Rerupture of intracranial aneurysms: a clinicoanatomic study

Transcription

1 J Neurosurg 67:29-33, 1987 Rerupture of intracranial aneurysms: a clinicoanatomic study ALBERT HIJDRA, M.D., MARINUS VERMEULEN, M.D., JAN VAN GIJN, M.D., AND HANS VAN CREVEL, M.D. Departments ~[ Neurology. University Hospital Dijkzigt, Rotterdam, Academisch Medisch Centrum, Amsterdam. and University Hospital, Utrecht, The Netherlands ~" In a series of 176 prospectively stu patients who survived for at least 24 hours after aneurysmal subarachnoid hemorrhage, 39 had at least one computerized tomography (CT)-proven rebleed within 4 weeks after the first rupture. There were peaks in the incidence of rebleeding at the end of the 2nd and 3rd weeks. Sudden loss of consciousness occurred in 35 patients, preceded in one-third of them by headache. A sudden increase in headache was a symptom of rebleeding in only one patient. Loss of brain-stem reflexes was recorded in 13 patients, respiratory, arrest in six, and both symptoms in eight patients. Apnea was temporary in 11 patients. Rebleeding occurred as gross intraventricular hemorrhage in 20 patients, as a space-occupying hematoma in four. as both types of hemorrhage in three, and as a purely subarachnoid hemorrhage in 12. The location of the rebleed could not be inferred from the clinical features. Rebleeding was fatal in 51% of cases (two of t 2 patients with a purely subarachnoid hemorrhage, and 18 of the other 27 patients (p < 0.005)). The risk of rebleeding could not be predicted from the patients' clinical condition on admission or from the amount of subarachnoid blood identified on the initial CT scan. The risk of further rebleeding was significantly increased in survivors of a first rebleed (47%: p < 0.01). Only seven (18%) of the 39 patients with rebleeding had survived at 3 months after the initial hemorrhage. KEY Wonos subarachnoid hemorrhage 9 rebleed tranexamic acid cerebral aneurysm T HE serious risk of rebleeding from an intracranial aneurysm is reflected in figures from the Cooperative Aneurysm Study, in which the incidence of rebleeding in unoperated patients within I month after the first rupture was approximately 35%, with a mortality rate of 42%. ~~ That study was performed before computerized tomography (CT) was introduced, in an era when rebleeding was diagnosed by the clinical features, by lumbar puncture, or by autopsy, and when the anatomical type of hemorrhage could be defined only at autopsy. It is difficult to distinguish rebleeding from other causes of secondary deterioration, particularly delayed cerebral ischemia (DCI), with clinical diagnosis or lumbar puncture, ~5~8-~9 and an autopsy (if performed) may not differentiate between an initial and a recurrent hemorrhage. The clinical features that are often thought indicative of rebleeding may also herald DCI, or may remain unexplained even after extensive investigation,~9 Lumbar puncture is often nondiagnostic because the time course of pigment changes in the cerebrospinal fluid (CSF) is unpredictable, TM and the study may be contraindicated because brain shift is often present after rebleeding. Identification of patients with an increased risk of rebleeding would be valuable for selecting the timing of surgery and the medical management. It has recently been suggested that rebleeding occurs more frequently in patients who are in a poor clinical condition after their initial bleed. 1,~cj4 However, in those studies the criteria for the diagnosis of rebleeding were insufficient or not given and, since the same has been demonstrated for vasospasm and DCI, 4"j~ the question is again raised as to whether the distinction between the two most common types of deterioration has been adequate. We therefore stu prospectively the clinical and anatomical features of recurrent aneurysmal subarachnoid hemorrhage (SAH), by means of close observation and serial CT scanning. We also assessed the influence of the patients' initial clinical condition and the amount of subarachnoid blood on the risk of rebleeding and on outcome. Clinical Material and Methods From October, 1977, to January, 1983, 176 patients with SAH were admitted in stable condition within 72 hours of the first clinical symptoms. Thirty-two patients J. Neurosurg. / Volume 67~July,

2 A. Hijdra, et al. TABLE 1 Clinical course in 39 patients with aneurysmal rerupture Brain- Outcome in Case Head- Level of Stem Apneaw Survivors No. ache* Consciousnesst Reflexes~ > 24 Hrs~l hematoma (4 cases) ~ 1--o death - ---*? * 54--, death ~ 3--* 2---~ death - ~ ~ further rebleeding, intraventricular hemorrhage (20 cases) , 1--o death - ~ , ~ death - ~ 7-4-* 1--~ 4--~ ~ ~ ~ death 9-5--* 1~ 1---~ death ~ 1---, 1 ~ death * ~ ~ ~ death? ~ 1---* ~ 2--0 death? ~ 2---, 2~ , hematoma progressive deterioration, severely disabled delayed cerebral ischemia, progressive deterioration, 16-5~ ~ death - ~ , 2---~ 5---~ death[[ - ~ ~ 1~ 4--0 deathll - ~ ~ l--*death - 20 ~ 1---~ 1---~ 5 - ~ - further rebleeding, 5---~ 1---~ progressive deterioration, 5---~ , 5 - ~ - further rebleeding, 54--* l---~ 5---~ , - further rebleeding, 5---~ 3---~ moderately disabled & intraventricular hemorrhage (3 cases) 5---~ ~ death - ~ * ~ death - ~ ~ 3---, 4--0 death * Data show whether headache was () or was not (-) the first symptom. i Numbers represent the motor part of the Glasgow Coma Scale:~3 1 = no reaction to pain; 2 = pain evokes extension movements; 3 = pain evokes flexion movements; 4 = patient localizes painful stimuli; 5 = patient obeys commands. First score: before rebleeding; second: immediately after rebleeding; third: 1 hour after rebleeding; fourth: 24 hours after rebleeding. :~ = pupillary reaction to light and oculocephalic and corneal reflexes were present; - = at least pupillary reaction to light was absent;? = brain-stem reflexes not documented. When brain-stem reflexes were initially absent, presence () or absence (-) after 1 hour is recorded. w Presence () or absence (-) of apnea or other respiratory abnormality calling for assisted ventilation. 82 Outcome 3 months after the initial hemorrhage, according to the Glasgow Outcome Scale: 8 1 = dead; 2 = persistent vegetative state; 3 = severely disabled; 4 = moderately disabled; 5 = recovered. II Case 17: death from third rebleed within 24 hours after first rebleed; Case 18: death from second rebleed within 24 hours after first rebleed; Case 32: nonfatal second rebleed. TABLE 1 (continued) Brain- Outcome in Case Head- Level of Stem Apneaw Survivors No. ache* Consciousnesst Reflexesz~ > 24 Hrs~l subarachnoid hemorrhage (12 cases) ,?---~?---~ 5? - recovered * ~ severely disabled ~ postoperative ischemia, ~ * 4 further rebleeding, , 3---~ 5 II - ~ persistent vegetative state * extracranial complication, 34 - ~3--*2---~death -~ * ~ 3 - extracranial complication, ~ 1---~ moderately disa-- bled ~ - further rebleeding, * 2---~ death - ~ - 39 ~ 3---~ 5---, 5 - severely disabled * Data show whether headache was () or was not (-) the first symptom. t Numbers represent the motor part of the Glasgow Coma Scale: ~3 1 = no reaction to pain; 2 = pain evokes extension movements; 3 = pain evokes flexion movements; 4 = patient localizes painful stimuli; 5 = patient obeys commands. First score: before rebleeding; second: immediately after rebleeding; third: 1 hour after rebleeding; fourth: 24 hours after rebleeding. $ = pupillary reaction to light and oculocephalic and corneal reflexes were present; - = at least pupillary reaction to light was absent;? = brain-stem reflexes not documented. When brain-stem reflexes were initially absent, presence () or absence (-) after 1 hour is recorded. w Presence () or absence (-) of apnea or other respiratory abnormality calling for assisted ventilation. 82 Outcome 3 months after the initial hemorrhage, according to the Glasgow Outcome Scale: 8 1 = dead; persistent vegetative state; 3 = severely disabled; 4 = moderately disabled; 5 = recovered. II Case 17: death from third rebleed within 24 hours after first rebleed; Case 18: death from second rebleed within 24 hours after first rebleed; Case 32: nonfatal second rebleed. who within 24 hours after admission were not included in this series. In 130 patients the offending aneurysm was demonstrated by angiography or subse- quent autopsy; in 46 patients angiography was not performed because of age or poor clinical condition, and in these patients the diagnosis was based on CT evidence of blood in the basal cisterns and fissures and on the absence of lesions other than aneurysms that might cause SAH. 16 The age of the 176 patients varied between 16 and 83 years (mean 52 years); 39 patients were over 65 years of age. All patients were stu prospectively, and were entered into a double-blind placebo-controlled trial of tranexamic acid as a means of preventing rebleeding. 17 The study period was 28 days, or less if the patient underwent surgery or. For patients selected for surgery, the operation was usually planned for Day J. Neurosurg. / Volume 67~July, 1987

3 Rerupture of intracranial aneurysms after the initial SAH; 62 patients actually underwent surgery. All patients were under continuous observation in an intensive care unit. Every suggestion of clinical deterioration was reported to one of us, in which case the patients were reexamined and the clinical signs documented. The level of consciousness was assessed according to the Glasgow Coma Scale.~3 A decrease of at least one point on the motor part of this scale was considered deterioration in the level of consciousness for the purposes of this study. Computerized tomography was performed on admission and was repeated weekly and after any clinical deterioration. The diagnosis of rebleeding was made when CT or autopsy showed an increase of blood (subarachnoid, intracerebral, or intraventricular) compared with the appearance on the previous CT scan. The amount of SAH on the initial CT scan was graded from 0 to 3 for each of the following 10 cisterns: the frontal interhemispheric fissure; the quadrigeminal cistern; both suprasellar cisterns; both ambient cisterns; both basal sylvian fissures; and both lateral sylvian fissures. All of the initial CT scans were graded separately in this manner by three of the authors. Differences in judgment were settled by majority vote or, in a few cases where this was not possible, by a joint review. Outcome was measured 1 and 3 months after the aneurysmal rupture. The results were recorded using the five-point Glasgow Outcome Scale? Results Of the 176 patients, 39 rebled within 4 weeks of the first rupture (22% of the total group: 13% of patients who were treated with tranexamic acid, and 32% of the placebo-treated group). Nine patients had more than one documented rebleed: two rebleeds in four patients, three in four patients, and five in one patient. Only the clinical and anatomical features of the first rebleed are described (Table 1). Twenty patients (51%) progressively deteriorated and from the rebleed (17 within 24 hours), 12 (31%) from further complications (two from another rebleed within 24 hours), and seven (18%) survived at least 3 months (four with severe brain damage). The time of rebleeding is given in Fig. 1. Of the 19 patients who survived the first rebleed, nine (47%) had a second rebleed. This increase in risk of rebleeding is statistically significant (chi-square 7.31; p < 0.01). Clinical Features Only one patient (Case 2) had a sudden increase in headache without initial decrease in level of consciousness. At the same time she developed a hemiparesis and within 24 hours from secondary effects of an intracerebral hematoma. Twelve patients had a sudden increase in headache (evident from sudden moans, cries, or groping for the head, and also with sudden vomiting in one) and then became unconscious within the next minute. Twenty-three patients suddenly lost m DAYS FROM PRESENTING HEMORRHAGE FIG. 1. Day after the first subarachnoid hemorrhage (SAH) on which 39 of 176 patients at risk suffered their first rebleed. The dashed bar at the left represents the risk of first rebleeding when 11 probable rebleeds in 32 patients who soon after the first SAH are included. consciousness without premonitory symptoms. In three cases the clinical signs were concealed by the patients' condition at the time of rebleeding: in one patient (Case 28) rebleeding occurred at surgery before the dura was opened, and in two others (Cases 33 and 35), who were in a poor condition from DCI, rebleeding was detected by a decrease in eye-opening response and on weekly CT scanning, respectively. In Case 15, rebleeding was evident not only from the clinical signs but also from the appearance of fresh blood in an external drainage system connected to a ventricular catheter. In 21 of the 35 patients who became unresponsive, this event was accompanied by initial loss of one or more brain-stem reflexes. All 21 patients lost the pupillary response to light (most had small pupils), and eight patients lost oculocephalic reflexes to passive movement of the head and the corneal reflexes. Fourteen patients who became unresponsive had initial respiratory abnormalities that called for assisted ventilation. This was needed for less than 1 hour in eight and between 1 and 24 hours in another three; in three patients spontaneous respiration did not return. 6 The different combinations of loss of consciousness, loss of brain-stem reflexes, apnea, and mortality in each group are presented in Fig. 2. Anatomical Types of Hemorrhage On CT scanning, seven of the 39 patients with a rebleed had an intracerebral or sylvian hematoma with signs of tentorial herniation: midline shift, torsion of the brain stem, or obliteration of basal cisterns.12 Three of these patients also exhibited gross intraventricular hemorrhage. Isolated intraventricular hemorrhage without a space-occupying hematoma was present in 20 patients: hemorrhage was seen only in the third ventricle in one patient, in one lateral ventricle in one, and in the lateral and third ventricles in one; slight hemorrhage was evident in all ventricles in one patient; and massive filling of all ventricles was seen in 16 patients. Four of these 20 patients also had a small intracerebral J. Neurosurg. / Volume 67~July,

4 A. Hijdra, et al. x os S of CONSC~C~&4.S3, ~OSS O ~ / 4/ /\ \ ( (6(3) (8(6) ) 13(9)~ ) \\ \/ /21 4"(1) FIG. 2. Combinations of loss of consciousness, loss of brain-stem reflexes, and apnea in 39 patients with aneurysmal rebleeding. Numbers in parentheses represent patients who from their rebleed. The number 4* includes two patients who were comatose before the rebleed and one anesthetized patient. hematoma without signs of brain shift. In the remaining 12 patients the recurrent hemorrhage was confined to the subarachnoid space. Clinicoanatornical Correlations The anatomical type of hemorrhage could not be identified by the initial clinical features. However, only two of the 12 patients with purely subarachnoid rebleeding from the rerupture; whereas 18 of 27 TABLE 2 Clinical grade on admission correlated with incidence of rebleeding and death Clinical No. of Rebleeds Mortality Grade* Cases No. % No. % 1 & & * Clinical grading was according to Hunt and Hess] TABLE 3 Level of consciousness on admission correlated with rebh, eding and death No. of Rebleeds Mortality GCS Score* Cases No. % No. % < * GCS = Glasgow Coma Scale.~3 patients with other types of hemorrhage from the rerupture (Fisher's exact test: p < 0.005). Predictive Factors There was no difference in the distribution of clinical grades (according to the classification of Hunt and Hess 7) among patients who had no rebleed compared with those who had (Tables 2 and 3). Based on our classification system for the amount of subarachnoid blood on the admission CT scan, the median score and interquartile range were 8 and 4 to 16, respectively, in patients with a rebleed and 9 and 2 to 15 in the other patients. This difference is not statistically significant (Mann-Whitney U-test: z = 0.55; p = 0.58). These figures were 10 and 7 to 17 in all patients who eventually within 3 months, and 4 and 1 to 13 in patients who survived. This difference is significant (Mann- Whitney U-test: z = 3.75; p < ). Discussion There was a 22% incidence of rebleeding in 176 patients who survived the first 24 hours after their initial SAH. The mortality rate due to rebleeding was 56%. The risk of rebleeding increases significantly after a first rebleed. Of the patients who were excluded from the antifibrinolytic drug trial, 32 within 24 hours after SAH. Eleven patients suffered sudden deterioration that suggested rebleeding, but this was not confirmed as CT scanning had not yet been performed. When 11 probable rebleeds in 32 patients at risk were added to the number of proven rebleeds to obtain the frequency of rebleeding correlated with the time after the first SAH (Fig. 1), the results confirm the peak incidence on the 1 st day, as reported in the 1983 Cooperative Aneurysm Study. 9 However, the exponential decline in the risk of rebleeding after the 1st day could not be confirmed in this study, since peaks were found in the 2nd week and at the end of the 3rd week after the first rupture (Fig. l). A sudden increase in headache without a decrease in the level of consciousness occurred in only one patient. All other patients suffered a sudden loss of consciousness, immediately preceded by headache in one-third of them. Sudden increase in headache occurred as an isolated symptom in two other patients, in whom no evidence of recurrent hemorrhage could be demonstrated. These patients had SAH scores of 2 and 0 on their previous CT scans, and it seems unlikely that an increase in intracranial blood would not have been visible. One of these patients had a rebleed associated with a sudden headache within hours after the episode. Sudden headache without CT evidence of rebleeding may be caused by a small hemorrhage confined to the aneurysmal wall. 2 Less dramatic fluctuations in the severity of headache often occurred and were not always systematically investigated. However, our protocol of serial CT scanning revealed asymptomatic rebleeding 32 J. Neurosurg. / Volume 67~July, 1987

5 Rerupture of intracranial aneurysms in only one (comatose) patient. A sudden loss of consciousness is much more suggestive of rebleeding, although (as demonstrated in a previous study 19) it may also herald the onset of DCI. As we have previously noted, 56 the initial loss of brain-stem reflexes with or without apnea may be followed by progressive deterioration and death, but it may also be followed by recovery. Furthermore, no specific clinical picture could be associated with the anatomical site of recurrent hemorrhage, either intracerebral, intraventricular, or subarachnoid. The mortality rate for patients with a purely SAH was lower than for those with intracerebral or intraventricular hemorrhages, but after 3 months only two of those 12 patients had made a good or moderately good recovery. There was no correlation between the clinical condition on admission and the incidence of rebleeding. Such a correlation has been found in several studies, L~~4 but in none of these was the diagnosis of rebleeding based on the CT or postmortem demonstration of an increase in blood. There also was no correlation between the amount of SAH on the initial CT and the occurrence of rebleeding. However, death after 3 months occurred significantly more often in patients with a poor initial clinical condition (Tables 2 and 3) or large amounts of subarachnoid blood on the initial CT scan, and this could be explained by a higher incidence of DCI in these groups (A Hijdra, el al., unpublished data). This study suggests that in previous studiesl ~ ~.~4 rebleeding and DCI may have been incompletely distinguished, because of imprecise criteria for both events. The finding that all patients are at equal risk of rebleeding again underlines the fact that measures to prevent recurrent hemorrhage should be taken as early as possible in all patients. Since antifibrinolytic agents increase the risk of DCI, L-, which is also an important complication of early surgery, 3 the development of measures that combat DCI should have priority in SAH research. References 1. Aoyagi N, Hayakawa l: Analysis of 223 ruptured intracranial aneu~sms with special reference to rerupture. Surg Neurol 21: , Ball M J: Pathogenesis of the "sentinel headache" preceding berry aneurysm rupture. Can Med Assoc J 112: 78-79, Drake CG: Management of cerebral aneurysm. Stroke 12: , Graf C J, Nibbelink DW: Cooperative Study of Intracra- nial Aneurysms and Subarachnoid Hemorrhage. Report on a randomized treatment study. III. Intracranial surge~'. Stroke 5: , Hijdra A, van Gijn J: Early death from rupture of an intracranial aneurysm. J Nearosarg 57: , Hijdra A, Vermeulen M, van Gijn J, et al: Respiratory arrest in subarachnoid hemorrhage. Neurology 34: , Hunt WE, Hess RM: Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 28:14-20, Jennen B. Bond M: Assessment of outcome after severe brain damage. A practical scale. Lancet 1: , Kassell NF. Torner JC: Aneurysmal rebleeding: a preliminary report from the Cooperative Aneurysm Study. Neurosurgery 13: , kocksley HB: Report on the Cooperative Study of lntracranial Aneurysms and Subarachnoid Hemorrhage. Section V, Part II. Natural history of subarachnoid hemorrhage, intracranial aneurysms and arteriovenous malformations. Based on 6368 cases in the Cooperative Study. J Neurosurg 25: , 1966 l 1. Maurice-Williams RS: Ruptured intracranial aneurysms: has the incidence of early rebleeding been overestimated? J Neurol Neurosurg Psychiatry 45: , Stovring J: Descending tentorial herniation: findings on computed tomography. Neuroradiology 14: , Teasdale G, Jennett B: Assessment of coma and impaired consciousness. A practical scale. Lancet 2:81-84, Torner JC, Kassell NF, Wallace RB, et al: Preoperative prognostic factors for rebleeding and survival in aneurysm patients receiving antifibrinolytic therapy: report of the Cooperative Aneurysm Study. Neurosurgery 9: , van Crevel H: Pitfalls in the diagnosis of rebleeding from intracranial aneurysm. Clin Neurol Neurosurg 82:1-9, van Gijn J. van Dongen KJ: Computerized tomography in subarachnoid hemorrhage: difference between patients with and without aneurysm on angiography. Neurology 30: , Vermeulen M, Lindsay KW, Murray GD, et al: Antifibrinolytic treatment in subarachnoid hemorrhage. N Engl J Med 311: , Vermeulen M, van Gijn J, Blijenberg BG: Spectrophotometric analysis of CSF after subarachnoid hemorrhage: limitations in the diagnosis of rebleeding. Neurology 33: , Vermeulen M, van Gijn J, Hijdra A, et al: Causes of acute deterioration in patients with a ruptured intracranial aneurysm. A prospective study with serial CT scanning. J Neurosurg 60: , 1984 Manuscript received October 24, Address reprint requests to. Albert Hijdra, M.D., Department of Neurology, Academisch Medisch Centrum, 1105 AZ Amsterdam, The Netherlands. J. Neurosurg. /Volurne 6 7 / July