(Stroke. 2000;31:901.)
© 2000 American Heart Association, Inc.
Original Contributions |
From the Department of Anaesthesiology and Intensive Care (S.N., E.E., S.Z.), Institute of Surgical Sciences; the Department of Neurosurgery (M.R., B.R.), Institute of Clinical Neurosciences; and the Department of Clinical Physiology (G.L., P.F.), Sahlgrenska University Hospital, Göteborg, Sweden.
Correspondence to Silvana Naredi, MD, Department of Anaesthesiology and Intensive Care, Umeå University Hospital, S-901 85 Umeå, Sweden. E-mail Peter.Naredi{at}surgery.umu.se
| Abstract |
|---|
|
|
|---|
MethodsWe used an isotope dilution technique to estimate the magnitude and time course of sympathoadrenal activation in 18 subarachnoid patients.
ResultsCompared with 2 different control groups, the patients with subarachnoid hemorrhage exhibited an approximately 3-fold increase in total-body norepinephrine spillover into plasma within 48 hours after insult (3.2±0.3 and 4.2±0.7 versus 10.2±1.4 nmol/L; P<0.05 versus both). This sympathetic activation persisted throughout the 7- to 10-day examination period and was normalized at the 6-month follow-up visit.
ConclusionsThe present study has established that massive sympathetic nervous activation occurs in patients after subarachnoid hemorrhage. This overactivation may relate to the well-known cardiac complications described in subarachnoid hemorrhage.
Key Words: norepinephrine subarachnoid hemorrhage sympathetic nervous system
| Introduction |
|---|
|
|
|---|
Although reports exist that attempt to delineate sympathetic nervous system involvement after SAH, the techniques used (which usually measure antecubital venous plasma or urinary catecholamines) lack the precision of isotope dilution methodologies. Hitherto, NE and epinephrine (EPI) kinetic studies have not been performed in patients after acute nontraumatic SAH. Estimation of NE spillover into the circulation by use of dilution of intravenously infused [3H]-NE with endogenous NE in plasma (ie, specific activity of plasma [3H]-NE) provides a useful index of NE release from sympathetic nerves.14 However, most NE released is efficiently removed by neuronal and extraneuronal uptake, so that only a small portion escapes into the circulation.15 Plasma concentrations of NE are the net result of bidirectional flux of the transmitter. Both removal and release processes must be considered to adequately assess the release of NE into plasma, because changes in the former may alter plasma concentrations of NE, irrespective of any change in NE release from sympathetic nerves. The radiotracer technique is based on steady-state infusion of tracer amounts of radiolabeled NE to establish rates of entry and removal of NE to and from the plasma compartment.16
In the present study, measurements of total-body NE and EPI spillover and clearance were performed on 3 separate occasions within the first 10 days after SAH insult and in 5 patients at an outpatient follow-up visit approximately 6 months subsequent. Findings in the SAH patient group were compared with data obtained in 2 control groups: (1) healthy subjects and (2) patients investigated invasively in the intensive care unit for refractory pain.
| Subjects and Methods |
|---|
|
|
|---|
Subjects
Eighteen patients, including 8 men and 10 women (median age, 51
years; range, 36 to 66), with acute SAH were included in the study.
Fifteen healthy, age-matched volunteers of median age 52 years (range,
38 to 64) with no history of neurological or
cardiovascular disease and with ECG and routine serum
chemistry within the normal range, served as 1 control group. Eleven
patients with no evidence of cardiovascular disease and
without demonstrable ECG changes who were undergoing clinical
investigation for refractory pain were included as a second control
group (median age, 68 years; range, 37 to 81). The latter group
underwent a catheterization procedure in the intensive
care unit. Thus, the "environmental milieu" for these patients was
similar to that of the SAH patients.
All SAH patients were treated at the neurointensive care unit of
Sahlgrenska University Hospital. All patients were admitted to the
neurointensive care unit within 24 hours after onset of bleeding.
Median Hunt/Hess classification of patients with intracranial
aneurysms17 was 3.5 (range, I through V; see
Table
). None of the patients had a
history of myocardial disease. Two patients were on antihypertensive
ß-blocker treatment.
|
Methods
Diagnosis of SAH was made by CT scan. Estimation of the amount
of blood in the subarachnoid space was determined according to
the criteria proposed by Fisher et al18
(Table
).
Monitoring
All patients had central venous and arterial lines
for continuous monitoring of central venous and systemic
arterial blood pressures, respectively. Oxygen saturation
was determined by pulsoximetry (Datex Cardiocap 2). In 7 patients, an
intraventricular catheter was inserted as a result
of hydrocephalus (Stille EDN v-kat 20-cm set; monitor ICP). The
catheter was used to monitor intracranial pressure and to drain
cerebrospinal fluid.
The limit for drainage of cerebral spinal fluid was normally set at 20 cm H2O above forehead level. Zero baseline for intracranial pressure was set at forehead level and for systemic pressure at heart level (pressure monitor, Datex Cardiocap 2). The drainage system was closed every hour to measure actual intracranial pressure and calculate cerebral perfusion pressure. The value for mean arterial pressure was given on the pressure monitor, and the cerebral perfusion pressure was calculated as mean arterial pressure minus intracranial pressure. The nursing staff recorded these variables at least hourly. Myocardial function was evaluated with serial ECGs, myocardial enzymes (CK-MB and troponin), and echocardiography (Sequia Acuson). Cerebral vasospasm was identified by clinical signs, angiogram, CT scan, or transcranial Doppler (MDX-TCD-7 DLW, Elechtronische System GmbH).
Treatment
In 17 patients, bleeding was associated with the presence of an
aneurysm. In 1 patient, the cause of the bleeding remained
unknown; neither an aneurysm nor an arteriovenous malformation
could be found on repeated angiograms. Treatment of the
aneurysms was either by surgical clipping (n=5) or embolization
with platina coils (n=12).
All patients received the calcium-blocker nimodipine as a continuous intravenous infusion to prevent cerebral vasospasm.19 Thirteen patients were mechanically ventilated (Siemens servoventilator 900D or 900C) for more than 24 hours. During that time they were sedated with continuous intravenous infusion of either midazolam or propofol.
Infusion of Tritium-Labeled CAT
Tracer doses of
L-2,5,6-[3H]-NE and
L-N-methyl-[3H]-EPI (40
to 60 and 65 to 75 Ci/mmol, respectively; New England Nuclear) were
delivered at 1.0 to 1.5 µCi/min through a peripheral
vein.
Experimental Protocol and Blood Sampling
Catecholamine kinetic determinations were made
within 48 hours after SAH, and follow-up studies were performed 24
hours subsequent and 7 to 10 days after SAH. Measurements of total-body
NE and EPI spillover were performed on admittance in all 18
patients (within 48 hours). Six patients were thereafter excluded
because they received a drug that interferes with
catecholamine release. Twelve patients were measured at 72
hours after insult. Two of these died before the end of the first week.
The sympathoadrenal function of the remaining 10 patients was examined
on 3 occasions (ie, within 48 hours, after 72 hours, and 7 to 10 days
after the insult). In addition, in 5 of 10 patients, a further study
was performed approximately 6 months after SAH.
Baseline arterial blood samples were taken at least 30 minutes after tritium-labeled catecholamine infusion was begun. This procedure is based on the fact that steady-state conditions of tritiated NE prevail approximately 13 minutes after infusion begins.20 Blood samples (10 mL) were collected into ice-chilled tubes containing EGTA and glutathione. Plasma was separated by centrifugation and stored at -80°C until it was assayed.
Assays
Catecholamines were extracted from plasma (1 mL) and
samples of infusate (10 µL) with alumina adsorption, separated, and
quantified by high-performance liquid
chromatography with coulometric
detection.21 Timed collection of
[3H] eluate that left the electrochemical cell
permitted separation of the [3H]-labeled NE and
EPI for subsequent counting by liquid scintillation spectroscopy.
Interassay coefficients of variation were 4.6% and 10% for
endogenous NE and EPI, respectively, and 3.2% and 7% for
[3H]-NE and [3H]-EPI,
respectively. Intraassay coefficients of variation were 2% and 10%
for endogenous NE and EPI, respectively, and 3% for both
[3H]-NE and
[3H]-EPI.
Calculations
Separate total-body NE and EPI spillover
(STB), respectively, were ascertained by
use of the isotope dilution method proposed by Esler et
al16 and calculated according to the following
formula (in pmol/min):
![]() | (1) |
![]() | (2) |
![]() | (3) |
Statistics
All results, unless otherwise specified, are expressed as
mean±SEM. Comparisons between groups were evaluated by use of
Kruskal-Wallis nonparametric ANOVA. Mann-Whitney
nonparametric test was used for comparison of NE and EPI
plasma concentrations and total-body catecholamine
spillover between healthy subjects and SAH patients. The null
hypothesis was rejected if the 2-tailed P value was <0.05.
The possible relation between total-body NE spillover and
intracranial and cerebral perfusion pressures was assessed with the
Spearman rank correlation.
| Results |
|---|
|
|
|---|
|
|
Clinical Results
Ten of 12 patients studied throughout the week-long study period
presented abnormal ECG recordings, including QT
prolongation, depressed or elevated S segments, T-wave abnormalities,
and arrhythmias. The 2 patients who displayed normal rates of
NE spillover to plasma presented normal ECGs throughout the
entire study period. In the patients examined approximately six months
after SAH, abnormalities in the previous ECG had normalized completely.
Three patients exhibited CK-MB and troponin plasma levels above the
diagnostic level for myocardial infarction, and 1 displayed
a pathological echocardiography with 2 areas of
hypokinesia and reduced ejection fraction. In this patient, the
echocardiogram was normal on examination 6 months after insult.
Five patients exhibited signs of cerebral vasospasm detected clinically or by transcranial Doppler, angiography, or signs of ischemia on CT scan. Mean arterial and central venous pressures for the 12 patients throughout the study period were 97±2 and 8±0.5 mm Hg, respectively. These pressures were stable throughout the intensive-care study period. Moreover, heart rate remained stable during the same period at 75±3 bpm. In the 7 patients with intracranial pressure monitoring, mean pressure was 18±5 mm Hg. If the 2 patients who died were excluded, this value was reduced to 11±2 mm Hg; mean cerebral perfusion pressure for the 7 patients with intracranial pressure monitoring was 77±4 mm Hg. All patients had normal serum electrolyte levels.
Overall outcome for the 18 patients according to the Glasgow outcome
score (GOS) 3 months after the bleeding was as follows: 13 patients
survived with good outcome or were moderately disabled, 1 was severely
disabled, and 4 had died (Table
).
| Discussion |
|---|
|
|
|---|
The idea that sympathoexcitation with associated elevations in systemic
blood pressure after nontraumatic SAH arises secondary to the
accompanying elevation in intracranial pressure stems from the
experiments conducted by Harvey Cushing early in the 20th century.
Although these experiments are not necessarily directly applicable to
the setting of SAH, Cushing22 noted that increments in
intracranial pressure resulted in a marked blood pressure elevation. In
an animal model of SAH, a marked elevation in systemic
arterial pressure occurred only when the rise in
intracranial pressure approached systemic blood
pressure.23 In rats instrumented with a catheter in the
cisterna magna, angiographically demonstrated constriction of the
vertebrobasilar arteries occurred approximately 10 minutes after
cisternal injection of either 70 or 300 µL of homologous
blood.24 The degree of vasoconstriction elicited by the
different volumes of blood did not vary considerably, but the
intracisternal injection of the larger volume elicited a marked blood
pressure elevation.24 The rise in systemic pressure as a
result of the Cushing response can be prevented by prior
-adrenergic
blockade25 and is believed to be mediated by means of
alterations in neuronal activity in the brain stem26 27 28
in response to local ischemia,29 or more
particularly, hypoxia.30 Taken together, the
sympathoexcitation that we observe in the clinical setting after
nontraumatic SAH would appear to originate in local constriction of
small vessels that supply the brain stem.
The pathophysiology behind cerebral vasospasm is not completely understood. Some studies suggest the possibility that elevated levels of circulating catecholamines,31 coupled with an abnormal sensitivity of the cerebral vasculature to these catecholamines,32 may be involved in the genesis of vasospasm. Development of delayed cerebral ischemia is a serious problem despite the prophylactic use of the calcium antagonist nimodipine. In the present study, the 3 patients with the highest rates of total-body NE spillover also experienced cerebral vasospasm. This supports the idea that catecholamines are involved in but are probably not the only factor in the pathogenesis of vasospasm.
A variety of other vasoactive substances, such as serotonin, endothelin, oxyhemoglobin, and nitric oxide, have also been proposed to play a role in the genesis of vasospasm.31 32 33 34 Catecholamines potentiate the action of endothelin.35 Hemoglobin and oxyhemoglobin are particularly efficient at producing spasm of cerebral vessels in vitro and in vivo. Part of this effect may relate to the ability of hemoglobin to bind the vasodilator nitric oxide, but oxyhemoglobin also produces free radicals, which could induce vasospasm by means of several mechanisms. The demonstration of the effectiveness of endothelin antagonists36 and nitric oxide donors37 38 in the management of vasospasm after experimental SAH holds further promise in the treatment of the condition. Furthermore, activation of the renin-angiotensin system seems to be of importance in the SAH condition, although no firm relationship could be found with regard to vasospasm and ECG changes in a small patient study.39 Interestingly, Fassot and collaborators40 found in experimental SAH that an intact renin-angiotensin system was a prerequisite for maintaining adequate blood pressure control. It is reasonable to assume that a high degree of sympathetic nervous system activation also gives rise to a secondary increase in renin release from the kidney.
One pertinent question is whether the blood-brain barrier is intact or disrupted in the presently investigated SAH patients. However, we are convinced that it is still intact, for the following reasons. First, brain NE at rest contributes only marginally to total-body NE spillover (on the order of approximately 3% if the principal central nervous system metabolite to NE, 3-methoxy-4-hydroxyphenylglycol (MHPG), is considered.41 In the present study, which shows such a dramatic increase in total-body NE spillover in the SAH patient group, central nervous system NE spillover would have to be unbelievably increased, which is highly unlikely. Second, we also measured MHPG concentrations in the cerebrospinal fluid in both SAH and refractory-pain patients. MPHG concentrations in the SAH patients were only half those found in the pain patients (who then are very likely to have an intact blood-brain barrier), 0.52±0.06 and 1.07±0.18 nmol/L for the former and latter groups, respectively. If the blood-brain barrier had been disrupted, the MHPG concentrations in the cerebrospinal fluid would have been much higher in the SAH patients, not lower. Third, other preliminary experiments that we performed have demonstrated that clonidine, a sympatholytic centrally acting drug, clearly reduced sympathetic activity short term, which indicates an augmented sympathetic nerve firing that can be reduced (unpublished data, 1999). Thus, we believe that we have a true elevation in sympathetic nerve traffic to a major part of the body.
Findings are controversial as to the connection between elevated levels of plasma catecholamines and ECG abnormalities.2 3 4 5 6 10 11 In the present study, the 2 patients with normal total-body NE spillover to plasma were the only subjects who did not develop ECG abnormalities. Although the number of observations was limited, these data support the contention of a link between increased sympathetic activity and cardiac abnormalities after SAH. Although equivocal in SAH, the linkage between sympathetic activation and cardiac consequences is not without precedent. For example, in patients with congestive heart failure, not only is sympathetic nervous activity increased, but the degree of cardiac sympathetic activation is the most reliable index of mortality.13 One may envisage that with the high prevailing rates of total-body NE spillover in the presently studied SAH patients, it is reasonable to assume that cardiac NE spillover is grossly elevated. Hence, it is tempting to speculate that high NE levels in the heart are associated with the observed ECG changes after SAH, in line with important evidence that shows high cardiac NE spillover in patients investigated for malignant arrhythmias.42 Moreover, high catecholamine concentrations both in man and pigs have been linked to myocardial damage.43 44 45 Consequently, reduction of the high degree of sympathetic activation present in the SAH patients should be considered favorable.
In response to the hypotension that accompanies the use of vasodilator drugs, the sympathetic nervous system is reflexly stimulated. Hypotension associated with calcium channelblocking drugs such as nimodipine could elicit activation of the sympathetic nervous system. However, the patients in the present study were not hypotensive, but were instead the reverse, with a mean arterial pressure close to 100 mm Hg. Therefore, it seems unlikely that the observed elevation in sympathetic nervous activity was due to a reflex response to hypotension.
Methodological Considerations
The bulk of circulating NE represents transmitter release
from sympathetic nerve varicosities and <10% originates from adrenal
medullary secretion.15 The released NE is mainly
recaptured into the axoplasm by an active transport mechanism (uptake
1) and transferred back to storage vesicles. Most studies that
investigate sympathetic nervous activation after SAH patients have
analyzed NE concentrations in plasma obtained from antecubital
venous blood. This procedure does not correctly describe the magnitude
of sympathetic activity. Research has clearly shown that the forearm
effectively extracts approximately 50% of inflowing NE.46
Thus, the forearm is not representative of total-body
sympathetic activity. Instead, arterial sampling should be
performed to determine endogenous NE. At least, this
procedure avoids any organ-extraction procedures.
Several studies have also tried to estimate sympathetic activity in SAH
patients from measurements of NE and metanephrines, as well as
vanillylmandelic acid, in urine. However, NE in urine has a
complex source. Only
10% of circulating NE appears unchanged in the
urine, which thereby renders urinary measurements of NE for
interpretation of the degree of sympathetic activity extremely
difficult and hazardous.
The metanephrines are extraneuronal O-methylated metabolites of NE and EPI. Thus, their plasma concentrations enable examination of the extraneuronal uptake and metabolism of catecholamines. The adrenals are the major source of circulating metanephrines,15 and consequently metanephrines either in plasma or urine are not a good index of sympathetic activity. Considered collectively, estimation of NE spillover into the circulation as performed in the present study, with the dilution of intravenous, infused, and tritiated NE with endogenous NE in plasma (ie, the specific activity of plasma-tritiated NE), provides a well-documented, robust, and useful index of NE release from sympathetic nerves.14
We conclude that nontraumatic SAH is associated with an extreme elevation in sympathetic nervous system activity that persists at least during the first week after insult. This increased sympathetic nervous activity may, at least in part, be related to the observed ECG aberrations, perhaps due to elevated cardiac sympathetic drive, which, in turn, may cause myocardial structural alterations.
| Acknowledgments |
|---|
Received October 28, 1999; revision received January 6, 2000; accepted January 18, 2000.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. K. S. Leow, K. C. Loh, T. Kiat Kwek, and P. Y. Ng Catecholamine and metanephrine excess in intracerebral haemorrhage: revisiting an obscure yet common "pseudophaeochromocytoma" J. Clin. Pathol., May 1, 2007; 60(5): 583 - 584. [Full Text] [PDF] |
||||
![]() |
A. Hays and M. N. Diringer Elevated troponin levels are associated with higher mortality following intracerebral hemorrhage Neurology, May 9, 2006; 66(9): 1330 - 1334. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. M. Banki, A. Kopelnik, M. W. Dae, J. Miss, P. Tung, M. T. Lawton, B. J. Drew, E. Foster, W. Smith, W. W. Parmley, et al. Acute Neurocardiogenic Injury After Subarachnoid Hemorrhage Circulation, November 22, 2005; 112(21): 3314 - 3319. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Mazighi, J.-P. Saint Maurice, A. Rogopoulos, and E. Houdart Extracranial vertebral and carotid dissection occurring in the course of subarachnoid hemorrhage Neurology, November 8, 2005; 65(9): 1471 - 1473. [Abstract] [Full Text] [PDF] |
||||
![]() |
W J Schuiling, P J W Dennesen, J T. J Tans, L M Kingma, A Algra, and G J E Rinkel Troponin I in predicting cardiac or pulmonary complications and outcome in subarachnoid haemorrhage J. Neurol. Neurosurg. Psychiatry, November 1, 2005; 76(11): 1565 - 1569. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. J. D. Andrews, B. Harris, and G. D. Murray Randomized controlled trial of effects of the airflow through the upper respiratory tract of intubated brain-injured patients on brain temperature and selective brain cooling Br. J. Anaesth., March 1, 2005; 94(3): 330 - 335. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Tung, A. Kopelnik, N. Banki, K. Ong, N. Ko, M. T. Lawton, D. Gress, B. Drew, E. Foster, W. Parmley, et al. Predictors of Neurocardiogenic Injury After Subarachnoid Hemorrhage Stroke, February 1, 2004; 35(2): 548 - 551. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2000 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |