sign in sign up
<<
Home , Chiasmatic cistern, Subarachnoid hemorrhage, Xanthochromia

Journal ofNeurology, , and Psychiatry 1990;53:365-372 365 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from OCCASIONAL REVIEW

The diagnosis of subarachnoid haemorrhage

M Vermeulen, J van Gijn

Lumbar puncture (LP) has for a long time been of 55 patients with SAH who had LP, before the mainstay of diagnosis in patients who CT scanning and within 12 hours of the bleed. presented with symptoms or signs of subarach- Intracranial haematomas with shift was noid haemorrhage (SAH). At present, com- proven by operation or subsequent CT scan- puted tomography (CT) has replaced LP for ning in six of the seven patients, and it was this indication. In this review we shall outline suspected in the remaining patient who stop- the reasons for this change in diagnostic ped breathing at the end of the procedure.5 approach. In the first place, there are draw- Rebleeding may have occurred in some ofthese backs in starting with an LP. One of these is patients. that patients with SAH may harbour an We therefore agree with Hillman that it is intracerebral haematoma, even if they are fully advisable to perform a CT scan first in all conscious, and that withdrawal of cerebro- patients who present within 72 hours of a spinal fluid (CSF) may occasionally precipitate suspected SAH, even if this requires referral to brain shift and herniation. Another disadvan- another centre.4 tage of LP is the difficulty in distinguishing It could be argued that by first performing between a traumatic tap and true subarachnoid CT the diagnosis may be delayed and that this haemorrhage. Secondly, the use of CT within may be dangerous if the patient has bacterial the first two or three days offers a wealth of . However, the likelihood of menin- information about the origin and extent of the gitis is extremely small in patients with an haemorrhage, and about the presence of early adequate history of or unconscious- complications requiring urgent treatment. An ness coming on within seconds and only rarely early brain scan also serves as a baseline against is the distinction more difficult, for instance, in which future changes can be measured. The patients who have been found in a- confused information about the location of the haemor- state, with marked neck stiffness and moderate rhage is especially valuable in patients with . SAH and a negative angiogram. This is a Ideally, patients with suspected SAH should

heterogeneous group of patients which will be not be admitted to a hospital without CT http://jnnp.bmj.com/ discussed separately. The review continues scanning facilities (and patients with confirmed with guidelines about the interpretation of the SAH not to hospitals without a neurosurgical CSF in patients with a negative CT scan, which unit). is the only remaining indication for LP in the diagnosis of SAH. The final paragraph points out the many pitfalls in the diagnosis of Traumatic tap or intracranial ? rebleeding by analysis of CSF samples, which Neck stiffness following SAH takes three to 12 makes serial CT scanning by far the preferred hours to develop-it is an unreliable test to on September 27, 2021 by guest. Protected copyright. method. perform on domiciliary calls. Patients arriving in casualty with a history of "" a few hours earlier and with equi- The dangers of vocal neck stiffness may either have a true SAH The possibility of a haematoma being present or a more usual type of headache with an in patients who present with the clinical unusual onset.' If an LP is performed and features of SAH should be considered. In a blood-stained CSF is found it is difficult to series of 100 consecutive patients who were distinguish between a traumatic tap and intra- suspected of aneurysmal rupture, fifteen had a cranial bleeding. It is a widely held belief that non-aneurysmal haematoma and eight of these this distinction can be made by collecting the were located in the .' It is obvious CSF in three test tubes and by counting the red that an LP could have done harm in these cells in the consecutive tubes. A decrease in red patients. In addition, a third of patients with a cells would indicate a traumatic tap. Indeed, a Department of , University ruptured harbour an intracerebral decrease in red cells occurs more often in Hospital Dijkzigt, haematoma,`'' and even in patients without traumatic punctures than in intracranial bleed- Rotterdam neurological signs other than neck stifEfess 10 ing, but it also occurs in patients with a M Vermeulen per cent may have a haematoma of at least previous bleed.9 Conversely, a constant num- University 30 mm.4 ber of cells can be seen with traumatic taps. In Department of Neurology, Utrecht, Deterioration after LP is not only a theor- the individual case therefore no firm diagnosis The Netherlands etical possibility as shown by Duffy, who can be made with this "three tube method". J van Gijn described severe clinical deterioration in seven The opinion of the doctor who carried out 366 Vermeulen, van Giin

the LP also appears unreliable. In a study by as possible. On the day of SAH, intracranial J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from Buruma et al doctors were often unable to blood is detected in about 95% of patients. decide for themselves whether the blood- This proportion declines to 90% after one day, stained CSF had resulted from the procedure 80% after five days, and 50% after one week.316 or not.9 The form ofthe red cells changes in the In the first place, if intracranial blood is CSF, and this so-called crenation of red cells detected, the site of the haemorrhage often has been advocated as a means of differentiat- indicates the likely source. In case of an ing a traumatic tap from a true haemorrhage. intracerebral haematoma, it is often possible to However, crenation of red cells occurs very distinguish primary intraparenchymal bleed- soon after red cells have entered the CSF and ing from SAH extending into the brain tissue may therefore be seen in many traumatic taps.'0 (fig 1).1' It is not always easy to determine Cytology of the CSF is also of limited value. whether a haematoma associated with The presence of erythrophages indicates aneurysm rupture is in fact intraparenchymal intracranial haemorrhage if an LP was not or whether it has only distended the subarach- carried out earlier, but it takes some time before noid space, usually the frontal interhemis- siderophages are demonstrable. Moreover, pheric or sylvian fissure.'8 However, this dis- negative cytology does not exclude a haemor- tinction is of limited importance for practical rhage.9 management, and follow up scans will usually The most reliable method for distinguishing resolve the issue. In case the extravasated blood a traumatic tap from a genuine bleed is to spin is confined to the basal cisterns, a ruptured down the CSF and to examine the supernatant aneurysm can be diagnosed ifthe subarachnoid fluid for the presence of . clots are most dense at one of the classical Examination ofthe supernatant should be dohe aneurysm sites (fig 2): the frontal interhemis- with spectrophotometry and not with the pheric fissure (aneurysm of the anterior com- naked eye, since direct vision is a rather in- municating ), the chiasmatic cistern on sensitive method. In a series of32 CSF samples one side (aneurysm of the internal carotid in which spectrophotometry had detected xan- artery, usually at the origin of the posterior thochromia, this was visible with the naked eye communicating artery), or the most lateral part in only half the samples." The pigments that of the sylvian fissure (aneurysm of the middle are responsible for the yellow colour ofthe CSF cerebral artery).' '9 after a haemorrhage result from lysis of red Bleeding from at more unusual cells. The average survival of red cells in the sites may be more difficult to recognise; CSF is much shorter than in the circulation. aneurysmal bleeding from a posterior inferior This rapid haemolysis has not been fully ex- cerebellar artery, for instance, may be easily plained. Recently, it has been suggested that missed by CT.202' Predicting the site of a red cells in the subarachnoid space lose their ruptured aneurysm from the CT scan is more identity as autologous cells and this results in than an intellectual game. This information is immune mediated haemolysis.'2 helpful in planning , in particular In the classic study by Barrows et al the the order and the extent of selective - development of xanthochromia was studied by isation, and it is vital ifmore than one aneurysm

means of spectrophotometry.'3 This technique is found.' 21-23 Even the uncommon event oftwo http://jnnp.bmj.com/ measures the light intensity in different regions aneurysms rupturing at the same time can be of the visible spectrum (400-700 nm), after correctly documented by CT.24 Visualisation of transmission of light through the coloured the aneurysm itself, after injection of contrast, CSF. Three different pigments can be distin- used to be the exception with CT, indicating guished by their characteristic absorption that the aneurysm is very large, with or without bands in the spectrum: oxyhaemoglobin, calcification.' High-resolution CT scanning, methaemoglobin, and . The wave- with 1-5 mm slices, is more sensitive but not length where absorption occurs permits quali- infallible.25 Magnetic resonance imaging on September 27, 2021 by guest. Protected copyright. tative determination, whereas the height of the (MRI) is not well suited for imaging subarach- peak makes quantification possible. noid haemorrhage in the acute stage.26 After The rate at which the pigments appear in the several days or weeks, however, when the CT CSF is of great practical importance, as their scan has become normal, MRI may detect presence excludes a traumatic puncture. Oxy- subpial deposition of haemosiderin near the haemoglobin has been detected as early as two source of the haemorrhage.2'26 In addition, hours after the bleeding,'3 but usually it takes a MRI allows visualisation of the aneurysm few hours or more for red cells to lyse and for itself. In two studies, a T2-weighted 2000/80 xanthochromia to develop. In two large series spin echo sequence, identified an aneurysm in ofpatients with SAH it was shown that xantho- about half the cases, by means of a signal-void chromia could be detected in all patients in area.27 28 whom the CSF was examined at least 12 hours Not all haemorrhage in the basal cisterns is after the haemorrhage. 4 15 of aneurysmal origin. A recently recognised form is the perimesencephalic haemorrhage,29 with extravasated blood mainly or only in the Extra information by CT scanning interpeduncular, ambient, or quadrigeminal Computed tomography not only obviates the cisterns (fig 3a). Not only is the pattern of need for a potentially hazardous LP, it offers haemorrhage on CT different from that with important information other than the mere aneurysmal haemorrhage, the clinical features presence or absence of intracranial blood. This may also differ; loss of consciousness occurs applies particularly if the scan is made as early less often and a gradual onset of headache The diagnosis ofsubarachnoid haemorrhage 367

Figure I (a) Bifrontal J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from haematoma,from ruptured aneurysm of the anterior communicating artery; (b) haematoma in the medial part of the temporal lobe, from ruptured aneurysm of the internal carotid artery, at thejunction with the posterior communicating artery; (c) haematoma in the lateral part of the temporal lobe,from ruptured aneurysm of the (the aneurysm itself is visible as a less dense area within the haematoma).

aneurysmal. After the exclusion by CT of intracerebral haemorrhages and arteriovenous malformations the remaining 10 are of non- aneurysmal perimesencephalic type and about half of these occur mainly in the interpedun- cular cistern. Thus, the chance of a ruptured basilar aneurysm with such a type of haemorr- hage is about one in six (fig 4). This is of course reason enough to perform a vertebral angio- gram if the patient is fit for operation, but if this is negative repeated angiography does not seem i ~~~~~~~~~ ~~~~~~~~y t ...~~~~~~~~~~~ ': warranted. A similar calculation applies to patients without blood in the basal cisterns or even http://jnnp.bmj.com/ & ~ ~~~~~~~~~:.4. ~ ~~~~~~~~without any evidence of extravasated blood on CT scanning, despite an interval of less than three days from the onset of symptoms, and yet with true SAH demonstrated by xantho- chromia of the CSF. With third or fourth generation CT scanners, and with adequate experience in the detection of subarachnoid blood, evidence of blood in the basal cisterns is on September 27, 2021 by guest. Protected copyright. lacking on the first day in about 5 %// of ruptured aneurysms,3 16 29 and in about a quarter of patients with SAH and a normal angiogram.29 Given that 800/ of patients with SAH have an aneurysm, the chance of an aneurysm in

(minutes rather than seconds) occurs more Table Patterns of bleeding on CT and results of often than with aneurysms. Together with the angiography in 120 consecutive patients with excellent this suggests a venous or subarachnoid haemorrhage. Patients with primary intracerebral haemorrhage have been excluded.24 CT was source of bleeding.29 This type of performed withinfive days in all patients, and within haemorrhage is not rare, occurring in about three days in most 1000 of all subarachnoid haemorrhages, and in half the patients with a negative angiogram Angiography (table). It is also rather specific. Basilar No an clot Aneurysm aneurysm aneurysms may also produce isolated (n = 92) (n = 28) in the , but these aneurysms constitute only one out of 10 rup- Pattern of haemorrhage on CT: Basal cisterns, compatible with tured aneurysms, and in turn only one in eight aneurysm 85 7 haemorrhages from basilar aneurysms do not Predominantly around 1 13 Aspecific or no evidence of extend into the chiasmatic and other basal haemorrhage 6 8 cisterns (fig 3b).29 For every 100 SAHs 80 are 368 Vermeulen, van Gijn

Figure 2 (a) J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from Subarachnoid haemorrhage, predominantly in the frontal part of the interhemispheric fissure, from a ruptured aneurysm of the anterior communicating aneurysm; (b) subarachnoid haemorrhage, with centre in the right suprasellar cisterns (to reader's left), from a ruptured aneurysm of the internal carotid artery, at thejunction with the posterior communicating artery; (c) subarachnoid haemorrhage, predominantly in the sylvianfissure,from a ruptured aneurysm of the middle cerebral artery.

not all of one kind. The distinction between these groups is difficult from the perspective of the neurosurgeon whose patients are referred after five days or more, without a CT scan, the diagnosis being based on a lumbar puncture of , r ... ,g _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~... F which there are no details. The relative propor- l tions ofthe different categories of patients with !...... ' . a negative angiogram may differ between centres, which makes it even more pointless to lump all these patients under one heading. It may be useful therefore to remember that there k~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .; ... are, at present, at least four subgroups of patients with "angio-negative SAH". The first and most unfortunate category is

that of patients who are brought to the emer- http://jnnp.bmj.com/ Fo~~~~~~~~~~~~~~~~~~~~~~~~~~.6.''... gency room with what in fact is nothing more than "crash " or a sudden exacerba- tion of , 8 and who undergo an ill-timed and traumatic lumbar puncture by young and eager physicians. They are then confined to bed with a diagnosis of SAH, subsequently transferred to a specialised unit on September 27, 2021 by guest. Protected copyright. for angiography, and they are finally told that patients with true SAH but without cisternal the weak spot in one of their brain vessels has blood on an early CT scan is about one in two. probably healed but that they would be wise to Early CT scanning not only gives important take things a bit easier in the future. diagnostic information on the source of The remaining groups of patients all have haemorrhage, it also demonstrates complica- a true SAH, but without a demonstrable tions that may require urgent treatment, aneurysm. Again, it is assumed that on the basis especially acute "0 or a life- of the CT scan primary intracerebral haemor- threatening haematoma." Finally, the initial rhages consistent with either the "hyper- CT scan provides an important baseline for the tensive" type or with arteriovenous malforma- diagnosis of future complications, particularly tions, cryptic or overt, have been ruled out rebleeding, or hydrocephalus. before. About halfofthese patients show blood on CT around the midbrain, unlikely to be "Angiogram-negative SAH": a mixed compatible with aneurysmal rupture (fig 3a). bag The localised nature of the extravasated blood, The fate of patients with the diagnosis of the less explosive onset of the symptoms in subarachnoid haemorrhage but a negative some cases, and the excellent prognosis all angiogram has been the subject of a spate of make it improbable that an occult arterial publications that have reported follow up aneurysm underlies these haemorrhages."9 The studies. We shall abstain from listing these third group of patients, making up about one studies, as it is abundantly clear from the quarter ofpatients with true SAH and a normal preceding paragraphs that these patients are angiogramj shows no evidence of blood in the The diagnosis ofsubarachnoid haemorrhage 369 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from Figure 3 (a) Non- aneurysmal r perimesencephalic haemorrhage, with extravasated blood predominantly in the interpeduncularfossa, and to a lesser extent in the ambient cisterns; (b) subarachnoid haemorrhage from ruptured aneurysm of the basilar artery, with extravasated blood predominantly in the interpeduncularfossa, but I also more anteriorly, in the chiasmatic cistern, the Sylvian fissures, and the frontal part of the interhemispheric fissure. *. .X.:~ ... A

I

basal cisterns on CT, or no blood at all.29 In or other intracranial haemorrhage. Yet, if the these patients the cause of bleeding is even patient is seen soon after the headache, we wish more elusive than in the perimesencephalic to emphasise again that it is vital to defer lumbar group, but it may be the same. The fate ofthese puncture to at least 12 hours after the onset of patients has not been separately investigated, symptoms and that xanthochromia should be but the impression is that they do well. examined not with the naked eye but with The fourth and last category are patients spectrophotometry. If LP is carried out too with a pattern ofbleeding on CT, that is typical early, xanthochromia may not have had enough of ruptured aneurysm but in whom angio- time to develop and it is then impossible to graphy fails to demonstrate the expected distinguish between a traumatic lumbar punc- aneurysm. This group constitutes the last ture and a genuine SAH. Blood artificially quarter of patients with SAH and normal introduced into the CSF will later lead to the angiography, and about 5%/ of all patients with appearance of pigments, and repeated CSF

cisternal haemorrhage (table). It may be that studies can never solve the problem that has http://jnnp.bmj.com/ occasional observations of local ,32 been created by an LP that has been performed vasospasm33 or rebleeding34 apply only to these too soon. patients, and that these harbour "occult" CSF examination in the diagnosis of SAH is aneurysms, despite a normal angiogram. It is also important when the patient presents late this category of patients in which repeated after a suspected SAH. The 90% probability3 16 angiography seems most indicated. ofdetecting blood in the basal cisterns on CT in the first two days after an SAH decreases on September 27, 2021 by guest. Protected copyright. Lumbar puncture after a negative CT rapidly thereafter. In contrast, all patients with scan a true SAH have CSF xanthochromia between Examination ofthe CSF is still important ifCT 12 hours and two weeks after the bleeding.'4 15 scanning provides no evidence ofsubarachnoid Xanthochromia is still present in over 700,, of patients after three weeks, and even after four Aneurysm Other weeks this proportion is still over 40%. " These (80) aneurysm (72) Blood also in findings contrast with another study35 where anterior cisterns xanthochromia was investigated with the naked L Basilar (7) eye, which explains the high proportion of 100 aneurysm (8) patients with SAH who had no xanthochromia. -(1) patients withi I Could an SAH have occurred in patients Blood mainily in 1 in 6 subarachnoid with symptoms suspected of SAH who have interpedu ncu lar haemorrhage blood-stained CSF but no xanthochromia? cistern Although one study suggested that this may (5) this is In a centre Peri - occur,35 extremely unlikely. mesencephalic - with approximately 80 patients with a proven (10) _ SAH admitted each year, only 12 patients in a elsewhere six year period had bloodstained CSF but no -No aneurysm I-BloodOther (10) around midbrain xanthochromia. In three of these 12 patients (20) (5) angiography was performed; all three were Figure 4 The chance of a ruptured aneurysm in patients with an isolated clot in the negative. All 12 patients survived without interpeduncular cistern on CT is 1 in 6. disability and were not admitted again for an 370 Vermeulen, van Gijn

SAH during a mean follow up period of four series of 62 episodes mimicking a rebleed had a J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from years.'5 Again, the most probable explanation sudden impairment of consciousness, followed for bloodstained CSF without xanthochromia by a period ofa fe hours in which there was no in patients suspected of SAH is that these change, and then had further gradual deter- patients had a non-haemorrhagic "thunderclap ioration while evidence of infarction developed headache" and a traumatic LP. on CT.4' Therefore, rebleeding cannot be diag- It has been suggested that angiography nosed on purely clinical grounds. These should be carried out in patients with a suspec- problems used to be often dismissed with the ted SAH but with bloodless CT or LP. This statement that rebleeding "was confirmed by suggestion is based on two reports of single lumbar puncture or ". patients suspected of SAH, who had a normal The diagnosis of rebleeding by analysis of CT and normal CSF but in whom subsequent CSF samples is, in fact, extremely difficult. angiography showed an aneurysm with vaso- Just as the diagnosis of a first haemorrhage by spasm.36 37 CSF analysis requires the demonstration of In contrast, a study of 71 patients with xanthochromia, the diagnosis of rebleeding sudden, severe and unusual headache but with requires the demonstration of "fresh" pig- normal CT and CSF showed that none ofthese ments. This raises the question of whether patients had complications related to SAH and blood pigments appear in and disappear from none later had a documented SAH, during a the CSF in a predictable fashion. Oxyhaemo- mean follow up period of 3-3 years.6 It is likely globin may be detected in the CSF a few hours that these patients have had benign "thunder- after the bleed.'3 In the early study by Barrows clap ", without bleeding. In both et al it remained the predominant pigment only reported cases with aneurysms, dn for the first few days, after which it was the angiogram was the feature believed to have converted to bilirubin, the iron-free derivative a causal relation with the headache. We think of haemoglobin.'3 This conversion does not that migraine is a more probable explanation take place in test tubes, as it is dependent upon for both the headache and the vasospasm.38 A the enzyme haem oxygenase, which is present small proportion of adults harbour asympto- in macrophages, the arachnoid membrane, and matic aneurysms,39 and indiscriminate angio- the .42 In a later study, however, graphy is bound to uncover some of these. We the relative concentration of oxyhaemoglobin agree that angiography should be done in was found to increase spontaneously after an patients without blood on CT and even without initial decrease, and oxyhaemoglobin could CSF xanthochromia if these patients present still be the predominant pigment as late as after more than two weeks of the suspected three weeks after the initial haemorrhage.43 The bleeding, since not only the blood on CT but same fluctuating pattern was observed when also xanthochromia of the CSF may have the haemoglobin concentration was serially disappeared by that time.'5 measured.43 Therefore, measurement of neither the proportion of oxyhaemoglobin nor the haemoglobin concentration can help in the The diagnosis of rebleeding diagnosis of rebleeding. It has also been sug- a If patient with a recently ruptured aneurysm gested that the total amount of pigments from http://jnnp.bmj.com/ complains of a sudden increase ofheadache and red cell lysis, the xanthochromic index, then abruptly becomes deeply unconscious, decreases after the first few days following the with wide, non-reacting pupils and apnoea, haemorrhage.44 Rebleeding would become there is no diagnostic problem. However, other apparent by the interruption of this normal complications may mimick rebleeding.Y resolution pattern. However, the xantho- Deterioration of consciousness and focal chromic index may continue to rise up to the neurological signs may have various causes, sixteenth day after SAH.43 The marked dif- particularly delayed cerebral ischaemia. ferences in the velocity of pigment clearance on September 27, 2021 by guest. Protected copyright. Furthermore, the clinical picture is often from the CSF have been attributed to advanced clouded by the previous haemorrhage. - age, , and atherosclerosis.45 Since most tose patients do not report an increase of rebleeds occur within this period' a suspected headache, and "increased neck stiffness" is rebleed cannot be confirmed merely by demon- often equivocal. Finally, close observation of strating an increase in xanthochromia. the patient around the clock for several weeks These findings make the often reported is not always practicable, with the result that statement that rebleeding was confirmed by LP the onset of any deterioration is not always barely tenable. If rebleeding is suspected and documented. Even if a sudden loss of con- LP is repeated, the demonstration of blood- sciousness (within minutes) is witnessed in a stained CSF does not prove rebleeding, as the patient with SAH, only two out of three turn CSF may remain blood-stained for as long as out to have a rebleed on CT. If the fatal cases four weeks after the initial bleed.'2 The blood- are excluded an even larger proportion of staining may also have resulted from a trau- abrupt episodes, nearly 50%, was not caused matic tap, and the distinction between a by rebleeding.4l Other causes of acute traumatic tap and recurrent bleeding is impos- deterioration were epilepsy, ventricular sible to make since the CSF supernatant was fibrillation (in only one of 62 episodes of already xanthochromic before the suspected deterioration) and, surprisingly, cerebral rebleed. In summary, examination of the CSF ischaemia. Cerebral ischaemia after SAH can only be of help in the diagnosis of rebleed- usually presents with a gradual onset of ing if two earlier LPs had demonstrated a deterioration. In contrast three patients in this decrease of the xanthochromic index, which is The diagnosis ofsubarachnoid haemorrhage 371 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from then followed by an increase.43 It is clear that in haemorrhage, and finally a few patients practice this method can seldom be applied. with "occult" aneurysms. Apart from the unpredictable clearance of 5 Lumbar puncture is indicated in SAH only pigments the diagnosis of rebleeding by CSF if CT scanning is negative. If LP is carried analysis is also difficult because LP cannot out between 12 hours and two weeks after always be performed after a rebleed. Since the haemorrhage, xanthochromia of the "fresh" xanthochromia needs time to develop, CSF can be demonstrated in all patients the LP has to be postponed to at least 12 hours with SAH, provided xanthochromia is after the rebleed. Meanwhile the patient may investigated by spectrophotometry. Even have died. In addition, the risk of LP in after three weeks xanthochromia is still patients with an unsuspected haematoma may present in over 70%0 of patients, and after be even greater than after a first bleed. In a four weeks in over 40%. consecutive series of 17 patients with CT 6 Rebleeding is the cause of acute deteriora- evidence of rebleeding, it was possible in only tion in only two out of three cases with six patients to carry out an LP and compare the sudden loss of consciousness; the diagnosis results ofspectrophotometry with two previous cannot be confirmed by CSF analysis, samples (the diagnosis was confirmed in five).43 except under exceptional circumstances. In short, the diagnosis of rebleeding by LP is Necropsy is also fallible in this respect. hardly possible. Serial CT scanning is mandatory for a Necropsy, if feasible, does not settle the confident diagnosis of rebleeding. matter either. The pathologist can confidently diagnose rebleeding only if the interval be- We received helpful comments from Dr E F M Wijdicks tween the first haemorrhage and rebleeding and D Hasan and we are grateful to Mrs Betty Mast and was not too short, and that between rebleeding Mrs Ellen Budelman for secretarial assistance. and was not too long.A' The present accepted standard in the diagnosis of rebleed- ing consists of the demonstration of a 1 Van Gijn J, Van Dongen KJ. Computed tomography in the fresh diagnosis of subarachnoid haemorrhage and ruptured haemorrhage on CT scanning or at necropsy, aneurysms. Clin Neurol Neurosurg 1980;82:11-24. not previously demonstrated on scan. 2 Pasqualin A, Bazzan A, Cavazzani P, Scienza R, Licata C, CT Such Da Pian R. Intracranial following aneurysmal comparisons are quite sensitive, as even three rupture: experience with 309 cases. Surgical Neurology or more successive haemorrhages can be relia- 1986;25:6-17. 3 Van Gijn J, Van Dongen KJ. The time course ofaneurysmal bly distinguished.4' Nevertheless, early haemorrhage on computed tomograms. Neuroradiol 1982; rebleeds cannot be diagnosed these 23:153-6. if occur 4 Hillman J. Should computed tomography scanning replace before the admission CT scan.' lumbar puncture in the diagnostic process in suspected subarachnoid haemorrhage? Surg Neurol 1986;26:547-50. 5 Duffy GP. Lumbar puncture in spontaneous subarachnoid Conclusions haemorrhage. Br Med J 1982;285:1 163-4. 1 Lumbar puncture after SAH may pre- 6 Wijdicks EFM, Kerkhoff H, Van Gijn J. Long-term follow up of 71 patients with thunderclap headache mimicking cipitate brain shift and herniation from an subarachnoid haemorrhage. Lancet 1988;ii:68-9. unsuspected haematoma, even in patients 7 Salloum A, Lebel M, Reiher J. Acces cephalalgiques simulant une hemorrhagic meningee. Rev Neurol who are fully conscious. 133:131-8. 1977; 2 Subarachnoid haemorrhage can be reliably 8 Pearce JMS, Pearce SHS. Benign paroxysmal cranial http://jnnp.bmj.com/ neuralgia or cephalgia fugax. Br Med J 1986;292:1015. diagnosed by analysis of the CSF only after 9 Buruma OJS, Janson HLF, Den Bergh FAJTM, Bots examination of the supernatant fluid with a GTHAM. Blood-stained : traumatic puncture or haemorrhage? J Neurol Neurosurg Psychiatry spectrophotometer. This implies that lum- 1981;44:144-7. bar puncture should be postponed until at 10 Matthews WF, Frommeyer WB. The in vitro behaviour of erythrocytes in human cerebrospinal fluid. J Lab Clin least 12 hours after the onset of symptoms. Med 1955;45:508-15. If lumbar puncture is performed earlier, a 11 Soderstrom CE. Diagnostic significance of CSF spectro- traumatic tap may unnecessarily photometry and computer tomography in cerebrovascular stigmatise disease. A comparative study in 231 cases. 1977; on September 27, 2021 by guest. Protected copyright. patients with benign forms of "thunderclap 5:606-12. 12 Peterson JW, Kwun BD, Teramura A, et al. Immunological headache". reaction against the aging human subarachnoid eryth- 3 Computed tomography (CT) scanning is rocyte. A model for the onset of cerebral vasospasm after a safe and . J Neurosurg 1989;71:718-26. sensitive method for demon- 13 Barrows LJ, Hunter FT, Banker BQ. The nature and clinical strating subarachnoid blood, but should significance of pigments in the cerebrospinal fluid. Brain 1955;78:59-80. preferably be performed within three 14 Walton JN. Subarachnoid haemorrhage. Edinburgh: Living- days. Important additional information stone, 1956:124. 15 Vermeulen M, Hasan D, Blijenberg BG, Hijdra A, Van Gijn provided by CT is: firstly, the demonstra- J. Xanthochromia after subarachnoid haemorrhage needs tion of non-aneurysmal sources of no revisitation. J Neurol Neurosurg Psychiatry 1989;52: haemor- 826-8. rhage ("hypertensive" haemorrhages, 16 Adams HP, Kassell NF, Torner JC, Sahs AL. CT and arteriovenous malformations, cryptic clinical correlations in recent aneurysmal subarachnoid hemorrhage: a preliminary report of the Cooperative haemorrhages around the midbrain); Aneurysm Study. Neurology 1983;33:981-8. secondly, the demonstration of early com- 17 Hayward RD, O'Reilly GVA. Intracerebral haemorrhage accuracy of computerized transverse axial scanning in plications requiring urgent treatment; predicting the underlying aetiology. Lancet 1976;i:1-4. thirdly, evidence about the probable site of 18 Schulder M, Hirano A, Elkin C. "Caval-septal" : does it exist? Neurosurgery 1987;21:239-41. ruptured aneurysms, and fourthly, estab- 19 Scotti G, Ethier R, Melancon D, Terbrugge K, Tchang S. lishment of a baseline for future changes. Computed tomography in the evaluation of intracranial aneurysms and subarachnoid haemorrhage. Radiology 4 Patients with negative angiograms are not 1977;123:85-90. a single category. This group consists of 20 Ruelle A, Cavazzani P, Andrioli G. Extracranial posterior inferior cerebellar artery aneurysm causing isolated patients without SAH but with a traumatic intraventricular hemorrhage: a case report. Neurosurgery lumbar puncture, patients with various 1988;23:774-7. 21 Stone JL, Crowell RM, Gandhi YN, Jafar JJ. Multiple types of non-aneurysmal subarachnoid intracranial aneurysms: magnetic resonance imaging for 372 Vermeulen, van Gijn

determination of the site of rupture-report of a case. 34 Barlow P. ofdelayed cerebral ischaemia following J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.53.5.365 on 1 May 1990. Downloaded from Neurosurgery 1988;23:97-100. subarachnoid haemorrhage of unknown cause. J Neurol 22 Kendall BE, Lee BCP, Claveria E. Computerized tomogra- Neurosurg P-sychiatry 1985;48:132-6. phy and angiography in subarachnoid haemorrhage. Br J 35 MacDonald A, Mendelow AD. Xanthochromia revisited: a Radiol 1976;49:483-501. re-evaluation of lumbar puncture and CT scanning in the 23 Lim ST, Sage DJ. Detection ofsubarachnoid blood clot and diagnosis of subarachnoid haemorrhage. J Neurol Neuro- other thin, flat structures by computed tomography. surg Psychiatry 1988;51:342-4. Radiology 1977;123:79-84. 36 Day JW, Raskin NH. Thunderclap headache: symptom of 24 Joslyn JN, Williams JP, White JL, White RL. Simultaneous unruptured cerebral aneurysm. Lancet 1986;ii: 1247-8. rupture of two intracranial aneurysms: CT diagnosis. 37 Clarke CE, Shepherd DI, Christi L, Victoratos G. Thunder- Stroke 1985;16:518-21. clap headache. Lancet 1988;ii:625. 25 Schmid UD, Steiger HJ, Huber P. Accuracy of high 38 Call GK, Flemming MC, Sealfon S, Levine H, Kistler JP, resolution computed tomography in direct diagnosis of Fisher CM. Reversible cerebral segmental vasoconstric- cerebral aneurysms. 1987;29:152-9. tion. Stroke 1988;19:1159-70. 26 Barkovich AJ, Atlas SW. Magnetic resonance imaging of 39 McCormick WF, Nofzinger JD. Saccular intracranial . Radiol Clin North Am 1988; aneurysms. An autopsy study. JNeurosurg 1965;22:155-9. 26:801-20. 40 Van Crevel H. Pitfalls in the diagnosis of rebleeding from 27 Jenkins A, Hadly DM, Teasdale GM, Condon B, Macpher- . Clin Neurol Neurosurg 1980;82: son P, Patterson J. Magnetic resonance imaging of acute 1-9. subarachnoid hemorrhage. J Neurosurg 1988;68:731-6. 41 Vermeulen M, Van Gijn J, Hijdra A, Van Crevel H. Causes 28 Saton S, Kadoya S. Magnetic resonance imaging of sub- of acute deterioration in patients with a ruptured arachnoid hemorrhage. Neuroradiology 1988;30:361-6. intracranial aneurysm. J Neurosurg 1984;60:935-9. 29 Van Gijn J, Van Dongen KJ, Vermeulen M, Hijdra A. 42 Roost KT, Pimstone NR, Diamond I, Schmid T. The Perimesencephalic hemorrhage: A nonaneurysmal and formation of cerebrospinal fluid xanthochromia after benign form of subarachnoid hemorrhage. Neurology subarachnoid hemorrhage: Enzyme conversion of 1985;35:493-7. to bilirubin by the arachnoid and choroid 30 Van Gijn J, Hijdra A, Wijdicks EFM, Vermeulen M, Van plexus. Neurology 1972;22:973-7. Crevel H. Acute hydrocephalus after aneurysmal sub- 43 Vermeulen M, Van Gijn J, Blijenberg BG. Spectro- arachnoid hemorrhage. J Neurosurg 1985;63:355-62. photometric analysis of CSF after subarachnoid hemor- 31 Brandt L, Sonesson 0, Ljunggren B, Saveland H. Ruptured rhage: Limitations in the diagnosis of rebleeding. middle cerebral artery aneurysm with intracerebral Neurology 1983;33:1 12-4. hemorrhage in younger patients appearing moribund: 44 Van der Meulen JP. Cerebrospinal fluid xanthochromia: An emergency operation? Neurosurgery 1987;20:925-9. objective inex. Neurology 1966;16:170-8. 32 Edner G, Forster DMC, Steiner L, Bergvall V. Spontaneous 45 Tourtellotte WW, Metz LN, Bryan ER, De Jong RN. healing of intracranial aneurysm after subarachnoid Spontaneous subarachnoid hemorrhage. Neurology 1964; hemorrhage. J Neurosurg 1978;48:540-4. 14:301-6. 33 Spetzler RF, Winestock D, Newton HT, Boldrey EG. 46 Hijdra A, Vermeulen M, Van Gijn J, Van Crevel H. Disappearance and reappearance of cerebral aneurysm in Rerupture of intracranial aneurysms: a clinico anatomic serial arteriograms. J Neurosurg 1974;48:508-10. study. J Neurosurg 1987;67:29-33. http://jnnp.bmj.com/ on September 27, 2021 by guest. Protected copyright.