraventricular haemorrhage and periventricular leukomalacia: ultrasound and autopsy correlation.’ Archives of Disease in Childhood, 61, 1203-1207. Young, R. S. K., Hernandez, M. J., Yagel, S. K. (1982) ‘Selective reduction of blood flow to white matter during hypotension in newborn dogs: a possible mechanism of periventricular leukomalacia.’ Annals of Neurology, 12, 445448.
An extended Family with a Domi nant ly Inherited Speech Disorder
family attended normal schools and had normal speech and language development. The unusual regular transmission of the speech disorder prompts us t o report this family, since it suggests that at least one type of dyspraxia is simply inherted.
J. A . Hurst
M. Baraitser E. Auger F. Graham S. Norell
352
Few speech disorders are inherited in a simple Mendelian way, and when familial clustering does occur, both genetic and environmental factors are involved. The genetic contribution in most cases involves more than one gene (polygenic inheritance), although a dominant gene with reduced penetrance is possible. In practice, it is unusual to find full penetrance of a speech disorder occurring in a regular manner over generations. The family reported here was referred to the genetics clinic from a speech and language unit, where seven affected members were pupils and other children in the same family had previously been pupils. The children had serious communication difficulties and their speech disorder was classified as a severe form of developmental verbal apraxia, since both speech and expressive language were involved. The unaffected members of the
-
Representative case histories CASE 1 R.N. (d.0.b. 27.2.72) is patient 1111 on the pedigree shown in Figure 1. This boy is the oldest of nine children born to unrelated parents. The pregnancy was normal and there were no neonatal problems. He spoke no words at 18 months, and by 2 years 5 months his mother was concerned about his speech delay. Other aspects of his development were normal. From the age of six years he attended a school for children with speech and language difficulties. A report by an educational psychologist at the age of 12 years 9 months concluded that he functioned over a wide range of levels, but most were clustered around his chronological age or slightly higher. At 16 years his height was on the 7Sth percentile, weight was just under the 50th percentile and occipitofrontal circumference (OFC) was 56cm (50th percentile). On examination, there was no facial weakness and there were full movements of the tongue and palate. He had no abnormal neurological signs. His speech was understandable, but not fluent because of searching for words. He used short, simple sentences, which were often telegrammatic. His comprehension of speech and his reading, spelling and mathematical abilities were adequate for most social occasions, but below those expected for his chronological age. Hearing and vision were normal. CASE 2 D.N. (d.0.b. 14.7.83) is patient 1119 on the pedigree (Fig. I). She is the youngest of nine children and was the second of dizygotic twins, with a birthweight of 36308. Apart from mild jaundice, there were no neonatal problems. She sucked and fed well. Leftsided plagiocephaly was noted and kyphoscoliosis developed. X-rays of the spine showed fusion of the 5th to 10th thoracic vertebrae. Early development
