Effects of Hypotension on Rhesus Monkeys Francis W. Gamache, Jr., MD, Ronald E.
Myers, MD,
Twenty-one late-juvenile rhesus monkeys were rendered profoundly hypotensive for 0-, 15-, or 30-minute periods by means of infusion of trimethaphan camsylate. Blood pressure was then restored to prehypotensive levels with phenylephrine infusions. Respiratory gas tensions and pH of arterial blood were maintained within their normal limits throughout experimental and recovery periods.
was
Animals either recovered and showed no sequelae or died 12 to 48 hours later of cardiorespiratory difficulties, often accompanied by brain swelling. Brain injury and death occurred in 64% of cases when arterial blood pressure was maintained at 25 mm Hg for up to 30 minutes. Multifocal
myoclonus, depressed electroencephalographic activity, rises in cisternal cerebrospinal (CSF) pressure, respiratory depression, and characteristic changes in serum and cisternal CSF glucose followed episodes of controlled hypotension. Hypoxia and acidosis occurring during insult or recovery periods rather than hypotension itself probably account for neuropathological sequelae described by others.
(Arch Neurol 32:374-380, 1975)
through which The mechanisms hypotension
ar¬
terial may cause brain injury have been studied in sev¬ eral laboratories, where blood pres¬ sure lowering has been produced by hemorrhage, hemorrhage and ganglionic blocking agents, and hemor¬
rhage, ganglionic blocking agents, and head tilt.15 In these
studies, the
hypotension was generally of gradual onset, the duration of hypotension
PhD
usually variable,
and the blood
restored
through grad¬ ual spontaneous recovery or by reinjection of withdrawn blood and infu¬
pressure
was
sion of vasopressor agents. The level of blood pressure reestablished also varied among the animals of these studies. Since the acid-base state and the respiratory gas tensions of the animals usually were not controlled, the possibility that hypoxia or sys¬ temic acidosis may have contributed to the resultant brain injury cannot be ruled out. During surgery on humans, the practice of lowering the arterial blood pressure is currently in wide use for the prevention or control of bleeding. Thus, an urgent need exists for a bet¬ ter knowledge of the body's toler¬ ances to blood pressure lowering of different degrees and for different durations. At present, conflicting re¬ ports still exist as to the injuriousness of different levels of hypotension, particularly when the blood pressure lowering is accomplished rapidly.5·6 Though the acid-base and respira¬ tory gas values of the blood should be maintained within narrow limits dur¬ ing any study of the injurious effects of hypotension, such a regulation has not been emphasized by others. The present study utilized a rhesus mon¬ key model where the blood pressure alone was varied while the blood com¬ position was well maintained to study the specific physiological and neuro¬ pathological effects of hypotension. MATERIALS AND METHODS
Accepted for publication Sept 5, 1974. From the Laboratory of Perinatal Physiology, National Institute of Neurological Diseases and Stroke, Bethesda, Md. Reprint requests to Room 110, Auburn Building, National Institute of Neurological Diseases and Stroke, National Institutes of Health, 4915 Auburn Ave, Bethesda, MD 20014 (Dr. Myers).
Twenty-one late-juvenile monkeys (Ma¬ mulata) weighing 2.8 to 8.8 kg were used in the present study. Each animal was anesthetized with pentobarbital (35 mg/kg, intravenously), intubated, and ventilated with a respirator (Harvard). Catheters were inserted into the right
caca
Downloaded From: http://archneur.jamanetwork.com/ by a University of Iowa User on 05/31/2015
femoral artery and vein and connected to pressure transducers (Hewlett-Packard model 267AC). Arterial blood samples (.2 to .4 ml) were withdrawn at intervals (see be¬ low) for the determination of pH, oxygen pressure, (Po2), and arterial carbon dioxide pressure (Pco,). Subcutaneous needle elec¬ trodes were employed to record lead II of the electrocardiogram and the right parie-
tooccipital electroencephalogram.
After all catheters and leads
were
placed, each animal was placed in the prone position. The head was secured in a stereotaxic headholder, and the spinal col¬ umn and external auditory canals were aligned in the same horizontal plane. A rectal thermistor
was
inserted to record
body temperature, which was maintained at 37.5 ± 5 C using a radiant-heat shield. A No. 18 spinal puncture needle (Yale) was introduced percutaneously into the cis¬ terna magna and fixed in position by at¬ tachment to a stage-housing. Polyethylene tubing connected the puncture needle to a pressure transducer (Hewlett-Packard model 267AC), which was positioned at the level of the external auditory canals. This system permitted the continuous recording of cisterna magna fluid pressure. Samples of cisternal fluid were periodically with¬ drawn for determination of glucose con¬ tent. The blood and cisternal pressures, the ECG and EEG, and the rectal temperature
all continuously recorded on a poly¬ graph (Sanborn model 7700). A biotachometer, driven by the electrocardiograph, provided a continuous recording of the were
heart rate. Each animal was followed for at least 75 minutes prior to initiating hypotension. During this time, control arterial blood and cisterna magna fluid (cisternal cerebrospi¬ nal fluid) samples were collected for analy¬ sis at 30-minute intervals. Among the group 3 animals (see below), all blood and cisterna magna fluid samples were ana¬ lysed for glucose as well as for pH, Po.,., and Pco,. Blood and cisternal cerebrospinal fluid (CSF) samples were also taken at 15minute intervals during the episodes of controlled hypotension and at 30-minute intervals during recovery to record any al-
teration in animal status. The respirator was also periodically (briefly) halted to evaluate each animal's ability to breath in¬ dependently of the respirator. The microelectrode unit (Copenhagen-Radiometer) was employed for all pH and respiratory gas analyses. Glucose determinations were carried out using a glucose analyzer (Beck-
man).
Animal Groups divided into three groups. Four animals constituted the control group (group 1). These animals were prepared as above, and the various physiological fac¬ tors measured over a two- to three-hour period. Thereafter, all catheters and leads were removed, and the animals were re¬ turned to their cages. Six animals consti¬ Animals
were
tuted experimental group 2. Each of these animals was prepared in a fashion similar to the control group, but was exposed to a single, closely regulated episode of hypo¬ tension that lasted 15 minutes. During this episode, the blood pressure was rapidly lowered to and generally maintained at a mean pressure in the range of 25 mm Hg. The 11 animals constituting experimental group 3 were similarly prepared and ex¬ posed to episodes of hypotension of the same magnitude but of 30-minute dura¬ tion.
Production and Maintenance of Hypotension
Trimethaphan camsylate (2 mg/ml of 5% dextrose solution) was rapidly infused in¬ travenously so that mean arterial blood pressure decreased to 25 mm Hg within 120 seconds (range, 30 to 180 seconds). The infusion rate was then reduced to a level sufficient to maintain the blood pressure at the new level throughout the 15- or 30-min¬ ute insult period. A total volume of 20 to 25 ml of trimethaphan solution was required for this blood pressure regulation and maintenance. No use was made of head or body tilt, nor was blood withdrawn in the process of reducing blood pressure.
Posthypotensive Recovery The episodes of hypotension were termi¬ nated by discontinuing the trimethaphan drip and infusing phenylephrine hydro¬ chloride, 2 mg/100 ml. Again, the phenyl¬
ephrine infusion
was
regulated
so as
to
maintain the blood pressure at a new level comparable to that exhibited by the same animal during the control period. Gener¬ ally (except as described below), this drip was discontinued after five to six hours. Throughout the hypotensive episode and the period of recovery, the gas tensions and pH of the arterial blood and the pH of the cisterna magna fluid were maintained
Fig 1 .—Blood pressure and EEG recordings from nonsurvivor H1 before, during, and after trimethaphan-induced hypotension. 1, Initiation of trimethaphan-induced hypotension. 2, After 25 minutes of hypotension. 3, At termination (30 min¬ utes) of hypotension. 4, Six minutes into recovery, with aid of phenylephrine. 5. Thirty minutes into recovery. 6, During agonal stages (about eight hours into re¬ covery). Chart speed: 5 sec 5 mm, except 6. =
prehypotensive levels by regulat¬ ing the respirator rate and administer¬ ing oxygen-enriched atmospheres or alkali or both (see below). Animals were weaned from the respirator only when they were fully capable of independently main¬ taining arterial gas tensions and pH at pre-insult values. Episodes of multifocal myoclonus were treated with pentobarbital (in 30- to 60-mg single intravenous in¬ jections) or diphenylhydantoin sodium (in 50-mg single intravenous doses). When at their
marked rises in cisternal fluid pressure ap¬ peared, furosemide (in 5- to 10-mg single intravenous injections) and urea (in 20- to 75-ml infusions) were given. All animals received procaine penicillin G, 6 to 1.2 mil¬ lion units, intramuscularly. Animals were killed (nonsurvivors, see below) late during their neurological deterioration (12 to 48 hours into recovery) when vasopressors had begun to fail to maintain arterial pres¬ sure at prehypotensive levels or when arterial gas tensions and pH were unsatis¬ factory despite continued mechanical ven¬ tilation or bicarbonate therapy.
Neuropathological Preparations The animals killed during the acute stage of the experiment (nonsurvivors) and those animals that survived the acute phase and were killed after a minimum of 14 days (survivors) were perfused with physiological saline solution followed by 4%
Downloaded From: http://archneur.jamanetwork.com/ by a University of Iowa User on 05/31/2015
formaldehyde solution (10% formalin). The brains were immediately removed and stored in large volumes of 4% formalde¬ hyde solution. After two weeks, the speci¬ mens were
removed from the solution,
briefly washed in tap water, sectioned coronally, and examined grossly for patholog¬ ical changes. Histological sections were prepared from the gross brain slices. The microscopical sections were stained with thionine (NissI technique). RESULTS
Electroencephalograms None of the group 1 (control) ani¬ mals showed changes in their electro¬ encephalograms. However, four group 2 (15 minute) animals manifested de¬
both in EEG amplitude and frequency late during their exposure to hypotension (Table 1). In every in¬ stance, the EEG activity improved within 5 to 15 minutes and was fully restored by 30 to 60 minutes into the recovery period. The group 3 animals showed significant decreases in their EEG frequencies and amplitudes. However, these changes recovered af¬ ter time delays only slightly longer
creases
than those observed among the group 2 animals, ie, improvement appeared 15 to 30 minutes into the recovery pe-
Table
1.—Summary of Physiological and Pathological Results* Cisternal
Spontaneous
Group (4)
Beats/min
ECG
EEG
Controls
Heart Rate,
Temperature
Clinical Status
Pressure, mm Hg
Respirations
change
Primarily low ampi
Regular wave
175-225
36.5-37.5 C
14-18/min
0-4
(4):
(2): little change; (3): mild D ampi,
(2):U (1): D in
(1): 10% D (5) : 20%-25%
All: V
(5) : spont S
(4):U (2) : I to high normal
(4): U (1): pro¬ longed hypo¬
no
characteristics
delta
15 Minutes
hypotension (6)
ORS amp!
(3): D
end of
T-wave
hypotension
ampi
&
effective (1): spont but ineffec¬
D
in
tension D in sensory accuity
tive
(1): flaccid
Cheyne-Stokes,
recovering
D in sensory accuity, multifocal
shortly later
(1): 50% D ampi &
myoclonus
in
frequency
(2): little change, mild irregu¬
30 Minutes
hypotension (11)
(4) :
(4):U (7): Din
25% D
All: V
collapse
ampi
frequency
±
(3): Pentobarbital
from 50% to
effective (1): vascular
T-wave
larities in
(7): lof
(5) : spont &
(7) : 33% D
ST segment
300% (1): I but
&
resp arrest (5): resp
changes
failure
sensitive (6) : 30% D in ampi
(4):U (7) : pro¬
longed hypotension
technical failure
flaccid, D
(1): D (2):U
sensory,
imperfect
multifocal
myoclonus
(2): recovered —
*
I
=
increase; D
terna magna
fluid
=
or
(4) : deteriorated decrease; V = varied in both directions; U
cisternal CSF
glucose;
resp
=
unchanged; ampi
=
respiratory.
amplitude; spont
=
riod and minutes Table 2.—A Comparison of Two Models of Categories Research animal Anesthesia Ventilation
Arterial gas tension
Production of
hypotension Hypotension: Onset Duration Reversal
Hypotension*
Another Major Study' Rhesus monkey Pentobarbital, 50mg/kg, IP Connected to pump when res¬ piratory slowing or failure was noted Assisted during recovery
This Study Rhesus monkey
Pentobarbital, 35mg/kg, IV Connected to pump at all times Maintained before, during, and after hypotension Infusion of trimethaphan IV
trimethaphan, plus head tilt, plus
IV bolus of
hemorrhage Mean arterial pressure of 25
mm
Hg
Abrupt: within minutes 15 or 30 minutes constant IV infusion of phenylephrlne in 5% dextrose in water, return to a
prehypotensive level within
30 to 60 seconds
Net cerebral perfusion pres¬ sure of 25 mm Hg Abrupt: within minutes 1 to 50 minutes, inconstant Reinjection of hemorrhage fluid, plus horizontal posi¬ tioning, and/or 0.2 ml metaraminol IV, and/or 6% dextran in saline, depending on given situation; re¬ turn to normotension minutes to hours
Rising intracranial Treated with IV furosemide and pressure Tissue: Fixation
Untreated
infúsate Physiological saline solution perfusion, followed by formaldehyde solution
Heparin physiological
Thionine
Nissl, hematoxylin-eosin
urea
saline perfusion, fol¬ lowed by formaldehydeglacial acetic acid-
methanol solution
Staining
IV
=
Intravenous; IP
=
resp,
Intraperitoneal.
Downloaded From: http://archneur.jamanetwork.com/ by a University of Iowa User on 05/31/2015
spontaneous; CMF glucose
was
complete
=
eis-
after 60 to 120
(Table 1). Three animals de¬
veloped bursts of increased EEG activity soon after the onset of hypo¬ tension, and their EEGs became isoelectric after a single 15- to 20-mg in¬ jection of pentobarbital, given as treatment for presumed seizure dis¬ charges. The EEG activity recovered in these animals after 15 to 30 min¬ utes. An example of the changes in EEG activity that occur during and after a 30-minute hypotensive epi¬ sode is shown in Fig 1.
ECG and Heart Rate The T-wave
amplitude decreased in
ten of the 17 group 2 and 3 animals. In addition, heart rate decreased by
25% immediately after the beginning of the hypotensive episodes in most of the group 2 and 3 animals. This bradycardia became more pronounced with time; group 3 animals experi¬ enced heart rate decreases of up to 33% by the end of their 30-minute epi¬ sodes of hypotension. The EEG, ECG, and heart rate changes of all animals are summarized in Table 1. Blood Gas Data All animals were maintained
on a
signifi¬
and nonsurvivors reached cance
Survivors/ Damage (4): alive/
Blood Glucose
ml
none
(4): alive/
Changes CMF
Changes 60-75
mg/100
Not recorded
Pathological Findings
Glucose
Gross Normal
50-60
Microscopic Normal
mg/100 ml Not recorded
none
(2) : short-
(4):U (2): brain swelling
(6) : Unre¬
(7):U (4): brain swelling (2): also had
(11): Unre¬
markable
term survi¬ vors
killed
(1): never regained consciousness
(4):U (1): cardio¬ vascular
collapse
(4): initial I, then plateau (7): marked
(4): initial I
persistent
and death
I (see
graph)
then plateau (7) : marked persistent I (see graph)
subarachnoid
hemorrhage
(6) : never regained
markable (1): did have a vascular abnormality in subthalamic area
consciousness
earlier, differing
into the recovery
at
one
hour
period (P < .001).
Respiration All animals became apneic after about 15 minutes of hypotension, ie, they failed to breath when tested by turning off the respirator. Those ani¬ mals stressed for only 15 minutes re¬ gained spontaneous breathing after an average delay of 220 minutes (range, 90 to 300 minutes) into the re¬ covery period. The group 3 animals were more depressed: about one-half of these animals recovered effective spontaneous breathing after an aver¬ age delay of 180 minutes (range, 100 to 390 minutes). The remainder failed to recover any effective breathing during the entire period of their sur¬ vival and could not be weaned from the respirator despite repeated ef¬ forts. Cisterna Magna Fluid Pressure
respirator from the beginning of the control period until well into the re¬ covery period. Whenever the arterial
blood oxygen pressure fell below 85 mm Hg, the gas mixture used to ven¬ tilate the animals was further en¬ riched with oxygen. Furthermore, the arterial blood carbon dioxide tension was maintained between 28 to 38 mm Hg by adjusting either the ventilator rate or its stroke volume or both. The arterial blood pH was maintained be¬ tween 7.35 and 7.50 (control range, 7.38 to 7.50) throughout the period of
study by making respirator adjust¬ ments as described above and by oc¬ casional 3- to 5-ml injections of sodium bicarbonate (.892 mEq/ml). Under these conditions, the cisterna magna fluid pH remained between 7.38 and 7.56 (control range, 7.40 to 7.52). Glucose
Analysis
and cisterna fluid carried out was glucose magna only on group 3 animals. Glucose in¬ creased by 10% to 25% in the serum and 0 to 20% in the cisterna magna fluid in the 11 animals that were sam¬ pled at the end of the first 15 minutes The
analysis of serum
hypotension. After 30 minutes of hypotension, the serum and cisternal CSF glucose values of the animals that were to be killed early (nonsurvi¬ vors) differed significantly (P < .01) from the corresponding values of the of
animals that were to survive the in¬ sult. Following the restoration of blood pressure, the serum and cister¬ nal CSF values of the survivors and nonsurvivors again no longer exhib¬ ited significant differences. At one to two hours into the recovery period, the glucose concentrations of the sur¬ vivors (four animals) plateaued at a level close to 200 mg/100 ml in the se¬ rum and 100 mg/100 ml in the cis¬ terna magna fluid (Fig 2 and 3). In contrast to this, the serum glucose concentrations of the nonsurvivors (seven animals) increased above and beyond the level exhibited by the sur¬ vivors, becoming significantly differ¬ ent from the values of the survivors two hours into the recovery period (P