Secondary Logo

Journal Logo

Hemodynamic Management and Outcome of Patients Treated for Cerebral Vasospasm with Intraarterial Nicardipine and/or Milrinone

Schmidt, Ulrich MD, PhD*†; Bittner, Edward MD, PhD*†; Pivi, Silvia MD*†; Marota, John J. A. MD, PhD*†

doi: 10.1213/ANE.0b013e3181cc9ed8
Neurosurgical Anesthesiology and Neuroscience: Research Reports

BACKGROUND: Vasospasm is a potentially devastating complication after aneurysmal subarachnoid hemorrhage. Although endovascular treatment with intraarterial nicardipine and milrinone is an accepted clinical treatment strategy, there is little information either on hemodynamic management during treatment or on outcome and consequences of the hemodynamic management. We tested 2 hypotheses: (1) intraarterial administration of nicardipine and milrinone to treat cerebral vasospasm would require increased administration of vasoconstrictor to support arterial blood pressure at target levels; and (2) high-dose vasopressors administered to increase blood pressure in these patients would lead to systemic acidosis and end-organ ischemic damage.

METHODS: We conducted a single-center, retrospective review of consecutive patients with clinically symptomatic vasospasm after aneurysmal subarachnoid hemorrhage that failed medical management with “triple H therapy” and subsequently received intraarterial nicardipine and/or milrinone between March 2005 and July 2007.

RESULTS: Of 160 endovascular interventions in 73 patients (aged 52 ± 10 years; 50 women), 96 received only nicardipine, 5 only milrinone, and 59 both drugs. General anesthesia with muscle relaxation was performed for 93% of procedures. During treatment, both the number and dose of vasopressors required to maintain arterial blood pressure at target levels increased; the median dose of phenylephrine increased from 200 (n = 121) to 325 μg/min (n = 122), norepinephrine increased from 12 (n = 60) to 24.5 μg/min (n = 87), and vasopressin infusions increased from 7 to 24. Nonetheless, arterial blood pressure decreased 13% during treatment. In >90% of procedures, the postprocedure angiogram showed improved vessel caliber. A single patient demonstrated troponin T increase; no patients had a decrease in renal function, bowel or peripheral ischemia, systemic acidosis, or acute stroke. Overall mortality was 11%.

CONCLUSIONS: Intraarterial administration of nicardipine and/or milrinone requires use of vasopressors to maintain arterial blood pressure. Despite high doses of vasoconstrictors, treatment has low mortality, minimal end-organ ischemic damage or systemic acidosis, and results in improved caliber of cerebral vessels affected by vasospasm.

From the *Department of Anesthesia and Critical Care, Massachusetts General Hospital; and †Department of Anaesthesia, Harvard Medical School, Boston, Massachusetts.

Accepted for publication November 5, 2009.

Address correspondence and reprint requests to John J.A. Marota, MD, PhD, Department of Anesthesia and Critical Care, Massachusetts General Hospital, 55 Fruit St., GRJ-422, Boston, MA 02114. Address e-mail to

Cerebral vasospasm is a potentially devastating complication that occurs in nearly half of all patients who survive the first 24 hours after subarachnoid hemorrhage (SAH) from a ruptured cerebral aneurysm. Subsequent reductions in cerebral perfusion downstream to the constricted arteries contribute significantly to both morbidity and mortality of these patients.1 With the exception of systemic administration of calcium channel blockers, pharmacologic interventions directed at decreasing both occurrence and severity of vasospasm have had only limited success.2,3 Traditionally, clinical management of cerebral vasospasm has focused on improving cerebral perfusion beyond the constricted vessels by increasing cardiac output, increasing perfusion pressure, and improving red blood cell rheology. The combination of induced hypertension, hypervolemia, and hemodilution has been advocated as a method to achieve these goals and is often referred to as triple H therapy.4 This approach frequently requires continuous infusions of vasopressors to maintain arterial blood pressure at target levels to preserve cerebral perfusion and reverse neurological deficits.4 Recently, however, vessel-specific, intraarterial infusion of potent vasodilators has been used to reverse the arterial spasm. Three different classes of vasodilating drugs are in current clinical practice: specifically, the opium alkaloid papaverine, the calcium channel antagonists verapamil and nicardipine,5 and, most recently, the phosphodiesterase inhibitor milrinone.6,7 Several nonrandomized trials have demonstrated the efficacy of intraarterial Ca2+ channel antagonists and milrinone to reverse cerebral vasospasm.5–11 Accordingly, these drugs have found widespread acceptance in clinical practice.

There is little information available, however, regarding hemodynamic management of patients undergoing selective intraarterial infusions of these drugs for treatment of cerebral vasospasm. A single report in the literature observed that intraarterial administration of nicardipine produced a significant decrease in arterial blood pressure.12 Although the hemodynamic consequences of intraarterial infusion of milrinone are not well described, the pharmacologic profile as a phosphodiesterase inhibitor would suggest a reduction in systemic vascular resistance, possibly inducing hypotension. In these patients, acute reduction in arterial blood pressure potentially could compromise cerebral perfusion and necessitate the use of vasopressors to support systemic pressure. High doses of catecholamines increase systemic resistance by vasoconstriction, a mechanism that can compromise blood flow to end organs other than the brain and thereby result in ischemic damage to other organs.13,14

Therefore, the purpose of this study was to review the hemodynamic management of patients undergoing treatment of cerebral vasospasm in our center. Specifically, we tested the hypothesis that intraarterial infusion of nicardipine or milrinone, the 2 pharmacologic drugs used at our institution to treat cerebral vasospasm, results in an increased administration of vasopressors to maintain arterial blood pressure at target levels during treatment. Second, we examined patient outcome after endovascular treatment to determine whether high-dose vasopressors administered to increase blood pressure in these patients would lead to reduced perfusion and end-organ ischemic damage. Specifically, we screened for systemic acidosis as a generalized index of reduced systemic perfusion and examined records for evidence of reduced renal function, bowel ischemia, peripheral ischemia, or myocardial ischemia. Finally, we assessed the 30-day mortality and disposition at discharge in all patients who underwent endovascular treatment for vasospasm. In addition, we reviewed the anesthetic management of these patients and the presentation at the time of the procedure to identify characteristics of anesthetic management at our institution.

Back to Top | Article Outline


A retrospective review of patient medical records was performed with approval by the IRB at the Massachusetts General Hospital, an academic, tertiary care center.

We collected and reviewed the electronic medical and anesthesia records of all patients, aged 18 years or older, admitted to our institution between March 2005 and July 2007 who underwent cerebral angiography to investigate suspected cerebral vasospasm after SAH after rupture of a cerebral aneurysm. The decision to perform angiography to confirm the diagnosis of cerebral vasospasm was made collectively by the neurosurgery service, neuro-intensive care team, and neuro-endovascular service caring for the patient. Cerebral vasospasm was suspected in patients based on daily examination by transcranial Doppler ultrasonography and clinical examination of the patient; patients were considered for endovascular treatment of vasospasm either when they did not show improvement 24 hours after initiation of triple H therapy or developed worsening neurological symptoms despite therapy. The target arterial blood pressure in patients selected for treatment was to maintain systolic blood pressure in a range of 160 to 200 mm Hg.

From this database of patients, we collected complete anesthetic and medical records for all patients who underwent either side-directed or cerebral vessel-specific intraarterial administration of nicardipine and/or milrinone to treat clinically and angiographically significant cerebral vasospasm. At the time of this study, nicardipine and milrinone were the only 2 drugs used for intraarterial infusion to treat refractory cerebral vasospasm at our institution. We excluded from the study group all patients in whom angiography did not identify significant vessel spasm to warrant treatment or in whom vasospasm was treated with angioplasty or another method. The decision to proceed with selective infusion of nicardipine and/or milrinone, dosages of drugs infused, and selection of vessels for treatment was made solely by the interventional endovascular trained physician performing the procedure based on vessel caliber as revealed by the cerebral angiogram at the time of the examination. The degree of spasm was graded based on the lumen size in comparison with normal segment of the artery either proximal or distal to the narrowing; spasm was graded as no spasm, mild (<40%), moderate (40%–75%), and severe (>75%). Once an affected vessel was identified for treatment, an initial dose of either nicardipine or milrinone was selected for administration; at the conclusion of the infusion, change in vessel caliber was assessed by repeat angiography. At all times, nicardipine and milrinone were administered individually to the affected vessels and never in combination. Based on the vessel caliber in posttreatment angiogram, an additional infusion of the same vasodilator was performed, a second drug was selected for administration, or another vessel was selected for treatment until all affected vessels had been treated.

Back to Top | Article Outline

Demographic and Clinical Factors

All information was collected from the electronic medical record. We recorded age and gender of all patients; ASA classification and emergent status had been designated by the anesthesiologist at the time of the procedure. Patient comorbidities were identified in the admitting history and physical examination; these included history or current manifestations of coronary artery disease, congestive heart failure, peripheral vascular disease, hypertension, chronic obstructive lung disease, and asthma. Renal function was assessed by comparison of serum creatinine at time of admission (baseline) and serum creatinine determined before and after each intervention.

Back to Top | Article Outline

Procedure Management

Anesthetic management was identified from the electronic anesthetic records for all patients and included invasive monitoring of arterial blood pressure. All patients received standard monitoring including electrocardiogram, pulse oximetry, end-tidal carbon dioxide and capnography, inspiratory pressure, tidal volume, and minute ventilation. A central line was present in all patients for infusion of vasoactive medications. Arterial blood pressure and heart rate were collected for analysis from the anesthetic record at 3 different time points during the procedure: 5 minutes immediately before infusion of intraarterial nicardipine or milrinone, at completion of the intraarterial infusion, and 5 minutes before emergence from general anesthesia or termination of the anesthetic for transport to the intensive care unit. Intraarterial administration of nicardipine and/or milrinone and dosage of drugs infused were identified retrospectively from the procedure records.

Back to Top | Article Outline


Success of treatment outcome was assessed angiographically immediately after completion of intraarterial vasodilator drug infusion; as with severity of spasm, outcomes were graded by a physician trained in interventional neuroradiology and endovascular procedures and were recorded in the procedure records as poor, good, or excellent depending on degree of improvement in vessel caliber. Compromise in end-organ function was identified from the medical records. Acute renal failure was characterized by Risk, Injury, Failure, Loss, and End-stage Kidney criteria.15 The criteria consist of 3 graded levels of injury: risk, injury, and failure. Postprocedure positive troponin T (>0.1 ng/mL), elevation of cardiac and liver function enzymes, ischemia on electrocardiogram, and new onset arrhythmia were recorded. Incidences of bowel ischemia and peripheral ischemia were identified from the medical records. Evidence of bowel ischemia included report in the patient record of exploratory laparotomy for resection of ischemic bowel or suspected ischemic bowel, general surgical or gastroenterology consult for problems with the bowel, or development of an acute abdomen. Intensive care unit mortality, hospital mortality, and discharge destination were identified from the discharge summary and postdischarge follow-up patient visits listed in the medical record.

Back to Top | Article Outline

Statistical Analysis

Data analysis was performed using Stata 10 (Stata Corp. LP, College Station, TX). Discrete variables are expressed as counts and percentages. We assessed normality of the distribution for all parameters by calculating the skew of the distribution. For parameters in which the skew exceeded 0.4, data are presented as median and first and third quartile; for parameters with normal distributions, skew <0.4, continuous variables are expressed as mean ± sd. Data collected for heart rate, systolic blood pressure, and diastolic blood pressure determined before and after treatment arrayed as a normal distribution. Accordingly, significance of differences between individual systolic and diastolic arterial blood pressure and heart rate determinations made immediately before administration of an intraarterial vasodilator and at completion of the vasodilator infusion were assessed using paired t tests. Similarly, individual arterial blood gas values for pH and Paco2 determined before and after treatment procedures arrayed as normal distributions, and significance of difference was assessed by paired t tests.

Back to Top | Article Outline


Of the 87 consecutive patients who underwent 197 angiographic examinations for suspected cerebral vasospasm, we collected and reviewed the complete electronic medical and anesthetic records of 73 patients who underwent 160 intraarterial, side-directed, or vessel-specific infusions of nicardipine and/or milrinone to treat both clinically symptomatic and angiographically documented cerebral vasospasm after SAH. In 13 angiographic examinations (6.6%) performed in 12 patients, cerebral vasospasm was deemed not significant enough to require treatment; 1 patient was studied twice on different days with negative results. Six of the 87 patients in this series did not receive an endovascular treatment for vasospasm; 1 patient was treated by another method. An electronic anesthesia record was not available for 23 angiographic examinations, and these treatments were not included in the analysis.

Table 1 summarizes patient demographics, comorbidities, and severity of cerebral vasospasm at time of treatment for the 73 patients included in the analysis. Of the patients treated for cerebral vasospasm, 57.5% received >1 endovascular treatment procedure over a period ranging from 2 to 23 days after SAH. The distribution for number of endovascular treatment sessions received by each of the 73 patients is shown in Figure 1A and ranged from 1 to 9 with a median of 2 treatment sessions per patient. Multiple vessels were treated in 85% of the 160 treatment sessions performed; the distribution of number of vessels treated during each treatment session is shown in Figure 1B. The total number of vessels requiring treatment in each session ranged from 1 to 9 with a median of 3 vessels treated per treatment session. The sites of infusion varied both among and within treatment sessions with only distal vessels treated in 66% of sessions, and only side-directed infusions made proximally in large vessels, the carotid and vertebral arteries, in 8% of treatments performed. Both proximal and distal infusions were used in 25% of treatment sessions. The date of first vasospasm treatment ranged from 2 to 18 days after SAH with a median number of days to first treatment of 7; the distribution for number of days after SAH to first treatment is shown in Figure 2A. The distribution of days after hemorrhage on which patients received treatment procedures is shown in Figure 2B; the median number of days after hemorrhage was 8. The median dose of nicardipine given as the sole treatment drug was 15 mg and ranged from 2 to 45 mg (n = 96); the median dose of milrinone given as the sole treatment drug was 15 mg and ranged from 10 to 17.5 mg (n = 5). The total dosage of both drugs administered in treatments in which both drugs were infused to treat spasm (n = 59) was slightly reduced; the median doses for nicardipine and milrinone were 12.3 and 10 mg, respectively.

Table 1

Table 1

Figure 1

Figure 1

Figure 2

Figure 2

More than 93% of procedures (n = 150) were performed under general anesthesia; 10 procedures (6%) were performed successfully under deep sedation; and 1 treatment was initiated under sedation and subsequently converted to general anesthesia because of lack of patient cooperation. The average case duration was 264 ± 76 minutes (range 61–537 minutes). Although 60% of patients (96 of 160) presenting for treatment were mechanically ventilated via an endotracheal tube on arrival, 83% of patients (124 of 150) remained intubated at completion of the treatment and were mechanically ventilated postprocedure. The majority of patients who underwent general anesthesia received an inhaled anesthetic (79% [n = 118]); propofol infusion was used as the sole drug to maintain anesthesia in 16% of patients (n = 24) and used in combination with a volatile anesthetic in 13% of cases (n = 20). During 7 procedures, propofol infusion was substituted as maintenance anesthesia after initiating anesthesia with a volatile anesthetic. For procedures in which anesthesia was maintained with a volatile drug, sevoflurane was selected in 74% of cases; isoflurane was selected in 23% of cases; in the remaining 3% of cases, either desflurane or >1 volatile had been used. Nitrous oxide was used in combination with a volatile drug in 7% of general anesthetics. Muscle relaxants were administered in all cases in which general anesthesia was performed.

In 144 of the 160 procedures, at least 1 vasopressor was administered by continuous infusion to support arterial blood pressure at the target level before initiation of intraarterial administration of a vasodilator. Vasopressors used to support arterial blood pressure included phenylephrine, norepinephrine, or vasopressin administered by continuous infusion; boluses of phenylephrine, norepinephrine, epinephrine, and ephedrine also were given to increase blood pressure as needed. Table 2 summarizes the vasopressors administered immediately before and during infusion of the vasodilating drug. Before initiation of treatment, only 1 vasopressor was administered in 56% of cases (n = 90), 2 vasopressors in 31% of cases (n = 49), and 3 or more vasopressors in 4% of cases (n = 6). As shown, initiation of intraarterial vasodilator infusion necessitated both an increase in the number of vasopressors used to support blood pressure and an increase in the dose of the infusion. The median dose of phenylephrine administered by infusion increased by 62.5%, whereas the dose of norepinephrine more than doubled. The number of vasopressin infusions more than tripled from 7 to 24. Finally, the number of cases requiring 2 drugs to support arterial blood pressure increased from 49 to 60, whereas the number requiring 3 or more drugs increased from 6 to 24.

Table 2

Table 2

Despite the increase in vasopressor administration, we observed a 13% reduction in both systolic and diastolic blood pressure during the period of intraarterial administration of nicardipine and/or milrinone. Heart rate increased 30% during this same period. Systolic blood pressure decreased from a preinfusion baseline average of 176 ± 23 mm Hg to 153 ± 17 mm Hg during infusion; similarly, diastolic blood pressure decreased from 83 ± 13 mm Hg preinfusion to 72 ± 11 mm Hg during treatment. Heart rate increased from 83 ± 19 bpm to 104 ± 21 bpm. Both the reduction in blood pressure and increase in heart rate achieved statistical significance with a P value <0.0001 (paired t test, normal distribution).

Table 3 summarizes the outcome measures assessed in all patients undergoing treatment. There was no decrement in renal function as assessed by Risk, Injury, Failure, Loss, and End-stage Kidney score in any patient. There was no significant difference in serum creatinine determined before and within 24 hours after each procedure (P > 0.1, paired t test). No patient developed bowel ischemia or any other sign of peripheral ischemia. One patient with a high cardiac risk profile had a periprocedural cardiac event as evidenced by an increase in troponin to 0.28; no other sequelae in cardiac function were evident. There was no evidence of systemic acidosis; pH and Paco2 did not change significantly from the pretreatment values of 7.43 ± 0.05 mm Hg and 36 ± 6 mm Hg (mean ± sd), respectively, to the posttreatment values of 7.40 ± 0.06 mm Hg and 37 ± 6 mm Hg, respectively (P > 0.1, paired t test). No patient showed evidence of a new or worsening arterial occlusion of a cerebral vessel after the procedure as evidenced at final angiography. In 91% of treatments (n = 145), the outcome assessed angiographically by increase in vessel caliber was classified as good or excellent improvement of vessel spasm. Finally, no patient died during the procedure. Mortality within 30 days of the last treatment was 11% (n = 8). Seven of the 8 patients who died had severe, diffuse vasospasm and progressed to multiple brain infarctions with deterioration in neurological function despite treatment. One patient developed heparin-induced thrombocytopenia and had a carotid artery thromboembolism and subsequent brain infarction. Disposition at the time of discharge from hospital for the surviving 65 patients who underwent treatment is listed in Table 3.

Table 3

Table 3

Back to Top | Article Outline


Although treatment of cerebral vasospasm that develops after SAH by vessel-selective, intraarterial administration of Ca2+ channel antagonists as well as milrinone has been used in clinical practice for several years and is an accepted treatment strategy for clinically manifested cerebral vasospasm,3 to our knowledge, anesthetic and hemodynamic management of these patients has not been described in detail. There are 4 major findings in this review of management of patients who underwent vessel-selective, intraarterial administration of nicardipine and/or milrinone to treat cerebral vasospasm. First, high doses of vasopressors are required to maintain arterial blood pressure during treatment with the vasodilator. Although the patients in this series were receiving large doses of vasopressors to induce hypertension before endovascular intervention, additional pressors were necessary to support arterial blood pressure to achieve target levels during the procedure. Second, despite administration of additional vasoconstrictors during treatment, arterial blood pressure was not supported to the pretreatment baseline but decreased significantly. Third, despite the extraordinarily large quantities and varied vasoconstrictors administered during treatment, often in combinations, we found no significant end-organ damage or development of systemic acidosis in these patients. Moreover, mortality of the procedure was low with 89% of treated patients discharged from hospital, and there was immediate improvement in vessel caliber in the majority of treatments. Lastly, general anesthesia with paralysis is the technique most frequently used at our institution.

After SAH, patients who develop cerebral vasospasm are routinely managed medically with triple H therapy consisting of induced hypertension, hypervolemia, and hemodilution.16 Phenylephrine and other vasoactive drugs are required to achieve the hemodynamic goal of increasing cerebral perfusion by increasing systemic blood pressure. Indeed, in this series, phenylephrine was administered by infusion before the initiation of treatment in 121 of the 160 interventions. The dose of phenylephrine used in this patient population was high and often exceeded the dose necessary to maintain blood pressure during sepsis.17 Previously, intraarterial nicardipine has been shown to decrease systolic, diastolic, and mean arterial blood pressure. Avitsian et al.12 reported a case series of 11 patients in whom intraarterial nicardipine produced a 23 mm Hg (14–32 mm Hg) reduction in systolic blood pressure. In contrast, Badjatia et al.5 reported no difference in either blood pressure or administration of vasopressor during treatment with nicardipine alone. In this report of 18 patients, the dose of nicardipine was lower than that reported in this study, ranging from only 1 to 6 mg total dose. Fraticelli et al.7 reported 22 patients undergoing selective intraarterial infusions of milrinone in which no decrease in blood pressure was observed. This is somewhat surprising considering that milrinone has potent vasodilating effects.18 In this series, we observed that intraarterial administration of nicardipine, either as the sole vasodilator or in conjunction with separate milrinone infusions, produced a significant decrease in blood pressure despite aggressive use of vasopressors. We found a sizable increase in both dose and number of pressors used to maintain blood pressure at target levels. Differences may reflect dose, repeated infusions, speed of injection, and the baseline hemodynamic state of patients. Nonetheless, the reduction in arterial blood pressure observed in this study is in good agreement with that reported by Avitsian et al.12

End-organ damage including cardiac ischemia, renal failure, gut ischemia, and skin necrosis have been associated with administration of high concentrations of vasoconstrictive drugs in critical care.19 Despite the large doses and variety of vasopressors administered in this series, we observed only a single patient who developed an increase in troponin T after treatment with nicardipine; of note, no additional cardiac sequelae indicative of significant decrement of cardiac function or arrhythmias were observed in this patient. On review, this specific patient did have increased cardiac risk factors, and increases in troponin are not unusual in this patient population under physiological stress. Surprisingly, we detected neither development of systemic acidosis nor end-organ damage in any patient in this series despite the large doses of vasopressors and long duration of administration. Several differences in patient characteristics could account for the difference in findings. In contrast to the patients reported in this series, critical care patients requiring vasopressors to support perfusion frequently experience hypotension as a result of ongoing sepsis or heart failure. In both conditions, inadequate cardiac output contributes to poor perfusion. In contrast, post-SAH patients have been reported to exhibit increased cardiac indices, particularly during triple H therapy.17,20 The patients reported in this study were maintained euvolemic to hypervolemic during the period of vasopressor administration; the central venous pressure recorded in the intensive care unit before treatment averaged 10 to 12 mm Hg (data not shown). Specifically, these patients underwent an endovascular procedure because triple H therapy failed to treat vasospasm. This high state of hydration may account for adequate perfusion despite the vasoconstrictive effects of the pressors. Finally, the acute reduction in blood pressure that required increased pressor administration most likely resulted from the direct vasodilatation produced by the systemic effects of the nicardipine and milrinone. Although the contribution of cerebral vasculature to overall systemic vascular resistance is not known, it is clear from the angiography that the resistance in the cerebral bed was decreased by the treatment. Accordingly, additional vasopressors administered during endovascular treatment may serve to increase vascular tone reduced by the intraarterial administration of vasodilators. Alternatively, nicardipine may act as a negative inotrope and reduce cardiac output. Invasive hemodynamic monitoring of filling pressures and cardiac output during intraarterial vasodilator therapy is necessary to resolve the contributions of individual variables to the hypotensive response.

Mortality from SAH has been reported to be as high as 50% with fully one-third of patients requiring long-term care.21 Vasospasm, although previously described as a potentially preventable complication, has been found to contribute significantly to this mortality and morbidity. As recently as 2003, the incidence of mortality and disabling ischemic stroke attributed to cerebral vasospasm was reported to be as high as 20%.22 A case series reported marked improvement of vasospasm in patients after endovascular interventions with intraarterial nicardipine23 or milrinone.7 We report herein, in a series larger than previously published, 73 patients undergoing 160 procedures with a 30-day mortality of 11% and good to excellent improvement in vessel caliber in >90% of patients as assessed by angiography. In 7 of the 8 mortalities reported herein, patients had multiple cerebral infarctions and recurrent cerebral vasospasm that did not resolve with treatment; subsequently, care was withdrawn. Clinically, symptomatic vasospasm resolved in the remaining 65 patients in this series who went on to hospital discharge. Because of the retrospective nature of this review and the lack of a control group that did not receive endovascular treatment, we cannot ascribe the resolution in vasospasm in the surviving patients to result from treatment.

This review identifies several relevant features of patient management. The timing of cases in days after SAH in which treatment was performed agrees well with the previously reported occurrence of vasospasm.1,3 Moreover, the majority of patients required >1 treatment, and multiple vessels required treatment in most sessions. In terms of case management, all patients were monitored with an arterial catheter and required a central line in place for administration of vasoactive drugs. In this series, >90% of patients received endovascular treatment of vasospasm under general anesthesia, whereas only 7% of procedures were performed under sedation. Although the decision to proceed with general anesthesia rather than sedation was made prospectively by the anesthesiologist at the time of procedure, several factors could have contributed to this decision. The majority of patients were classified as ASA physical status III or IV and considered emergent at the time of treatment. The average case duration was in excess of 4 hours. Of the 150 treatments performed under general anesthesia, almost two-thirds of patients had an endotracheal tube in place, and their lungs were mechanically ventilated before the decision to investigate vasospasm. All of these factors favored selection of general anesthesia over sedation. Furthermore, of the 53 patients who were intubated for their procedure, it is noteworthy that less than half of these were extubated at the conclusion of the procedure. This most likely reflects the depressed level of consciousness in these patients presenting for treatment of vasospasm. As noted above, angiography is performed in our institution to investigate failure of triple H therapy to resolve neurological deficits, including declining mental status.5,16 In a single patient, angiography was initiated under sedation but was converted to general anesthesia because of the inability to maintain sufficient immobility; treatment in a previous session had been performed successfully under sedation.

In theory, performing these procedures under sedation has the specific advantage that alterations in neurological status during treatment are either immediately recognizable or can be elicited from examination. Under general anesthesia, however, the neurological examination is not possible and deferred until after recovery. In such cases, monitoring of intracranial pressure serves as a “surrogate” for neurological well-being. There are several advantages specific to general anesthesia as well. First, it provides the anesthesiologist with control of the airway in the event of progressive or sudden deterioration in mental status that can occur with intracranial hemorrhage, an increase in intracranial pressure, cerebral vessel occlusion, seizures, or potential side effects of the treatment procedure including profound hypotension. General anesthesia with muscle relaxation provides for a motionless field during advancement of the microcatheter distally into smaller vessels and imaging. Furthermore, hypercarbia resulting from sedation may further increase intracranial pressure in patients with reduced intracranial compliance. General anesthesia permits control of ventilation and provides an easy method for short-term management of intracranial pressure.

Inherent to the methodology of a retrospective chart review of management performed in a single center are several limitations on conclusions that can be obtained from the data. Specifically, from these data, we cannot determine either 1 anesthetic technique or results from 1 vasodilator to be superior over another. In a nonrandomized, nonprospective selection process, as reported in this study, biases introduced by the physicians performing the procedures may influence the results. In this study, the cerebral vasodilator was selected by the endovascular physician based on a number of factors including individual preference, degree of spasm, and hemodynamic stability of the patient. Frequently during procedures in which it was difficult to maintain arterial blood pressure at a target value, milrinone was selected as the treatment drug. Therefore, in the setting of this bias, it is not possible to compare nicardipine and milrinone in terms of efficacy or hemodynamic effect. Although the results presented in this study are descriptive of the events, by definition, all hemodynamic data were recorded prospectively during treatments. Similarly, the anesthesiologist made all decisions prospectively concerning both selection and dose of vasoconstrictors and rates of infusions as necessary to maintain target blood pressure without any influence that the data were subject to future review.

No conclusions can be obtained from selections of drug for maintenance anesthesia in this study. Although volatile anesthetics were selected more often than propofol infusion, there was no obvious reason for selection bias. Nitrous oxide was used in only 7% of general anesthesia cases. This reflected the preference of the endovascular physicians over concern for expansion of small, intraarterial bubbles that might be injected inadvertently during the procedure. An expanding arterial air embolism potentially could obstruct distal flow and potentially produce ischemia.

In conclusion, intraarterial administration of nicardipine and milrinone for treatment of cerebral vasospasm after aneurysmal SAH requires both close hemodynamic monitoring and administration of high doses of vasopressors to maintain an increased arterial blood pressure at target levels. Despite the intense vasoconstriction produced by these drugs, the procedure has a low mortality and results in increased vessel caliber within affected segments of vessels. Furthermore, there is minimal end-organ ischemic damage and systemic acidosis.

Back to Top | Article Outline


We thank Johnny C. Pryor, MD, of the Department of Neurosurgery and Department of Interventional Neuroradiology and Endovascular Neurosurgery at Massachusetts General Hospital for his many helpful discussions and encouragement in the preparation of the manuscript.

Back to Top | Article Outline


1.Harrod CG, Bendok BR, Batjer HH. Prediction of cerebral vasospasm in patients presenting with aneurysmal subarachnoid hemorrhage: a review. Neurosurgery 2005;56:633–54
2.Barker FG II, Ogilvy CS. Efficacy of prophylactic nimodipine for delayed ischemic deficit after subarachnoid hemorrhage: a metaanalysis. J Neurosurg 1996;84:405–14
3.Weyer GW, Nolan CP, Macdonald RL. Evidence-based cerebral vasospasm management. Neurosurg Focus 2006;21:E8
4.Lee KH, Lukovits T, Friedman JA. “Triple-H” therapy for cerebral vasospasm following subarachnoid hemorrhage. Neurocrit Care 2006;4:68–76
5.Badjatia N, Topcuoglu MA, Pryor JC, Rabinov JD, Ogilvy CS, Carter BS, Rordorf GA. Preliminary experience with intra-arterial nicardipine as a treatment for cerebral vasospasm. AJNR Am J Neuroradiol 2004;25:819–26
6.Arakawa Y, Kikuta K, Hojo M, Goto Y, Yamagata S, Nozaki K, Hashimoto N. Milrinone reduces cerebral vasospasm after subarachnoid hemorrhage of WFNS grade IV or V. Neurol Med Chir (Tokyo) 2004;44:393–400
7.Fraticelli AT, Cholley BP, Losser MR, Saint Maurice JP, Payen D. Milrinone for the treatment of cerebral vasospasm after aneurysmal subarachnoid hemorrhage. Stroke 2008;39:893–8
8.Biondi A, Ricciardi GK, Puybasset L, Abdennour L, Longo M, Chiras J, Van Effenterre R. Intra-arterial nimodipine for the treatment of symptomatic cerebral vasospasm after aneurysmal subarachnoid hemorrhage: preliminary results. AJNR Am J Neuroradiol 2004;25:1067–76
9.Arakawa Y, Kikuta K, Hojo M, Goto Y, Ishii A, Yamagata S. Milrinone for the treatment of cerebral vasospasm after subarachnoid hemorrhage: report of seven cases. Neurosurgery 2001;48:723–8
10.Biondi A, Le Jean L, Puybasset L. Clinical experience of selective intra-arterial nimodipine treatment for cerebral vasospasm following subarachnoid hemorrhage. AJNR Am J Neuroradiol 2006;27:474
11.Hui C, Lau KP. Efficacy of intra-arterial nimodipine in the treatment of cerebral vasospasm complicating subarachnoid haemorrhage. Clin Radiol 2005;60:1030–6
12.Avitsian R, Fiorella D, Soliman MM, Mascha E. Anesthetic considerations of selective intra-arterial nicardipine injection for intracranial vasospasm: a case series. J Neurosurg Anesthesiol 2007;19:125–9
13.Bomzon L, Rosendorff C. Renovascular resistance and noradrenaline. Am J Physiol 1975;229:1649–53
14.Hollenberg SM, Ahrens TS, Annane D, Astiz ME, Chalfin DB, Dasta JF, Heard SO, Martin C, Napolitano LM, Susla GM, Totaro R, Vincent JL, Zanotti-Cavazzoni S. Practice parameters for hemodynamic support of sepsis in adult patients: 2004 update. Crit Care Med 2004;32:1928–48
15.Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) group. Crit Care 2004;8:R204–R212
16.Treggiari MM, Walder B, Suter PM, Romand JA. Systematic review of the prevention of delayed ischemic neurological deficits with hypertension, hypervolemia, and hemodilution therapy following subarachnoid hemorrhage. J Neurosurg 2003;98:978–84
17.Miller JA, Dacey RG Jr, Diringer MN. Safety of hypertensive hypervolemic therapy with phenylephrine in the treatment of delayed ischemic deficits after subarachnoid hemorrhage. Stroke 1995;26:2260–6
18.Chatterjee K. Newer oral inotropic agents: phosphodiesterase inhibitors. Crit Care Med 1990;18:S34–S38
19.Lollgen H, Drexler H. Use of inotropes in the critical care setting. Crit Care Med 1990;18:S56–S60
20.Mayer SA, Lin J, Homma S, Solomon RA, Lennihan L, Sherman D, Fink ME, Beckford A, Klebanoff LM. Myocardial injury and left ventricular performance after subarachnoid hemorrhage. Stroke 1999;30:780–6
21.Suarez JI, Tarr RW, Selman WR. Aneurysmal subarachnoid hemorrhage. N Engl J Med 2006;354:387–96
22.Dumont AS, Dumont RJ, Chow MM, Lin CL, Calisaneller T, Ley KF, Kassell NF, Lee KS. Cerebral vasospasm after subarachnoid hemorrhage: putative role of inflammation. Neurosurgery 2003;53:123–33
23.Tejada JG, Taylor RA, Ugurel MS, Hayakawa M, Lee SK, Chaloupka JC. Safety and feasibility of intra-arterial nicardipine for the treatment of subarachnoid hemorrhage-associated vasospasm: initial clinical experience with high-dose infusions. AJNR Am J Neuroradiol 2007;28:844–8
© 2010 International Anesthesia Research Society