involve partial trisomy [4], deletion with pericentric inversion [5], reciprocal translocation [6], and ring chromosome [7, 8]. The incidence of epileptic seizures in
patients with ring 14 chromosome is reported to be 100%
[3]. Seizures in patients with chromosomal abnormalities
tend to be considered as primary generalized ones, and
the seizures described in the literature are mainly of the
primary generalized type including the ones in patients
with ring 14 chromosome [7,9]. However, Lippe and
Sparkes noted that seizures might also be associated with
focal cerebral atrophy in ring 14 chromosome patients
[8] ; in addition, some patients actually exhibited partial
onset seizures [8, 10] and unbalanced developmental subnormality [7,9] indicating the presence of a focal disturbance in the CNS. Thus, in ring 14 chromosome patients,
focal CNS disturbances can be present.
In our patient, complex partial seizures were diagnosed
on the basis of EEG findings and clinical manifestations,
and the CPS were considered to be a symptom of the
focal lesions in the CNS. There remains the possibility
that her seizures were caused not only by genetic factors,
but by some disturbances during pregnancy, however, it is
likely that the genetic defect does enhance the appearance
of CPS. Reviewing the reports of 25 patients with ring 14
chromosome, our case report is the first one involving a
patient with ring 14 chromosome and CPS and shows that
focal CNS lesions can also occur in ring 14 chromosome
patients.

REFERENCES
1. Doose H, Baier W. Genetic aspects of chidhood epilepsy.
Clev Clin J Med 1989;56(suppl1):S105-1O.
2. Delgado-Escueta AV, Greenberg D. The search for epilepsies
ideal for clinical and molecular genetic studies. Ann Neural
1984 ;16(suppl):Sl-l1.
3. Ieshima A, Takeshita K. Chromosome abnormalities and epileptic seizures. Jpn J Human Genet (Tokyo) 1988 ;33:49-60.
4. Smith A, Dulk GD, Elliott G. A severely retarded 18-yearold boy with tertiary partial trisomy 14. J Med Genet 1980;
17:230-2.
5. Nielsen J, Homma A, Rasmussen K, Ried E, Sorensen K,
Saldana-Garcia P. Deletion 14q and pericentric inversion 14.
J Med Genet 1978; 15: 236-8.
6. Hattori H, Hayashi K, Okuno T, et al. De novo reciprocal
translocation t(6;14) (q27;q13,3) in a child with infantile
spasms. Epilepsia 1985;26:310-3.
7. Howard PJ, Clark D, Dearlove 1. Retinal/macular pigmentation in conjunction with ring 14 chromsome. Hum Genet
1988;80:140-2.
8. Lippe BM, Sparkes RS. Ring 14 chromosome: association
with seizures. Am J Med Genet 1981 ;9:301-5.
9. Abe T, Misawa S, Nishioka K, Okuno T, Nakagome Y. Formation of a ring chromosome 14 subsequent to the de novo
13/14 reciprocal translocation: a new cytogenetic evidence
obtained by the nucleolus-organizer staining. Ann Genet
1978;21: 109-12.
10. Amarose AP, Dorus E, Huttenlocher PR, Csaszar S. A ring
14 chromsome with deleted short arm. Hum Genet 1980;
54:145-7.

Cortical Reflex Myoclonus Associated with
Mitochondrial Myopathy~ Encephalopathy~ Lactic Acidosis
and Stroke-like Episodes (MELAS):
A Case Report
Shinji Saitoh, MD, Shinobu Kohsaka, MD, Susumu Mizukami, MD,
and Naofumi Kajii, MD

A 9-year-old female MELAS patient with myoclonus is reported, with emphasis on the results of electrophysiological
studies of the myoclonus. At age 5 years she experienced a stroke-like episode, and a diagnosis of MELAS was made at
age 6 years on the basis of muscle biopsy findings. At age 9 years spontaneous and segmental myoclonus, predominantly
affecting the upper extremities, developed because of complications. Electrophysiological examination, including of
somatosensory-evoked potentials (SEPs) and averaged EMG for long loop reflexes, revealed the so-called "giant SEP" and
enhanced long loop reflexes reflecting cortical hyperexcitability. Jerk-locked averaging yielded no myoclonus related

260 Brain & Development, Vol 14, No 4,1992

spikes, but myoclonus-contingent 4-5 Hz theta bursts appeared. These findings suggest that some types of MELAS may
be associated with cortical types of myoclonus. Key words: MELAS, cortical reflex myoclonus, giant SEP, jerk-locked
averaging, C-reflex.
Saitoh S, Kohsaka S, Mizukami S, Kajii N.
Cortical reflex myoclonus associated with mitochondrial myopathy, encephalopathy,
lactic acidosis and stroke-like episodes (MELAS): a case report.
Brain Dev 1992; 14: 260-3

Mitochondrial encephalomyopathy is present in several
syndromes, which can be mainly subdivided into three
major ones: Kearns-Sayre syndrome (KSS), myoclonus epilepsy with ragged-red fibers (MERRF), and mitochondrial
myopathy, encephalopathy, lactic acidosis and stroke-like
episodes (MELAS) [1]. It is well known that several intermediate forms also exist [2]. Besides these three major
syndromes, there are other mitochondrial encephalomyopathies, such as subactue necrotizing encephalomyelopathy, Alper's syndrome and trichopoliodystrophy. Myoclonus is one of the major manifestations of MERRF, but
it is not usually associated with either MELAS or KSS.
Hence, the nature of the myoclonus in MELAS has remained unclear.
In this paper, we report a 9-year-old female MELAS patient with myoclonus. On the basis of electrophysiological
findings, the myoclonus was deduced to be of corticalorigin.
CASE REPORT
A Japanese girl was born to healthy and unrelated parents
after a normal pregnancy and delivery at term. She first
experienced loss of consciousness, vomiting and a generalized tonic clonic convulsion at age 5 years. At that
time, no other manifestations were noted except for a
short stature (-2.2 SD). The patient had no siblings and
no family history of neuromuscular disease. Because
of the recurrence of similar episodes twice within one
year, she was hospitalized for further examination. Laboratory studies showed increased serum transaminase and
creatine kinase levels, hyperlactatemia and metabolic
acidosis. A CT scan revealed multiple, asymmetric lowdensity areas in the occipital, parietal and temporal lobes
bilaterally. On MRI, these lesions exhibited low-intensity

in Tl-weighted images and high-intensity in T2 weighted
images. These neuroradiographic findings indicated multiple cerebral infarctions. A biopsy of the right quadriceps
femoris muscle was performed at age 6 years. Modified
Comori trichrome staining revealed 40-50% ragged-red
fibers. Biochemical studies on the muscle revealed a reduction in cytochrome c oxidase activity to 63.2 nmolj
min/mg mitochondrial protein (normal mean ± SD: 270 ±
133.0). Other mitochondrial respiratory enzyme activities
were within normal limits. Based on these findings, a diagnosis of MELAS was made. Treatment with co-enzyme
Q (70 mg/day) and dichloroacetate (350 mg/day), however, failed to prevent the recurrence of the stroke-like
episodes, and the patient's condition gradually deteriorated in association with complex partial seizures.
At age 9 years the existence of spontaneous myoclonus,
predominantly affecting the upper extremities, but also
involving the face, trunk and lower extremities, was first
noted. The myoclonus did not increase with voluntary
movement and almost completely disappeared during
sleep. Because of the development of myoclonus, the
patient was again hospitalized and electrophysiological
studies were performed. On admission, physical examination revealed generalized mild muscle atrophy, proximal
dominant muscle weakness and a positive Cowers' sign.
The patient could walk without support, but unsteadily.
Neither ataxia nor extrapyramidal signs were present.
Deep tendon reflexes were normal and no pathological reflexes could be elicited, however, there was associated
mild flexion contracture of the right ankle jOint. The IQ
was 49 (the Tanaka-Binet test). An EEC showed marked
slowing of the background activity (3.5-4 Hz theta range),
predominantly recorded in the parieto-occipital region,
but no paroxysmal discharges were detected.
METHODS

From the Department of Pediatrics, Hokkaido University School
of Medicine, Sapporo.
Received for publication: March 20, 1991.
Accepted for publication: April 26, 1992.
Correspondence address: Dr. Shinji Saitoh, Department of Pediatrics, Hokkaido University School of Medicine, N1S-W7, Kita-ku,
Sapporo 060, Japan.

Somatosensory-evoked potentials (SEPs) were recorded
from C3 with the reference electrode at Al and stimulation of the right median nerve at the wrist joint, with
an intensity of 10% above the threshold, while the patient
was awake in the resting state (no of responses averaged:
200; band-pass filter: 15-1,500 Hz; signal processor,
model 7S12, San-Ei Co, Ltd, Japan). A surface EMC was

Saitoh et al: Myoclonus with MELAS 261

-8. 8

-58.8

-\88.8

-\58. 8

-288.8

- 258.8

-388.8

·a . 888

os

+9.589

+1 . B88

+1 .58R

+2.08e

+2.580

+3.88 8

sec

Av.EEG

SEP
(C3-A1)

C3-A1
---.J10jJV
50msec

Av.EMG
(Rt.Thenar m.

C4-A2

C Reflex
Av.EMG
(Rt.thenar m.)

500msec
myoclonus

Fig 1 SEPs and averaged EMG from the right thenar muscle following right median nerve stimulation (n = 200). Note the large
amplitudes of the P25 and N34 components at the C3 electrode,
and the presence of the C-reflex.

Fig 2 Jerk-locked averaging (n =50) of myoclonus of the right
thenar muscle as a trigger. Note the 4-5 Hz theta bursts at the C3
and C4 electrodes.

simultaneously recorded from the right thenar muscle.
Two-channel EEGs (C3 with reference to AI, and C4
with reference to A2) and I-channel surface EMGs (right
thenar muscle) were recorded using an electroencephalograph (band-pass filter: 15-200 Hz; model IA94, San-Ei
Co, Ltd, Japan) and stored on magnetic tape. Jerk-locked
averaging was performed with a I-channel EMG as a trigger while reproducing and digitalizing these data (no of responses averaged: 50; sampling clock: 2 msec; personal
computer model PC98XL, NEC Co, Ltd, Japan). EMG
potentials longer than 100 msec were ignored.

drial encephalopathies including MERRF [3], therefore,
the precise mechanism underlying myoclonus associated
with mitochondrial encephalopathies has remained unclear. So et al [4] reported five MERRF patients with
"giant SEPs," and assumed that the cortex plays an important role in the pathogenesis of myoclonus in MERRF,
although it is still uncertain whether the mechanism underlying myoclonus associated with other mitochondrial
encephalopathies, such as MELAS, is the same as that in
the case of MERRF.
In this case, the significantly enhanced amplitude and
the prolonged latency of the cortically generated SEPs are
consistent with the characteristics of the so-called "giant
SEP," most commonly associated with progressive myoclonus epilepsy [5]. Although the precise pathophysiological basis of the "giant SEP" has not been elucidated
in detail, there is a general agreement that these characteristics reflect a state of abnormal excitability, with
enhancement of the cortical reflex pathway [6]. The definition of "giant SEP" is controversial, particularly in
childhood. In the case of adult patients, Shibasaki et al
reported that a "giant SEP" comprises a PI-N2 amplitude of more than 804 J1.V [7], and Obesso et al proposed
7 J1.V as the threshold [5]. For childhood patients, the
definition of a "giant SEP" has not yet been established.
According to our experience with 27 control epilepsy patients aged from 1 to 18 years (mean ± SD: 8.0 ± 5.1), the
average and SD of the P25-N34 amplitude were 4.5 J1.V
and 3.8 J1. V, respectively. Other than the high amplitudes
of early cortical segments of SEPs, the changes in later
segments of SEPs are common to "giant SEPs." In "giant
SEPs," the later cortical segments are usually difficult
to recognize, being replaced by high voltage slow waves
[8]. The patterns of later cortical segments in the present patient were similar to these patterns. Therefore, we
think that the SEPs in this patient have sufficient characteristics of a "giant SEPs" and thus indicates cortical hyperexci tab ility.

RESULTS
The amplitude ofP25-N34, regarded as the initial segment
of cortically generated SEPs, was 16.4 J1.V, which was
significantly enhanced (control range, 4.5 ± 3.8 J1.V (27
subjects examined)). The peak latency of the second negative component of SEPs, so-called N55, was significantly
prolonged, and the later components of SEPs were not
discernible, being replaced by a negatively-deflected,
dome-shaped prolonged potential (Fig 1). The averaged
EMG from right thenar muscle following electrical stimulation of the right median nerve revealed an enhanced
long-latency potential. The latency of the long-latency
potential observed on the stimulation was 47 msec, and
the duration from P25 of SEPs to the long-latency potential was 2004 msec (Fig 1).
Jerk-locked averaging revealed no myoclonus-related
spikes, but 4-5 Hz theta bursts appeared at both C3 and
C4 (Fig 2). These theta bursts were not observed upon
random averaging.
DISCUSSION
Myoclonus is a major symptom of MERRF, but it is usually not associated with MELAS. Until now, only a few
evoked potential studies have been reported on mitochon-

262 Brain & Development, Vol 14, No 4,1992

The averaged EMG, recorded simultaneously with the
SEPs, revealed an enhanced ~ong-latency potential. This
latency characteristics, which is believed to reflect passage
through the cortical reflex pathway, implies that this potential is identical to the C-reflex described by Sutton and
Mayer [9] , who advocated that the C-reflex represents cortical hyperexcitability. On the basis of these findings, we
concluded that the myoclonus in this patient was caused
by cortical hyperexcitability and thus consistent with the
cortical reflex myoclonus described by Hallett et al [10].
Shibasaki et al [11] described the myoclonus-contingent spikes on the basis of jerk-locked averaging in patients
with cortical reflex myoclonus. We also performed jerklocked averaging in the present MELAS patient and found
that a theta burst was closely related to the myoclonus
instead of a spike. Considering the fact that patients with
cortical reflex myoclonus do not always exhibit myoclonus-related cortical spikes [11], the early examination of
this patient's myoclonus might be one explanation for the
unusual myoclonus-related cortical activity observed.
ACKNOWLEDGMENTS
The authors are grateful to Prof. Shuzo Matsumoto, Department
of Pediatrics, Hokkaido University School of Medicine, for reviewing the paper.
REFERENCES
1. Di Mauro S, Bonilla E, Zeviani M, Nakagawa M, De Vivo DC.

Mitochondrial myopathies. Ann Neurol 1985;17:521-38.
2. Byrne E, Trounce I, Dennett X, Gilligan B, Morley JB,
Marzuki S. Progression from MERRF to MELAS phenotype
in a patient with combined respiratory complex 1 and 4
deficiencies. J Neurol Sci 1988;88:327-37.
3. Tsuji S, Uozumi T, Nakano S, Ohnishi A, Murai Y. Evoked
potential studies in mitochondrial encephalomyopathy. Arch
Neurol 1990;47:465-7.
4. So N, Berkovic S, Andermann F, Kuzniecky R, Gendron D,
Quesney LF. Myoclonus epilepsy and ragged-red fibers
(MERRF). 2. Electrophysiological studies and comparison
with other progressive myoclonus epilepsies. Brain 1989;
112:1261-76.
5. Obeso JA, Rothwell JC, Marsden CD. Somatosensory evoked
potentials in myoclonus. In: Fahn S, Marsden CD, Van Woert
MH, eds. Myoclonus. Advances in Neurology, Vol 43. New
York: Raven Press, 1986:373-84.
6. Kakigi R, Shibasaki H. Generator mechanisms of giant somatosensory evoked potentials in cortical reflex myoclonus.
Brain 1987 ;110: 1359-73.
7. Shibasaki H, Yamashita Y, Kuroiwa Y. Electroencephalographic studies of myoclonus. Brain 1978;101:447-60.
8. Yoshikawa H, Suzuki H, Kaga M, Sekuragawa N. Giant somatosensory evoked potentials in Rett syndrome (in Japanses). No To Hattatsu (Tokyo) 1990;22:186-8.
9. Sutton GG, Mayer RF. Focal reflex myoclonus. J Neural
Neurosurg Psychiatry 1974;37:207-17.
10. Hallett M, Chadwick D, Marsden CD. Cortical reflex myoclonus. Neurology 1979;29: 1107-25.
11. Shibasaki H, Yamashita Y, Tobirnatsu S, Neshige R. Electroencephalographic correlates of myoclonus. In: Fahn S,
Marsden CD, Van Woert MH, eds. Myoclonus. Advances in
Neurology, Vol 43. New York: Raven Press, 1986:357-72.

Cerebral Aneurysms in Children
Masanori Ito, MD, Masaaki Y oshihara, MD, Minoru Ishii, MD,
Akihiko Wachi, MD and Kiyoshi Sato, MD

Childhood intracranial saccular aneurysms are rare, accounting for only 0.6-4.6% of all aneurysms. This paper presents
two such cases. A 12-year-old boy developed sudden severe headaches. CT indicated subarachnoid hemorrhage in the bilateral sylvian fissure. Angiography failed to demonstrate any vascular lesions. Repeat angiography, however, revealed a
2 mm aneurysm on the right A-l. A miniclip was successfully applied to obliterate the aneurysm. The postoperative
course was uneventful. In the second case, a 3-year-old boy, CT incidentally revealed a bullet-shaped high density area.
Angiography demonstrated a large aneurysm arising from the insular portion of the prefrontal branch of middle cerebral
artery. This large aneurysm with a broad-based neck was successfUlly clipped, employing the angioplastic and/or tandem
clipping, despite the fact the prefrontal artery unfortunately was compromised in postoperative angiography. The patient
was dOing well postoperatively. Spontaneous subarachnoid hemorrhage (SAH) in childhood is uncommon. However,
when children with SAH is encountered, angiography should be performed repeatedly even if the first angiogram does
not demonstrate any vascular abnormality. Childhood aneurysms are increasingly being found incidentally in CT scans.
The differences between child and adult aneurysms in regard to site, size, sex predominance, clinical features, and sur-

263