Nimotop

Patients (n = 127) with aneurysmal subarachnoid hemorrhage (SAH) were examined by transcranial Doppler ultrasonography (TCD) in a prospective study to follow the time course of the posthemorrhagic blood flow velocity in both the middle cerebral artery (MCA) and in the anterior cerebral artery (ACA). Results were analysed to reveal their relationship and predictive use with respect to the occurrence of delayed ischemic deficits. Mean flow velocities (MFV) higher than 120 cm sec^sup -1^ in MCA and 90 cm sec^sup -1^ in ACA were interpreted as indicative for significant vasospasm. In 20 of our 127 patients (16%) a delayed ischemic deficit (DID) was subsequently diagnosed clinically (DID+ group). Patients in the DID+ group can be characterized as those individuals who presented early during the observation period post-SAH with highest values of MFV, a faster increase and longer persistence of pathologically elevated MFV-values (exceeding 120 cm sec^sup -1^ in MCA and 90 cm sec^sup -1^ in ACA). They also show a greater difference in MFV-- values if one compares the operated to the nonoperated side. Differences in MFV-values obtained in MCA or ACA were statistically significant (p 120 cm sec^sup -1^ in MCA. If pathological values were obtained in ACA, this ratio increases to about four times, if DID+ patients presented with MFV > 90 cm sec^sup -1^ versus patients with MFV

Keywords: Delayed ischemic deficit; vasospasm; transcranial Doppler ultrasonography

INTRODUCTION

Delayed ischemic deficits (DID) can frequently be observed in patients during the time course of aneurysmal SAH. They present as newly appearing neurological deficits at clinical follow- up examination. Time of their appearance corresponds to the observation of new ischemic zones as found in computer tomography scans (CT). DID are assumed to be sequeale of massive artery vasospasm which follow SAH. Besides secondary cerebral haemorrhage, they are the leading cause of disability and death in patients with a history of ruptured cerebral aneurysms. Such cerebral artery vasospasm causes vascular stenosis with a subsequent increase in blood flow resistance. During the initial phase of vasospasm, this change in flow dynamics may remain tolerable for the patient due to mechanisms of cerebral autoregulation which effect both regional cerebral blood flow (rCBF) and global cerebral blood flow (CBF).However, further increase of such resistance due to continued vasospasm may gradually exceed the limitations of any autoregulatory mechanism and may then lead to a subsequent drop in rCBF and exhaustion of the circulatory reserves as an ultimate consequence rCBF collapses1,2. A fall in CBF leads to impaired tissue oxygenation with the corresponding appearance of ischemic zones in computed tomography (CT) scans. This becomes apparent as delayed neurological deficits in clinical examinations3-6.

The influence of vasospasm on CBF has previously been investigated7-10. Vasospasm also contributes to the formation of intravascular blood clots which can occlude the vessel or may result in embolism, both of which can lead to cerebral infarction11,12. In extremis, generalised vasospasm with collapse of CBF can ultimately lead to a patient's death1,2,5,7,11-13. However, cerebral vasospasm need not necessarily lead to cerebral infarction even in circumstances where it coincides with other rCBF disorders. This supports the notion that the phenomenon is 'reversible' during the initial period following SAH.

The aim of this prospective study was therefore:

1. To characterise the time-course of CBF changes as demonstrated by consecutive TCD examination following SAH.

2. To identify typical CBF-parameters in the subgroup of patients with subsequent development of a DID as compared to the patient group without such DID to predict the risk for DID to occur.

To precisely work out any subtle differences between SAH patients who are at increased risk for DID and those who are not is very important for the establishment of a therapeutic regimen that takes into account the prevention of DID as one of the most serious complications following SAH.

MATERIALS AND METHODS

The patient group numbered 127 individuals, all of whom suffered from aneurysmal SAH and all of which underwent open aneurysm surgery between January 1995 and July 1997 at the Department of Neurosurgery, Wroclaw Medical Academy, Poland. There were 63 women and 64 men in this group, with a mean age of 45 years (range 16 to 69 years). An initial computed tomography (CT) scan was performed in each case at admission. The amount of blood clots seen in the subarachnoid space was classified according to the Fisher grading14. In patients without signs of SAH in CT scans a spinal tap was performed to define blood in the CSF and confirm SAH. The existence and location of a cerebral aneurysm was confirmed by angiography. In all patients included in this study the ruptured aneurysm had to be located in the anterior cerebral circulation. The neurological status of patients upon admission was classified according to the criteria of Hunt and Hess15 and was repeatedly re-evaluated. Patients who presented with intracerebral hematoma causing a mass effect and/or with pathologically elevated intracranial pressure (ICP > 25 mmHg) were excluded from this study. Patients in poor neurological state corresponding to group V of the Hunt-Hess classification were also not included in this analysis. All patients underwent surgery within 72 h after SAH and underwent daily follow up TCD-examinations for clinical monitoring during a period of up to 21 days after admission. The appearance of a delayed ischemic deficit in some patients prompted us to evaluate the TCD-data for vasospasm in the course of treatment of patients for SAH. Patients were assigned to one of two groups depending on the occurrence of a delayed ischemic deficit within the post-operative period: patients with DID are assigned to group DID+ (n=20), all patients without any DID are assigned to group DID- (n= 107). All patients who presented with DID had to undergo another CT scan, which was performed without further delay on the day of the first appearance of neurological symptoms. This was done in order to exclude other reasons, which could account for the worsening of their clinical status, such as hydrocephalus, secondary bleeding, severe edema or hematoma. Blood gas analysis, electrolyte-screening, serologic and cardiologic tests were routinely performed to exclude systemic causes of the deterioration of a patient's condition. The following criteria have been applied to define delayed ischemic deficits in this setting:

1. Sudden appearance of new focal neurological symptoms more than 48 h after surgery.

2. Confident exclusion of other clinical causes which could account for the development of neurological symptoms

3. Simultaneous presence of a hypodense supposedly ischemic zone in a follow-up CT scan performed within 24 h after first presenting signs for a newly developed delayed ischemic deficit.

Sequential cerebral blood flow (CBF) measurements were made using standard transcranial Doppler ultrasound equipment employing a 2 MHz probe (TCD 2000 S/N; Eden Medizinische Electronic, Ueberlingen, Germany). In order to assess the patients for arterial stenosis or arterial occlusion, all cerebral arteries were examined and documented in each patient upon admission. In each patient, TCD examinations were performed daily until the 21st day after SAH with complete documentation of mean flow velocities (MFV) in both MCA and ACA on the operated as well as the nonoperated side. A mean flow velocity (MFV) higher than 120 cm sec^sup -1^ was considered to be diagnostic of vasospasm in MCA. In agreement with previous studies , we suggest that, value of MFV > 90 cm sec^sup -1^ were indicative for the diagnosis of vasospasm when obtained in ACA. In patients with MFV >120 cm sec^sup -1^ in MCA an additional measurement of MFV were performed in the extracranial part of ICA to eliminate a possible influence of hyperemia on cerebral blood flow. According to Lindegaard et al. the ratio of values of MFV in MCA to values of MFV obtained in the extracranial portion of ICA can be used as an index of hyperemia. If this ratio is above three the result suggests vasospasm, if it is lower than three it suggests hyperemia. TCD measurements were performed only by highly experienced personnel.

All patients were treated with calcium channel blocker-nimodipin (Nimotop, Bayer, Germany) applied at an initial i.v. dose of 2 mg h^sup -1^. This treatment regimen was maintained for an average period of 14 days and was then changed to an oral dose of 6x 80 mg day^sup -1^. As a prerequisite for data inclusion in this study the following criteria had to be met: patients had to present with normal pCO^sub 2^ and pO^sub 2^ and had to have a stable cardiovascular status with an arterial blood pressure within the range of 140-160/80 mmHg, a heart rate of 65-90/min as well as a normal hematocrit (38%-48%). An intensive care protocol with 'HHH' therapy (induced hypertension, hypervolemia, hemodilution) was mandatory for those patients in whom an increase of MFV > 120 cm sec^sup -1^ in MCA was observed. The intravascular blood volume was raised by an increase of the i.v. fluid supply up to 3000-5000ml day^sup -1^ (using dextran, albumin and plasma preparations) to maintain a slightly elevated central venous pressure of 10-15 mmH2O. Systolic arterial blood pressure (ABP) was controlled to stay around 160 mmHg by administering ephedrine in a dose of 6-8x25 mg day^sup -1^ and/or dopamine at 5-- 15 g kg^sup -1^ min^sup -1^ when needed.

Statistical analysis was performed using a package of computer programs Instat. Mean MFV measured values from each patient were used for statistical analysis and compared to values obtained from all other patients. Comparison of the MFV values in MCA and ACA in patients with (DID+) and without DID (DID-) was performed in all patients classified according to the Fisher's group or Hunt-Hess group's. Analysis was performed using the Student t-test and with the exact Fisher's test with an approximation of Katz.

RESULTS

The neurological condition at admission and delayed ischemic deficits

A clinical neurological examination was performed on each patient on the day of admission and classified according to the Hunt-Hess scale. The distribution of these patients in H-H groups was ex-post segregated into DID+ and DID- groups and this is summarised in Table 1. As can readily be seen, most of the patients (40%) from group DID+ were classified as H-H group II, 35% belonged in H-H group III, and those in a serious condition upon admission (H-H group IV) constituted 25% of the DID + patients. There were no patients from H-H group I in the DID + group. In the DID- group the largest number of patients also belonged to H-H group II (37%) and fewer categorized in H-H group III (30%). Patients in a serious condition upon admission (H-H group IV) made up for 21 % of the DID- group and only a relatively small number of patients were in a good neurological condition (10% in H-H group I). Statistical analysis shows that the distribution of patients with DID is not statistically different (v^sup 2^ = 2.27, p= 0.518) across the H-H groups. The neurological status of a patient at admission has no predictive value with respect for the risk of DID development with the exception for all those patients categorized in H-H group I. The distribution of DID+ patients according to the H-H classification is not significantly different from distribution of DID-- patients classified accordingly.

Delayed ischemic deficits and extension of SAH

The distribution of patients in a group with and without DID according to the Fisher's grading is summarised in Table 2. The number of our patient population categorised in Fisher's groups II and IV was similar in both DID+ and DID- subgroups. They accounted for 60% (4 + 8 = 12/20) in the DID + group and 55% (14+45=59/107) in the DID- group respectively. This contrasts with the appreciable percentage of patients without SAH (Fisher's group I) in the DID- group (10/107=9%) which compares to a small number only (1/20= 5%) in the DID + group. Expressed as a percentage, the former number is almost twice as high. However, statistical analysis has not shown any significant difference (v^sup 2^ = 1,13, p= 0.76) in the correlation between the amount of blood clots in CT scan and occurrence of DID symptoms.

Occurrence of the delayed ischemic deficits

In our study, 20 patients (16%) among the total of 127 patients with aneurysmal SAH developed a delayed ischemic deficit. Since other causes could be confidently ruled out and based on our corroborative TCD results, we conclude that this symptomatology is caused by vasospasm following SAH. In the DID+ group 14 patients were female (70%) and 6 male (30%) with ages ranging from 24 to 62 years (mean 45 years). In the DID- group were 107 patients, 52 of them were women (49%) and 55 men (51%). Age in this group ranged from 16 to 72 years with a mean of 46 years. Angiography, which was performed directly upon admission, revealed vasospasm in only four cases (20%) among the 20 patients of group DID+ and none in DID- group. Of those, segmental vasospasm of the middle cerebral artery (MCA) was observed in two patients and generalised vasospasm in another two patients. In nine patients suspected of nondefinite clipping of aneurysm three weeks after SAH a control angiography was performed. Eight of them were further classified in DID + group and one in DID- group. The angiography shows segmental cerebral arteriostenosis in three of the nine patients: in MCA in two patients, and in ACA in one patient. In all patients included in this study, a control CT was performed routinely during 72 h after operation. In patients suspected of DID symptoms control CT were performed during first 24 h after new post-operation neurological deficits were demonstrated. An ischemic zone was observed in each of the DID+ patients and its localization correlated to the supply area of the respective artery, which was also identified and categorized as severely vasospastic by our TCD examination. Delayed ischemic deficits occurred in the DID+ group almost always between day 7 and 14 after SAH (on average 11th day after SAH). This corresponds well to the period within which the highest values of flow velocities were observed. In most DID+ patients (75% = 15/20) ischemic symptoms were preceded by a phase of maximal MFV seen on TCD examination. A MFV decrease to 50% of the normal was observed on the first day after the appearance of ischaemic symptoms. (Figures 1 and 2).

Comparison of the MFV changes in the DID + and DID- groups

Values of MFV in MCA and ACA were always higher in the DID+ group compared to those obtained in the DID- group and this will be well demonstrated every consecutive day of TCD-examination after SAH (Figures 1 and 2). Differences of MFV in MCA between the DID+ and DID- patient group were 27 cm sec^sup -1^ on the first day after SAH and 35 cm sec^sup -1^ on the last day. The greatest mean difference (about 50 cm sec^sup -1^) was noticed on the sixth day after SAH. The highest values of MFV were observed in both groups in between the seventh to 14th day after SAH. In the DID+ group the MFV reached maximal values of 181 +/- 17 cm sec^sup -1^ on the 10th day after SAH, while they remained lower in the DID- group being only at 138 +/- 11 cm sec^sup -1^ on the 11th day after SAH. Pathological MFV values persisted in DID+ patients throughout the entire observation period to the last day (day 21). A small decrease of MFV values was registered during the third week in the DID+ group when compared to patients of the DID-- group. MFV differences between both groups were statistically significant (p

MFV values in ACA were also higher in the DID+ group than in the DID- group (Figure 2). Significant differences between MFV obtained in ACA of both groups were found on the first day after SAH (17 cm sec^sup -1^) and on the last day (20 cm sec^sup -1^) of this study period. The highest difference in MFV values (25 cm sec^sup -1^) was noticed on the sixth day after SAH. Highly elevated MFV values were found in both groups from the seventh to the 13th day post SAH and these reached a maximum value of 119 +/- 14 cm sec^sup -1^ on the ninth day post SAH in the DID+ group and only 100 +/- 7 cm sec^sup -1^ on the 10th day after SAH in the DID-- group. The differences in MFV values between the DID+ and DID- groups were statistically significant (p

Vasospasm in TCD examination

Values of MFV > 120 cm sec^sup -1^ were accepted as indicative for vasospasm in MCA. Such pathologically elevated values were registered twice as often in the DID + group (18 days = 80% of monitoring time) as in the DID- group (9 days= 43% of monitoring time). The occurrence of DID symptoms in patients with a MFV > 120 cm sec^sup -1^ as opposed to MFV 90 cm sec^sup -1^ and MFV 90 cm sec^sup -1^ were taken as indicative for vasospasm in ACA. Based on this threshold, pathological values were observed for 18 days (= 86% of monitoring time) in the group DID + while they accounted for only seven days (=33% of monitoring time) in the DID-- group.

It can therefore be shown that the probability of developing a delayed ischemic deficit in a patient with SAH is higher when MFV values above 120 cm sec^sup -1^ in MCA are registered. This is surprising, since other patients who did fall into the same clinical categorization initially, remained stable throughout their hospital stay with their MFV values remaining below 120 cm sec^sup -1^. The reason for this difference in clinical course remains unclear. A subsequent relative risk (RR) analysis of DID occurrence demonstrates accordingly that delayed ischemic deficits in patients with MFV > 120 cm sec^sup -1^ in MCA occur more than three times as often (RR = 3.33) as for patients with MFV 90 cm sec^sup -1^ in ACA this relative risk increases even more to about four fold (RR= 4.08).

The numbers of patients presenting with a MFV > 120 cm sec^sup -1^ and MFV 120 cm sec^sup -1^ was observed in patients from both the DID+ and DID-- group. However, during our observation interval of 21 days, MFV values > 120 cm sec^sup -1^ were noted in up to 95% of all patients in the DID+ group whereas such values from the DID- group were registered in only 60% of patients. The highest number of patients with a MFV > 120 cm sec^sup -1^ was observed in the DID + group on the ninth day after SAH comparing to a peak on the 12th day in the DID- group. Thus, MFV values above 120 cm sec^sup -1^ occurred earlier in patients of group DID+ if compared to patients of group DID-. The fraction of patients with MFV >120 cm sec^sup -1^ was almost twice as high in the group DID+ than in the DID- group. The v^sup 2^ test was used to evaluate the occurrence and distribution of DID symptoms in patients with MFV > 120 cm sec^sup -1^ and MFV 120 cm sec^sup -1^ and those with MFV

Comparison of the MFV changes on the operated and nonoperated side

We finally compared MFV values obtained in the operated side to the nonoperated side. As can be seen those differences were higher in the DID+ group than in the DID- group. This result holds true for examinations of both the MCA (Figure 4A,B) as well as of the ACA (Figure 5A,B). Such differences were greater for those values obtained in MCA as compared to values from ACA and this result could be shown for patients of both the DID+ and DID- groups. The biggest MFV difference was 63 cm sec^sup -1^ in the DID+ group and 40 cm sec^sup -1^ in the DID- group respectively if we look at the MCA examination. Corresponding values for the ACA were: 24 cm sec^sup -1^ in the group DID+ and 23 cm sec^sup -1^ in the group DID-. Values from both groups differed the most within the initial 14 days after SAH with a peak discrepancy between days 6 and 14 after SAH. In the third week after SAH the MFV differences between those groups became smaller and no longer reached significant levels. Statistical analysis showed that the differences in MFV values between the operated and nonoperated sides are significant (p

DISCUSSION

The most serious complication after vasospasm is the occurrence of a delayed ischemic deficit. It accounts for the majority of cases where disability and death occur in patients operated after aneurysmal SAH. The frequency of DID has been reported to be as low as 2% but up to 39% in this group of patients14,16-18. The higher numbers from the earlier years have dropped considerably with the introduction of new methods of preventing and treating vasospasm, e.g. the application of selective calcium blockers.

What do we know so far? Eighteen patients out of 34 patients described by Davis et al.19 developed clinical signs of delayed ischemia. Of those, 10/34 showed a lateralizing deficit (29%) whereas a non lateralizing neurological deterioration was observed in the remaining 8/34 patients (24%). Grosset et al.20 reported in their study that DID occured in 112 out of a total of 323 patients (39%) after SAH. Harders et al.1 3 observed DID in 7 of 50 patients (14%) after SAH, but this number dropped to only 3 of 40 patients (7%) in those who were treated with a hypertensive regimen. In another study, Harders et al.21 reported that DID symptoms appeared in only 1 out of 47 SAH patients in a clinical setting in which patients were treated with nimodipine (2 mg h^sup -1^).

The precise definition and hence predispositions of delayed ischemic deficits remains controversial. According to Grosset et a1.20 DID symptoms can confidently be diagnosed in patients after SAH if other reasons (such as hydrocephalus, secondary bleeding, electrolytic, gazometric and circulatory-respiratory disorders) have been eliminated in a patient with worsening neurological condition, and none of the DID symptoms were observed in this patient previously. It is furthermore required that the onset of DID symptoms corresponds with new ischemic zones as demonstrated in CT examinations. Kassel et al.22 defined DID symptoms as a suddenly appearing syndrome of impaired neurological functions including consciousness disorders with a most frequent occurrence between days 4 and 9 after SAH. They also pointed to the characteristic presence of ischemic zones seen on accompanying CT-scans. Laumer et al.23 have paid attention to the differentiation of initial changes after hemorrhage, appearing within the first 48 h after SAH and those true ischemic symptoms (defined as DID) which appear perioperative. Davis et al.19 suggested that once other causes were excluded a clinical diagnosis of DID can also be made when patients present with a depressed state of consciousness only with or without focal neurological deficits (such as frontal dysfunction with disturbance of the senses, incontinence and changes in reflexes).

Grotte and Hassler 24, Romner et al.25 and McCormick et al.11 observed among others an impairment of the cerebral circulation due to secondary hemorrhage in the follow-up of SAH patients who did not undergo aneurysm surgery ('nonclipped aneurysm patients'). A sudden increase of ICP caused by further hemorrhage may also influence the occurrence of acute ischemia compromising an already impaired cerebral circulation. Ischemic symptoms may show as early as within 24 h after SAH as a direct consequence of the initial hemorrhage and may not be a delayed complication due to some developing vasospasm. Neurological deficits found in SAH patients within the 24 h after a SAH or in cases involving the ventricular system or with parenchymal hematoma have also to be discussed in the light of direct damage to cerebral tissue as a result of the SAH. For this reason they are not included in our study group. Because a clear differentiation between such primary and secondary ischemic deficits due to vasospasm is not always possible, it had been postulated to discard the analysis of DID symptoms in the clinical setting of acute post-hemorrhage hydrocephalus or in patients who present with intracerebral hematoma20,23,26. Furthermore the influence of intra-operative and peri-operative complications (e.g. intraoperative hemorrhage, prolonged use of temporary clips, incorrect positioning of a clip leading to a iatrogenic or induced stenosis, disruption of arterial blood supply, spatula retraction-injury of the brain or more general sequelae of anesthesia) should be considered in the evaluation of neurological symptoms within the first 24 h after surgery27-30. Doubts have to be raised whether it may be harmful for a patient to undergo surgery in a 'period at risk for vasospasm', i.e. between the third and 1 Oth day after SAH. During that vulnerable period surgical manipulation on the blood vessels may more easily provoke an even stronger vasospasm and may cause amplification of ischemic complications. Cerebral edema, intracranial hematoma and secondary hemorrhage may further cause worsening of the neurological condition of patients after the operation12,20,22.

At present it remains difficult to distinguish so called 'early ischemic symptoms' from 'delayed ischemic symptoms'. Therfore a diagnosis of 'delayed ischemic deficits' should be made with great caution and only after careful exclusion of other causes for a worsening neurological status. With this precaution in mind, we recommend the following approach which we adopted in this study: A DID is only a DID if:

1. It shows a sudden onset of new focal neurological symptoms along with consciousness disorders later than 48 h after operation.

2. It withstands confident exclusion of any other reasons than vasospasm which may account for the symptoms.

3. A new ischemic zone can be seen in follow up CT scans within 24 h after the onset of clinical symptoms.

Among our total of 127 patients a DID was unequivocally diagnosed in 20/127 (16%) cases, a finding which is within the range reported by other authors3,13,20,21,31,32. All of our patients in whom DID symptoms occurred had undergone surgery for SAH. Some authors emphasized that ischemic symptoms can be observed more often in patients who were operated on early (i.e. within 72 h after SAH as in our group) if compared to those patients operated later (i.e. after 10 days)9,17,18,33,34. However, this is most probably related to the fact that the greatest risk of vasospasm lies within the first two weeks after SAH. In this study higher MFV values were generally observed in patients from the DID+ group when compared to those in the DID-- group. The differences between our two groups remained statistically significant during the entire observation period after SAH. The maximum difference (50 cm sec^sup -1^) between the MFV values in both groups was observed on day 6 post-SAH. Values of MFV increased steadily within the first seven days post-SAH in both groups with the steepest increase to be found in the DID+ group. This is in good agreement to the reports of other authors1,13,21,33 . The maximal value of MFV for DID+ patients was about 180 cm sec-1 whereas it was about 138 cm sec^sup -1^ in group DID-. These results are in good agreement with those reported by Grosset et al.20 who showed that their DID + patients (n=47) had a maximal MFV of about 186.6 cm sec^sup -1^ while their DID- patients (n=74) had a maximal MFV of about 149.5 cm sec^sup -1^. Most of the authors5,13,27,35,37 considered MFV values exceeding the threshold of 200 cm sec^sup -1^ as diagnostic for vasospasm which may subsequently lead to DID. In more recent studies a differing opinion has gained ground1,33. Some considerations are of interest here: Pucher et al.38 have shown by means of a computer model that narrowing of the proximal part of MCA to 60% causes an increase in MFV to as high as 350 cm sec^sup -1^ while overall cerebral blood flow (CBF) remains constant. Further narrowing of the vascular lumen to as much as 80% leads to a sudden drop in CBF, which in itself causes an inevitable drop in MFV below 100 cm sec^sup -1^. Comparably, elevated MFV values as observed in cases of severe vessel stenosis would reflect a drop in rCBF only if they exceeded such a critical value, and this would parallel the formation of focal ischemic zones as found in CT scans. Current models however do not include the cerebral autoregulation as a participating active regulatory mechanism following SAH. However, disruption of cerebral autoregulation has been observed in various cases by some authors1,12,39,40. In patients after SAN this dysregulation causes most frequently a drop in CBF and only rarely a transient rise, which is generally described as general hyperemia.

Autoregulation disturbances after SAH should provide a sufficiently large signal reflecting the increased overall blood flow velocity (BFV) that should also be susceptible to transcranial Doppler examination. This should be sustained even during HHH therapy and/or in a setting of equivocal TCD data. Mano et al.41 found in 10 out of 19 patients that BFV changes by more than 15% in TCD examinations with concomitant changes in the mean arterial blood pressure (MABP). Similarity, Levy et al.42 described a 'paradoxical increase' in MFV (when measured by TCD). Interestingly they showed a reversal of neurological symptoms in SAH patients who were treated with dobutamine. The authors therefore recommended that all TCD examinations should be carefully documented including vital data representing blood pressure manipulation since only then is a meaningful interpretation of TCD measurements possible. A reduction of CBF during the initial phase after SAH (day 1) has been corroborated in paralleled SPECT and PET studies12,32. Lower MFV values in the first day after SAH are most likely related to a drop in CBF as a result of the initial hemorrhage and its subsequent increase in intracranial pressure (ICP) Another interesting phenomenon related to the occurrence DID is a sudden fall in MFV observed often within 24 h after DID symptoms development. In our DID+ group, an average drop in MFV by 28 cm sec^sup -1^ was observed within the first 24 h after DID in four patients. The mechanism explaining that interesting observation may be explained as follows46: critical narrowing as a results of vasospasm can lead to the formation of a cerebral ischemic zone with a subsequent fall in regional CBF. Simultaneously an increase in intracranial pressure takes place, which furthermore diminishes rCBF. As a result a pronounced drop in MFV in the already vasospastic artery may be observed. Davis et al.19 found a similar incidence of MCA vasospasm in TCD examination and hypoperfusion in a SPECT study of a group of 20 patients examined after SAH who developed delayed neurological deterioration. However, it seems that in patients with reversible DID such MFV reduction is not observed1,36. Possibly a reversible drop in CBF takes place in those cases but is normalised after initiation of intensive care treatment (e.g. HHH).

Romner et al.25 examined a large group of 500 patients after SAH and described reduced MFV values in 57% patients within the first 12 days after SAH. The normal MVF range could be demonstrated in 38% of their patients and a slightly increased value in 5%. The authors observed in the days following a gradual rise in MFV also representing developing vasospasm, which was then confirmed by angiography. Most authors1,13,20,35,36 accept a MFV value of 120 cm sec^sup -1^ in MCA as indicative for vasospasm. This correlates well with an increase in DID occurrence in patients presenting with critically high MFV, who suffered at the same time from autodysregulation6,7,12,28,40. However, some authors (e.g. Compton et al.35) did not observe ischemic symptoms in patients in whom MFV above 120 cm sec^sup -1^ were found. Similarly, Fahmy et al.43 found no substantial disorders of the cerebral circulation even in patients with MFV exceeding 200 cm sec^sup -1^. In this study we were able to clearly demonstrate that values of MFV > 120 cm sec^sup -1^ can be registered twice as long in our DID+ group compared to group DID-- (86% of our monitoring time versus 43% of monitoring time). Within this period of 21 days, MFV values above 120 cm sec^sup -1^ were measured in up to 95% of patients in group DID+ while only 60% of patients fell above this critical mark in the DID- group. The DID symptoms appeared over three times more often in patients with such increased MFV > 120 cm sec^sup -1^ than in patients with MFV 120 cm sec^sup -1^ was found on the ninth day after SAH whereas this peak was delayed to the 12th day in DID- patients. These findings are in good agreement to reports by other researchers1,13,20,36. Analysis of the relative risk of DID occurrence in patients with MFV > 120 cm sec^sup -1^ have shown that between the third and the 13th day after SAH the risk increases substantially, reaching its maximum value around the 13th day after SAH in our patient cohort. At this particular time post SAH the risk of DID appearance is almost 10 times higher in patients with MFV >120 cm sec^sup -1^ than in those with MFV

How can we assess this risk, when we state above that the maximum risk is at day 13 after SAH, and here after about one week we find only detrimental complications? In patients in whom a DID had developed, a greater daily increase of MFV is found within the first week after SAH if values are compared to patients who did not develop a DID1,10,43,44.

In our group of patients the occurrence of DID symptom was preceded by a daily increase of MFV ranging from 8 to 540 cm sec^sup -1^ day^sup -1^ (mean 150 cm sec^sup -1^ day^sup -1^) observed between the third and ninth day after SAH. Harders et al.33 as well as Seiler et al.37 have shown that a daily MFV increase of more than 200 cm sec^sup -1^ (between the third and sixth day after SAH) reflects the patient as being at a substantial threat for vasospasm and thus subsequent DID occurrence. Taken to the extreme and according to Grosset et al.4 an increase in MFV above 500 cm sec^sup -1^ day^sup -1^ is associated with a more than 60% risk of developing DID. Some authors have reported a sudden increase of MFV during the days before DID1,36. In 65% of our DID+ patients (n = 13/20) the daily MFV increment was greater than 200 cm sec^sup -1^ day^sup -1^ on the day preceding the DID and even higher than 500 cm sec^sup -1^ day^sup -1^ in 15% of the patients (n = 3/20). Grosset et al.3 have investigated a group of 121 patients after SAH to assess for a possible correlation between rapidly rising MFV values as obtained in TCD measurements and a perfusion deficit as imaged by SPECT. They found in 14 out of their 15 patients with a rapid rise MFV (more than 500 cm sec^sup -1^ day^sup -1^) an abnormal regional CBF pattern in the SPECT examination and this was demonstrable before patients developed any DID. In all cases examined, the sites of SPECT perfusion deficits coincided with the location of CT region where vasospasm artery was found. A different finding is reported by Laumer et a1.23 who observed such a steep increase of MFV in only 3 out of 11 patients the day before a DID developed. Furthermore, they did not find DID symptoms in some patients with MFV > 2500 cmsec^sup -1^. It has therefore to be assumed that even a daily increase of MFV as steep as 20-500 cm sec^sup -1^ day^sup -1^ may only be a relative predictor for the risk of DID occurrence but is not an absolute criterion to indicate imminent manifestation. Brent et al.45 mentioned that increased MFV values as revealed by TCD examination correlated well with an increased rCBF as seen in Xenon enhanced computed tomography (XE/CT), but it did not correlate with signs of ischemia. No difference in CBF was found in MCA territories with rapid MFV increase (> 50 cm day^sup -1^) if compared to territories without such increase in MFV. Though most of the authors conclude that a high MFV predisposes to DID, there is still discussion about the factors leading to a symptomatic clinical course1,36, Even serious vasospasm does not necessarily lead to the creation of an ischemic zone, particularly if a well-- developed collateral circulation is present to compensate. In young individuals a tendency towards higher MFV can be observed but also a higher tolerance to vasospasm is found, which probably relates to a relatively high CBF in those patients1. Furthermore, several authors have pointed out that besides vasospasm, an essential role in the manifestation of DID has to be played by other factors: disorders of autoregulation, state of collateral circulation, current CBF, hematocrit, pCO^sub 2^, blood pressure and volume status, patient age and possible coexisting metabolic and/or cardiovascular diseases 1,12,22,27. Raab et al.47 identified younger age as a positive predictor for the risk of symptomatic vasospasm. Brint et al.45 also suggested that age is a significant factor in predicting normalisation of MCA velocities after SAH. They considered older age to be associated with shorter time-spans to MFV normalisation. Multivariate analysis in another group of 244 SAH patients undergoing either surgical or endovascular treatment revealed that the probability of symptomatic vasospasm decreased with age > 50 years and severe neurological condition at admission, but increased with hyperglycaemia48. The influence of pharmacological treatment (mannitol, nimodipine, and dextran) should also not be ignored. It has to be kept in mind that TCD examination does not register flow in the periphery of cerebral arteries where ischemic zones will form first.

This study supports the notion that careful TCD examination if frequently performed may help to identify early those patients at risk for a delayed neurological deficit after SAH. We therefore suggest that the following criteria should be included in any algorithm that initiates the treatment of vasospasm post-SAH to further lower the risk of DID occurrence:

1. MFV increase over 1200cm sec^sup -1^ on the fourth day post-SAH.

2. MFV reaching 160 to 2000cm sec^sup -1^ between seventh and 15th day post-SAH.

3. Average daily MFV increase within the first week > 150 cm sec^sup -1^ day^sup -1^.

4. Difference in MFV between operated and nonoperated side > 400 cm sec^sup -1^.

REFERENCES

1 Sloan MA. Detection of vasospasm following subarachnoid haemorrhage. In: Babikan VL, Wechsler LR, eds. Mosby-Year Book, St. Louis, MO: Mosby, 1993: pp. 105-127

2 Suzuki J, Komatsu S, Sato T, Sakurai Y. Correlation between CT findings and subsequent development of cerebral infarction due to vasospasm in subarachnoid haemorrhage. Acta Neurochir (Wien) 1980; 55:63-78

3 Grosset DG, Straiton J, du Trevou M, Bullock R. Prediction of symptomatic vasospasm after subarachnoid haemorrhages by rapidly increasing transcranial Doppler velocity and cerebral blood flow changes. Stroke 1992; 23: 105-214

4 Grosset DG, Straiton J, McDonald I, Bullock R. Angiographic and Doppler diagnosis of cerebral artery vasospasm following subarachnoid haemorrhage. Br] Neurosurg 1993; 7: 291-298

5 Handa Y, Kubota T, Tsuchida A, Kaneko M, Cancer H, Kobayashi H. Effect of systemic hypotension on cerebral energy metabolism during chronic cerebral vasospasm in primates. J Neurosurg 1993; 78: 112-119

6 Mickey B, Vorstrup S, Voldby B, Lindewald H, Harmsen A, Lassen NA. Serial measurement of regional cerebral blood flow in patients with SAH using Xe inhalation and emission computerized tomography. J Neurosurg 1984; 60: 916-922

7 Ishii R. Regional cerebral blood flow in patients with ruptured intracranial aneurysms. J Neurosurg 1979; 50: 587-594

8 Kamiya K, Kuyama H, Symon L. An experimental study of the acute stage of subarachnoid hemorrhage. Neurosurg 1983; 59: 917-924

9 Meyer CHA, Lowe D, Meyer M, Richardson PL, Neil-Dwyen G. Progressive change in cerebral blood flow during the first three weeks after subarachnoid hemorrhage. Neurosurgery 1983; 12: 58-76

10 Hayashi S, Nehls DG, Kieck ChF, Vielma J, de Girolami U, Crowell RM. Beneficial effects of induced hypertension on experimental stroke in awake monkeys. J Neurosurg 1984; 60: 151-157

11 McCormick PW, McCormick J, Zabramski JM, Spetzler RF. Hemodynamics of subarachnoid hemorrhage arrest. J Neurosurg 1994; 80: 710-715

12 Diehl RR, Linden D, Lucke D, Berlit P. Phase relationship between cerebral blood flow velocity and blood pressure. A clinical test of autoregulation. Stroke 1995; 26: 1802-1804

13 Harders AG, Gilsbach J. Haemodynamic effectiveness of Nimodipine on spastic brain vessels after subarachnoid haemorrhage evaluated by the transcranial Doppler method. A review of clinical studies. Acta Neurochir Suppl 1988; 45: 21-28

14 Fisher CM, Kistler JP, Davis JM. Relation of cerebral vasospasm to subarachnoid hemorrhage visualised by computerized tomographic scanning. Neurosurgery 1980; 6: 1-9

15 Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair of intracranial aneurysms. J Neurosurg 1968; 28: 14-20

16 Allen GS, Ahn HS, Preziosi TJ. Cerebral arterial spasm - a controlled trial of nimodipine in patients with subarachnoid hemorrhage. N Engl I Med 1983; 308: 619-624

17 Ljunggren B, Brandt L, Kagstrom E, Sundbarg G. Results of early operations for ruptured aneurysms. J Neurosurg 1984; 54: 473-479

18 Ljunggren B, Brandt L, Saveland H, Nilsson PE, Conqvist ST, Andersson KE, Ninge E. Outcome in 60 consecutive patients

treated with early aneurysm. J Neurosurg 1984; 61: 864-873

19 Davis SM, Andrews JT, Lichtenstein M, Sossiter SC, Kaye AH, Hopper J. Correlation between cerebral arterial velocities, blood flow, and delayed ischaemia after subarachnoid hemorrhage. Stroke 1992; 23: 492-497

20 Grosset DG, Straiton J, McDonald I, Cockburn M, Bullock R. Use of transcranial Doppler sonography to predict development of a delayed ischaemic deficit after subarachnoid hemorrhage, J Neurosurg 1993; 78: 183-187

21 Harders AG, Gilsbach J. Time course of blood velocity changes related to vasospasm in the circle of Willis measured by transcranial Doppler ultrasound. J Neurosurg 1987; 66: 718-728

22 Kassel NF, Shaffrey ME, Shaffrey Cl. Cerebral vasospasm following aneurysmal subarachnoid haemorrhage. In: Apuzo MLJ, ed. Brain Surgery, I, Churchill-Livingstone, 1993: pp. 847-856

23 Laumer R, Steinmeier R, Gonner F, Vogtmann T, Priem R, Fahlbusch R. Cerebral hemodynamics in subarachnoid haemorrhage evaluated by Transcranial Doppler Sonography, Part 1. Reliability of flow velocities in clinical management. Neurosurgery 1993; 33:1-9

24 Grotte E, Hassler W. The critical first minutes after subarachnoid hemorrhage. Neurosurgery 1967; 27: 265-267

25 Romner B, Ljunggren B, Brandt L, Saveland H. Transcranial Doppler sonography within 12 hours after subarachnoid hemorrhage. I Neurosurg 1989; 70: 732-736

26 Kassel NF, Boarini DJ. Patients with ruptured aneurysm: Preand postoperative management. In: Wilkins RH, Rengachary EE, eds. Neurosurgery, Vol 2, New York: McGraw-Hill, 1985: pp. 1367-1371

27 Enblad P, Persson L. Impact on clinical outcome of secondary brain insults during the neurointensive care of patients with subarachnoid haemorrhage: A pilot study. J Neurol Neurosurg Psychiatry 1997;62:512-516

28 Find lay JM, MacDonald RL, Weir BKA, Grace MG. Surgical manipulation of primate arteries in established vasospasm. I Neurosurg 1991; 75: 425-432

29 Gotoh 0, Tamura A, Yasui N, Suzuki A, Hadeishi H, Sano K. Glasgow coma scale in the prediction of outcome after early aneurysm surgery. Neurosurgery 1996; 39: 19-26

30 Le Roux PD, Elliott JP, Newell DW, Grady MS, Winn HR. Predicting outcome in poor-grade patients with subarachnoid hemorrhage: A retrospective review of 159 aggressively managed cases. J Neurosurg 1996; 85: 39-49

31 Gromulid P, Seller RW, Aaslid R, Huber P, Zurbruegg H. Evaluation of cerebrovascular disease by combined extracranial and transcranial Doppler sonography: Experience in 1039 patients. Stroke 1987; 18: 1018-1024

32 Grosset DG, Bone I, Bullock R. Vasospasm after subarachnoid haemorrhage: Transcranial Doppler monitoring predicts focal blood flow changes on SPECT and neurological deficit. In: Sano K, Takakaura K, Kassel NF, Sasaki T, eds. Cerebral Vasospasm, Tokyo, Japan: Tokyo Press, 1990: pp. 18-20

33 Handers AG, Gilsbach J, Hornyak ME. Incidence of vasospasm in transcranial Doppler sonography and its clinical significance: A prospective study in 100 consecutive patients who were given intravenous Nimodipine and who underwent early aneurysm surgery. In: Wilkins RH, ed. Cerebral Vasospasm, New York: Raven Press, 1988: pp. 33-52

34 Sloan MA, Haley EC, Kassel NF, Henry ML, Stewart SR, Beskin RP, Sevilla EA, Torner IC. Sensitivity and specificity of transcranial Doppler ultrasonography in the diagnosis of vasospasm following subarachnoid haemorrhage. Neurology 1989; 39: 1514-1518

35 Compton GS, Redmond S, Symon N. Cerebral blood velocity in subarachnoid haemorrhage: A transcranial Doppler study. I Neurol Neurosurg Psychiatry 1987; 50: 1499-1503

36 Kilic T, Pamir N, Ozek MM, Zirh T, Erzen C. A new, more dependable methodology for the use of transcranial Doppler ultrasonography in the management of subarachnoid haemorrhage. Acta Neurochir (Wien) 1996; 138: 1070-1078

37 Seiler RW, Newell DW. Subarachnoid haemorrhage and vasospasm. In: Newell DW, Aaslid R, eds. Transcranial Doppler, New York: Raven Press, 1992: 101-107

38 Pucher RK, Auer LM. Effects of vasospasm in the middle cerebral

artery on flow velocity and volume flow. A computer simulation. Acta Neurochir (Wien) 1988; 93: 123-128

39 Aaslid R, Huber P, Nornes H. Evaluation of cerebrovascular spasm with transcranial Doppler ultrasound. J Neurosurg 1984; 60: 37-41 40 Powers WJ, Grubb RL, Baker RP, Mintun MA, Raichle ME. Regional

cerebral blood flow and metabolism in reversible ischemia due to vasospasm. Determination by positron emission tomography. J Neurosurg 1985; 63: 539-546

41 Manno EM, Gress Dr, Schwamm LH, Diringer MN, Ogilvy CS. Effects of induced hypertension on transcranial Doppler ultrasound velocities in patients after subarachnoid hemorrhage. Stroke 1998; 29:422-428

42 Levy ML, Rabb CH, Zelman V, Giannotta SL. Cardiac performance enhancement from dobutamine in patients refractory to hypervolemic therapy for cerebral vasospasm. J Neurosurg. 1993; 79: 494-499

43 Fahmy MA, Smith RR. Identification of presymptomatic vasospasm by transcranial Doppler sonography. Stroke 1992; 23: 156

44 German JW, Gross CE, Giclas P, Watral W, Bednar NM. Systemic complement depletion inhibits experimental cerebral vasospasm. Neurosurgery 1996; 39: 141-146

45 Brint SU, Yoon WB, Hier DB, Ausman JI, Charbel F. Normalisation of transcranial Doppler middle cerebral artery velocities after aneurysmal clipping. Surg Neurol 1997; 47: 541-546

46 Newell DW, Grady MS, Eskrige JM, Winn HR. Distribution of angiographic vasospasm after subarachnoid hemorrhage: Implications for diagnosis by TCD. Neurosurgery 1990; 27: 574-577

47 Rabb CH, Tang G, Chin LS. A statistical analysis of factors related to symptomatic cerebral vasospasm. Acta Neurochor (Wien) 1994; 127: 27-31

48 Charpentier C, Audibert G, Guillemin F, Civit T, Ducrocq X, Bracard S, Hepner H, Picard L, Laxenaire MC. Multivariate analysis of predictors of cerebral vasospasm occurence after aneurysmal subarachnoid hemorrhage. Stroke 1999; 30: 1402-1408

49 Jarus-Dziedzic K, Zub W, Wronski J, Juniewicz H, Kasper E. The relationship between cerebral blood flow velocities and the amount of blood clots in computed tomography after subarachnoid haemorrhage. Acta Neurochir (Wien) 2000; 142: 309-318

50 Jarus-Dziedzic K, Bogucki J, Zub W. The influence of ruptured cerebral aneurysm localization on the blood flow velocity evaluated by transcranial Doppler ultrasonography. Neurol Res 2001; 23: 23-28

Katarzyna Jarus-Dziedzic*, Henryk Juniewicz^, Jerzy Wron-ski^^, Wojciech Leslaw Zub^^, Ekkehard Kasper$, Mariusz Gowacki* and Janusz Mierzwa^^

*Department of Neurosurgery, Medical Research Centre, Polish Academy of Sciences, Warsaw

^Faculty of Electronics, University of Technology, Wroclaw

^^Department of Neurosurgery, Wroclaw Medical Academy, Wroclaw, Poland

$Department of Neurosurgery, Massachussetts General Hospital, Boston, MA, USA

Correspondence and reprint requests to: Katarzyna Jarus-Dziedzic, MD, Department of Neurosurgery, Medical Research Centre, Polish Academy of Sciences, Ceglowska 80, 01-809 Warsaw, Poland. [neuropan@cmdik.pan.pl] Accepted for publication April 2002.

Copyright Forefront Publishing Group Sep 2002
Provided by ProQuest Information and Learning Company. All rights Reserved