1442 Letters to the Editor J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.11.1442 on 1 November 1994. Downloaded from nerve palsy were present. Perimetry showed recognition of this "non-apoplectic" mode (pH 6-0) with a 30% methanol gradient. full visual fields and was 6/9 in of presentation allows prompt neurosurgical Standards of aspartate, glutamate, taurine, both eyes. Magnetic resonance imaging management, hopefully averting permanent and alanine in concentrations of 9, 90, and showed appreciable tumour recurrence. ophthalmoplegia and the possibility of life 900 pmol (Wakojunyaku, Tokyo, Japan) Foci of haemorrhage were evident through- threatening upward extension of the pitu- were measured every eight to 12 CSF sam- out the lesion, which was compressing both itary mass, while permitting definitive treat- ple analyses to obtain calibration curves. cavernous sinuses, although there was no ment of the underlying adenoma. The retention time was identified from the chiasmal compression. A further endonasal DOMINIC THOMPSON standard chromatograms and the area was MICHAEL POWELL transphenoidal was per- Department ofSurgical Neurology, used for the calculation of CSF amino acid formed. Overtly haemorrhagic and necrotic The National Hospitalfor Neurology neurotransmitters. We estimated the maxi- was identified at opera- and , mal error of the glutamate concentration to Queen Square, London, UK tion and later verified histologically. Two OLIVER FOSTER be 30% in some of our samples based on months after this procedure the left third Department ofNeurology, the data reported by Ferrarese et al.4 We nerve palsy had begun to resolve. The Middlesex Hospital, London, UK expressed the results as mean (SD) and Haemorrhage within pituitary adenomas compared the concentrations between 1 Brougham M, Heusner AP, Adams RD. is a common finding either at the time of Acute degenerative changes in adenomas of patients and control subjects by Mann- operation or on subsequent histopathologi- the pituitary body with special reference to Whitney U test. cal examination and is often subclinical.23 It . J Neurosurg 1955;7: Symptomatic cerebral vasospasm was has been suggested that in cases of subclini- 421-39. defined with and, 2 Mohr G, Hardy J. Haemorrhage, necrosis, cal haemorrhage, tissue necrosis is an and apoplexy in pituitary adenomas. Surg together with neurological deficits, was uncommon finding whereas it is common in Neurol 1982;18:181-9. graded as mild, moderate, or severe. The those presenting with clinical apoplexy. 3 Wakai S, Fukishima T, Teramoto A, Sano K. outcome was assessed at three months after minor haemorrhage within vascu- Pituitary apoplexy: Its incidence and clinical to the Repeated significance. Y Neurosurg 1981;55: 187-93. subarachnoid haemorrhage according lar adenomas probably accounts for the 4 Onesti ST, Wiseniewski T, Post KD. Clinical Glasgow outcome scale. The patients were clinically occult group, whereas swelling versus subclinical pituitary apoplexy: pre- divided into two groups; good outcome, if secondary to ischaemic oedema or haemor- sentation, surgical management, and out- the Glasgow outcome score showed good come in 21 patients. Neurosurgery 1990; rhagic in an adenoma which has 26:980-6. recovery or moderately disabled, and poor outgrown its blood supply, is proposed as 5 Rovitt RL, Fein JM. Pituitary apoplexy: a outcome, if it showed severely disabled or the more likely sequence of events in cases review and reappraisal. Y Neurosurg 1972; dead. The amino acid neurotransmitters of clinical apoplexy.4 The second mecha- 37:280-8. were compared between the two groups by nism seems to have been responsible for the Student's t test. three cases presented here, where extensive Patients and normal controls were com- necrosis was uniformnly seen. The absence pared. Mean values of glutamate, aspartate, of severe , obtundation, and and alanine increased some threefold (p < altered visual acuity accords with the find- 0-05). The taurine increased 1-7-fold but ing that whereas ophthalmoplegia is com- Excitotoxic amino acid neurotrans- the difference was not significant. Five patible with an alert state, defects in visual mitters are increased in human patients made a good recovery, one was acuity are more often associated with after subarachnoid moderately disabled, two severely disabled, impaired consciousness.5 haemorrhage and two died of delayed ischaemia due to The mortality and morbidity of untreated vasospasm (table). Differences in amino classical pituitary apoplexy is high, whereas Experimental evidence' suggests a role for acid neurotransmitter concentrations were the results of expedient surgical intervention glutamate neurotoxicity in many neurode- compared between those who had good for pituitary apoplexy are generally good. generative disorders, but few clinical studies outcomes and those with poor outcomes. Misdiagnosis of pituitary apoplexy even in have been conducted to substantiate this Concentrations of the amino acid neuro- its more typical presentation is common, hypothesis.23 We have measured excitatory transmitters tended to be higher in the poor however, and it is therefore important that and inhibitory amino acid neurotransmitters outcome group, but these differences were those involved in the initial evaluation of including glutamate, aspartate, taurine, and not significant. patients are aware of the variability in clini- alanine in human CSF in the acute stage of Persson et al evaluated the cortical extra- cal presentation and the need for early and subarachnoid haemorrhage in relation to cellular glutamate concentration in a patient appropriate radiological evaluation. clinical outcome. with severe subarachnoid haemorrhage by a Magnetic resonance imaging combines high The study protocol was approved in microdialysis method5 and found a 25-fold http://jnnp.bmj.com/ diagnostic accuracy with excellent anatomi- advance by the local ethics committee and increase in glutamate, aspartate, and taurine cal detail particularly in the coronal plane informed consent was obtained from a rela- under conditions of energy perturbation as (figure). tive. Ten patients who had a subarachnoid indicated by the simultaneously measured The case for urgent surgical decompres- haemorrhage due to a ruptured cerebral lactate:pyruvate ratio and suggested that a sion in the presence of obtundation and were included in this study massive accumulation of glutamate and actual or threatened visual loss is generally (table) and their ruptured aneurysm was aspartate in the extracellular space may supported. In the alert patient presenting clipped within 48 hours of ictus. A ventricu- contribute to neuronal injury in subarach- with ophthalmoplegia in the absence of lar catheter was placed, CSF was intermit- noid haemorrhage.5 We measured concen- deficit or impaired acuity the tently drained to maintain the intracranial trations of amino acids in the ventricular on September 26, 2021 by guest. Protected copyright. surgical role is less clearly defined and many pressure below 20 mm Hg, and CSF sam- CSF, which can be regarded as a pool into advocate conservative management in this ples were collected from day 3 to day 8 after which substances diffuse from the extra- instance. Steroid treatment and close clini- the haemorrhage and for at least two days cellular space. As the amino acids were cal observation is certainly associated with after operation. Samples were deproteinised diluted in the CSF in the ventricles, their spontaneous improvement in some cases, and stored at -80°C until analysis. Control concentrations in the extracellular space although there do not seem to be any reli- CSF samples were collected and stored in may be higher than those measured in the able predictive criteria on which such the same way from patients without neuro- CSF. patients can be identified. The clinical logical disorders who underwent spinal taps The concentration of extracellular gluta- course is unpredictable and the possibility for spinal anaesthesia. mate that is toxic to neurons is poorly of rapid deterioration with blindness or The CSF samples were diluted 10 times defined particularly as any such effects are death must be borne in mind and weighed with distilled water and 30,l aliquots were enhanced by compromising energy genera- against the surgical alternative. The used. Amino acids were measured three tion.6 In vitro experiments with cultured transphenoidal approach in experienced times for each CSF sample by high perfor- neurons have suggested that concentrations hands is extremely well tolerated and mance liquid chromatography with an elec- of 5-100 ,umol/l glutamate are toxic to neu- ensures immediate decompression trochemical detector (ECD-100; EICOM, rons.' Olney et al reported that direct injec- of the parasellar structures and affords the Kyoto, Japan) after a 4-6 mm (diameter) x tion of 500 nmol (as 11l of a 500 mmol/l opportunity to obtain tissue for histological 150 mm reverse phase precolumn glutamate solution over five minutes) into evaluation to aid subsequent management. (Eicompak, MA-50DS, EICOM, Kyoto, the rat striatum was necessary to cause Extraocular muscle palsies may occur as Japan) derivatisation procedure with o- delayed local neuronal degeneration.3 The the sole manifestation of infarction or phthaldialdehyde and ethylmercaptan. The relation between the amount of increase of haemorrhage in pituitary adenomas. Early liquid phase was 0-1 mol phosphate buffer amino acid neurotransmitters in CSF and Letters to the Editor 1443 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.11.1442 on 1 November 1994. Downloaded from

Summary ofpatients with subarachnoid haemorrhage: clinical profiles, and amino acid neurotransmitter (AANT) concentrations in CSF AANT concentrations (piM) Age H and K Day of Day of Day of Case Isex grade GCS Operation Deterioration sampling Vasospasm GOS ASP GLU TAU ALA 1 53/M 2 15 1 4 + GR 5-26 9-28 7-68 60-3 2 56/M 2 14 1 3 + GR 0-22 2-57 5-74 64-43 3 62/F 3 11 1 8 + GR 2 13 9-32 15-6 129-1 4 52/M 3 13 0 3 + + MD 1-02 0-88 4-65 155 4 5 64/M 3 14 1 3 5 + + SD 2-02 3-11 8-94 132-4 6 48/M 3 13 1 6 + GR 0 54 3 76 5-98 60-94 7 41/F 4 7 1 5 + GR 1-48 0 77 4-35 110-0 8 79/F 4 6 1 5 5 + + + D 0-24 9 45 5-33 40 3 9 55/M 4 5 0 3 2 + + + D 2-00 3-24 12-9 145-4 10 65/F 4 11 1 4 5 + + + SD 4-29 7-85 14-8 151-0 Mean (SD) 1-9(1 7)* 5 0(3-6)* 8 6(4 3) 105-3(44) Control (n = 16) 0 6(0 5) 1-6(0 7) 5-8(1-8) 31-9(6-2) *p < 0 05 v controls. H and K = Hunt and Kosnik grade; GOS = Glasgow outcome scale; GCS = Glasgow scale; ASP = aspartate; GLU = glutamate; TAU = taurine; ALA = alanine; Day = postsubarachnoid haemorrhage day; + = mild (100-149 cm/s); ++ = moderate (150-199 cm/s); +++ = severe (>200 cm/s); GR = good recovery; MD = moderately disabled; SD = severely disabled; D = dead. outcome requires study of a larger popula- Correspondence to: Dr Kashiwagi, Department of Excitotoxic mechanisms of epileptic Neurosurgery, Yamaguchi University, 1144 damage. In: Delgado-Escueta A, Ward A, tion of patients but our preliminary results Kogushi, Ube Yamaguchi, 755 Japan. Woodbury D, Porter R, eds. Basic mecha- provide a rational basis for such studies and for nisms of the epilepsies, molecular and cellular trials of excitotoxic amino acid antagonists. approaches, advances in neurology. New York: Raven Press, 1986:857-77. S KASHIWAGI 4 Ferrarese C, Pecora N, Frigo M, et al. H FUJISAWA Assessment of reliability and biological sig- T YAMASHITA nificance of glutamate levels in cerebrospinal H ITO fluid. Ann Neurol 1993;33:316-9. Department ofNeurosurgery, 1 Choi DW, Maulucci-Gedde M, Kriegstein A. 5 Persson L, Hillered L. Chemical monitoring Yamaguchi University School ofMedicine, Glutamate neurotoxicity in cortical cell of neurosurgical intensive care patients Ube, Japan culture.7 Neurosci 1987;7:357-68. using intracerebral microdialysis. J Neurosurg T MAEKAWA 2 Baker AJ, Moulton RJ, MacMillan VH, 1992;76:72-80. Y KURODA Shedden PM. Excitatory amino acids in 6 Novelli A, Reilly JA, Lysko PG, Henneberry A TATEISHI cerebrospinal fluid following traumatic brain RC. Glutamate becomes neurotoxic via the Critical Care Medical Center, injury in humans. J Neurosurg 1993;79: N-methyl-D-aspartate receptor when intra- Yamaguchi University School ofMedicine, 369-72. cellular energy levels are reduced. Brain Res Ube, Japan 3 Olney JW, Collins RC, Sloviter RS. 1988;451:205-12.

Incidence of internal carotid dis- cranial nerve palsy in one. Isolated Minnesota, 1987-92 gave a similar inci- section in the community of Dijon headache was the single symptom in three dence of 2-6 per 100 000.2 cases (8-3%). The diagnosis of Headache, ischaemic cerebrovascular dis- Dissection of the cervical internal carotid was suspected with ultrasound in 34 cases ease, and oculosympathetic palsy were the artery is one of the major causes of (with demonstration of a double lumen in most common manifestations of sponta- ischaemic in patients under 50 years eight cases), and proved with angiography neous internal carotid artery dissection. old.' Reliable epidemiological data are not in all cases, with MRI in only the last four Some internal carotid artery dissections do numerous. Only one defined population patients. No necropsy cases of spontaneous not present with ischaemic stroke (10% in study on internal carotid artery dissection cervical internal carotid artery dissection the study of Shievinck et al,2 8-3% in the has been published.2 were found during this period. No bilateral present study). cervical internal carotid artery dissections, M GIROUD We examined the occurrence of sponta- H FAYOLLE neous internal carotid artery dissection in or associated vertebral artery dissection N ANDRE the population of the city of Dijon from were found. Arterial was seen R DUMAS 1985 to 1993, by means of the stroke in 19 cases. Dysplastic cervical or renal Service de Neurologie, H6pital Gineral, http://jnnp.bmj.com/ Registry of Dijon3 which records all first were found in 15 cases. Association 21000 Dijon, France , and also asymptomatic , with tobacco and pill consumption was pre- F BECKER occlusion, and dissection of internal carotid sent in 15 out of 21 female cases. Evolution Service d'Angiologie, was marked by slight neurological sequelae H6pital du Bocage, artery occurring in the population of the 21000 Dijon, France city of Dijon. The diagnosis was ascertained in four cases. Headache resolved in all D MARTIN by an ultrasonic Doppler examination of the patients. Recurrent arterial dissections were N BAUDOIN cervical carotid arteries, and a carotid not found at follow up ranging from three Service de Neuroradiologie, H6pital Giniral, angiography. Patients with a history of cer- months to nine years. 21000 Dijon, France vical trauma were not included. For the cal- These 36 cases represent 2% of 1784 D KRAUSE on September 26, 2021 by guest. Protected copyright. culation of incidence, the whole population patients with stroke collected in this popula- Service de radiologie, was tion of Dijon and 10-1% of the 356 stroke H6pital du Bocage, of Dijon considered to be at risk, with 21000 Dijon, France 135 000 inhabitants in 1985 and 150 000 patients under 50 years old. This confirms Correspondence to: Dr M Giroud. inhabitants in 1992, and was used as the that internal carotid artery dissection is a denominator for the study. major cause of cerebral infarction in those 1 Mokri B, Sundt TM Jr, Houser OW, Piepgras DG. Spontaneous dissection of the cervical From 1985 to 1993, 36 intemal carotid under 50 years old.' internal carotid artery. Ann Neurol 1986; artery dissections were diagnosed. The The average annual incidence of sponta- 19:126-38. average annual incidence for all age groups neous internal carotid artery dissection for 2 Schievink NI, Mokri B, Whisnant JP. Internal was 2-9 per 100 000 con- all age groups was 2-9 per 100 000 inhabi- carotid artery dissection in a community. population (95% Rochester, Minnesota, 1987-1992. Stroke fidence interval 1 9-3 9). There were no tants, similar to that in the city of 1993;24: 1678-80. statistically significant annual variations Rochester.2 This compares with 2-2 per 100 3 Giroud M, Milan C, Beuriat P, et al. (three cases in 1985, four in 1986, three in 000 for aneurysmal subarachnoid haemor- Incidence and survival rates during a 2-year period of intracerebral and subarachnoid 1987, five in 1988, four in 1989, five in rhage in Dijon.3 haemorrhages, cortical infarcts, lacunes and 1990, three in 1991, four in 1992, five in Before ultrasound was introduced transient ischaemic attacks. The Stroke 1993). The mean age was 39 9 (7 5) years many cases remained undiagnosed. Biller Registry of Dijon: 1985-1989. Int for the 21 women, 43-7 years for the et al 4 reported a spontaneous cervical Epidemiol 1991;20:892-9. (5 7) 4 Biller J, Hingtgen WL, Adams HP Jr, Smoker 15 men. All had headache or neck pain; 19 internal carotid artery dissection in 0-13% WRK, Godersky JC, Toffol GJ. patients presented with cerebral ischaemic of 4531 patients with acute stroke and Cervicocephalic arterial dissections: a ten- symptoms, four with retinal ischaemic Bogousslavsky et al5 reported dissection year experience. Arch Neurol 1986;43: 1234-8. symptoms; one patient had had a sub- in 2-5% of 1200 patients with acute 5 Bogousslavsky J, Despland PA, Regli F. arachnoid haemorrhage. Oculosympathetic stroke. Spontaneous carotid dissection with acute palsy was noted in 12 patients, and 12th A community based study in Rochester, stroke. Arch Neurol 1987;44:137-40.