Journal of Clinical InvestigationVol. 43, No. 11, 1964

Role of the Autonomic Nervous System in the CirculatoryResponse to Acutely Induced Anemia in Unanesthetized

Dogs *GERALDGLICK,t WILLIAM H. PLAUTH, JR., AND EUGENEBRAUNWALDWITH

THE TECHNICAL ASSISTANCE OF HOPECOOKANDROBERTM. LEWIS(From the Cardiology Branch, National Heart Institute, Bethesda, Ald.)

Studies in patients with chronic anemia and inanesthetized animals rendered anemic have dem-onstrated that, within wide limits, cardiac outputrises as the hematocrit falls (1-5). The funda-mental importance of the autonomic nervous sys-tem in circulatory regulation in the intact organ-ism (6) influenced us to determine its role in theprofound hemodynamic changes that occur dur-ing the stress of acutely induced isovolemic anemia.The basic objective of this study was to comparethe circulatory responses to acutely induced iso-volemic anemia in normal dogs with the responsesobserved in animals in which the control exertedby the autonomic nervous system had been im-paired, either by chronic total cardiac denerva-tion or by reserpine. Most previous experi-ments on acutely induced anemia have been car-ried out on anesthetized animals (2-5), in whichmany compensator) reflex mechanisms must, ofnecessity, have been either obtunded or ablated.To obviate this problem, the present investigationwas carried out on trained, unanesthetized dogs.

Methods

The studies were performed on dogs weighing be-tween 13.0 and 22.3 kg. With the dog under local anes-thesia, catheters were introduced into the right atriumvia the jugular vein and into the aorta through thefemoral artery. Isovolemic anemia, to an average he-matocrit of 13.6% ± 0.3%7o (standard error of the meanSEM), was produced by withdrawing blood from theaortic catheter and simultaneously replacing it with anequal volume of 6%o dextran 1 injected into the rightatrium. This exchange required 10 to 15 minutes, andparticular care was taken to ensure that no changesin blood volume occurred. Cardiac output was meas-

* Submitted f~or publication April 13, 1964; acceptedJuly 9, 1964.

t Established Investigator, American Heart Association.1 Cutter Laboratories, Berkeley, Calif.

ured during the control period and 15 minutes after thecompletion of the blood-for-dextran exchange by theindicator-dilution technique, with injection of indocya-nine green into the right atrium and sampling from theaorta. Blood was drawn through a cuvette densitometer 2

at the rate of 37.5 ml per minute, while the dilutioncurves were inscribed on a linear chart recorder. Theblood was returned to the animal immediately afterwithdrawal. All determinations of cardiac output wereperformed in duplicate, the paired values differing byan average of 12.3% ± 9.1% (SD) from one another.Separate dye calibration curves were constructed for thenormal and for the anemic blood. Pressures were re-corded from the right atrium and aorta immediately af-ter each measurement of cardiac output. Arterial bloodsamples were collected during each period for measure-ment of hematocrit and for determination of 02 contentand capacity and of C02 content by the method of VanSlyke and Neill (7). The hematocrit values were notcorrected for trapped plasma. Arterial blood pH andPco2 were determined with a glass electrodes Totalsystemic resistance (TSR) in dynes-sec-cm-5 was cal-culated as follows:

[systemic arterial mean pressure(millimeters Hg) - right atrial mean

TSR = pressure (millimeters Hg)] X 1,332cardiac output (milliliters per second)

Left ventricular minute work (LVMW) in gram-me-ters per minute per kilogram was derived as follows:LVMW=[systemic arterial mean pressure (centimetersH20) - 6.8] X cardiac index (liters per minute perkilogram) X 10, where 6.8 cm H20 (5 mmHg) is as-sumed to be the left ventricular end-diastolic pressure.Left ventricular stroke work in gram-meters per beatper kilogram was obtained by dividing left ventricu-lar minute work by the heart rate. Oxygen delivery inmilliliters per minute per kilogram to the tissues wascalculated as follows: 02 delivery = arterial 02 con-tent (volumes per 100 ml) X cardiac index (millilitersper minute per kilogram)..

Seven animals served as controls and will be re-ferred to as "intact" dogs. Eight dogs had been sub-jected to cardiac denervation by the regional neural

2 Gilford Instrument Laboratories, Oberlin, Ohio.3 Instrumentation Laboratory, Boston, Mass.

2112

ROLE OF THE AUTONOMICNERVOUSSYSTEMIN RESPONSETO ACUTE ANEMIA

ablation technique of Cooper, Gilbert, Bloodwell, andCrout (8) 4 to 6 months before study. This method ofdenervation isolates the heart from both the sympatheticand parasympathetic nervous systems. Three of theeight dogs that had been subj ected to cardiac denerva-tion were studied on another occasion after blockadeof beta-adrenergic receptors by the intravenous ad-ministration of 5 mg per kg nethalide4 (2-isopropyl-amino-i-[2-naphthyl] ethanol hydrochloride). The stud-ies with nethalide were performed on the animals 5 to 7months after the initial. experiment. As indicated inTable III, the dogs were no longer anemic. Both thecontrol measurements and those after anemia had beeninduced were carried out immediately after the ad-ministration of nethalide.

At the time of each experiment the completeness ofthe denervation was verified by studying the responseto the intravenous injections of tyramine (40 I.Lg per kg)and norepinephrine (0.5 ,ug per kg). Tyramine, whichacts by releasing norepinepherine from myocardialstores (9, 10), did not produce a rise either in bloodpressure or in heart rate in the dogs with denervatedhearts, indicating that the cardiac catecholamine storeshad been depleted as a consequence of denervation (8,11). Norepinephrine produced marked hypertension andtachycardia instead of the usual bradycardia. It was

therefore concluded that reflex slowing resulting fromvagal stimulation or withdrawal of sympathetic ac-

tivity had been eliminated; rather, the direct chrono-tropic effect of norepinephrine was observed (11).Finally, after the conclusion of this series of experi-ments, the right atrium was biopsied in seven of theeight dogs whose hearts had been denervated, and thenorepinephrine concentrations were determined by thetrihydroxyindole photofluorometric method, as detailedelsewhere (12). No detectable levels of norepinephrinewere found in any specimen, confirming that completeadrenergic denervation had indeed been achieved.

In four dogs, generalized catecholamine depletion was

produced by the intravenous administration of reserpine,0.15 mg per kg, for 2 days before the study. Analysisof atrial tissue for catecholamine content in these dogsconfirmed that complete catecholamine depletion hadbeen produced. In two normal dogs and in five dogswith denervated hearts, plasma catecholamine levelswere measured by the trihydroxyindole fluorometricmethod during the control period and 1 and 15 minutesafter the end of the exchange transfusion. The re-

ported statistical analyses of absolute changes were per-

formed by Student's t test (13). Similar results alsowere obtained when comparisons of percentage changeswere made.

Results

During the control period the arterial hemato-crit in the 22 studies on 19 dogs averaged 42.2%o

4 Supplied by Dr. Alex Sahagian-Edwards, AyerstLaboratories, New York, N. Y.

± 1.0 (SEM). Fifteen minutes after the blood-for-dextran exchange the hematocrit averaged13.6% + 0.3. No significant differences amongthe hematocrit values after the production ofanemia in the different groups of dogs were ob-served. During the control period the intact dogsand the dogs with denervated hearts did not differsignificantly from each other in respect to cardiacindex, °2 delivery, right atrial mean pressure,systemic arterial mean pressure, systemic vascularresistance, or left ventricular minute work. How-ever, in the dogs with denervated hearts, theheart rates were significantly higher (p < 0.05),while the stroke index and the left ventricularstroke work index were each significantly lower(p < 0.05) than in the intact dogs.

Although after the production of anemia thedogs with denervated hearts showed a substantialincrease in cardiac index that averaged 77%,from a mean value of 173 to 309 ml per minuteper kg, this rise was significantly less (p < 0.02)than that noted in the intact dogs, in which it in-creased by an average of 119%, from a meanvalue of 175 to 379 ml per minute per kg (TablesI and II, Figure 1). The manner in which theincrease in cardiac output was mediated differedbetween the two groups of animals. Whereas inthe intact animals the heart rate rose by an av-erage of 83% from 108 to 196 beats per minute, itincreased significantly less (p < 0.01) in the dogswith denervated hearts, rising by an average ofonly 23 %, from a mean of 132 to 161 beats perminute (Figure 1). In contrast, the changes instroke index were more prominent in the dogswith denervated hearts. Thus, while in the in-tact dogs the stroke index increased by an averageof only 21%o, from a mean value of 1.62 to 1.93ml per beat per kg, it rose significantly more(p < 0.04) in the dogs with denervated hearts,increasing by an average of 44%o, from a meanvalue of 1.32 to 1.90 ml per beat per kg.

A comparison of the relative importances ofthe augmentation of heart rate and stroke volumein the elevation of the cardiac output induced byanemia is represented graphically in Figure 2.In the intact dogs the increase in cardiac outputresulted primarily from an augmentation of heartrate and to a lesser extent from an elevation ofstroke volume. In the dogs with denervated

2113

G. GLICK, W. H. PLAUTH, JR., AND E. BRAUNWALD

X Ir u o0 01C tr, OIl0) In-.0 I - It m el)

O0 - f,_IM

Ir, N 0% 0-00 t-0 0,Fm eq CS' C0)CN cS

02O CN t- M C O IN2100t- m0) Cs rlcCt - e c :-S-_* - CS

O O c uo ) r ul 0O S r,0- 0 0000 0 - a, 00

C

e 0e 000a, 0c0n t- 00

_ en

IZ4I-0

C4 t t0 000 _( N to

2 C. )10e .0 - -

F>

0000£ O U0) - O + OYn " t d t 't t

*- *- I t- t- r_ t- t- t.- r- X. x x; Z1

\ 0e 000O O.q1 00 C40 e en) xF F

e M m

000 00on 00

0el 0O000 00 ol 00 00 00

000 00 00 0-0

0 0o to Ln t - 00

0o\ )C0 F. a 00

t- 0C-O 000en inM _ -t _ 't _ -- _ en _ 't _

°O m NO in+0_~

e

-4 el 1' In 0o Cd () C.)

¢..

2114

=

E

COI¢43

I-N

-H0,N IC,

u

-

>

0000

C ~ e

-0)

.Q _ CN

_z

0 t-- O6

-H 0H2002C

0.1

Ir. _

-H -H00

00%'

*Cl)

t0

.)

.It

.t4

0%t_ t-

q0 0

4 _H(1

t- t-

t4

mC$

- rv

2)

-COr.cdC

40C

.

0

oi

0 -

-0o-H-

-

,.

5

-

-

11

2

_--

r-

::

-

-

b

w

rW4zr,

11

2,,)

ntJ

vr.r.

t̂j

:4,>

._l--

x

fru

11=rX

E<.

Q_;

r.;

.Y

*-L.

11U

i;

<

X._

u

oYEQ

;

R

=._@At,._Ea"mro

._

@AoC)U¢F

r@

C14 _

-1 cCS; (1

-H00I'l O.

_" n1I4

t-: el

ROLE OF THE AUTONOMICNERVOUSSYSTEMIN RESPONSETO ACUTE ANEMIA

C4 -0 to ") eq (1 ") m -- I- m - o 1"

_~~~~~~~~4

4 o, m -X__ m e dq C14 0 -H --o

"_o C

(

-H-H

06~~ ~ _

ree4 oo Su-4 o e.4 i c +0'0 0'O m C-C' '00 '~0 0 C-4_ -_ __ __ 0

+1

'+

m to

00 _4 U) so t- 't in 00 ) *- to m -M C-0_e_M-4 MEc,4 44-4 C4I N -- .4 a

t

o. o

bbt-t- b I- b b \0

t-

"') ON -. t- .dq ..4 r- ON

r r btZ .Z vz t- t

'o 00

u 00o. C0(*0*u) uto~ >o 0-otr, 'D zO

00uo C-U( cU 00-d en 00 - ' q eiCcl cli C' Lf) t: el; cl C; z hON00(00) 00ON0O -C 00 C- co-0( 0( 0 -U

00)00

e 10 -0 X0 to " In !~ --t m if) \e t 0 O _Ne H +M _4 dq -4 -4 .4 -4 -- - dq -4 10 -4 Tr

,d -

0

C Crz o u; t-N _ I

CO 0) 0 -4 C- 0-m L0 Co(U(I

cooo~~~~- -4 -4 1

"

¢-H ¢

.0

CU

H

r.

cCdv5

w-,!.

1!00w

E g04

1.-404

2115

1; W. tn:74 .4 C14.1.1J.

b4-k

rl., 'i oq1. In 14,

, C4 -,

(S

P.

Y m 4 _4)

6 0_En- 1-1

$t

-bAXe 0 r-r. -

be

be

be

le

.0

cl)

to

e 'C el000);

0LCU

5

d>.4

0)

be.

Pu

0

00

_ Z;ecd Cd

'j:o

4iC.)

A. o =

._

be0

-be 0

G. GLICK, W. H. PLAUTH, JR., AND E. BRAUNWALD

CONTROL ANEA

Int-wao

CONTROL ANEMIA

CONTROL ANEMIA

OXYGEN DELIVERY

50

40

0

z2 30

0

20

10

CONTROL

FIG. 1. EFFECT OF ACUTELY INDUCED ISOVOLEMIC ANEMIA ON CARDIAC INDEX, HEART RATE, STROKE INDEX, AND

OXYGENDELIVERY TO THE TISSUES. In each panel the values during the control period are plotted on the left, andthe values after the production of anemia are on the right. The values in each individual experiment are connectedby a solid line for the intact dogs and by a broken line for the cardiac denervated dogs. The horizontal lines on

each side of the individual panels represent the mean values for each group of dogs.

2116

S00CARDIAC INDEX

450 F

400X

350 _

HEARTRATE

"I,I

3001-

250 _

200 _

150 _

100

436

44wdo

ANEMIA

lr T

ROLE OF THE AUTONOMICNERVOUSSYSTEMIN RESPONSETO ACUTE ANEMIA

hearts there was a greater tendency for the in-crease in cardiac output to result primarily froman elevation of stroke volume than in the intactdogs.

Oxygen delivery to the tissues fell in all dogsin both of these groups (Figure 1). Thus, in noanimal was the increase in cardiac output duringanemia sufficient to maintain oxygen delivery tothe tissues at control levels. In the intact dogsthe oxygen delivery declined by an average of22%, from a mean value of 31.2 to 23.8 ml °2per minute per kg, whereas in the dogs with de-nervated hearts it decreased significantly more(p <0.01), falling by an average value of 42%,from a mean value of 33.2 to 18.9 ml °2 per min-ute per kg.

Anemia did not result in any consistent or sig-nificant changes in arterial mean pressure in eitherthe intact dogs or in the dogs with denervatedhearts (Tables I and II). Calculated systemicvascular resistance fell in each animal, decliningin the intact animals by an average of 52%, froma mean value of 3,991 + 246 to a mean value of1,902 ± 107 dynes-sec-cm-5. The fall in resistancewas not significantly different in the dogs withdenervated hearts, in which it fell by an average of46%, from a mean value of 3,344 + 192 to 1,831+ 194 dynes-sec-cm-5 (Figure 3).

Anemia resulted in an increase of left ventricu-lar stroke work that averaged 28% in the intactdogs, stroke work rising from a mean value of2.89 +.23 to 3.63 ± 0.20 g-m per kg. In thedogs with denervated hearts stroke work rosefrom a mean value of 2.28 +.13 to one of 3.27 ±0.33 g-m per kg, an average rise of 41 %, whichwas not significantly different from that notedin the intact dogs (Figure 3). Left ventricularminute work in the intact animals increased by anaverage of 131%, from a mean value of 312 + 31to 711 + 56 g-m per minute per kg, a significantlygreater (p < 0.01) elevation than in the cardiacdenervated dogs, in which it rose an average of74%, from 304 + 26 to 536 + 67 g-m per minuteper kg (Figure 3).

The right atrial mean pressure rose from anaverage of 2.6 to 5.9 cm H2O in the intact dogs.Although the elevation in the dogs with cardiacdenervation tended to be greater, the changes inright atrial mean pressure resulting from anemia

were not significantly different in the two groupsof animals. The changes in the pH and Pco2 ofthe arterial blood were small and comparable inthe two groups of animals (Tables I and II).The changes in plasma epinephrine and norepi-nephrine concentrations induced by anemia in thetwo intact dogs and in the five dogs with dener-vated hearts in which they were determined weresmall and inconsistent.

The response to anemia in three dogs withdenervated hearts was not prevented by blockadewith nethalide (Table III). Thus, cardiac index,heart rate, and stroke index increased by aver-ages of 95%o, 20%, and 60% in the presence ofbeta-blockade and cardiac denervation, whereasthese variables rose by averages of 77%, 23%,and 44%o in dogs with cardiac denervation butwithout beta-blockade. The absolute values andthe changes induced by anemia in these variablesalso were similar in these two groups of animals.Thus, although the limited number of experi-ments precludes a statistical comparison, beta-receptor blockade did not appear to modify theresponse to anemia of the dogs with denervatedhearts.

The absolute values for cardiac index andheart rate were uniformly lower in the dogs thathad been given reserpine than in the intact dogsor in the dogs with denervated hearts, but noapparent differences in stroke index were noted(Table III). The absolute increases in cardiacindex and heart rate resulting from anemia wereless in the animals given reserpine than in theintact group, whereas the changes in stroke indexin these two groups of dogs were of comparablemagnitudes. When the responses to anemia ofthe dogs given reserpine were compared to thoseof the animals with cardiac denervation, however,no apparent differences in cardiac index, heartrate, or stroke index were observed (Tables IIand III).

Discussion

The results of this investigation demonstratethat the cardiac output of the normal unanesthe-tized dog rises strikingly when the animal is sub-jected to acutely induced isovolemic anemia andthus are in agreement with the work of other in-vestigators who have studied this state in anes-

2117

G. GLICK, W. H. PLAUTH, JR., AND E. BRAUNWALD

I I I I

0

00

00

0

0

0o 5

_ O

_ O

O'~~~~~~~~~~~~~~- O o

3nf Lndino OV1G8VONI 3S

L./MIN.o 0 w - WU) it w

0 0 0 0 0

o to It r4) CN

(4OG8/1VW) 3vwn-10A 3> L081S

* O < ,,,.0I U F

0 00 C

o1-~~~~h

o : '-U ok Z , = S oUZ-

0-

I- ° -0QQ

00~~~~

0%

D V

0 GU C

Z ~ ~~~~~~~U°EtQ

U~ U

0 CC_n 0ito V

O ~~~I. O X

_ OVd 0 c4 c*-)U)1-

0 C)

0 r. 40

0 4 ',

~~o0~~~

o 0 0

w CZ- CZ

o ~~ 1-.b U

-o 00

o~~~~~4- ~~~~E- U

4-0)~~~~~5o a~Z

0~~~~~~> 0

CZ c bo

00 0

2118

3VYM0A3)I081S 01

ROLE OF THE AUTONOMICNERVOUSSYSTEMIN RESPONSE

*9)4/ NI V/VW 9)

, ,\r \

9)4/IA *WI*W9

I . .

o 0 0

~0 0 8§ 9 ~~0 I 0ol Oe on i

S- " 3'3S 3Nla

, ,z

I, . <

4 CZ

o o

4~~~i

>4

0 0 0

n I I _

4> <

cm

14~~~~~~~~~~T

14 ~~~-~

I' ~ z

I ~ ~U 114~ (120

44~*0O Q4

'n4 O I

TO ACUTE ANEMIA 2119

IiX ''\\Nx~' '4

,~~~~~~II

s'

N~~~~~~~'-

x4k2I%4

Y-

z

4

r

zLJ

Li.4)

0

w

0I~-U)

4-J

y

cr

I--zw

I.-ILLIA4i

z

ox

U)

I-

I.-

4-

I I/

I I

I a //'

I I ~ ~ I

I , ,

K I I

I

G. GLICK, W. H. PLAUTH, JR., ANDE. BRAUNWALD

Un00 0 0 0n O 0LOU) 0ON _ N -Se C S_ C-4

-_ __ -+000 0'( 000 z00 1000O 0000, Q 0 0 uO oo t rN o CN - 0

_ _ _

I'l ei oomV~r t me "l It*_ 0_ 0_

04

00(N 0 0- O. N Z b o00CS _0 0- 00 _ _0I

0400000e N ( N00U tU) 0 0 t0 00 0 00. .000 ..

-(N --(N_ -E --_ __ -

.o~~~~~orO

o. m0 00'4 - O 0'- (N .cNO 00 O o' 0-0 - -

0. -0 _N - - -c -

mmm+

00~0C\ a, 00a,000 00 00

0 '4'0 '40 000 00'n

0(N- 0 .-0 00r, uo 0000

e)4

00 (No" 000 00' 00X XX

'4 '4 00-> ) e- '4

00 00 00 00

_

be

00 00 00

_ - cd (2d0.

4)

(N' - 00- I',00 000en

N

(NI- 't- ('N- 00-

0000 en _t 0 0000 00

,:: .~ .q .~ .'CS _ - CS4 _ - _

0 000 s000 Cc00 0-0 000 - ---..+.

CO

0

0

0Q0

00

ON C

'- -d

0_ N

000

o0000

0* o

4 m CN Itr (1 -I Cq

ICO 00 '-0 01O t0 I(0 N NX o4 CN - C -C( +

00

0o000

0000

C1 _

¢0 m

.~~~~-

.~~~C -:t(N 00

~00 00(N~

r r -

r

CN OU

0 N(N (N d CIS Q¢. 0.

e

.0!4 <

2120

0.

04

(-)

(U

0.< o

0-Co

T)2

C0.

204

co

.4)

cd

0)"20z

be

0.00

0

t00

._.2.0

cd.0

o00

Cd

.0*d

ROLE OF THE AUTONOMICNERVOUSSYSTEM IN RESPONSETO ACUTE ANEMIA

thetized animals (2-5). Although the mecha-nism responsible for this increase in cardiac out-put has not been elucidated, a number of possi-bilities have been suggested. These include adecrease in peripheral resistance resulting froma fall in blood viscosity and arteriolar dilatation,an elevation of right heart filling pressures, stimu-lation of chemoreceptors sensitive to a decreasedpartial pressure of oxygen, the action of a non-catecholamine humoral mediator, and the activityof the adrenergic nervous system (14-17). Thepresent study was designed to examine the im-portance of the autonomic nervous system in thecirculatory response to anemia.

It is well recognized that the use of generalanesthesia and the depth and specific agent em-ployed can profoundly alter the response of theautonomic nervous system. To obviate theseproblems it was necessary to perform these stud-ies on trained. unanesthetized animals. To de-termine the role of the autonomic nervous sys-tem, the effects of anemia on dogs subjected tototal chronic cardiac denervation were determined.Thoracic sympathectomy has been the traditionalmethod of producing cardiac denervation (18,19), but complete denervation and total myo-cardial catecholamine depletion are rarely pro-duced by this operation (8, 20). Moreover, por-tions of the animal other than the heart are neces-sarily denervated by such a procedure, and theparasympathetic fibers of the vagus nerve are notincluded in the operation. Although cardiacdenervation by pharmacologic agents that depletethe entire animal of norepinephrine have alsobeen utilized, the entire circulatory system is de-pleted of catecholamines in this manner, makingit difficult to assess the effects of the sympatheticcardiac nerves per se. The regional neural abla-tion technique employed in this study makes itpossible to overcome these problems.

Dogs deprived of both the sympathetic andparasympathetic cardiac nerves were still capableof substantially increasing their cardiac outputand left ventricular work when rendered acutelyanemic. However, since the increases in cardiacindex and minute work were significantly less inthe dogs with denervated hearts than in the in-tact animals, the autonomic nervous system ap-parently plays a significant role in the circula-

tory response to anemia. These observations arenot consonant with the hypothesis that the in-crease in cardiac output produced by anemia maybe attributable entirely to the fall in blood vis-cosity (14). It should be emphasized, however,that four of the eight dogs subjected to cardiacdenervation behaved in an apparently normalmanner, as regards the changes in cardiac index,although as a group their increases in output weresignificantly less than the control (p < 0.02).Quite conceivably the changes noted in thosedogs with denervated hearts manifesting in-creases in cardiac output in the normal rangewere nonetheless smaller than they would havebeen had the autonomic innervation to the heartbeen intact. Additional evidence indicating thesignificance of the autonomic nervous system isprovided by an analysis of the manner in whichthe augmentation of cardiac output was achieved.In all of the intact dogs the rise of heart rate con-tributed more than the elevation of stroke volumeto the augmentation of cardiac output. In con-trast, in the dogs with denervated hearts, the aug-mentation of heart rate produced by isovolemicanemia was significantly less and the contributionof an elevated stroke volume was usually ofgreater importance than in the intact dogs.

The small increases in heart rate that occurredin the dogs with denervated hearts cannot, ofcourse, be explained by any neural mechanism.One possibility for the augmentation of heart rateand cardiac index in the dogs with denervatedhearts is a response to circulating catecholaminesreleased from the adrenal medulla. Although theplasma concentrations of epinephrine and nor-epinephrine in the five denervated dogs in whichthey were measured did not increase, it couldstill be argued that since the denervated heartsare supersensitive to circulating catecholamines(21), elevations in plasma levels, too small to bedetected, may actually have played a role in thecirculatory response to anemia. This possibilitywas investigated in three dogs with denervatedhearts which were also treated with a beta-adrenergic blocking agent before the inductionof anemia. Inasmuch as receptor blockade wouldprevent the cardiac action of circulating catechola-mines, and since the blockade did not prevent therise of heart rate and cardiac output consequent

2121

G. GLICK, W. H. PLAUTH, JR., AND E. BRAUN'WALD

to anemia, it appears likely that the elevations ofthese variables observed in the dogs with dener-vated hearts did not result from circulating cate-cholamines. In this connection it is also pertinentthat the increase in cardiac output produced byanemia is not reduced by prior treatment withdichloroisoproterenol (4), and that after adrenal-ectomy and ganglionic blockade the cardiac outputstill rises markedly with the production of anemia(17). Other possible explanations for the eleva-tion of heart rate in the dogs with enervatedhearts include: 1) elevation of right atrial pres-sure, which by increasing the tension of the atrialwall may, as postulated by Blinks (22), increasethe rhythmicity of the sinoatrial node, 2) localmetabolic changes resulting from the changes intissue Po2 produced by anemia, and 3) non-catecholamine humoral substances released dur-ing anemia.

A possible explanation for the augmentation ofcardiac index, which may be applicable to boththe intact dogs and to those with denervatedhearts, is reflex venoconstriction occurring duringanemia. Accordingly, a group of intact dogs waspretreated with reserpine, which has been showncapable of blocking reflex venoconstriction (23)as well as depleting myocardial norepinephrinestores. Although reserpine lowered the controlvalues for cardiac index, heart rate, and minutework, the changes in these variables produced byanemia were similar to those observed in the dogsin which only the hearts had been denervated.Thus, it does not appear that reflex venoconstric-tion, nor indeed other extracardiac reflexes me-diated through the adrenergic nervous system,played a prominent role in the circulatory re-sponse to anemia.

The possibility exists that the elevation in car-diac output was the result of stimulation ofchemoreceptors consequent to the production ofanemia. Since these receptors are sensitive to

changes in the partial pressure of oxygen, andsince the arterial blood in the anemic period isfully saturated or nearly so, there is no reason

to believe that the arterial chemoreceptors are

stimulated at this time. Therefore, if chemore-ceptors are of importance, they are probably lo-cated in the capillary or venous beds. This ideais consistent with the work of Gowdey, who ob-

served large increases in cardiac output after theproduction of methemoglobinemia (24), and ofBartlett and Tenney, who demonstrated that ane-mia profoundly reduces the tissue oxygen tensionin rats (25'). Such a reduction in tissue oxygentension mar. however, lower the total systemicresistance through local effects. Thus, arteriolardilation mIay result either directly from the hy-poxic, milieu or from axon reflexes, causing asecondary increase in cardiac output to ensue.These changes might be considered analogous tothe opening of a systemic arteriovenous fistula.

The rise in right atrial filling pressure observedafter the production of anemia in 13 of 15 ani-mals (Tables I and II) may also play a role inthe augmentation of cardiac output. although thecause of this rise in atrial pressure remains un-certain. It seems reasonable to assume, how-ever, that this rise in filling pressure reflects theheart's utilizing the Frank- Starling mechanismand moving up on its ventricular function curveto meet the demands for increased stroke work.

The observations described herein are also ofinterest in considering the manner in which theheart responds to demands for an increased cardiacoutput. In the presence of an intact autonomicnervous system the heart of the unanesthetizeddog increases output primarily by elevating rate,with relatively small elevations in stroke volume.This response to anemia is similar to that ob-served when the cardiac output is elevated by mus-cular exercise (26), and it is likely that the in-crease in heart rate results largely from altera-tions in the activity of the autonomic nervoussystem during stress. As observed in the presentinvestigation, when the organism is deprived ofits autonomic nervous system and only minor in-creases in heart rate can occur, the heart tendsto fall back on what might be considered to bea reserve mechanism, namely, an increase instroke volume.

Summary

The role of the autonomic nervous system inthe circulatory adaptation to acutely induced iso-volemic anemia was evaluated by comparing theresponses of seven intact unanesthetized dogs withthe responses of eight dogs which had been sub-jected to chronic total cardiac denervation. After

2 122

ROLE OF THE AUTONOMICNERVOUSSYSTEMIN RESPONSETO ACUTE ANEMIA

the induction of severe isovolemic anemia, pro-duced by the exchange of dextran for blood, car-diac output rose markedly in both groups of ani-mals. The rise was significantly greater, how-ever, in the intact dogs than in the cardiac de-nervated animals. In the intact dogs the in-crease in cardiac output stemmed predominantlyfrom a rise in heart rate, elevations in stroke vol-ume playing a less important role. In contrast,in the cardiac denervated dogs, the increase incardiac output tended to be more the result of anaugmentation in stroke volume. In an additionalfour dogs, catecholamine depletion was producedby prior administration of reserpine. These dogsresponded in a fashion similar to the cardiac de-nervated animals in respect to the changes in car-diac index, heart rate, and stroke volume that oc-curred during anemia, but they displayed smallerincreases in cardiac output and in heart rate thanthe intact dogs. From these results it is con-cluded that an intact autonomic nervous systemis necessary for the total circulatory response toanemia. In the absence of a functioning auto-nomic nervous system, the organism appears tofall back on what may be looked upon as a reservemechanism for the augmentation of cardiac out-put-the elevation of stroke volume.

Acknowledgment

We would like to thank Dr. Charles A. Chidsey forperforming the catecholamine determinations.

References1. Brannon, E. S., A. J. Merrill, J. V. Warren, and

E. A. Stead, Jr. The cardiac output in patientswith chronic anemia as measured by the techniqueof right atrial catheterization. J. clin. Invest. 1945,24, 332.

2. Sunahara, F. A., and L. Beck. Cardiovascular effectsof acutely produced anemia in the normal dog.Amer. J. Physiol. 1954, 176, 139.

3. Richardson, T. Q., and A. C. Guyton. Effects ofpolycythemia and anemia on cardiac output andother circulatory factors. Amer. J. Physiol. 1959,197, 1167.

4. Gowdey, C. W. Anemia-induced changes in cardiacoutput in dogs treated with dichloroisoproterenol.Circulat. Res. 1962, 10, 354.

5. Murray, J. F., P. Gold, and B. L. Johnson, Jr. Sys-temic oxygen transport in induced normovolemicanemia and polycythemia. Amer. J. Physiol. 1962,203, 720.

6. Heymans, C., and E. Neil. Reflexogenic Areas ofthe Cardiovascular System. Boston, Little, Brown,1958.

7. Van Slyke, D. D., and J. M. Neill. The determina-tion of gases in blood and other solutions by vac-uum extraction and manometric measurement. J.biol. Chem. 1924, 61, 523.

8. Cooper, T., J. W. Gilbert, Jr., R. D. Bloodwell, andJ. R. Crout. Chronic extrinsic cardiac denervationby regional neural ablation. Description of theoperation, verification of the denervation, and itseffects on myocardial catecholamines. Circulat.Res. 1961, 9, 275.

9. Burn, J. H., and M. J. Rand. The action of sym-pathomimetic amines in animals treated with re-serpine. J. Physiol. (Lond.) 1958, 144, 314.

10. Chidsey, C. A., D. C. Harrison, and E. Braunwald.Release of norepinephrine from the heart by vaso-active amines. Proc. Soc. exp. Biol. (N. Y.)1962, 109, 488.

11. Gaffney, T. E., E. Braunwald, and T. Cooper.Analysis of the acute circulatory effects of gua-nethidine and bretylium. Circulat. Res. 1962, 10,83.

12. Harrison, D. C., C. A. Chidsey, R. Goldman, and E.Braunwald. Relationships between the releaseand tissue depletion of norepinephrine from theheart by guanethidine and reserpine. Circulat.Res. 1963, 12, 256.

13. Snedecor, G. W. Statistical Methods Applied to Ex-periments in Agriculture and Biology, 4th ed.Ames, Iowa, Iowa State College Press, 1946.

14. Guyton, A. C., and T. Q. Richardson. Effect of he-matocrit on venous return. Circulat. Res. 1961, 9,157.

15. Sharpey-Schafer, E. P. Cardiac output in severeanaemia. Clin. Sci. 1944, 5, 125.

16. Justus, D. W., R. W. Cornett, and J. D. Hatcher.A humoral influence on cardiovascular adjust-ments to acute and chronic posthemorrhagic anemiain dogs. Circulat. Res. 1957, 5, 207.

17. Lovegrove, T. D., C. W. Gowdey, and J. A. F. Steven-son. Sympathoadrenal system and response ofheart to acute exchange anemia. Circulat. Res.1957, 5, 659.

18. Harrison, T. R., A. Blalock, C. Pilcher, and C. P.Wilson. The regulation of circulation. VIII. Therelative importance of nervous, endocrine and vas-cular regulation in the response of the cardiac out-put to anoxemia. Amer. J. Physiol. 1927-28, 83,284.

19. Nahas, G. G., G. W. Mather, J. D. M. Wargo, andW. L. Adams. Influence of acute hypoxia onsympathectomized and adrenalectomized dogs.Amer. J. Physiol. 1954, 177, 13.

2123

2124 G. GLICK, W. H. PLAUTH, JR., AND E. BRAUNWALD

20. Goodall, McC., and N. Kirshner. Effect of cervico- 23. Gaffney, T. E., W. M. Bryant, and E. Braunwald.thoracic ganglionectomy on the adrenaline and Effects of reserpine and guanethidine on venousnoradrenaline content in the mammalian heart. J. reflexes. Circulat. Res. 1962, 11, 889.clin. Invest. 1956, 35, 649. 24. Gowdey, C. W. Cardiac output in acute experimental

21. Cooper, T., V. L. Willman, L. T. Potter, and C. R. methemoglobinemia. Canad. J. Biochem. 1960, 38,Hanlon. Responses to sympathomimetic amines 1411.after chronic cardiac denervation. Physiologist 25. Bartlett, D., Jr., and S. M. Tenney. Tissue gas ten-1963, 6, 160. sions in experimental anemia. J. appl. Physiol.

22. Blinks, J. R. Positive chronotropic effect of in- 1963, 18, 734.creasing right atrial pressure in the isolated mam- 26. Marshall, R. J., and J. T. Shepherd. Exercise andmalian heart. Amer. J. Physiol. 1956, 186, 299. the circulation. Circulation 1963, 27, 323.