AMERICAN
JOURNAL
OF
Vol. 231, No. 3, September
PHYSIOLOGY
1976.
Printed
in U.S.A.
Autoregulation of cerebral blood flow and its relation to cerebrospinal fluid pH MILTON J. HERNANDEZ-PEREZ AND DOUGLAS K. ANDERSON Division of Neurology, Pennsylvania State University, Milton S. Hershey Medical Center, Hershey, Pennsylvania 17033; and Research Service, Tampa Veterans Administration Hospital and Department of Physiology, University of South Florida College of Medicine, Tampa, Florida 33620 hERNANDEz-PEREZ, MILTON J., AND DOUGLAS K. ANDERSON. Autoregulation of cerebral bZood fZow and its relation to cerebrospinal fluid pH. Am. J. Physiol. 231(3): 929-935. 1976. -Internal carotid artery blood flow (ICBF) was determined in each of nine Macaca muZatta by means of a flow transducer implanted around an internal carotid artery. The monkeys were lightly anesthetized, intubated, and paralyzed. Normoxia and normocarbia were maintained stable throughout the experiment. ICBF was monitored while mean arterial blood pressure (MABP) was lowered by withdrawal of blood. MABP was kept within the known limits of autoregulation in order not to compromise CBF. Cerebrospinal fluid (CSF) from the cisterna magna was analyzed for pH, Pco2, and PO, before and after a 30-min hypotensive period in which MABP was lowered from 116 t 4 to 70 t 2 mmHg (mean t SE). Corresponding HCO, concentrations were calculated. The decrease in MABP did not result in a signficant reduction in ICBF but elicited a 37% reduction in calculated cerebrovascular resistance, indicating normal autoregulation. Mean CSF pH was not significantly decreased (P < 0.05); it changed from 7.320 t 0.010 to 7.317 t 0.010 after the induced hypotensive period. Thus CSF pH does not appear to have a significant role in cerebral blood flow autoregulation.
hypertension, the increments in CVR necessary to keep CBF constant are brought about by cerebral vascular smooth muscle contracting in response to increased transmural pressure. Likewise, autoregulation is preserved when PP is lowered. Reducing PP decreases cerebral vascular smooth muscle contraction, resulting in cerebral vasodilation and the appropriate reduction in CVR. Autonomic factors are also known to influence cerebral circulation, but their role in the control of autoregulation appears only to be one of subtle modulation (11). Previous findings that even brief periods of lowered PP are capable of decreasing cerebrospinal fluid (CSF) pH and CVR have led to the explanation of CBF autoregulation by metabolic factors (6, 7, 15, 17, 18). It has been proposed that the acidity resulting from the accumulation of vasodilator metabolites, such as carbon dioxide or lactic acid, in CSF during periods of hypotension might cause compensatory decreases in CVR thus keeping blood flow constant. This study was conducted to ascertain whether decreased CVR, within the autoregulation range, is accompanied by decrements in CSF
Doppler
PH.
flow transducers;
rhesus monkeys
METHODS
may be defined as the phenomenon whereby adjustments in cerebrovascular resistance (CVR) tend to keep blood flow to the brain constant during increases or decreases in perfusion pressure (PP). In a broader context, the term autoregulation has also been employed to describe the regulation of blood flow in accordance with the metabolic needs of the brain. Yet constancy of CBF during changes in PP and regulation of CBF by metabolic demand are possibly two different phenomena, particularly if they occur in different segments of the cerebral vasculature (30). It is possible that true autoregulatory changes in CVR are not brought about by the metabolic activity of the surrounding brain tissue, but may be due to other factors. Myogenic (4, 31, 32), autonomic (3, 10, ZZ), and metabolic (20) elements have been identified in relation to autoregulation. According to the theory of autoregulation based on myogenic control, the adjustments in CVR needed to keep flow constant during variations in PP originate in the basic contractile properties of the cerebral arteries and arterioles. During periods of systemic CEREBRAL
BLOOD FLOW (CBF) AUTOREGULATION
Nine young adult monkeys (Macaca mulatta, 3-4 kg), were anesthetized with halothane and Doppler ultrasonic flow transducers were implanted around either the left or the right common carotid artery. The external carotid and superior thyroid arteries were doubly ligated, diverting common carotid artery flow through the internal carotid artery and ensuring that flow to extracranial vasculature was minimal (11). Before any experimentation, the monkeys were allowed to recover for 15-20 days, during which the transducers, by fibrosis, acquired a fi’xed position around the common carotid artery. Following the recovery period, the monkeys were lightly anesthetized with cyclohexylamine (lo-14 mg/ kg) and intubated with a cuffed endotracheal tube. The lead wires from the flow transducer were retrieved from a subcutaneous pouch above the shoulder and connected with Doppler electronics to measure carotid blood velocity. The animals were paralyzed with gallamine triethiodide (1 mg/kg) and artificially ventilated to maintain normal blood pH, Pco2, and PO,. These blood gas and pH
929
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
930
M.
determinations were made every 30-60 min except during the 30-min hypotensive period when they were measured every lo-15 min. Control PcoZ values were arbitrarily set in the normocarbic range but, once established, did not deviate more than 2 mmHg from control levels with the exception of monkey 9, which exhibited a variation of 5 mmHg. Cyclohexylamine and gallamine were supplemented hourly. Expired CO, was continuously monitored with a Beckman gas analyzer. Body temperature was continuously measured with a rectal thermometer and was maintained at 37°C by an electric heating pad. The abdominal aorta was cannulated by way of a femoral arteriotomy so that arterial blood and blood samples obpressu .re could be monitored tained . A femoral vein was cannulated for withdrawal of blood and administration of drugs. The monkeys were and a shortthen placed in a stereotaxic apparatus beveled, hubless, 22-gauge sta inless steel cannula was introduced percutaneously into the cisterna magna for the measurement of CSF pressure (PCSF) and for withdrawal of CSF samples. Hypotension was induced utilizing an infusion-withdrawal pump to remove 30-50 ml of venous blood into a heparinized loo-ml glass syringe. Blood withdrawal prostate w as ceeded at a rate of 8 ml/min. The hypotensive of maintained for 30 min, wih continuous adjustments the pump and the respirator to keep mean arterial blood pressure (MABP) close to 70 mmHg and blood gas and pH levels near control values. Cisternal CSF pH, Pco2, and PO, were determined before and after the 30-min period of hypovolemic hypotension. Continuous measurement of arterial blood pressure, ICBV, and PCSF were made on a direct-writing recorder. In addition, arterial blood pressure and ICBV were recorded in analog form on magnetic tape before and during the hypotensive period. Autoregulation was monitored by connecting arterial blood pressure and ICBV outputs from the recorder to the axes of an x-y plotter. All CSF and blood samples were drawn anaerobically into glass syringes and analyzed for pH, PcoZ, and PO, within 3-5 min with appropriate glass electrodes thermostated to 37°C. CSF samples in which air bubbles or traces of blood were evident were discarded. Repeated measurements of CSF pH were made without intermittent rinsing of the glass electrode. CSF pH was measured with a Radiometer pH meter, model 22, equipped with a scale expander. Triplicate measurements varied by 0.007 pH U or less. CSF Pco2 and PoZ, and blood pH, Pco~, and PO, were determined with an Instrumentation Laboratory model 313 pH/blood gas analyzer. CSF bicarbonate concentrations were calculated from the Henderson-Hasselbalch equation with use of a CO, molar solubility (S) of 0.0318 mmol/liter mmHg Pco2 for CSF and the appropriate pK1, according to the method of Mitchell et al. (24). A pK, of 6.10 and S of 0.0306 mmol/liter mmHg Pco2 were used to calculate blood bicarbonate concentration (28). Following each experiment, the common carotid artery encased in the flow transducer was cannulated, filled under pressure with an acrylic casting material which, when hardened, provided the acrylic casts for
J. HERNANDEZ-PEREZ
AND
D.
K. ANDERSON
determination of the diameter of the vessel within each flow transducer. Internal carotid blood flow (ICBF) was calculated as the product of ICBV and the cross-sectional area of the vessel. Since CSF pressure and cerebral venous pressure have been considered to be identical (8), perfusion pressure was defined as the difference between MABP and mean PCSF. CVR was then calculated according to the equation CVR
--
(MABP ICBF
PCSF)
Statistical analysis was accomplished by use of the paired-t test and linear regression. The level of significance chosen was 0.05.
RESULTS
Pressure-flow Relationships During After Hypovolemic Hypotension
and
Determination of changes in ICBF by means of flow transducers permits analysis of cerebral circulatory changes a) as instantaneous, dynamic phenomena occurring concomitantly with alterations in CVR (such hypotenas those induced in this study by hypovolemic in which sion); and b) durin g steady-state conditions CVR remains constant. Table 1 includes individual animal and mean values for ICBF and CVR during normotension (N), immediately after establishing hypotension (Hi,,) 9 and after 30 min in the hypotensive ). In this study, we will consider ICBF and state (Hsomin CVR values obtained at H~~~min as steady-state values. Table 1 and Figs. 1 and 2 show the results of reducing mean PP (t SE) from 111 t 4 to 66 t 2 mmHg on ICBF and CVR by the controlled withdrawal of venous blood. Mean ICBF decreased significantly from 24 t 2 ml/min at N to 19 t 2 ml/min at Himmyrising within 814 min to 23 t 2 ml/min and remaining at this level for the duration of the 30-min hypotensive period (Fig. 1). The difference in mean ICBF values from N to HSOrnin was not significant. The delayed autoregulatory response of ICBF to alterations in PP was present in all animals studied. CBF autoregulation was exhibited by all nine monkeys reported in this study; this was particularly evident under steady-state conditions. Mean calculated CVR decreased significantly with PP, dropping from 4.94 t 0.51 to 3.81 ? 0.38 mmHg ml-’ min-’ and finally to 3.03 t 0.24 mmHg ml+ min-’ at at Himm H 30min (Fig. 2). The d ecrement in mean CVR from N to H 30min was 37%. Cisternal CSF and Arterial Blood Acid-base Balance Before and After Hypovolemic Hypotension
l
CSF samples were obtained at normotension and at H 30min Withdrawal of l-ml sample resulted in CSF pressure dropping to near 0 mmH,O. In order not to introduce the additional variables of decreased intracranial pressure and its effects on CVR, following removal of the control CSF sample at N, we did not proceed with experimentation until CSF pressure had returned to pre-withdrawal 1evels. l
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
CBF
AUTOREGULATION
AND
TABLE 1. Internal carotid establishing hypotension,
CSF
931
pH
artery blood flow and resistance during or after 30 min in hypotensive state
normotension,
immediately
after
Himma Monkey
Mean Mean
MABP,’ mmHg
Condition
PP,” mmHg
PCSF,”
mmHg
ICBV, cm/s
H 30mln CVR,* mmHg ml-’ min ’
ICBF, * ml/min
f
ICBV, cm/s
b
ICBF, ml/min
CVR, mmHg ml-’ mine1
7 7
N’ H’
125 78
4 5
121 73
12 9
19 14
6.37 5.21
12 13
19 20
6.37 3.65
9 9
N H
94 60
6 5
88 55
16 13
25 20
3.52 2.75
16 14
25 22
3.52 2.50
IO 10
N H
113 78
5 4
108 74
20 16
29 23
3.72 3.22
20 19
29 28
3.72 2.64
12 12
N H
120 80
6 4
114 76
15 12
17 14
6.71 5.43
15 16
17 18
6.71 4.22
17 17
N H
100 70
5 4
95 66
20 17
26 22
3.65 3.00
20 18
26 23
3.65 2.87
19 19
N H
124 70
4 2
120 68
30 24
39 31
3.08 2.19
30 28
39 36
3.08 1.89
20 20
N H
132 60
4
126 56
11 8
19 14
6.63 4.00
11 11
19 19
6.63 2.95
22 22
N H
126 70
6 4
120 66
16 12
19 14
6.32 4.71
16 15
19 18
6.32 3.67
23 23
N H
112 66
4 2
108 64
21 15
24 17
4.50 3.70
21 19
24 22
4.50 2.91
N H
116 2 4 70 t 2
24 2 2 19 + 2”
4.94 + 0.51 3.81 + 0.38”
24 2 2 23 2 2
4.94 5 0.51 3.03 f 0.24”
+ SE + SE
(1 Immediately sure. f Internal
111 2 4 66 ” 2
after establishing hypotension. carotid artery blood velocity.
’ After B Internal
30 min carotid
in hypotensive artery blood
state. flow.
c Mean arterial h Cerebrovascular
I’ ’ ’ ’ ’ ’ ’ ‘I z
--
#---
+
-----
Himm
1
r
L,
’
50
I
I
I
60
70
80
I
I 90
I
I
I
100
110
120
As shown in Table 2, control CSF pH, PCo,, PO, and HC03- values did not differ significantly from those obtained at H:jomin.Mean CSF pH was 7.320 t 0.010 at N and 7.317 t 0.010 at Hsomin.In addition, the regression line related CSF pH to CVR (Fig. 3) was not significantly different from zero, indicating that CSF pH was independent of CVR within the normal autoregulatory range. Throughout the experiment, blood gases and pH were kept as close as possible to control values (Table 2). Particular care was taken to maintain arterial Pcoz at the established control levels. There was no change in individual arterial pH, Pco~, and PO, at N or at Haomin with the exception of monkey 9. Mean arterial pH was 7.417 * 0.021 at N and 7.392 t 0.017 Hsomin.The range of
I
I
I
” Perfusion li P < 0.05. I
pres-
I
-
TN
-;-- 5. E. CrT 4. 2E .
hmm I- /--F
f 3. -E
PP (mmHg) FIG. 1. Cerebral blood flow immediately after achieving the desired hypotensive state (H imm), and after 30 min of hypotension carotid artery blood flow (ICBF; U-3 :)~)mi~~). Mean values for internal ordinate) are plotted as a function of perfusion pressure (PP, abscissa). SE for ICBF and PP are represented by vertical and horizontal bars, respectively.
I
fluid pressure. j Hypotension.
6
N
H30 min
blood pressure. ” Cerebrospinal resistance. i Normotension.
H30min
2 t,
’ 50
I 60
I 70
I
I
1
80 90 100 PP (mmHg)
I
II0
I
120-
2. Cerebrovascular resistance (CVR, ordinate) at control, and H 30min plotted as a function of perfusion pressure (PP, abscissa). SE of CVR and PP are represented by vertical and horizontal bars, respectively. FIG.
Hirnrn9
initial pH values at N, from 7.294 in monkey 9 to 7.494 in monkey 19, is probably attributable to metabolic or dietary factors existing prior to the experiments. The mean values for PCO~ at N and Hzominwere 34.9 t 1.0 and 36.0 t 0.8 mmHg. Individual arterial PO, values were kept within normoxic levels; mean values ranged from 84.6 t 2.1 mmHg at N to 83.5 t 1.7 mmHg at H 30min* Arterial blood and CSF lactates were determined in two monkeys (Table 2). The modest rise elicited in CSF lactate at H3ominin both monkeys did not appear to be sufficient to change CSF pH in either monkey. Arterial blood pH, on the other hand, was lowered by 0.033 and
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
932
M.
TABLE 2. Arterial blood and cerebrospinal after 30 min in hypotensive state
fluid
Arterial Monkey
Condition
PC029 mmHg
PH
acid-base
status,
J. HERNANDEZ-PEREZ
during
AND
normotension
Blood HC03-, meq/liter
Lactate mg/lOO ml
PH
Pcoz, mmHg
K. ANDERSON
and
Cerebrospinal
PO*, mmHg
D.
PO,, mmHg
Fluid HCO,-, meq/liter
7 7
N H
7.367 7.360
37.3 37.8
83.9 75.1
21.1 21.0
7.276
55.4 55.5
52.8 55.5
24.0 24.5
9 9
N H
7.294 7.334
29.7 34.7
86.1
83.4
14.2 18.2
7.297 7.294
49.7 50.5
48.3 45.8
23.0 23.2
10 10
N H
7.470
33.5 34.6
92.4 86.7
24.0 23.8
7.300 7.298
50.4
7.451
51.8
47.9 48.6
24.1
12 12
N H
7.384 7.346
31.9 31.8
74.8 80.4
18.8 17.1
7.327
44.8
7.313
44.1
17 17
N H
7.454 7.446
34.7 35.0
88.8 84.8
24.0 23.8
7.362 7.353
19 19
N H
7.494 7.469
38.9 38.8
83.9 83.6
29.5 27.8
7.315 7.319
20 20
N
7.468 7.382
37.6 39.4
81.7 93.6
26.8
H
23.1
7.297 7.289
22 22
N H
7.409 7.376
34.1
77.2 88.0
21.3
16.2
20.7
25.8
23 23
N
7.416
7.365
36.2 36.5
93.1
H
22.9 20.6
43.7
Mean SE
N
Mean SE
H
35.8
7.417
N, normotensive;
82.5
34.9
22.5
84.6
20.021
21.0
22.1
7.392 kO.017
36.0 +0.8
k1.7
83.5
I
I
I
I
I
22.3
48.7 47.9
43.7
26.4 25.4
52.5
35.6 49.7
25.4
50.7 50.2
22.9 23.1
51.3 49.2
50.8
48.1
21.3
25.1
7.360 7.366
16.0 17.9
7.350 7.346
13.2 14.3
50.1
46.6
23.9
kO.010
k1.2
t2.1
kO.6
21.8 21.1
7.317 +O.OlO
50.3
50.2
21.3
a.2
23.8 kO.5
I
I
I
7.2OOc
-I
I I
1 2 CVR
I 3
I
I
I
4 5 (mmHg.ml-‘.rnirri)
I
6
7
8
FIG. 3. Cisternal cerebrospinal fluid pH (CSF pH; ordinate) plotted as a function of cerebrovascular resistance (CVR, abscissa). Equation for line was calculated by linear regression and is: CSF pH = -0.003(~0.005)CVR + 7.320.
0.051 U in each of the two monkeys, in part the result of an increase in lactate of 9.6 and 26.3 mg/lOO ml, respectively. DISCUSSION
Methodological
47.3 53.3
21.5
7.400-
7.ooot
17.4
23.5
7.320
H, hypotensive. I
0
7.269
Lactate, mg/lOO ml
Considerations
Flow measurements. The arterial inflow method has been employed by us and other investigators as a sensitive index of cerebral circulation (11, 13, 21, 23, 33, 35). This method, employing electromagnetic or Doppler flow transducers to measure arterial inflow, reflects small changes occurring in total CBF (16, 29) and is
uniquely suited to determine rapid shifts in global cerebral circulatory patterns. Measurements during steadystate conditions can also be made with equal validity. A chronic, rather than acute preparation was chosen for this study in order to minimize the effects of surgical trauma and transducer movement artifacts on ICBF measurements during the experimental period. In man and nonhuman primates, both common carotid arteries bifurcate to form the internal and external carotid arteries. The internal carotid arteries do not branch before entering the skull. After entering the skull, but before joining the circle of Willis, the internal carotid arteries contribute a few minute branches to the dura and other structures too small to be of significance. In addition, the ophthalmic artery (1) also originates from this portion of the internal carotid artery. The ophthalmic artery supplies neural and nonneural tissues of the globe and, in addition, is distributed to the lacrimal apparatus, the ocular muscles, and part of the nasal mucosa (1, 34). Since the ophthalmic artery remains intact in our method, flow through this vessel to extracerebral tissue is incorporated into our ICBF measurements and must be taken into consideration. In our chronic monkey preparation, we have studied the anatomical relationships that exist between the intra- and extracerebral vasculatures via the ophthalmic arteries by rapid sequence cerebral angiography, India ink infusions, and examination of acrylic casts after
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
CBF
AUTOREGULATION
AND
CSF
pH
corrosion with sodium hydroxide. These studies were carried out on three monkeys and suggested that the anastomoses with the extracranial vasculature were minimal. The results of our India ink infusions were identical to those of Dumke and Schmidt (Z), who found that only small twigs in and close to the orbits showed evidence of the injected dye. They concluded that these channels were negligible. Additionally, during the development of our model, experiments were carried out on chronically implanted monkeys in which one Doppler flow transducer had been placed on the middle cerebral artery, with a second on either the internal carotid or the common carotid artery. In some instances, all three arteries were implanted. From the internal carotid or from the common carotid with the external branches chronically ligated, changes in blood flow induced by carbon dioxide inhalation, hypoxia, or changes in arterial blood pressure correlated closely, both in degree and direction, with simultaneous measurements made from the middle cerebral artery transducer (12). If the external carotid branches were left unoccluded, measurements from the common carotid artery did not, as expected, correlate closely with those from the flow transducer on the middle cerebral artery. Meyer et al. (23) attempted to determine the amount of extracerebral contamination present in their arterial internal carotid inflow preparations. They measured artery flow in two monkeys before and after enucleation of the ipsilateral eye and removal of the orbital contents. These investiga .tors concluded th at extracerebral contamination was minimal (less than 10%) and did not significantly affect their results. We have measured ICBF in three monkeys, by the arterial inflow method, as described in the present report, before and after careful removal of the ipsilateral orbital contents. At a mean arterial blood pressure of 100 mmHg, orbital enucleation resulted in only a 6-10% decrease in our resting ICBF values. Our observations agree with those of Meyer et al. (23) and, together with our other anatomical and physiological observations, suggest that the extracerebral component of our arterial inflow ICBF measurements is not significant. In the macaque monkey, blood flow through one internal carotid artery represents 30-35% of the entire cerebral circulation. Therefore, our CVR values are necessarily higher than comparable ones elsewhere in the literature. They reflect, however, as does ICBF, true changes in cerebral circulation. Under the conditions of our experiment, the normal were preserved. responses of the cerebral vasculature Autoregul .ation was present in all nine monkeys. To minimize the effects of changes in arterial Pco2 on ICBF, this variable was maintained constant in all experiments. CSF sumpZing. If the appropriate steady-state conditions exist (5), cisternal CSF should reflect changes in brain extracellular fluid acid-base status. Since arterial blood pH and gases were maintained constant throughout the experiment and CBF was not compromised, steady-state conditions should have existed between all cerebral fluids (i.e., plasma, brain extracellular fluid,
933 and CSF). In addition, in these studies, the animals were maintained hypotensive for 30 min before CSF was sampled. Other investigators (6, 7, 14, 15, 17, 18) have previously demonstrated decreased CSF pH values following shorter intervals (3-20 min) of lowered PP. For further evidence that a steady state had been achieved at H 30min 9 two monkeys were maintained at hypotensive pressure for 4 h; no CSF pH changes were detected with repeated sampling throughout this time period. Experimental considerations. One of the mechanisms proposed for the maintenance of autoregulation involves the effects of carbon dioxide evolved in normal aerobic metabolism on the opening and closing of small intraparenchymal vessels. According to this theory (27), brief periods of lowered PP delay the washout of metabolically produced carbon dioxide in any given segment of the brain’s microcirculation. The slight, localized accumulation of carbon dioxide and its effects on brain extracellular fluid pH bring about the opening of the vessels in question. Since carbon dioxide is highly diffusible into the bloodstream, it rapidly disappears from brain extracellular fluid. Thus there would probably not be any change in CSF pH or Pco2 at H:
JOURNAL
OF
Vol. 231, No. 3, September
PHYSIOLOGY
1976.
Printed
in U.S.A.
Autoregulation of cerebral blood flow and its relation to cerebrospinal fluid pH MILTON J. HERNANDEZ-PEREZ AND DOUGLAS K. ANDERSON Division of Neurology, Pennsylvania State University, Milton S. Hershey Medical Center, Hershey, Pennsylvania 17033; and Research Service, Tampa Veterans Administration Hospital and Department of Physiology, University of South Florida College of Medicine, Tampa, Florida 33620 hERNANDEz-PEREZ, MILTON J., AND DOUGLAS K. ANDERSON. Autoregulation of cerebral bZood fZow and its relation to cerebrospinal fluid pH. Am. J. Physiol. 231(3): 929-935. 1976. -Internal carotid artery blood flow (ICBF) was determined in each of nine Macaca muZatta by means of a flow transducer implanted around an internal carotid artery. The monkeys were lightly anesthetized, intubated, and paralyzed. Normoxia and normocarbia were maintained stable throughout the experiment. ICBF was monitored while mean arterial blood pressure (MABP) was lowered by withdrawal of blood. MABP was kept within the known limits of autoregulation in order not to compromise CBF. Cerebrospinal fluid (CSF) from the cisterna magna was analyzed for pH, Pco2, and PO, before and after a 30-min hypotensive period in which MABP was lowered from 116 t 4 to 70 t 2 mmHg (mean t SE). Corresponding HCO, concentrations were calculated. The decrease in MABP did not result in a signficant reduction in ICBF but elicited a 37% reduction in calculated cerebrovascular resistance, indicating normal autoregulation. Mean CSF pH was not significantly decreased (P < 0.05); it changed from 7.320 t 0.010 to 7.317 t 0.010 after the induced hypotensive period. Thus CSF pH does not appear to have a significant role in cerebral blood flow autoregulation.
hypertension, the increments in CVR necessary to keep CBF constant are brought about by cerebral vascular smooth muscle contracting in response to increased transmural pressure. Likewise, autoregulation is preserved when PP is lowered. Reducing PP decreases cerebral vascular smooth muscle contraction, resulting in cerebral vasodilation and the appropriate reduction in CVR. Autonomic factors are also known to influence cerebral circulation, but their role in the control of autoregulation appears only to be one of subtle modulation (11). Previous findings that even brief periods of lowered PP are capable of decreasing cerebrospinal fluid (CSF) pH and CVR have led to the explanation of CBF autoregulation by metabolic factors (6, 7, 15, 17, 18). It has been proposed that the acidity resulting from the accumulation of vasodilator metabolites, such as carbon dioxide or lactic acid, in CSF during periods of hypotension might cause compensatory decreases in CVR thus keeping blood flow constant. This study was conducted to ascertain whether decreased CVR, within the autoregulation range, is accompanied by decrements in CSF
Doppler
PH.
flow transducers;
rhesus monkeys
METHODS
may be defined as the phenomenon whereby adjustments in cerebrovascular resistance (CVR) tend to keep blood flow to the brain constant during increases or decreases in perfusion pressure (PP). In a broader context, the term autoregulation has also been employed to describe the regulation of blood flow in accordance with the metabolic needs of the brain. Yet constancy of CBF during changes in PP and regulation of CBF by metabolic demand are possibly two different phenomena, particularly if they occur in different segments of the cerebral vasculature (30). It is possible that true autoregulatory changes in CVR are not brought about by the metabolic activity of the surrounding brain tissue, but may be due to other factors. Myogenic (4, 31, 32), autonomic (3, 10, ZZ), and metabolic (20) elements have been identified in relation to autoregulation. According to the theory of autoregulation based on myogenic control, the adjustments in CVR needed to keep flow constant during variations in PP originate in the basic contractile properties of the cerebral arteries and arterioles. During periods of systemic CEREBRAL
BLOOD FLOW (CBF) AUTOREGULATION
Nine young adult monkeys (Macaca mulatta, 3-4 kg), were anesthetized with halothane and Doppler ultrasonic flow transducers were implanted around either the left or the right common carotid artery. The external carotid and superior thyroid arteries were doubly ligated, diverting common carotid artery flow through the internal carotid artery and ensuring that flow to extracranial vasculature was minimal (11). Before any experimentation, the monkeys were allowed to recover for 15-20 days, during which the transducers, by fibrosis, acquired a fi’xed position around the common carotid artery. Following the recovery period, the monkeys were lightly anesthetized with cyclohexylamine (lo-14 mg/ kg) and intubated with a cuffed endotracheal tube. The lead wires from the flow transducer were retrieved from a subcutaneous pouch above the shoulder and connected with Doppler electronics to measure carotid blood velocity. The animals were paralyzed with gallamine triethiodide (1 mg/kg) and artificially ventilated to maintain normal blood pH, Pco2, and PO,. These blood gas and pH
929
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
930
M.
determinations were made every 30-60 min except during the 30-min hypotensive period when they were measured every lo-15 min. Control PcoZ values were arbitrarily set in the normocarbic range but, once established, did not deviate more than 2 mmHg from control levels with the exception of monkey 9, which exhibited a variation of 5 mmHg. Cyclohexylamine and gallamine were supplemented hourly. Expired CO, was continuously monitored with a Beckman gas analyzer. Body temperature was continuously measured with a rectal thermometer and was maintained at 37°C by an electric heating pad. The abdominal aorta was cannulated by way of a femoral arteriotomy so that arterial blood and blood samples obpressu .re could be monitored tained . A femoral vein was cannulated for withdrawal of blood and administration of drugs. The monkeys were and a shortthen placed in a stereotaxic apparatus beveled, hubless, 22-gauge sta inless steel cannula was introduced percutaneously into the cisterna magna for the measurement of CSF pressure (PCSF) and for withdrawal of CSF samples. Hypotension was induced utilizing an infusion-withdrawal pump to remove 30-50 ml of venous blood into a heparinized loo-ml glass syringe. Blood withdrawal prostate w as ceeded at a rate of 8 ml/min. The hypotensive of maintained for 30 min, wih continuous adjustments the pump and the respirator to keep mean arterial blood pressure (MABP) close to 70 mmHg and blood gas and pH levels near control values. Cisternal CSF pH, Pco2, and PO, were determined before and after the 30-min period of hypovolemic hypotension. Continuous measurement of arterial blood pressure, ICBV, and PCSF were made on a direct-writing recorder. In addition, arterial blood pressure and ICBV were recorded in analog form on magnetic tape before and during the hypotensive period. Autoregulation was monitored by connecting arterial blood pressure and ICBV outputs from the recorder to the axes of an x-y plotter. All CSF and blood samples were drawn anaerobically into glass syringes and analyzed for pH, PcoZ, and PO, within 3-5 min with appropriate glass electrodes thermostated to 37°C. CSF samples in which air bubbles or traces of blood were evident were discarded. Repeated measurements of CSF pH were made without intermittent rinsing of the glass electrode. CSF pH was measured with a Radiometer pH meter, model 22, equipped with a scale expander. Triplicate measurements varied by 0.007 pH U or less. CSF Pco2 and PoZ, and blood pH, Pco~, and PO, were determined with an Instrumentation Laboratory model 313 pH/blood gas analyzer. CSF bicarbonate concentrations were calculated from the Henderson-Hasselbalch equation with use of a CO, molar solubility (S) of 0.0318 mmol/liter mmHg Pco2 for CSF and the appropriate pK1, according to the method of Mitchell et al. (24). A pK, of 6.10 and S of 0.0306 mmol/liter mmHg Pco2 were used to calculate blood bicarbonate concentration (28). Following each experiment, the common carotid artery encased in the flow transducer was cannulated, filled under pressure with an acrylic casting material which, when hardened, provided the acrylic casts for
J. HERNANDEZ-PEREZ
AND
D.
K. ANDERSON
determination of the diameter of the vessel within each flow transducer. Internal carotid blood flow (ICBF) was calculated as the product of ICBV and the cross-sectional area of the vessel. Since CSF pressure and cerebral venous pressure have been considered to be identical (8), perfusion pressure was defined as the difference between MABP and mean PCSF. CVR was then calculated according to the equation CVR
--
(MABP ICBF
PCSF)
Statistical analysis was accomplished by use of the paired-t test and linear regression. The level of significance chosen was 0.05.
RESULTS
Pressure-flow Relationships During After Hypovolemic Hypotension
and
Determination of changes in ICBF by means of flow transducers permits analysis of cerebral circulatory changes a) as instantaneous, dynamic phenomena occurring concomitantly with alterations in CVR (such hypotenas those induced in this study by hypovolemic in which sion); and b) durin g steady-state conditions CVR remains constant. Table 1 includes individual animal and mean values for ICBF and CVR during normotension (N), immediately after establishing hypotension (Hi,,) 9 and after 30 min in the hypotensive ). In this study, we will consider ICBF and state (Hsomin CVR values obtained at H~~~min as steady-state values. Table 1 and Figs. 1 and 2 show the results of reducing mean PP (t SE) from 111 t 4 to 66 t 2 mmHg on ICBF and CVR by the controlled withdrawal of venous blood. Mean ICBF decreased significantly from 24 t 2 ml/min at N to 19 t 2 ml/min at Himmyrising within 814 min to 23 t 2 ml/min and remaining at this level for the duration of the 30-min hypotensive period (Fig. 1). The difference in mean ICBF values from N to HSOrnin was not significant. The delayed autoregulatory response of ICBF to alterations in PP was present in all animals studied. CBF autoregulation was exhibited by all nine monkeys reported in this study; this was particularly evident under steady-state conditions. Mean calculated CVR decreased significantly with PP, dropping from 4.94 t 0.51 to 3.81 ? 0.38 mmHg ml-’ min-’ and finally to 3.03 t 0.24 mmHg ml+ min-’ at at Himm H 30min (Fig. 2). The d ecrement in mean CVR from N to H 30min was 37%. Cisternal CSF and Arterial Blood Acid-base Balance Before and After Hypovolemic Hypotension
l
CSF samples were obtained at normotension and at H 30min Withdrawal of l-ml sample resulted in CSF pressure dropping to near 0 mmH,O. In order not to introduce the additional variables of decreased intracranial pressure and its effects on CVR, following removal of the control CSF sample at N, we did not proceed with experimentation until CSF pressure had returned to pre-withdrawal 1evels. l
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
CBF
AUTOREGULATION
AND
TABLE 1. Internal carotid establishing hypotension,
CSF
931
pH
artery blood flow and resistance during or after 30 min in hypotensive state
normotension,
immediately
after
Himma Monkey
Mean Mean
MABP,’ mmHg
Condition
PP,” mmHg
PCSF,”
mmHg
ICBV, cm/s
H 30mln CVR,* mmHg ml-’ min ’
ICBF, * ml/min
f
ICBV, cm/s
b
ICBF, ml/min
CVR, mmHg ml-’ mine1
7 7
N’ H’
125 78
4 5
121 73
12 9
19 14
6.37 5.21
12 13
19 20
6.37 3.65
9 9
N H
94 60
6 5
88 55
16 13
25 20
3.52 2.75
16 14
25 22
3.52 2.50
IO 10
N H
113 78
5 4
108 74
20 16
29 23
3.72 3.22
20 19
29 28
3.72 2.64
12 12
N H
120 80
6 4
114 76
15 12
17 14
6.71 5.43
15 16
17 18
6.71 4.22
17 17
N H
100 70
5 4
95 66
20 17
26 22
3.65 3.00
20 18
26 23
3.65 2.87
19 19
N H
124 70
4 2
120 68
30 24
39 31
3.08 2.19
30 28
39 36
3.08 1.89
20 20
N H
132 60
4
126 56
11 8
19 14
6.63 4.00
11 11
19 19
6.63 2.95
22 22
N H
126 70
6 4
120 66
16 12
19 14
6.32 4.71
16 15
19 18
6.32 3.67
23 23
N H
112 66
4 2
108 64
21 15
24 17
4.50 3.70
21 19
24 22
4.50 2.91
N H
116 2 4 70 t 2
24 2 2 19 + 2”
4.94 + 0.51 3.81 + 0.38”
24 2 2 23 2 2
4.94 5 0.51 3.03 f 0.24”
+ SE + SE
(1 Immediately sure. f Internal
111 2 4 66 ” 2
after establishing hypotension. carotid artery blood velocity.
’ After B Internal
30 min carotid
in hypotensive artery blood
state. flow.
c Mean arterial h Cerebrovascular
I’ ’ ’ ’ ’ ’ ’ ‘I z
--
#---
+
-----
Himm
1
r
L,
’
50
I
I
I
60
70
80
I
I 90
I
I
I
100
110
120
As shown in Table 2, control CSF pH, PCo,, PO, and HC03- values did not differ significantly from those obtained at H:jomin.Mean CSF pH was 7.320 t 0.010 at N and 7.317 t 0.010 at Hsomin.In addition, the regression line related CSF pH to CVR (Fig. 3) was not significantly different from zero, indicating that CSF pH was independent of CVR within the normal autoregulatory range. Throughout the experiment, blood gases and pH were kept as close as possible to control values (Table 2). Particular care was taken to maintain arterial Pcoz at the established control levels. There was no change in individual arterial pH, Pco~, and PO, at N or at Haomin with the exception of monkey 9. Mean arterial pH was 7.417 * 0.021 at N and 7.392 t 0.017 Hsomin.The range of
I
I
I
” Perfusion li P < 0.05. I
pres-
I
-
TN
-;-- 5. E. CrT 4. 2E .
hmm I- /--F
f 3. -E
PP (mmHg) FIG. 1. Cerebral blood flow immediately after achieving the desired hypotensive state (H imm), and after 30 min of hypotension carotid artery blood flow (ICBF; U-3 :)~)mi~~). Mean values for internal ordinate) are plotted as a function of perfusion pressure (PP, abscissa). SE for ICBF and PP are represented by vertical and horizontal bars, respectively.
I
fluid pressure. j Hypotension.
6
N
H30 min
blood pressure. ” Cerebrospinal resistance. i Normotension.
H30min
2 t,
’ 50
I 60
I 70
I
I
1
80 90 100 PP (mmHg)
I
II0
I
120-
2. Cerebrovascular resistance (CVR, ordinate) at control, and H 30min plotted as a function of perfusion pressure (PP, abscissa). SE of CVR and PP are represented by vertical and horizontal bars, respectively. FIG.
Hirnrn9
initial pH values at N, from 7.294 in monkey 9 to 7.494 in monkey 19, is probably attributable to metabolic or dietary factors existing prior to the experiments. The mean values for PCO~ at N and Hzominwere 34.9 t 1.0 and 36.0 t 0.8 mmHg. Individual arterial PO, values were kept within normoxic levels; mean values ranged from 84.6 t 2.1 mmHg at N to 83.5 t 1.7 mmHg at H 30min* Arterial blood and CSF lactates were determined in two monkeys (Table 2). The modest rise elicited in CSF lactate at H3ominin both monkeys did not appear to be sufficient to change CSF pH in either monkey. Arterial blood pH, on the other hand, was lowered by 0.033 and
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
932
M.
TABLE 2. Arterial blood and cerebrospinal after 30 min in hypotensive state
fluid
Arterial Monkey
Condition
PC029 mmHg
PH
acid-base
status,
J. HERNANDEZ-PEREZ
during
AND
normotension
Blood HC03-, meq/liter
Lactate mg/lOO ml
PH
Pcoz, mmHg
K. ANDERSON
and
Cerebrospinal
PO*, mmHg
D.
PO,, mmHg
Fluid HCO,-, meq/liter
7 7
N H
7.367 7.360
37.3 37.8
83.9 75.1
21.1 21.0
7.276
55.4 55.5
52.8 55.5
24.0 24.5
9 9
N H
7.294 7.334
29.7 34.7
86.1
83.4
14.2 18.2
7.297 7.294
49.7 50.5
48.3 45.8
23.0 23.2
10 10
N H
7.470
33.5 34.6
92.4 86.7
24.0 23.8
7.300 7.298
50.4
7.451
51.8
47.9 48.6
24.1
12 12
N H
7.384 7.346
31.9 31.8
74.8 80.4
18.8 17.1
7.327
44.8
7.313
44.1
17 17
N H
7.454 7.446
34.7 35.0
88.8 84.8
24.0 23.8
7.362 7.353
19 19
N H
7.494 7.469
38.9 38.8
83.9 83.6
29.5 27.8
7.315 7.319
20 20
N
7.468 7.382
37.6 39.4
81.7 93.6
26.8
H
23.1
7.297 7.289
22 22
N H
7.409 7.376
34.1
77.2 88.0
21.3
16.2
20.7
25.8
23 23
N
7.416
7.365
36.2 36.5
93.1
H
22.9 20.6
43.7
Mean SE
N
Mean SE
H
35.8
7.417
N, normotensive;
82.5
34.9
22.5
84.6
20.021
21.0
22.1
7.392 kO.017
36.0 +0.8
k1.7
83.5
I
I
I
I
I
22.3
48.7 47.9
43.7
26.4 25.4
52.5
35.6 49.7
25.4
50.7 50.2
22.9 23.1
51.3 49.2
50.8
48.1
21.3
25.1
7.360 7.366
16.0 17.9
7.350 7.346
13.2 14.3
50.1
46.6
23.9
kO.010
k1.2
t2.1
kO.6
21.8 21.1
7.317 +O.OlO
50.3
50.2
21.3
a.2
23.8 kO.5
I
I
I
7.2OOc
-I
I I
1 2 CVR
I 3
I
I
I
4 5 (mmHg.ml-‘.rnirri)
I
6
7
8
FIG. 3. Cisternal cerebrospinal fluid pH (CSF pH; ordinate) plotted as a function of cerebrovascular resistance (CVR, abscissa). Equation for line was calculated by linear regression and is: CSF pH = -0.003(~0.005)CVR + 7.320.
0.051 U in each of the two monkeys, in part the result of an increase in lactate of 9.6 and 26.3 mg/lOO ml, respectively. DISCUSSION
Methodological
47.3 53.3
21.5
7.400-
7.ooot
17.4
23.5
7.320
H, hypotensive. I
0
7.269
Lactate, mg/lOO ml
Considerations
Flow measurements. The arterial inflow method has been employed by us and other investigators as a sensitive index of cerebral circulation (11, 13, 21, 23, 33, 35). This method, employing electromagnetic or Doppler flow transducers to measure arterial inflow, reflects small changes occurring in total CBF (16, 29) and is
uniquely suited to determine rapid shifts in global cerebral circulatory patterns. Measurements during steadystate conditions can also be made with equal validity. A chronic, rather than acute preparation was chosen for this study in order to minimize the effects of surgical trauma and transducer movement artifacts on ICBF measurements during the experimental period. In man and nonhuman primates, both common carotid arteries bifurcate to form the internal and external carotid arteries. The internal carotid arteries do not branch before entering the skull. After entering the skull, but before joining the circle of Willis, the internal carotid arteries contribute a few minute branches to the dura and other structures too small to be of significance. In addition, the ophthalmic artery (1) also originates from this portion of the internal carotid artery. The ophthalmic artery supplies neural and nonneural tissues of the globe and, in addition, is distributed to the lacrimal apparatus, the ocular muscles, and part of the nasal mucosa (1, 34). Since the ophthalmic artery remains intact in our method, flow through this vessel to extracerebral tissue is incorporated into our ICBF measurements and must be taken into consideration. In our chronic monkey preparation, we have studied the anatomical relationships that exist between the intra- and extracerebral vasculatures via the ophthalmic arteries by rapid sequence cerebral angiography, India ink infusions, and examination of acrylic casts after
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on November 13, 2018. Copyright © 1976 American Physiological Society. All rights reserved.
CBF
AUTOREGULATION
AND
CSF
pH
corrosion with sodium hydroxide. These studies were carried out on three monkeys and suggested that the anastomoses with the extracranial vasculature were minimal. The results of our India ink infusions were identical to those of Dumke and Schmidt (Z), who found that only small twigs in and close to the orbits showed evidence of the injected dye. They concluded that these channels were negligible. Additionally, during the development of our model, experiments were carried out on chronically implanted monkeys in which one Doppler flow transducer had been placed on the middle cerebral artery, with a second on either the internal carotid or the common carotid artery. In some instances, all three arteries were implanted. From the internal carotid or from the common carotid with the external branches chronically ligated, changes in blood flow induced by carbon dioxide inhalation, hypoxia, or changes in arterial blood pressure correlated closely, both in degree and direction, with simultaneous measurements made from the middle cerebral artery transducer (12). If the external carotid branches were left unoccluded, measurements from the common carotid artery did not, as expected, correlate closely with those from the flow transducer on the middle cerebral artery. Meyer et al. (23) attempted to determine the amount of extracerebral contamination present in their arterial internal carotid inflow preparations. They measured artery flow in two monkeys before and after enucleation of the ipsilateral eye and removal of the orbital contents. These investiga .tors concluded th at extracerebral contamination was minimal (less than 10%) and did not significantly affect their results. We have measured ICBF in three monkeys, by the arterial inflow method, as described in the present report, before and after careful removal of the ipsilateral orbital contents. At a mean arterial blood pressure of 100 mmHg, orbital enucleation resulted in only a 6-10% decrease in our resting ICBF values. Our observations agree with those of Meyer et al. (23) and, together with our other anatomical and physiological observations, suggest that the extracerebral component of our arterial inflow ICBF measurements is not significant. In the macaque monkey, blood flow through one internal carotid artery represents 30-35% of the entire cerebral circulation. Therefore, our CVR values are necessarily higher than comparable ones elsewhere in the literature. They reflect, however, as does ICBF, true changes in cerebral circulation. Under the conditions of our experiment, the normal were preserved. responses of the cerebral vasculature Autoregul .ation was present in all nine monkeys. To minimize the effects of changes in arterial Pco2 on ICBF, this variable was maintained constant in all experiments. CSF sumpZing. If the appropriate steady-state conditions exist (5), cisternal CSF should reflect changes in brain extracellular fluid acid-base status. Since arterial blood pH and gases were maintained constant throughout the experiment and CBF was not compromised, steady-state conditions should have existed between all cerebral fluids (i.e., plasma, brain extracellular fluid,
933 and CSF). In addition, in these studies, the animals were maintained hypotensive for 30 min before CSF was sampled. Other investigators (6, 7, 14, 15, 17, 18) have previously demonstrated decreased CSF pH values following shorter intervals (3-20 min) of lowered PP. For further evidence that a steady state had been achieved at H 30min 9 two monkeys were maintained at hypotensive pressure for 4 h; no CSF pH changes were detected with repeated sampling throughout this time period. Experimental considerations. One of the mechanisms proposed for the maintenance of autoregulation involves the effects of carbon dioxide evolved in normal aerobic metabolism on the opening and closing of small intraparenchymal vessels. According to this theory (27), brief periods of lowered PP delay the washout of metabolically produced carbon dioxide in any given segment of the brain’s microcirculation. The slight, localized accumulation of carbon dioxide and its effects on brain extracellular fluid pH bring about the opening of the vessels in question. Since carbon dioxide is highly diffusible into the bloodstream, it rapidly disappears from brain extracellular fluid. Thus there would probably not be any change in CSF pH or Pco2 at H:
