clinical vasospasm after subarachnoid hemorrhage: …improvement by at least 1 neurologic grade, 24%...

365

Clinical Vasospasm After SubarachnoidHemorrhage: Response to HypervolemicHemodilution and Arterial Hypertension

Issam A. Awad, L. Philip Carter, Robert F. Spetzler,

Marjorie Medina, and Fred W. Williams Jr.

Delayed neurologic deterioration from vasospasm remains the greatest cause of morbidity and mortal-ity following subarachnoid hemorrhage. The authors assess the incidence and clinical course ofsymptomatic vasospasm following subarachnoid hemorrhage using a uniform management protocolover a 24-month period. One hundred eighteen consecutive patients were admitted to the neurovascu-lar surgery service within 2 weeks of subarachnoid hemorrhage not attributed to trauma, tumor, orvascular malformation (113 patients had aneurysms). Early surgery was performed whenever possi-ble, and hypertensive hypervolemk hemodilution therapy was instituted at the first sign of clinicalvasospasm. Forty-two patients (35.6%) developed characteristic signs and symptoms of clinical vaso-spasm with angiographic verification of spasm in 39 cases. All patients with clinical vasospasmreceived hypervolemic hemodilution therapy aiming for a hematocrit of 33-38%, a central venouspressure of 10-12 mm Hg (or a pulmonary wedge pressure of 15-18 mm Hg), and a systolic arterialpressure of 160-200 mm Hg (120-150 mm Hg for undipped aneurysms) for the duration of clinicalvasospasm. Over the course of treatment, 60% of patients with clinical vasospasm had sustainedimprovement by at least 1 neurologic grade, 24% maintained a stable neurologic status, and 16%continued to worsen. At the end of hyper volemic hemodilution therapy, 47.6% had become neurologi-cally normal, 33.3% had a minor neurologic deficit, and 19% had a major neurologic deficit or weredead. There were 3 Instances of cardiopulmonary deterioration (7%), all of which were in patientswithout Swan-Ganz catheters, and all resolved with appropriate diuresis. One patient rebled and diedwhile on hypervolemic hemodilution therapy. Death or major neurologic deficit from clinical vaso-spasm occurred in <7% of all patients with subarachnoid hemorrhage. This compares favorably withthe morbidity and mortality attributed to vasospasm in recent reports. The authors conclude thatearly surgery and aggressive management of clinical vasospasm with hypervolemic hemodilutiontherapy can be accomplished with minimal morbidity. This management strategy may lower theincidence of death and disability from vasospasm after subarachnoid hemorrhage. (Stroke1987;18:365-372)

DELAYED neurologic deterioration from vaso-I spasm remains the greatest cause of death and

major disability following subarachnoid hem-orrhage (SAH).1-* It has recently been estimated that7% of patients reaching referral centers will die or beseverely disabled by recurrent hemorrhage, while 14%will succumb to or be maimed by ischemic complica-tions of vasospasm.3 While early surgery can be ex-pected to effectively minimize rebleeding, there con-tinues to be concern that such intervention mightaggravate and worsen the outcome from vasospasm.Others suggest that early surgery might allow moreaggressive medical management of clinical vasospasm(in the absence of an undipped aneurysm), and might

From the Neurovascular Surgery Service, Division of Neurosur-gery, Barrow Neurological Institute, Phoenix, Arizona.

Presented at the 11th International Joint Conference on Strokeand Cerebral Circulation, San Francisco, California, February 6-8,1986.

Address for reprints: L. Philip Carter, MD, Editorial Office,Barrow Neurological Institute, 350 West Thomas Road, Phoenix,AZ 85013.

Received September 15, 1986; accepted October 31, 1986.

prevent vasospasm in some patients by clearing spas-mogenic substances from the subarachnoid space.7"10

Since 1983, a uniform protocol for the managementof SAH has been adopted at the Barrow NeurologicalInstitute. Early surgery is performed whenever possi-ble, and patients are aggressively treated with hyper-tensive hypervolemic hemodilution at the first sign ofclinical vasospasm. The objective of this report was toassess the incidence of clinical vasospasm with thismanagement strategy, to evaluate the response of thiscondition to hypervolemic hemodilution therapy, andto define and discuss management complications.

Subjects and MethodsFrom January 1984 to December 1985, 118 con-

secutive patients (ages 20-77; median age 49) wereadmitted to the Neurovascular Surgery Service at theBarrow Neurological Institute within 2 weeks of SAHnot attributed to trauma, tumor, or vascular malforma-tion. The diagnosis was confirmed by computed to-mography (CT) or lumbar puncture in all cases. Allpatients had complete cerebral angiography via retro-grade femoral artery catheterization on the day of ad-

by guest on April 12, 2017

http://stroke.ahajournals.org/D

ownloaded from

http://stroke.ahajournals.org/

366 Stroke Vol 18, No 2, March-April 1987

mission. An aneurysm was demonstrated on the initialangiogram in 112 cases and on a subsequent angio-gram 1 week later in 1 case. In the remaining 5 pa-tients, no aneurysm was demonstrated.

Patients were managed according to a uniform pro-tocol (Figure 1). Craniotomy was performed within 24hours of admission unless the patient already exhibitedsigns and symptoms of clinical vasospasm or was oth-erwise a poor surgical candidate (comatose withoutmass lesion, unstable hemodynamic status, etc.). Atsurgery, it was uniform policy to open the arachnoidcisterns widely and to clear all accessible subarachnoidclot in addition to clipping or wrapping the aneurysm.

Clinical vasospasm was defined as delayed neuro-logic deterioration that could not be attributed torebleeding, hydrocephalus, intracerebral hematoma,electrolyte abnormalities, or toxic and metabolic fac-tors.'•3-311 It usually followed a stereotyped course,with gradual blunting of the level of consciousnessstarting 3-10 days after SAH, followed within hoursby focal neurologic deficits. At the first suggestion ofclinical vasospasm, a central venous catheter was in-troduced; a pulmonary artery catheter (Swan-Ganzcatheter) was used if the patient had preexisting cardiacor pulmonary disease. Simultaneously, necessary di-agnostic tests were obtained to rule out other causes ofdelayed neurologic deterioration. Once the diagnosisof clinical vasospasm was established, hypervolemichemodilution was initiated under the guidance of aconsulting cardiologist.

The hemodynamic objectives consisted of raising

the central venous pressure to 10-12 mm Hg (or thepulmonary wedge pressure to 15-18 mm Hg), andlowering the hematocrit to 33-38%. This was accom-plished using colloid agents (Plasmanate, dextran, He-tastarch, etc.) in volumes not exceeding 1500 ml perday. To accomplish the desired hemodynamic objec-tives, whole blood transfusion or phlebotomy was usedas needed in anemic and polycythemic patients, re-spectively. In addition, the systolic arterial pressurewas maintained in the 160-200 mm Hg range if theaneurysm had been clipped (hypertensive hypervole-mic hemodilution) and in the 120-150 mm Hg range inthe presence of an undipped aneurysm (normotensivehypervolemic hemodilution). Since many patientswith SAH have a spontaneous tendency toward arterialhypertension, the desired blood pressure objectiveswere usually accomplished passively by withdrawingsome or all antihypertensive agents. Only 16 patientsneeded additional active intervention to reach thedesired hypertension. In these cases, i.v. dopamine(Intropin) or other pressors were used at the discretionof the consulting cardiologist.

Intensive monitoring and hypervolemic hemodilu-tion therapy were maintained until the resolution ofclinical vasospasm. This was typically signaled by agradual clearing of the sensorium, disappearance orstabilization of focal neurologic deficits, and simulta-neous resolution of angiographic spasm. Adjunctivetreatment consisted of ventricular or lumbar cerebro-spinal fluid drainage at the first sign of ventriculome-galy or external hydrocephalus.

SUBARACHNOID HEMORRHAGE

CT scan, start hydratlon with crystalloidsEARLY ANGIOGRAPHY

SEVERE ANGIOGRAPHICSPASM

NO SEVERE ANGIOGRAPHICSPASM

SEVERECLINICAL SPASM

CVP or Swan-Ganz line

NORMOTENSIVEHEMODILUTION

Resolution of spasm

DELAYED SURGERYFOR CLIPPING

OF ANEURYSM

NO SEVERECLINICAL SPASM

EARLY SURGERYClipping of aneurysm and

clearing of subarachnoid clot

Continue hydralion with crystalloids

POSTOPERATIVECLINICAL SPASM

NO POSTOPERATIVECLINICAL SPASM

CVP or Swan-Ganz line

HYPERTENSIVEHYPERVOLEMICHEMODILUTION

| Resolution of spasm [

FIOURE 1. Management protocol for subarachnoid hemorrhage, Neurovascular Surgery Service, Barrow Neurological Institute,

January 1984-December 1985.



ownloaded from


Awad et al Clinical Vasospasm 367

Table 1. Modified Hunt-Hess Grading Scheme of NeurologicStatus After Subarachnotd Hemorrhage '

Grade I No neurologic deficits, no headache or mild headache

Grade II Cranial nerve palsy, severe headache, roeningismus,irritability

Grade III Arm or leg drift, confusion, drowsiness (arousable byverbal stimulation)

Grade IV Paresis of one or more limbs, aphasia (expressive orreceptive), stupor (arousable by painful stimulation)

Grade V Plegia of one or more limbs, posturing in one or morelimbs, coma (unarousable by painful stimulation),death

Neurologic status was graded using a modifiedHunt-Hess scale (Table 1). This was determined onadmission in all patients with SAH and on a daily basisin all patients with clinical vasospasm. The progress ofthe latter patients, including pertinent hemodynamicparameters and laboratory data, was closely monitoredin a prospective fashion by a full-time clinical researchnurse.

ResultsIncidence of Clinical Vasospasm

Of the 118 patients admitted within 2 weeks ofSAH, 42 (35.6%) developed clinical vasospasm.These cases represented 41 of 113 cases (36.3%) withproven aneurysm, and 1 of 5 cases (20%) in which no

aneurysm could be identified. Patients with tumor,trauma, or vascular malformations had been excluded.Clinical vasospasm developed in 9 of 28 patients(32.1%) who were Grade I at the time of admission, in16 of 44 patients (36.4%) who were Grade II on admis-sion, in 10 of 22 patients (45.5%) who were Grade IIIon admission, and in 7 of 16 patients (43.8%) whowere Grade IV on admission. There was no delayedneurologic deterioration attributable to vasospasm inany of the 8 patients who were comatose (Grade V) onadmission.

Clinical vasospasm developed and was treated aftercraniotomy in 27 cases operated in the acute state(within 24 hours of admission). It was present andtreated prior to delayed surgery in 10 cases. An addi-tional 5 patients with clinical vasospasm never under-went definitive craniotomy (4 because of death fromvasospasm and 1 because no aneurysm was identified).Twenty-two of the 42 patients with clinical vasospasmrequired adjunctive cerebrospinal fluid diversion.

Response to Hypervolemic Hemodilution TherapyNeurologic grades on admission, at initiation of hy-

pervolemic hemodilution therapy (onset of clinical va-sospasm), and at conclusion of hypervolemic hemodi-lution therapy (resolution of clinical vasospasm) areillustrated in Figure 2. Forty-one of the 42 patients haddeteriorated to Grade IH or worse at the onset of clini-cal vasospasm. Sustained improvement by at least 1neurologic grade was observed in 60% of the patients,

NEUROLOGICAL GRADES ON ADMISSION AND DURINGHYPERVOLEMIC HEMODILUTION

FOR CLINICAL VASOSPASM(42 Patients)

Ifa 1

O ~6? U

100-

80-

6 0 -

4 0 -

2 0 -

III

1

II

III

IV

1 II

IV

li1

II

III

: " ' • • • • • .

' : " ' • " ' . :

/ : ; / • ;

ON ADMISSION AT INITIATIONOF H/H THERAPY

(ONSET OFCLINICAL SPASM

AT CONCLUSIONOF H/H THERAPY(RESOLUTION OFCLINICAL SPASM)

FIGURE 2. Neurologic grades on admission, at the initiation of hypervolemic hemodilution therapy, and at the conclusion ofhypervolemic hemodilution therapy in 42 patients with clinical vasospasm after subarachnoid hemorrhage. All but 1 patient haddeteriorated to Grade III or worse at the initiation of therapy.



ownloaded from



RESPONSE TO HYPERVOLEMIC HEMODILUTIONAS A FUNCTION OF

ADMISSION NEUROLOGICAL GRADE

• Sustainedimprovement

No change

3 T3

»> sc oi* cn

100-

9 0 -

60-

4 0 -

2 0 -

I

Worsening

I(N = 9)

II(N = 16)

III(N n 10)

IV(N = 7)

NEUROLOGICAL GRADE ON ADMISSION

FIGURE 3. Response to hypervolemic hemodilution therapy as a function of admission neurologic grade in 42 patients with clinicalvasospasm after subarachnoid hemorrhage. Patients who were Grade I or II on admission were least likely to continue to worsendespite therapy. Patients who were Grade IV on admission were most likely to worsen despite therapy.

while 24% had no change in neurologic grade, and16% worsened or died during hypervolemic hemodilu-tion therapy. Sustained improvement was most likelyto occur in clinical vasospasm patients who wereGrades I or II on admission, while continued deteriora-tion was most frequently encountered among patientswho were Grade IV on admission (Figure 3). Uponresolution of clinical vasospasm, 48% of the patientswere neurologically normal (Grade I), 33% had a mi-nor neurologic deficit (Grades II or HI), and 19% had amajor neurologic deficit or died (Grades IV or V).Overall, death or major neurologic deficit from vaso-spasm occurred in 6.7% of all patients with SAH.

Diagnostic and Laboratory DataThe desired hemodynamic objectives were ap-

proached or achieved in all patients. Mean hematocritwas 38.4 ± 3.3% on admission, 37.1 ± 3.1% at on-set of hypervolemic hemodilution therapy, and33.0 ± 1.9% at the conclusion of hypervolemic hemo-dilution therapy. Mean sodium concentration was137.7 ±2.0 mEq/dl on admission, 141.6 ±2.6mEq/dl at the onset of hypervolemic hemodilutiontherapy, and 140.7 ± 3.1 mEq/dl at the conclusion ofhypervolemic hemodilution therapy. In some patients,neurologic deficits fluctuated markedly with changesin arterial pressure despite optimal volume status (Fig-ure 4). In other patients, neurologic deficits fluctuatedmarkedly with changes in volume status despite maxi-mal hypertension (Figure 5). Close titration of all pa-rameters was therefore essential throughout the dura-

tion of clinical vasospasm. Improvement in clinicalstatus frequently occurred within minutes of the initi-ation of therapy.

Cerebral angiography was performed at least onceduring the period of clinical vasospasm in 39 patients.It revealed definite angiographic spasm (to a blindedradiologist) in all cases. Regional cerebral blood flowmeasurements (using the stable xenon-CT technique)

3 4 5 6 7

TIME (Hours)8 9 10

CASE #31; Day 4 post SAH; Hct 36%

FIGURE 4. Hemodynamic parameters in Case 31. Aphasia ap-peared whenever the systolic arterial pressure dropped below170 mm Hg despite optimal central venous pressure and hemat-ocrit.



ownloaded from



4 5 8 7 8 S 10

TIME (Hours)

CASE #18; Day 5 post SAH; Hct 36%

FIGURE 5. Hemodynamic parameters in Case 18. Hemipare-sis appeared whenever the central venous pressure droppedbelow 10 mm Hg despite elevated systolic arterial pressure andoptimal hematocrit.

were performed in 10 patients and showed focal areasof hypoperfusion in all cases (Figure 6).

Cases with Poor Outcome and ManagementComplications

Eight patients with clinical vasospasm had a pooroutcome (worse than Grade IE upon resolution ofclinical vasospasm). One patient was Grade III on ad-mission to the hospital and had clinical vasospasm atthat time; he was receiving normotensive hypervole-mic hemodilution therapy while awaiting delayed clip-ping of an anterior communicating artery aneurysmwhen he rebled and died 5 days after SAH. Coagula-tion parameters and platelet count were normal at alltimes, and systolic arterial pressure never exceeded150 mm Hg prior to the rebleed.

One patient was Grade II on admission 3 days afterSAH; he developed a dense hemiparesis within hoursof admission and died of progressive vasospasm 1week later. An additional 4 patients (including 2 withmycotic aneurysms) were already in fulminant clinicalvasospasm (Grade IV) on admission to the hospital; 3died of progressive spasm before surgical intervention,and 1 recovered with residual left hemiparesis andhomonymous hemianopsia. Another 2 patients in poorneurologic condition on admission (Grade IV) had fur-ther deterioration from vasospasm after an early oper-ation for evacuation of a hematoma (and aneurysmclipping); both died despite hypertensive hypervole-mic hemodilution therapy. No Grade I or II patientoperated in the acute state sustained a severe perma-nent neurologic deficit or died as a result of vaso-spasm.

Three patients developed respiratory failure withfrank pulmonary edema during hypervolemic hemodi-lution therapy. All 3 had central venous pressure moni-toring. In 1 case (a 47-year-old healthy woman) thecentral venous pressures were > 15 mm Hg for severalhours before decompensation, and she received nodiuretics or fluid restriction. This patient deviated from

our management protocol. The other 2 cases (a 34-year-old severely asthmatic woman and a 69-year-oldman with cardiac disease) decompensated despite nor-mal central venous pressures. The latter required endo-tracheal intubation. All 3 patients responded to appro-priate diuresis, and all made excellent recovery. These3 complications could have been prevented by judi-cious attention to the hemodynamic parameters andthrough prophylactic use of a Swan-Ganz catheter inthe presence of preexisting pulmonary or cardiac dis-ease. There were no complications attributed to cath-eter insertion or care. There were no instances of wors-ened ischemic brain edema or hemorrhagic infarction.

DiscussionAngiographic evidence of vasospasm is encountered

in 60-80% of patients with SAH.3-512-14 It is mostfrequently seen between the fourth and tenth daysposthemorrhage and is most severe around the sixthday.13 The presence and extent of angiographic vaso-spasm correlate strongly with the severity of the hem-orrhage as determined by the amount of blood in thesubarachnoid cisterns on CT scans.1516 The preciseincidence of angiographic spasm in a given series isdifficult to determine and obviously depends on theparticular mix of patients (mild vs. severe hemorrhage,type of aneurysm, etc.), the timing and quality of an-giography, and the definition of spasm.

While many patients with angiographic vasospasmdo not seem to suffer any adverse consequences, othersdevelop a stereotyped clinical syndrome apparentlyrelated to brain ischemia.3~6'"'13 Symptomatic or clini-cal vasospasm most frequently occurs during the peri-od of most intense angiographic spasm and is charac-terized by gradual worsening of the sensorium,followed within hours by fluctuating focal neurologicdeficits. The syndrome resolves several days (rarelyweeks) later, with gradual improvement in the sensori-um. Delayed neurologic deterioration from vasospasmaffects over one-third of patients with SAH and kills orseverely disables nearly half of these patients.336 It hasbeen estimated that clinical vasospasm kills approxi-mately 7% and severely disables another 7% of SAHpatients currently reaching referral centers.23

While many pharmacologic approaches have beensuggested in an effort to prevent or reduce arterialnarrowing, the experimental evidence remains con-flicting, and no agent has achieved widespread accep-tance in clinical circles.17 The mainstay of currentmanagement of vasospasm has remained aggressive,supportive care and optimization of the hemodynamicstatus. While most aneurysm surgeons accept thesebroad objectives, there is no uniform agreement on thespecific use of hypertension, hypervolemia, or hemo-dilution. Furthermore, the literature is limited regard-ing the effect of these modalities on patient outcome.

In 1967, Farhat and Schneider published their obser-vations on the improvement of ischemic neurologicdeficits with increased systemic blood pressure.18

Since that time, Simeone and others have emphasizedthe theoretical rationale for hypertensive therapy.19"22



ownloaded from



FIGURE 6. Stable xenon computed regional cerebral flow tomogram and corresponding computed tomographic scan in Case 7. Thepatient was Grade 1 on admission and had an uneventful clipping of a left posterior communicating artery aneurysm that day. Five dayslater, the patient became stuporous and developed a moderate right hemiparesis. The scan done at that time shows areas of decreasedperfusion in the left middle cerebral artery and both anterior cerebral artery territories. Angiography confirmed severe spasm in thecorresponding vessels.

Regional cerebral blood flow is depressed and pressureautoregulation is clearly impaired in clinical vaso-spasm.20'2'123 Cerebral blood flow is directly dependenton perfusion pressure under these circumstances. De-spite this compelling rationale, clinicians have beenreluctant to use this modality on a large scale for fear ofincreasing intracranial pressure, precipitating hemor-rhage into an area of infarction, worsening ischemicedema, or rupturing an undipped aneurysm. Howev-er, judicious control of intracranial pressure, timelytherapy prior to frank infarction, and early clipping ofthe aneurysm can circumvent most of these concerns.There were no instances of hemorrhagic infarction orintractable cerebral edema in this series. The only pa-tient with aneurysmal rebleed was not on hypertensivetherapy.

Hypervolemia with or without hypertension hasbeen advocated by several surgeons."2224-26 The theo-retical rationale for hypervolemia is based on evidencethat S AH results in contraction of the extracellular andvascular spaces, perhaps because of cerebral salt wast-ing or other mechanisms.2728 Furthermore, hypervole-mia increases cardiac output and facilitates hyperten-

sive therapy.24 In 1976, Kosnik and Hunt reported thereversal of ischemic symptoms in 6 of 7 patients usingvasopressors and hypervolemia.22 In 1977, Gianotta etal reported success with these modalities in 15 of 17patients.23 In 1978, Pritz and coworkers reported anadditional 4 such cases and emphasized the importanceof aggressive monitoring of hemodynamic parametersduring such therapy.24 The largest series to date waspublished by Kassell et al in 1982, reporting on theresponses of 58 patients with the strict diagnosis ofclinical vasospasm.26 The patients represented con-secutive cases treated with hypertensive hypervolemiaat 2 large centers. While the management protocolaimed for a hematocrit of 40%, some patients receivedcolloid agents and may have been subjected to an ele-ment of adjunctive hemodilution. Sustained neurolog-ic improvement was documented in 74% of the pa-tients. There were 10 cases of pulmonary edema and 3rebleeds during the period of therapy. There were nocases of hemorrhagic infarction or increased brainswelling associated with the treatment. The problemwith these and other reports has been the retrospectivenature of patient identification. Also, there was no



ownloaded from



information regarding other vasospasm cases (if any)or the overall number of patients with SAH managedduring the same time period (Table 2).

The use of hemodilution as an adjunct to hyperten-sive hypervolemia remains controversial. Hemodilu-tion results in a dramatic decrease in blood viscos-ity.29-30 This change occurs at hematocrit levels wellwithin the physiologic range and is more pronouncedat lower shear rates (i.e., in constricted vessels and atlow flow rates).30 Hemodilution has been shown toincrease regional cerebral flow in the ischemic core(even in the absence of adjunctive hypervolemia) andto decrease infarct size in experimental focal cerebralischemia.2931"33 More recently, Kee and Wood havedemonstrated an increase in blood flow and oxygentransport across tight stenoses by hemodilution to ahematocrit of 34%.M Other clinical reports have em-phasized the harmful sequelae of an elevated hemato-crit after cerebral infarction.3336 Conversely, critics ofhemodilution caution that blood oxygen content de-creases with decreasing hematocrit, and that some ofthe improvement in regional cerebral blood flow maytoe the result of such hypoxemia. The optimal hemato-crit thought to maximize oxygen delivery to tissue hasbeen estimated at 33%, but may be higher in the ische-mic brain.293337 In the Barrow Neurological Instituteprotocol, either extreme was avoided by aggressivelylowering the hematocrit in a patient when it was above40%, and by aiming for a hematocrit in the mid-30's inall patients.

Sustained neurologic improvement by at least 1grade occurred in 60% of all patients with vasospasm.Improvement was most likely to occur with hypervole-mic hemodilution therapy if vasospasm followed earlyclipping of the aneurysm (patient in good neurologiccondition on admission and deterioration from vaso-spasm in the postoperative period). None of 25 suchpatients sustained a major permanent neurologic defi-

cit or died from vasospasm. Continued deteriorationfrom vasospasm was most likely to occur in poor grade(Grades IV and V) patients already in fulminant vaso-spasm on admission to the hospital. Management com-plications were limited to 3 cases of pulmonary con-gestion. All 3 should have been avoided by closermonitoring of hemodynamic parameters. As in Kas-sell's series, there were no instances of worsened brainedema or hemorrhagic infarction. The formal evalua-tion of late outcome, including possible neuropsycho-logical sequelae, was not within the scope of thisstudy.

While early operative intervention might be expect-ed to increase the incidence of clinical vasospasm,such was not the case in this series. The 35.6% inci-dence of clinical vasospasm following SAH is compa-rable to the 33.5% incidence reported in the recentcooperative study on the timing of aneurysm surgery.It is only slightly higher than the 26.7% incidence inthe placebo limb of the nimodipine cooperative trial(which included only Grade I patients admitted within96 hours of hemorrhage).' Outcome from vasospasmcompared favorably with either series (Table 2). Deathor major neurologic deficit attributed to vasospasm(defined quite liberally as worse than Grade III uponresolution of vasospasm) afflicted only 6.7% of allSAH patients in this institution. In contrast, vasospasmkilled or severely disabled 16% of patients in the coop-erative study and 13.3% of patients in the placebo limbof the nimodipine study.12 The difference would begreater if rebleeding that occurred during the period ofvasospasm were excluded (as it was in these otherseries), and if patients with mycotic aneurysms wereexcluded. Since the SAH populations were not concur-rent and were not necessarily comparable, no firmstatistical conclusions can be drawn. However, theresults in this prospectively surveyed SAH population(subjected to a uniform management protocol at a sin-

Table 2. Outcome of Patients With Clinical Vasospasm After Subarachnoid Hemorrhage

Therapy (reference)

Standard supportivetherapy (Cooper-ative study, Kas-sell 1984)

Placebo limb of ni-modipine study(AUen 1983)

Hypertensive hyper-volemia (Kassel1982)

Hypertensive hyper-volemic hemodi-lution (BNI 1986)

Pts with SAH

1,272(all grades)

60(neurologically

normal pts only)

118(all grades)

Pts withCV

425

16

58

42

% withCV

33.5%

26.7%

?

35.6%

ResponseCV to t

Worse

10%

16%

of pts withreatment*

Same

16%

24%

Im-proved

74%

60%

Outcome of pts

Mild ormoderate

Normal disability

? ?

37.5% 12.5%

? ?

47.6% 33.3%

with CV

Severedisabilityor death

50%t

50%t

?

19%(16.6%)t

% deathand/or

disabilityfrom

vasospasm

16%t

13.3%t

6.7%(5.9%)t

Pts, patients; SAH, subarachnoid hemorrhage; CV, clinical vasospasm; %, percent of total patients with subarachnoid hemorrhage.•Percent of total patients with clinical vasospasm.tExcluding death or disability due to rebleeding during period of clinical vasospasm.



ownloaded from



gle institution) should provide a reference for compar-ing outcomes with other management modalities andserve as a guideline for more controlled randomizedtrials in the future.

In conclusion, the authors propose that early opera-tive intervention for aneurysmal SAH and subsequentaggressive medical management of vasospasm withhypervolemic hemodilution therapy can be accom-plished safely and without worsening the outcomefrom vasospasm. It is hoped that such a managementstrategy might lower the incidence of death and dis-ability from vasospasm subsequent to SAH.

References1. Allen GS, Ahn HS, Preziosi TJ, Battye R, Boone SC, Chou

SN, Kelly DL, Weir BK, Crabbe RA, Lavik PJ, RosenbloomSB, Dorsey FC, Ingram CR, Mellits DE, Bertsch LA, BoisvertDPJ, Hundley MB, Johnson RK, Strom JA, Transou CR: Ce-rebral arterial spasm—A controlled trial of nimodipinc in pa-tients with subarachnoid hemorrhage. N Engl J Med 1983;308:619-624

2. Kassell NF, Tomer JC: The International Cooperative Study ontuning of aneurysm surgery—An update. Stroke 1986; 15:566—570

3. Kassell NF, Sasaki T, Colohan ART, Nazar G: Cerebral vaso-spasm following aneurysmal subarachnoid hemorrhage. Stroke1985;16:562-572

4. Chyatte D, Sundt TM: Cerebral vasospasm after subarachnoidhemorrhage. Mayo Clin Proc 1984;59:498-5O5

5. Heros RC, Zervas NT, Varsos V: Cerebral vasospasm aftersubarachnoid hemorrhage—An update. Ann Neurol 1983;14:599-608

6. Fisher CM, Roberson GH, Ojemann RG: Cerebral vasospasmwith ruptured saccular aneurysm—The clinical manifestations.Newoswgery 1977; 1:245-248

7. Hashi K, Nin K, Shimotake K: Surgery in the prevasospasticinterval. Acta Neurochir 1982;63:141-145

8. Mizukami M, Kawase T, Usami T, Kawase T: Prevention ofvasospasra by early operation with removal of subarachnoidblood. Newosurgery 1982; 10:301-307

9. Ohta H, Ito Z, Yasui N, Suzuki A: Extensive evacuation ofsubarachnoid clot for prevention of vasospasm—Effective ornot? Acta Neurochir 1982;63:111-116

10. Suzuki J, Onuma T, Yoshimoto T: Results of early operationson cerebral aneurysms. Surg Neurol 1979; 11:407-412

11. Peerless SJ: Pre- and postoperative management of cerebralaneurysms. Clin Neuroswg 1979;26:209-231

12. Fletcher TM, Taveras JM, Pool JL: Cerebral vasospasm inangiography for intracranial aneurysms. Arch Neurol

13. Weir B, Grace M, Hansen J, Rothberg C: Time course ofvasospasm in man. J Neurosurg 1978;48:173—178

14. Black PM: Hydrocephalus and vasospasm after subarachnoidhemorrhage from ruptured intracranial aneurysms. Neurosur-gery 1986;18:12-16

15. Fisher CM, Kistler JP, Davis JM: Relation of cerebral vaso-spasm to subarachnoid hemorrhage visualized by computer-ized tomographic scanning. Newosurgery 1980;6:l-9

16. Mizukami M, Takemae T, Tazawa T: Value of computed to-mography in the prediction of cerebral vasospasm after aneu-rysm rupture. Newosurgery 198O;7:583-586

17. Wilkins RH: Attempted prevention or treatment of intracranialarterial spasm: A survey. Newoswgery 1980;6:198-210

18. Farhat SM, Schneider RC: Observations on the effect of sys-temic blood pressure on intracranial circulation in patients withcerebrovascular insufficiency. J Newoswg 1967;27:441-445

19. Simeone FA, Trepper PJ, Brown DJ: Cerebral blood flowevaluation of prolonged experimental vasospasm. J Newoswg1972;37:302-311

20. Hashi K, Meyer JS, Shinmaru S, Welch KM A, Teraura T:Changes in cerebral vasomotor reactivity to CO? and autoregu-lation following experimental subarachnoid hemorrhage. JNeurol Sci 1972;17:15-22

21. Ishii R: Regional cerebral blood flow in patients with rupturedintracranial aneurysms. J Newoswg 1979;50:587-594

22. Kosnik EJ, Hunt WE: Postoperative hypertension in the man-agement of patients with intracranial arterial aneurysms. JNewoswg 1976;45:148-154

23. Pitts LH, Macpherson P, Wyper DJ, Jennett WB: Cerebralblood flow, angiographic cerebral vasospasm, and subarach-noid hemorrhage. Acta Neurol Scand 1977;56(suppl 64):334-335

24. Pritz MB, Gianotta SL, Kindt GW, McGillicuddy JE, PragerRL: Treatment of patients with neurologic deficits associatedwith cerebral vasospasm by intravascular volume expansion.Newoswgery 1978^:364-368

25. Gianotta SL, McGillicuddy JE, Kindt GW: Diagnosis andtreatment of postoperative cerebral vasospasm. Surg Newol1977;8:286-29O

26. Kassell NF, Peerless SJ, Durward QJ, Beck DW, Drake CG,Adams HP: Treatment of ischemic deficits from vasospasmwith intravascular volume expansion and induced arterial hy-pertension. Newosurgery 1982; 11:337-343

27. Kassell NF, Boarini DJ: Patients with ruptured aneurysm: Pre-and postoperative management, in Wilkins RH, RengacharySS (eds): Newoswgery, vol. 2. New York, McGraw-HillBook Co., 1985, pp 1367-1371

28. Maroon JC, Nelson PB: Hypovolemia in patients with subar-achnoid hemorrhage: Therapeutic implications. Newoswgery1979;4:223-226

29. Wood JM, Kee DB: Hemorrheology of the cerebral circulationin stroke. Stroke 1985; 16:765-772

30. Stone HO, Thompson HK, Schmid-Nielson K: Influence oferythrocytes on blood viscosity. Am J Physiol 1968;214:913-918

31. Wood JH, Simeone FA, Fink EA, Golden MA: Correlativeaspects of hypervolemic hemodilution: Low-molecular weightdextran infusions after experimental cerebral artery occlusion.Newology 1984^4:24-34

32. Wood JH, Simeone FA, Kron RE, Snyder LL: Experimentalhypervolemic hemodilution: Physiological correlations of cor-tical blood flow, cardiac output, and intracranial pressure widifresh blood viscosity and plasma volume. Newosurgery 1984;14:709-723

33. Kiyohara Y, Fujishima M, Ishitsuka T: Effects of hematocriton brain metabolism in experimentally induced cerebral ische-mia in spontaneously hypertensive rats (SHR). Stroke1985;16:835-84O

34. Kee DB, Wood JH: Experimental hemodynamic and rheologicalterations of blood flow through carotid stenosis followinghemodilution. Newology (in press)

35. Tohgi H, Yamanouchi H, Murakami M, Kameyama M: Impor-tance of the hematocrit as a risk factor in cerebral infarction.Stroke 1978;9:369-374

36. Harrison MJG, Pollock S, Kendell BE, Marshall J: Effect ofhematocrit on carotid stenosis and cerebral infarction. Lancet1981;2:114-115

37. Thomas DJ: Hemodilution in acute stroke. Stroke 1985;16:763-764

KEY WORDS • clinical vasospasm • subarachnoid hemor-rhage • hypervolemia • hemodilution • hypertension •early surgery



ownloaded from


I A Awad, L P Carter, R F Spetzler, M Medina and F C Williams, Jrand arterial hypertension.

Clinical vasospasm after subarachnoid hemorrhage: response to hypervolemic hemodilution

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1987 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.18.2.365

1987;18:365-372Stroke.

http://stroke.ahajournals.org/content/18/2/365World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Stroke Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:



ownloaded from

http://stroke.ahajournals.org/content/18/2/365

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://stroke.ahajournals.org//subscriptions/


clinical vasospasm after subarachnoid hemorrhage: …improvement by at least 1 neurologic grade, 24%...

Documents