Archives of Disease in Childhood, 1986, 61, 1096-1099

Sonographic finding of the bright thalamusE Y SHEN, C C HUANG, S C CHYOU, H Y HUNG, C H HSU, AND F Y HUANG

Department of Pediatrics, Mackay Memorial Hospital, Taipei, Taiwan

SUMMARY Six of 83 asphyxiated neonates showed a diffuse increase of echodensity in bilateralthalami with or without other lesions in the basal ganglia on sonographic examination. Thethalamic image still had a fairly high echogenecity compared with the surrounding brainparenchyma on follow up examination and was hence termed the bright thalamus. These sixpatients had a poor neurological outcome, including psychomotor retardation, spastic diplegia orquadriplegia, microcephaly, failure to thrive, seizures, and one death. The bright thalamus is notan occasional or an isolated sonographic finding of hypoxic-ischaemic encephalopathy, but it mayserve as a landmark of severe hypoxic brain damage with adverse outcome.

Neonatal asphyxia is one of the main problemsoccurring in neonatal intensive care units. Braininsult during the neonatal period accounts for theadverse neurological outcome in many handicappedchildren with non-progressive neurological deficits.'The introduction of sector sonography providesan efficient, non-invasive diagnostic technique toevaluate these asphyxiated neonates at the bedside,even if the subject is being mechanically ventilated.The sonographic findings of post-asphyxial encepha-lopathy have been described by several authors.2 4An additional finding of hyperechodense bilateralthalami has been reported by one of us.5 Thisfinding had been termed the 'bright thalamus'. Toclarify this finding and its prognostic value, thesonographic features of six cases with bright thala-mus are presented and their clinical features as wellas short term outcome are described.

Patients and methods

Eighty three patients with fetal distress or perinatalasphyxia were admitted to the neonatal intensivecare units of our hospital from October 1983 toSeptember 1985. All of them received sonographicexamination of the brain either once or sequentiallyover an indefinite period. Perinatal asphyxia wasdefined according to the criteria of Mulliga et al andFiner et al.6 7 Timing of the examination wasdecided by the attending physician according to theclinical state of the patient. The sonographic scan-ning was performed by the physician subspecialisedin paediatric neurology and brain sonography. Since

the first report of high amplitude echogenecity ofbilateral thalami in our unit,5 we have paid muchmore attention to the thalamic image obtained bysonography of asphyxiated neonates. Thus we havecollected six cases with this finding in the past twoyears. We present here the study of these cases withtheir clinical manifestations, sonographic findings,results of computed tomography, and short termoutcome.

Brain sonography was performed with the AlokaSSD-720 5-0 MHz sector scanner in both coronaland sagittal plane through the anterior fontanelleapproach.Computed tomograms were obtained in five

cases, using a Toshiba TCT 80A and scan times of2-7 seconds.

Results

The summary of clinical information and computedtomogram and sonographic findings of the six caseswith bright thalamus are presented in the Table. Allof the subjects were born at term with profoundperinatal asphyxia. The delay of initial crying rangedfrom 10 to 20 minutes, with a mean of 16-3 minutes.

All six cases showed clinical features of hypoxic-ischaemic encephalopathy with seizures, disturb-ance of level of consciousness, and hypotonia orhypertonia during the neonatal period.

Timing of detection of the bright thalamus variedfrom 4 days to 4 months. One case, who had hadserial scanning follow up since birth, had a brightthalamus image detected at 89 hours of age.

1096

copyright. on January 3, 2022 by guest. P

rotected byhttp://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.61.11.1096 on 1 N

ovember 1986. D

ownloaded from

Sonographic finding of the bright thalamus 1097

Table 1 Summary of clinical, sonographic, and computed tomogramfindings ofsix cases with the bright thalamus

Case Sex Birth Gesta- Clinical Sonographicfindings Computed tomogram Outcome (age)No weight (g) tional age presentation (age atfindings) findings

(weeks) (age atfindings)

1 M 3400 40 DOIC 15 min; seizure; High echogenetic Dilatation of lateral Psychomotor retardation;hypotonia; SIADH thalami; basal ventricles and basal microcephaly; spastic

ganglia (15d; 3-5 m) cisterna; isodense diplegia (13 m)thalami (25 d)

2 F 2350 40 Fetal distress; meconium Ventricular dilatation; Ventricular dilatation; Seizure; spastic quadriplegia;stain; DOIC 20 min; high echogenetic slightly hyperdense microcephaly (18 m)hypertonia; opisthotonos thalami (4 m; 5-5 m; 75 m) thalami (5-5 m)

3 F 3400 42 Meconium aspiration; High echogenetic thalami; Brain oedema; slightly Microcephaly; infantileDOIC 15 min; seizure; basal ganglia (3 d; 7 m) hypodense thalami (7 d) spasm (10 m)hypertonia

4 M 3900 40 DOIC 15 min; seizure; Brain oedema; high Died (8 d)hypertonia; acute echogenetic thalamirenal failure (89 h, 1 h after death)

5 F 1960 38 DOIC 15 min; seizure; High echogenetic Multiple infarction, Failure to thrive;hypotonia thalami; basal involving basal psychomotor retardation;

ganglia (3 m; 6 m) ganglia, thalami, and infantile spasm (8 m)periventricular whitematter (6 m)

6 F 3500 41 Fetal distress, meconium High echogenetic Periventricular Psychomotor retardation;stain; DOIC 18 min; thalami; leucomalacia; microcephaly; spasticseizure; hypotonia; periventricular slightly hyperdense quadriplegia; infantileSIADH hyperechodense (7 d; 1 m) thalami (1 m) spasm (4 m; 8 m)

High echogenetic thalami;multicystic leucomalacia(3 m; 4 m)

DOIC=Delay of initial crying; SIADH=syndrome of inappropriate antidiuretic hormone secretion.

He died at the age of 8 days. The sonography wasperformed again one hour after death and showedmild obliteration of the hyperechoes. Unfortu-nately, we failed to obtain permission for autopsyfrom the parents. The other five cases receivedsonography follow up and showed persistence of

hyperdense thalamus at the age of 3, 3, 31/2, 6, and 7months, respectively (Fig. 1).Computed tomography was performed in the five

remaining cases within 10 days after sonographicexamination and all showed isodense or slightlyhyper- or hypodense thalami (Fig. 2).

Fig. 1 The bright thalamus. (a) Increased echogenecity ofthalamus (A) associated with multicystic periventricularleucomalacia (B). (b) Diffuse echogenecity ofthalamus and caudate nucleus (arrow).

copyright. on January 3, 2022 by guest. P

rotected byhttp://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.61.11.1096 on 1 N

ovember 1986. D

ownloaded from

1098 Shen, Huang, Chyou, Hung, Hsu, and Huang

Fig. 2 Case 2. (a) High echogenecity of the bright thalamus in the angle sagittal plane at the age of5½1 months (arrow).(b) Computed tomogram performed at the same time, showing slight hyperdensity of the thalami (arrow).

Discussion

In normal sonography of the neonatal brain thethalamus appears as areas of medium levelechogenecity.=-° The present study showed a diffuseincrease of echogenecity over bilateral thalami inseverely asphyxiated infants, and this image per-sisted on follow up sonography if the patient survived.We believe this may be a pathological featureof hypoxic-ischaemic encephalopathy because it wasnot found in our previous experience when examin-ing 500 normal Chinese neonates,"1 210 prematureChinese neonates (when attempting to determinethe incidence of subependymal/intraventricularhaemorrhage in premature Chinese infants. Unpub-lished data.), and more than 1200 scans of sickinfants who were not asphyxiated.

Previous investigators have succeeded in detect-ing parenchymal echogenecity in the neonatal brainwith post-asphyxial encephalopathy.2-4 They sug-gested that the hyperechogenecity was presumablydue to widespread ischaemic change, cerebraloedema, or focal ischaemia in the distribution of oneof the major cerebral vessels.4

Skeffington et al reported 13 cases of perinatalasphyxia with diffuse high amplitude echogenecitythroughout the brain and suggested cerebraloedema. 12 Others have described an echogenicappearance that corresponded to brain oedemaduring an intraoperative procedure.13 14

In the present six cases two who had been given asonographic examination in the early neonatal

period showed diffuse increase of echogenecity inwhole brain parenchyma and obscured the echoes ofnormal structures as described by Skeffington et al. 12The high amplitude echoes attenuated withtime, but the thalamic echoes gradually becamemore obvious. The computed tomograms remainedisodense or slightly hypodense compared with otherbrain parenchyma (Fig. 2). They became slightlyhyperdense if the scanning was performed in lateinfancy.

Primary thalamic haemorrhage is a rare findingduring the neonatal period.'1-7 It may mimic thebright thalamus sonographically. An early com-puted tomogram may differentiate the bright thala-mus from the thalamic haemorrhage by the higherresolution density on the scan of the haemorrhage.From the sonographic point of view, the brightthalamus showed a diffuse homogenous echogenec-ity all over the thalamus and the margin was sharpand regular. A previous report of acute haemor-rhagic necrosis of the thalamus and basal gangliawith neuropathological confirmation showed thestriking feature of haemorrhage with strong echo-genecity and irregularity of the echoic margin.'5 Theprimary thalamic haemorrhage reported by Trounceet al also showed heterogenous echoes of thehaemorrhage.16 The bright thalamus tends to be adiffuse parenchymal change of the thalamus ratherthan a focal destructive lesion.

Hypoxic-ischaemic insults may cause diffuse braindamage, in particular to the cerebral cortex, subcor-tical and periventricular white matter, basal ganglia,

copyright. on January 3, 2022 by guest. P

rotected byhttp://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.61.11.1096 on 1 N

ovember 1986. D

ownloaded from

Sonographic finding of the bright thalamus 1099

brain stem, and cerebellum.'819 The thalamus isone of the nervous tissues most vulnerable tohypoxia. Sensory nuclei, including the thalamus,lateral geniculate body, and colliculi, are the subcor-tical structures that bear the brunt of asphyxic braindamage because of their high metabolic require-ment.18 Rorke suggested that selective damage tothalamus or deep grey nuclei, or both, probablyresults from diminution of blood flow in branches ofbasal penetrating vessels or from the phenomenonof selective neuronal vulnerability.19

Sonographically, the bright thalamus was just onefeature of hypoxic-ischaemic encephalopathy inour cases. Increased echoes of the basal ganglia(cases 1, 3, and 5) and parenchymal change, such asmulticystic periventricular leucomalacia (case6), were also noted. Whether there is any lesionin the brain stem or cerebellum is difficult toascertain due to the limited resolution of themachine. The outstanding thalamic echogenecity is,however, much more easily identified than others.Thus it may represent a critical sign of hypoxic braininsult.The neuropathology of the persistence of the

bright thalamus in late infancy remains obscure. Wepresume this could be a feature of status marmor-atus and look forward to a pathological confirma-tion. This disorder is more common in term than inpremature infants, with premature birth being notedin fewer than 5% of cases.21 21 The neurologicalmanifestations of status marmoratus in the neonatalperiod are unknown.' A further study of thecorrelation between the bright thalamus and statusmarmoratus will be helpful in identifying thisdisorder early.

Five of our cases survived and all showed pro-nounced psychomotor retardation and other com-plications, such as microcephaly, seizures, spasticdiplegia or quadriplegia, and failure to thrive. Thethalamic damage per se may not account for all ofthese complications. From the present study, wefound that only 7% (6/83) of the asphyxiatedneonates had such complications, and all of themhad longer duration of asphyxia than others. Thissuggests that the bright thalamus is only found inprofoundly asphyxiated neonates.

In summary, the bright thalamus should bedistinguished from thalamic haemorrhage or otherthalamic desease. It is not an occasional or anisolated finding in hypoxic-ischaemic encephalo-pathy. The association of increased echogenecity ofthe basal ganglia or other parts of the brain is notrare. But because of its easy identification, wesuggest that the bright thalamus may serve as alandmark of profound asphyxic brain damage withadverse neurological outcome.

We thank the sisters and nursing staff of the neonatal intensive careunits for their help and cooperation and Miss H J Chang fortechnical help.

References

Volpe JJ. Neurology of the newborn. Philadelphia:W B Saunders, 1981.

2 lill A, Martin DJ, Daneman A, Fitz CR. Focal ischemiccerebral injury in the newborn: diagnosis by ultrasound andcorrelation with computed tomographic scan. Pediatrics1983;71:790-3.

3 Babcock DS, Ball W. Post asphyxial encephalopathy in full-terminfants: ultrasound diagnosis. Radiology 1983;148:417-23.

4 Siegel MJ, Shackelford GD, Perlman JM, Fulling KH. Hypoxic-ischemic encephalopathy in term infants: diagnosis and prog-nosis evaluated by ultrasound. Radiology 1984;152:395-9.

5 Shen EY. The 'bright thalamus'. Arch Dis Child 1984;59:695.6 Mulliga JC, Painter MJ, O'Donoghue PA, MacDonald HM,

Allen AC, Taylor PM. Neonatal asphyxia. II. Neonatalmortality and long-term sequela. J Pediatr 1980;96:903-7.

7 Finer NN, Robertson CM, Richards RT, Pinnen LE, Peters KL.Hypoxic-ischemic encephalopathy in term neonates: perinatalfactors and outcome. J Pediatr 1981:98:112-7.Kossoff G, Garrett WJ, Radovanovich G. Ultrasonic atlas ofnormal brain of infant. Ultrasound Med Biol 1974;1:259-66.

9 Pigadas A, Thompson JR, Grube GL. Normal infant brainanatomy: correlated real-time sonograms and brain specimens.AJR 1979;137:815-20.Shuman WP. Rogers JV, Mack LA, Alvord EC, Christie DP.Real-time sonographic sector scanning of the neonatal cranium:technique and normal anatomy. AJR 1981;137:821-8.Shen EY, Huang FY. Subependymal cysts in normal neonates.Arch Dis Child 1985;60:1072-4.

12 Skeffington FS, Pearse RG. The 'bright brain'. Arch Dis Child1983;58:509-1 1.

'3 Smith SJ, Vogelzang RL, Marzano Ml, Cerullo LJ, Gore RM,Neiman HL. Brain edema: ultrasound examination. Radiology1985;155:379-82.

14 Williams JL. Intracranial vascular pulsations in pediatric neuro-sonology. Journal of Ultrasound Medicine 1983;2:485-8.

s5 Kreusscr KL, Schmidt RE, Shackelford GP, Volpe JJ. Value ofultrasound for identification of acute hemorrhagic necrosis ofthalamus and basal ganglia in an asphyxiated term infant. AnnNeurol 1984;16:361-3.

'6 Trounce J1, Fawer CL, Punt J, Dodd KL, Fielder AR,Levene MI. Primary thalamic haemorrhage in the newborn: anew clinical entity. Lancet 1985;i:190-2.

17 Kotagal S, Toce SS, Kotagal P, Archer CR. Symmetricbithalamic and striatal hemorrhage following perinatal hypoxiain a term infant. J Comput Assist Tomogr 1983;7:353-5.

18 Vannucci RC, Voorhies TM. Perinatal cerebral hypoxia-ischemia: pathogenesis and neuropathology. In: Sarnat HB, ed.Topics in neonatal neurology. Orlando: Grune and Stratton,1984:27-59.

19 Rorke LB. Pathology of perintatal brain injury. New York:Raven Press, 1982.

21) Friede RL. Developmental neuropathology. New York:Springer-Verlag, 1975:64-75.

21 Malamud N. Status marmoratus: a form of cerebral palsyfollowing either birth injury or inflammation of the centralnervous system. J Pediatr 1950;37:610-9.

Correspondence to Dr E Y Shen, Mackay Memorial Hospital, 92,Scction 2, Chung-San N Road, Taipei, Taiwan, Republic ofChina.

Received I July 1986

copyright. on January 3, 2022 by guest. P

rotected byhttp://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.61.11.1096 on 1 N

ovember 1986. D

ownloaded from