136
The Role of Short Latency Somatosensory Evoked Responses in Infants with Rapidly Progressive Ventricular Dilatation* v. Pierratl, T. Minami 1, M. Smet2 and P. Casaer1
1Department of Paediatric
Neurology and 2Department of Paediatric Radiology, Katholieke Universiteit, B-3000 Leuven
Abstract The effeet of hydrocephalus on somatosensory evoked potentials was studied in nine infants. An inerease in N llateney was found in five infants studied longitudinally during aperiod of progressive ventrieular dilatation. A marked decrease in NI lateney was noted in 7 infants, within one week following shunt insertion and in two infants who showed spontaneous arrest of ventrieular growth. A eorrelation was found between eerebrospinal fluid pressure and the delay in NI latency, but the number of infants studied is still small. SEPs appear to be a useful additional test when assessing infants with progressive ventrieular dilatation. Onee a baseline value for NI has been obtained following shunt insertion, SEPs may subsequently be useful when assessing a ehild with possible shunt dysfunetion. Keywords Hydroeephalus, newborn - Intraeranial pressure - Somatosensory evoked responses (SEP) - Progressive ventrieular dilatation
Introduction Evoked potentials are more and more used during the neonataI period and earlyehildhood (1, 4, 7,. 11). In infants with hydrocephalus, both visual as wen as auditory evoked potentials have been studied. Following shunt insertion, a signifieant deerease in lateneies has been noted for visually evoked potentials in both neonates and infants, and for auditory brainstem evoked potentials in older ehildren (2, 8). In premature infants, no eorrelation has been found between eerebrospinal fluid pressure and prolonged interpeak intervals of auditory brainstem responses (10). Most of the studies of somatosensory evoked potentials (SEPs) reported so far have looked at maturation
and eontributed to establish normative data (1, 4,7,11). A limited number of studies have been done to assess whether there is an effeet of pathologieal eonditions on SEPs (3, 6, 9, 13). The aim of the present study was to evaluate the effeet of progressive ventrieular dilatation and drainage of eerebrospinal fluid on SEPs. Patients and methods
Patients Clinieal details of the nine infants studied are summarised in the table.
Posthaemorrhagic
ventricular
dilatation
(PHVD) was present in 6 of the 9 cases. Five were born prematurely with a gestational age between 29 and 33 weeks. Intraventrieular haemorrhage was diagnosed during the first week of life, without involvement of the brain parenehyma in two, with unilateral parenehymal involvement evolving into a poreneephalie eyst in two and assoeiated with eystie periventrieular leukomalaeia in one patient. The sixth infant was born at 38 weeks gestational age. Maternal trauma had oeeurred at 32 weeks and ventrieular dilatation with resolving intraventrieular elots were present at birth. A lumbosacral myelomeningocele (MMC) was present in three full term infants. Rapidly progressive ventrieuJar dilatation oeeurred following surgical elosure of the skin defeet.
Cranial ultrasound illtrasound examinations were performed using a UM 4 maehine with a multifrequeney seanhead (5-7.510 MHz). The diagnosis of PHVD was made when two or more measurements of ventrieular width, on ultrasonography, exeeeded by at least 4 mm the 97th eentiles of the measurements appropriate for gestational age as determined by Levene (12). Infants were seanned twiee a week to follow both spontaneous evolution of the ventrieular dilatation as weIl as the effeet of lumbar punetures. All infants had an ultrasound sean 1 day before and 4 days after insertion of the shunt.
Received August 25, 1989; accepted October 31, 1989 Neuropediatrics 21 (1990) 136-139 © Hippokrates Verlag Stuttgart
* This study is supported by agrant from The Medical Research Council, Belgium (PGWO) and by a grant from the Janssen Research Foundation, Beerse, Belgium.
Downloaded by: University of British Columbia. Copyrighted material.
By L. s. De Vries 1,
Short Latency Somatosensory Evoked Responses in Progressive Ventricular Dilatation msec
N
80
1
All infants had SEPs done one day before and five days after shunt insertion and in five infants at least two recordings were obtained both before as weIl as after shunt insertion. Cerebrospinal fluid pressure was measured in six of the nine infants using a capillary column. In two, by means of a lumbar puncture, and in four by means of a ventricular puncture during anaesthesia. Six mmHg were taken as the upper limit of normal cerebrospinal fluid pressure (5). Depending on the pressure obtained, a Hakim system was inserted with an opening pressure between 60 and 100 mmH 20. The three aspects studied were: the evolution of SEPs in relation to the occurrence of progressive ventricular dilatation (n = 5); the association between SEPs and cerebrospinal fluid pressure (n = 6); the changes in SEPs in response to withdrawal of cerebrospinal fluid by either shunt insertion, or periodic drainage by lumbar puncture (n = 9). Results
120
i i i i
~""'. .: .x
60
·_·-MMC
i
-ti
70
:::::: PHVO
S
Sp
..~
SEPs were done using a Nihon Kohden Neuropack 11 model 5100. Ag-AgC1disk electrodes were applied; the negative electrode on C3', the positive on Fz and the neutral on the lower arm. C3' was marked with permanent ink. The impedance was usually between 2 and 5 and never more than 10 kOhms. Electrical stimulation of 0.1 msec duration was delivered using a hand held device, at a rate of 1 per 2 seconds. Stimulation intensity was that necessary to produce a minimal thumb twitch. As the shunt system was inserted in the right ventricle, all recordings were obtained from the left somatosensory cortex, following stimulation of the right median nerve. A total of 128 responses was averaged through a bandpass of 2-100 Hz. Each trial was repeated two or three times to ensure reproducibility. The neonates were not sedated and the test was done during natural sleep following a feed. The first negative wave (NI), which signals the arrival of the afferent impulse at the cerebral cortex (1, 4), was measured in all infants.
\
S
i~~
'l}\
i
S
i i
i\
............. 40
; \\
"*
5
.
\ \
S 'V
S
~~--~ ~
i
\
-\ !
I \\
•
s\
*~\ \
.).
\
\
I \
\
,,!
i
i
30
\
'\
,~
50
c;\
f!,-L.J--...:b\
\.
I
\\
'\r
\.
\~~
/ \\
i i i
\ \
\\1/ o
20
\ \
\\
, 0-'-0
i
I
34
32
i
3&
I
I
40
38
42
44
4&
wks PMA
Fig. 1 Data of all nine infants one day before and five days after shunt insertion (S) or spontaneous stabilisation (Sp). The normal range and the 95 percent tolerance limits are those according to Klimach and Cooke (7) (with permission).
Six infants had cerebrospinal fluid pressures measured 24 hours of less before shunt insertion; in two by means of a lumbar puncture and in four by means of a ventricular puncture during anaesthesia. The pressures were raised in all infants studied. The direct cerebrospinal fluid pressure measurements were correlated with the decrease in NI latency observed following shunt insertion in five and spontaneous stabilisation in one (p < 0.05) (Fig. 2).
Effect of cerebrospinal fluid drainage on the SEPs In three infants, on five occasions, SEPs were done one hour before one hour after and 24 hours after removal of 5 to 10 mVkg of cerebrospinal fluid by means of a lumbar puncture. No difference was noted between measurements taken one hour before and one hour after the lumbar puncture, and in only one of the three infants a decrease (4 msec) was noted 24 hours after the lumbar puncture.
Changes in cranial ultrasound jindings All infants showed an increase in ventricular width during the period preceeding shunt insertion or spontaneous stabilisation, varying between 3 and 9 mm. Five days following shunt insertion or spontaneous stabilisation a decrease in ventricular width of 2 to 5 mm was noted.
mmHg
24
•
20
•
Effect of progressive ventricular dilatation of theSEPs In four infants with PHVD and in one with MMC, several SEPs were done over aperiod of rapidly progressive ventricular dilatation, varying between 1 and 3 weeks. An increase in NI latency, varying between 4 and 20 msec was noted in these five infants (Fig. 1).
• • ~ 8 z
a:
'" o
•
~
~
:v
4
Effect of cerebrospinal fluid pressure on the SEPs All infants had clinical signs of raised intracranial pressure, i. e. a full fontanelle, suture diastasis and a rapid increase in head circumference.
137
I
90msec
DECREASE IN N1 LATENCY
Fig. 2 Relationship between cerebrospinal fluid pressure and decrease in NI latency noted five days after shunt insertion in five infants and after spontaneous stabilisation in one infant.
Downloaded by: University of British Columbia. Copyrighted material.
SEPs
Neuropediatrics 21 (1990)
138 N europediatrics 21 (1990)
L. S. De Vries et al
VI=22 mm Before shunt
(- lday)
VI=17mm After shunt
N1
26
2. 5., V
....... , --,I
1..........-.._.-..-.......
100 msec
Four of the six infants with PHVD and the three infants with MMC had a ventrieuloperitoneal shunt inserted. A eomparison between the Nllateney 1 day before and 5 days after shunt insertion showed a marked reduetion in NI lateney. In six infants this eoineided with an inerease in amplitude. The deerease in Nllateney was espeeially marked in the infants with MMC (Fig. 3). In five of the nine infants, the SEPs were reeorded on at least two oeeasions following shunt insertion showing a further deerease in NI lateney within the range expeeted on the basis of maturation. Shunt dysfunetion oeeurred in two infants. In one of the infants with MMC (Case 8)
Aetiology hydrocephalus
ICP Gestational Age at shunt age (wks) insertion (wkPMAjmonth)
1. PHVO + PC
29 31 38 33 33 30 38 40
2. 3. 4. 5. 6. 7. 8.
PHVO PHVO PHVO PHVO PHVO MMC MMC
9. MMC
+ PVL
+ PC
40
41 PMA 46 PMA 42.5 PMA 46
PMA
40 43 45 41
PMA PMA PMA PMA
7 mmHg'
7 12 17 >20 11 23
mmHg" mmHg' mmHg" mmHg mmHg" mmHg"
shunt dysfunetion oeeurred 10 days after shunt insertion. The Nllateney, whieh had normalised after insertion ofthe first system, rose sharply to preshunt values. After shunt revision a marked deerease of Nllateney onee again oeeurred with subsequent stabilisation of the NI lateney (Fig. 1). In one of the infants with PHVD (Case 5) shunt dysfunetion oeeurred 10 weeks following shunt insertion. The NI lateney, whieh had been 36 msee at diseharge, was 38 msee on readmission and eame down to 28 msee following revision. In two infants with PHVD shunt insertion was not required as spontaneous stabilisation oeeurred. The fontanelle beeame normotonie, having. been tense previously. In
VI before -after (mm) I 19 27 22 22 29 18 22 18
I 16
26
21
25 19 22 24 14 17 16
Fig. 3 Effect of shunt insertion on somatosensory evoked potentials in an infant with MMC (Case 7). Ultrasonography and SEPs done one day before and five days after shunt insertion, showing a reduction in ventricular width coinciding with a reduction in NI latency.
N 11atency before -after
Increase in amplitude
I
40 48 57 60 44 75 53 47 48 1207
1
33 39 46 38 36 61 26 22 19 34
+ +
+ + + +
ICP = Intracranial pressure; VI = ventricular index; , = pressure measured during lumbar puncture; " = ventricular pressure measured during anaesthesia.
Tab. 1
Downloaded by: University of British Columbia. Copyrighted material.
(+5 days)
Neuropediatrics 21 (1990)
Short Latency Somatosensory Evoked Responses in Progressive Ventricular Dilatation
Discussion The effect of progressive ventricular dilatation on somatosensory evoked potentials was studied in nine infants. In the infants who had sequential SEPs during aperiod of rapidly progressive ventricular dilatation, an increase in NI latency was noted instead of a decrease, as would be expected on the basis of maturation. A marked decrease in Nllatency was seen in seven infants following insertion of a shunt, and in two infants who showed a spontaneous stabilisation of ventricular growth. The decrease coincided with an increase in NI wave amplitude in six infants. Both Hrbek et al (4) and Klimach and Cooke (7) showed that the latency of the NI peak decreases linearly with postmenstrual age, between 28 and 42 weeks gestational age at a rate of about 3 msec/week, and our own crosssectional and longitudinal data are in agreement with these findings. Following the neonatal period, the decrease has been noted to be much slower, but exact normative data during the first months of age are still sparse (11). The actual decrease in NI latency noted after shunt insertion, varied from 7-86 msec, which is at least twice the decrease expected on the basis of maturation. The decrease in NI latency varied considerably from one child to another and was noted to be more pronounced in the infants with MMC, who showed a rapid increase in ventricular size and head circumference following surgical closure of the skin defect. We were able to show a correlation between the cerebrospinal fluid measurements and the decrease in NI latency seen after shunt insertion or spontaneous stabilisation. This relationship should be interpreted with caution as the number of infants who had intracranial pressure measurements is still small and as the cerebrospinal fluid pressure measurements were obtained under different circumstances. Animal studies reported so far show a reduction in amplitude ofthe negative wave rather than an increase in latency following an artificially induced increase in intracranial pressure and ventricular size (14, 15). The changes observed for the SEPs in the infants studied are comparable to changes shown previously by Ekle and Sklar (2). They studied visual evoked responses before and after shunt insertion in a similar population. Lary et al (10), however, were unable to show a correlation between cerebrospinal fluid pressure and I-V interval of the ABR in a group of predominantly premature infants. Most of the cerebrospinal fluid pressures obtained were within the normal range and only four infants eventually required insertion of a shunt. 1t is possible that the somatosensory and visual tracts are more readily affected by an increase in ventricular size due to the proximity of the tracts to the ventricular system. SEPs appear to be a useful additional test when assessing infants with progressive ventricular dilatation. As the
normal range for SEPs in premature infants is wide, only sequential SEPs will provide additional information, when assessing the child with ventricular dilatation. The data obtained from the two infants during two consecutive periods due to shunt dysfunction do imply that once a baseline value has been obtained following shunt insertion, SEPs may subsequently be useful in detecting shunt dysfunction.
References 1
Desmedt,j. E., E. Brunko,j. Debecker: Maturation ofthe somatosensory
evoked potentials in normal infants and children with special reference to the early NI component. Electroencephalogr. Clin. Neurophysiol. 40 (1974) 43-58 2 Ehle, A., F. Sklar: Visual evoked potentials in infants with hydrocephalus. Neurology 29 (1979) 1541-1544 3 Gorke, W.: Somatosensory evoked cortical potentials indicating impaired motor development in infancy. Dev. Med. Child Neurol. 28 (1986) 633641 4 Hrbek, A., P. Kalberg, T. Olsson: Development of visual and somatosensory evoked responses in the preterm newborn infant. Electroencephalogr. Clin. Neurophysiol. 34 (1973) 225-232 5 Kaiser, A. M., A. G. L. Whitelaw: Normal cerebrospinal fluid pressure in the newborn. Neuropediatrics 17 (1986) 100-102 6 Klimaeh, V. j., R. W. 1. Cooke: Short latency somatosensory cortical evoked responses of preterm infants with ultrasound abnormalities. Dev. Med. Child Neurol. 30 (1988) 215-221 7 Klimaeh, V.]., R. W. 1. Cooke: Maturation of the neonatal somatosensory evoked response in preterm infants. Dev. Med. Child Neurol. 30 (1988) 208-214 8 Kraus, N., O. Ozdaman, P. T. Heydermann, L. Stein, N. L. Reed: Auditory brainstem responses in hydrocephalic patients. Electroencephalogr. Clin. Neurophysiol. 59 (1984) 310-31 7 9 Laget, P., R. Salbreux,j. Raimbault, A. M. D'Allest,j. Mariani: Relationship between changes in somesthetic evoked responses and electroencephalographic findings in the child with hemiplegia. Dev. Med. Child Neurol. 18 (1976) 620-631 10 Lary, 5., L. S. De Vries, A. Kaiser, L. M. S. Dubowitz, V. Dubowitz: Auditory brainstem responses in infants with posthaemorrhagic ventricular dilatation. Arch. Dis. Child. 64 (1989) 17-23 11 Laureau, E., A. Majnemer, B. Rosenblatt, P. Riley: A longitudinal study of short latency somatosensory evoked responses in healthy newborns and infants. Electroencephalogr. Clin. Neurophysiol. 71 (1988) 100108 12 Levene, M. 1.: Measurement of the growth of the lateral ventricles in preterm infants with real-time ultrasound. Arch. Dis. Child. 56 (1981) 416-424 13 Majnemer, A., B. Rosenblatt, P. Riley, E. Laureau, A. M. O'Gorman: Somatosensory evoked responses abnormalities in high risk newborns. Pediatr. Neurol. 3 (1987) 350-355 14 Nagao, 5., P. Roccaforte, R. A. Moody: Acute intracranial hypertension and auditory brainstem response. Part 1: Changes in the auditory brainstem and somatosensory evoked responses in intracranial hypertension in cats. J. Neurosurg. 51 (1970) 669-676 15 Sutton, L. N., B. Cho,]. jaggi, P. M. joseph, D. A. Bruce: Effects of hydrocephalus and increased intracranial pressure on auditory and somatosensory evoked responses. Neurosurgery 18 (1986) 756-761
Professor Paul Casaer Paediatric Neurology Section and Developmental Neurology Research Unit Department of Paediatrics and Neonatal Medicine Gasthuisberg B-3000 Leuven
Downloaded by: University of British Columbia. Copyrighted material.
one of the two infants (Case 6) this was noted to coincide with a decrease in cerebrospinal fluid pressure from 12 mmHg to 6 mmHg. A decrease in Nllatency after spontaneous stabilisation was seen in both infants (Fig. 1).
139
The Role of Short Latency Somatosensory Evoked Responses in Infants with Rapidly Progressive Ventricular Dilatation* v. Pierratl, T. Minami 1, M. Smet2 and P. Casaer1
1Department of Paediatric
Neurology and 2Department of Paediatric Radiology, Katholieke Universiteit, B-3000 Leuven
Abstract The effeet of hydrocephalus on somatosensory evoked potentials was studied in nine infants. An inerease in N llateney was found in five infants studied longitudinally during aperiod of progressive ventrieular dilatation. A marked decrease in NI lateney was noted in 7 infants, within one week following shunt insertion and in two infants who showed spontaneous arrest of ventrieular growth. A eorrelation was found between eerebrospinal fluid pressure and the delay in NI latency, but the number of infants studied is still small. SEPs appear to be a useful additional test when assessing infants with progressive ventrieular dilatation. Onee a baseline value for NI has been obtained following shunt insertion, SEPs may subsequently be useful when assessing a ehild with possible shunt dysfunetion. Keywords Hydroeephalus, newborn - Intraeranial pressure - Somatosensory evoked responses (SEP) - Progressive ventrieular dilatation
Introduction Evoked potentials are more and more used during the neonataI period and earlyehildhood (1, 4, 7,. 11). In infants with hydrocephalus, both visual as wen as auditory evoked potentials have been studied. Following shunt insertion, a signifieant deerease in lateneies has been noted for visually evoked potentials in both neonates and infants, and for auditory brainstem evoked potentials in older ehildren (2, 8). In premature infants, no eorrelation has been found between eerebrospinal fluid pressure and prolonged interpeak intervals of auditory brainstem responses (10). Most of the studies of somatosensory evoked potentials (SEPs) reported so far have looked at maturation
and eontributed to establish normative data (1, 4,7,11). A limited number of studies have been done to assess whether there is an effeet of pathologieal eonditions on SEPs (3, 6, 9, 13). The aim of the present study was to evaluate the effeet of progressive ventrieular dilatation and drainage of eerebrospinal fluid on SEPs. Patients and methods
Patients Clinieal details of the nine infants studied are summarised in the table.
Posthaemorrhagic
ventricular
dilatation
(PHVD) was present in 6 of the 9 cases. Five were born prematurely with a gestational age between 29 and 33 weeks. Intraventrieular haemorrhage was diagnosed during the first week of life, without involvement of the brain parenehyma in two, with unilateral parenehymal involvement evolving into a poreneephalie eyst in two and assoeiated with eystie periventrieular leukomalaeia in one patient. The sixth infant was born at 38 weeks gestational age. Maternal trauma had oeeurred at 32 weeks and ventrieular dilatation with resolving intraventrieular elots were present at birth. A lumbosacral myelomeningocele (MMC) was present in three full term infants. Rapidly progressive ventrieuJar dilatation oeeurred following surgical elosure of the skin defeet.
Cranial ultrasound illtrasound examinations were performed using a UM 4 maehine with a multifrequeney seanhead (5-7.510 MHz). The diagnosis of PHVD was made when two or more measurements of ventrieular width, on ultrasonography, exeeeded by at least 4 mm the 97th eentiles of the measurements appropriate for gestational age as determined by Levene (12). Infants were seanned twiee a week to follow both spontaneous evolution of the ventrieular dilatation as weIl as the effeet of lumbar punetures. All infants had an ultrasound sean 1 day before and 4 days after insertion of the shunt.
Received August 25, 1989; accepted October 31, 1989 Neuropediatrics 21 (1990) 136-139 © Hippokrates Verlag Stuttgart
* This study is supported by agrant from The Medical Research Council, Belgium (PGWO) and by a grant from the Janssen Research Foundation, Beerse, Belgium.
Downloaded by: University of British Columbia. Copyrighted material.
By L. s. De Vries 1,
Short Latency Somatosensory Evoked Responses in Progressive Ventricular Dilatation msec
N
80
1
All infants had SEPs done one day before and five days after shunt insertion and in five infants at least two recordings were obtained both before as weIl as after shunt insertion. Cerebrospinal fluid pressure was measured in six of the nine infants using a capillary column. In two, by means of a lumbar puncture, and in four by means of a ventricular puncture during anaesthesia. Six mmHg were taken as the upper limit of normal cerebrospinal fluid pressure (5). Depending on the pressure obtained, a Hakim system was inserted with an opening pressure between 60 and 100 mmH 20. The three aspects studied were: the evolution of SEPs in relation to the occurrence of progressive ventricular dilatation (n = 5); the association between SEPs and cerebrospinal fluid pressure (n = 6); the changes in SEPs in response to withdrawal of cerebrospinal fluid by either shunt insertion, or periodic drainage by lumbar puncture (n = 9). Results
120
i i i i
~""'. .: .x
60
·_·-MMC
i
-ti
70
:::::: PHVO
S
Sp
..~
SEPs were done using a Nihon Kohden Neuropack 11 model 5100. Ag-AgC1disk electrodes were applied; the negative electrode on C3', the positive on Fz and the neutral on the lower arm. C3' was marked with permanent ink. The impedance was usually between 2 and 5 and never more than 10 kOhms. Electrical stimulation of 0.1 msec duration was delivered using a hand held device, at a rate of 1 per 2 seconds. Stimulation intensity was that necessary to produce a minimal thumb twitch. As the shunt system was inserted in the right ventricle, all recordings were obtained from the left somatosensory cortex, following stimulation of the right median nerve. A total of 128 responses was averaged through a bandpass of 2-100 Hz. Each trial was repeated two or three times to ensure reproducibility. The neonates were not sedated and the test was done during natural sleep following a feed. The first negative wave (NI), which signals the arrival of the afferent impulse at the cerebral cortex (1, 4), was measured in all infants.
\
S
i~~
'l}\
i
S
i i
i\
............. 40
; \\
"*
5
.
\ \
S 'V
S
~~--~ ~
i
\
-\ !
I \\
•
s\
*~\ \
.).
\
\
I \
\
,,!
i
i
30
\
'\
,~
50
c;\
f!,-L.J--...:b\
\.
I
\\
'\r
\.
\~~
/ \\
i i i
\ \
\\1/ o
20
\ \
\\
, 0-'-0
i
I
34
32
i
3&
I
I
40
38
42
44
4&
wks PMA
Fig. 1 Data of all nine infants one day before and five days after shunt insertion (S) or spontaneous stabilisation (Sp). The normal range and the 95 percent tolerance limits are those according to Klimach and Cooke (7) (with permission).
Six infants had cerebrospinal fluid pressures measured 24 hours of less before shunt insertion; in two by means of a lumbar puncture and in four by means of a ventricular puncture during anaesthesia. The pressures were raised in all infants studied. The direct cerebrospinal fluid pressure measurements were correlated with the decrease in NI latency observed following shunt insertion in five and spontaneous stabilisation in one (p < 0.05) (Fig. 2).
Effect of cerebrospinal fluid drainage on the SEPs In three infants, on five occasions, SEPs were done one hour before one hour after and 24 hours after removal of 5 to 10 mVkg of cerebrospinal fluid by means of a lumbar puncture. No difference was noted between measurements taken one hour before and one hour after the lumbar puncture, and in only one of the three infants a decrease (4 msec) was noted 24 hours after the lumbar puncture.
Changes in cranial ultrasound jindings All infants showed an increase in ventricular width during the period preceeding shunt insertion or spontaneous stabilisation, varying between 3 and 9 mm. Five days following shunt insertion or spontaneous stabilisation a decrease in ventricular width of 2 to 5 mm was noted.
mmHg
24
•
20
•
Effect of progressive ventricular dilatation of theSEPs In four infants with PHVD and in one with MMC, several SEPs were done over aperiod of rapidly progressive ventricular dilatation, varying between 1 and 3 weeks. An increase in NI latency, varying between 4 and 20 msec was noted in these five infants (Fig. 1).
• • ~ 8 z
a:
'" o
•
~
~
:v
4
Effect of cerebrospinal fluid pressure on the SEPs All infants had clinical signs of raised intracranial pressure, i. e. a full fontanelle, suture diastasis and a rapid increase in head circumference.
137
I
90msec
DECREASE IN N1 LATENCY
Fig. 2 Relationship between cerebrospinal fluid pressure and decrease in NI latency noted five days after shunt insertion in five infants and after spontaneous stabilisation in one infant.
Downloaded by: University of British Columbia. Copyrighted material.
SEPs
Neuropediatrics 21 (1990)
138 N europediatrics 21 (1990)
L. S. De Vries et al
VI=22 mm Before shunt
(- lday)
VI=17mm After shunt
N1
26
2. 5., V
....... , --,I
1..........-.._.-..-.......
100 msec
Four of the six infants with PHVD and the three infants with MMC had a ventrieuloperitoneal shunt inserted. A eomparison between the Nllateney 1 day before and 5 days after shunt insertion showed a marked reduetion in NI lateney. In six infants this eoineided with an inerease in amplitude. The deerease in Nllateney was espeeially marked in the infants with MMC (Fig. 3). In five of the nine infants, the SEPs were reeorded on at least two oeeasions following shunt insertion showing a further deerease in NI lateney within the range expeeted on the basis of maturation. Shunt dysfunetion oeeurred in two infants. In one of the infants with MMC (Case 8)
Aetiology hydrocephalus
ICP Gestational Age at shunt age (wks) insertion (wkPMAjmonth)
1. PHVO + PC
29 31 38 33 33 30 38 40
2. 3. 4. 5. 6. 7. 8.
PHVO PHVO PHVO PHVO PHVO MMC MMC
9. MMC
+ PVL
+ PC
40
41 PMA 46 PMA 42.5 PMA 46
PMA
40 43 45 41
PMA PMA PMA PMA
7 mmHg'
7 12 17 >20 11 23
mmHg" mmHg' mmHg" mmHg mmHg" mmHg"
shunt dysfunetion oeeurred 10 days after shunt insertion. The Nllateney, whieh had normalised after insertion ofthe first system, rose sharply to preshunt values. After shunt revision a marked deerease of Nllateney onee again oeeurred with subsequent stabilisation of the NI lateney (Fig. 1). In one of the infants with PHVD (Case 5) shunt dysfunetion oeeurred 10 weeks following shunt insertion. The NI lateney, whieh had been 36 msee at diseharge, was 38 msee on readmission and eame down to 28 msee following revision. In two infants with PHVD shunt insertion was not required as spontaneous stabilisation oeeurred. The fontanelle beeame normotonie, having. been tense previously. In
VI before -after (mm) I 19 27 22 22 29 18 22 18
I 16
26
21
25 19 22 24 14 17 16
Fig. 3 Effect of shunt insertion on somatosensory evoked potentials in an infant with MMC (Case 7). Ultrasonography and SEPs done one day before and five days after shunt insertion, showing a reduction in ventricular width coinciding with a reduction in NI latency.
N 11atency before -after
Increase in amplitude
I
40 48 57 60 44 75 53 47 48 1207
1
33 39 46 38 36 61 26 22 19 34
+ +
+ + + +
ICP = Intracranial pressure; VI = ventricular index; , = pressure measured during lumbar puncture; " = ventricular pressure measured during anaesthesia.
Tab. 1
Downloaded by: University of British Columbia. Copyrighted material.
(+5 days)
Neuropediatrics 21 (1990)
Short Latency Somatosensory Evoked Responses in Progressive Ventricular Dilatation
Discussion The effect of progressive ventricular dilatation on somatosensory evoked potentials was studied in nine infants. In the infants who had sequential SEPs during aperiod of rapidly progressive ventricular dilatation, an increase in NI latency was noted instead of a decrease, as would be expected on the basis of maturation. A marked decrease in Nllatency was seen in seven infants following insertion of a shunt, and in two infants who showed a spontaneous stabilisation of ventricular growth. The decrease coincided with an increase in NI wave amplitude in six infants. Both Hrbek et al (4) and Klimach and Cooke (7) showed that the latency of the NI peak decreases linearly with postmenstrual age, between 28 and 42 weeks gestational age at a rate of about 3 msec/week, and our own crosssectional and longitudinal data are in agreement with these findings. Following the neonatal period, the decrease has been noted to be much slower, but exact normative data during the first months of age are still sparse (11). The actual decrease in NI latency noted after shunt insertion, varied from 7-86 msec, which is at least twice the decrease expected on the basis of maturation. The decrease in NI latency varied considerably from one child to another and was noted to be more pronounced in the infants with MMC, who showed a rapid increase in ventricular size and head circumference following surgical closure of the skin defect. We were able to show a correlation between the cerebrospinal fluid measurements and the decrease in NI latency seen after shunt insertion or spontaneous stabilisation. This relationship should be interpreted with caution as the number of infants who had intracranial pressure measurements is still small and as the cerebrospinal fluid pressure measurements were obtained under different circumstances. Animal studies reported so far show a reduction in amplitude ofthe negative wave rather than an increase in latency following an artificially induced increase in intracranial pressure and ventricular size (14, 15). The changes observed for the SEPs in the infants studied are comparable to changes shown previously by Ekle and Sklar (2). They studied visual evoked responses before and after shunt insertion in a similar population. Lary et al (10), however, were unable to show a correlation between cerebrospinal fluid pressure and I-V interval of the ABR in a group of predominantly premature infants. Most of the cerebrospinal fluid pressures obtained were within the normal range and only four infants eventually required insertion of a shunt. 1t is possible that the somatosensory and visual tracts are more readily affected by an increase in ventricular size due to the proximity of the tracts to the ventricular system. SEPs appear to be a useful additional test when assessing infants with progressive ventricular dilatation. As the
normal range for SEPs in premature infants is wide, only sequential SEPs will provide additional information, when assessing the child with ventricular dilatation. The data obtained from the two infants during two consecutive periods due to shunt dysfunction do imply that once a baseline value has been obtained following shunt insertion, SEPs may subsequently be useful in detecting shunt dysfunction.
References 1
Desmedt,j. E., E. Brunko,j. Debecker: Maturation ofthe somatosensory
evoked potentials in normal infants and children with special reference to the early NI component. Electroencephalogr. Clin. Neurophysiol. 40 (1974) 43-58 2 Ehle, A., F. Sklar: Visual evoked potentials in infants with hydrocephalus. Neurology 29 (1979) 1541-1544 3 Gorke, W.: Somatosensory evoked cortical potentials indicating impaired motor development in infancy. Dev. Med. Child Neurol. 28 (1986) 633641 4 Hrbek, A., P. Kalberg, T. Olsson: Development of visual and somatosensory evoked responses in the preterm newborn infant. Electroencephalogr. Clin. Neurophysiol. 34 (1973) 225-232 5 Kaiser, A. M., A. G. L. Whitelaw: Normal cerebrospinal fluid pressure in the newborn. Neuropediatrics 17 (1986) 100-102 6 Klimaeh, V. j., R. W. 1. Cooke: Short latency somatosensory cortical evoked responses of preterm infants with ultrasound abnormalities. Dev. Med. Child Neurol. 30 (1988) 215-221 7 Klimaeh, V.]., R. W. 1. Cooke: Maturation of the neonatal somatosensory evoked response in preterm infants. Dev. Med. Child Neurol. 30 (1988) 208-214 8 Kraus, N., O. Ozdaman, P. T. Heydermann, L. Stein, N. L. Reed: Auditory brainstem responses in hydrocephalic patients. Electroencephalogr. Clin. Neurophysiol. 59 (1984) 310-31 7 9 Laget, P., R. Salbreux,j. Raimbault, A. M. D'Allest,j. Mariani: Relationship between changes in somesthetic evoked responses and electroencephalographic findings in the child with hemiplegia. Dev. Med. Child Neurol. 18 (1976) 620-631 10 Lary, 5., L. S. De Vries, A. Kaiser, L. M. S. Dubowitz, V. Dubowitz: Auditory brainstem responses in infants with posthaemorrhagic ventricular dilatation. Arch. Dis. Child. 64 (1989) 17-23 11 Laureau, E., A. Majnemer, B. Rosenblatt, P. Riley: A longitudinal study of short latency somatosensory evoked responses in healthy newborns and infants. Electroencephalogr. Clin. Neurophysiol. 71 (1988) 100108 12 Levene, M. 1.: Measurement of the growth of the lateral ventricles in preterm infants with real-time ultrasound. Arch. Dis. Child. 56 (1981) 416-424 13 Majnemer, A., B. Rosenblatt, P. Riley, E. Laureau, A. M. O'Gorman: Somatosensory evoked responses abnormalities in high risk newborns. Pediatr. Neurol. 3 (1987) 350-355 14 Nagao, 5., P. Roccaforte, R. A. Moody: Acute intracranial hypertension and auditory brainstem response. Part 1: Changes in the auditory brainstem and somatosensory evoked responses in intracranial hypertension in cats. J. Neurosurg. 51 (1970) 669-676 15 Sutton, L. N., B. Cho,]. jaggi, P. M. joseph, D. A. Bruce: Effects of hydrocephalus and increased intracranial pressure on auditory and somatosensory evoked responses. Neurosurgery 18 (1986) 756-761
Professor Paul Casaer Paediatric Neurology Section and Developmental Neurology Research Unit Department of Paediatrics and Neonatal Medicine Gasthuisberg B-3000 Leuven
Downloaded by: University of British Columbia. Copyrighted material.
one of the two infants (Case 6) this was noted to coincide with a decrease in cerebrospinal fluid pressure from 12 mmHg to 6 mmHg. A decrease in Nllatency after spontaneous stabilisation was seen in both infants (Fig. 1).
139
