THE EFFECT OF ALTERATIONS IN BLOODVOLUMEONTHEANEMIA ANDHYPOPROTEINEMIAOF HUMANMALARIA'

By HARRYA. FELDMAN2 AND FRANKLIN D. MURPHY8

(From the Division of Preventive Medicine, University of Tennessee, Memphis)

(Received for publication May 31, 1945)

There is general agreement that humans withactive malarial infections 4 develop amenia andhyproproteinemia. The anemia is usually con-sidered to be secondary to erythrocyte destructionby the intracellular parasites, but no adequate ex-planation has been offered for the hypoprotein-emia. Since laboratory techniques commonlyused measure only the relative amounts of chem-ical and cellular components of the blood, it can-not be assumed that these relative measurementsalways reflect fluctuations in the total amounts ofthese constituents.

In order to determine alterations in the totalamounts of circulating plasma proteins and redblood cells, this study of the blood volume changesassociated with clinically active, untreated malariawas undertaken. Simultaneous determinations ofthe plasma proteins, plasma bilirubin, and ery-throcyte concentrations were made so that therelative and total amounts of these elements mightbe calculated and compared.

METHODSAND MATERIALS

Nine patients with central nervous system syphilis, butotherwise relatively normal, served as experimental sub-jects. The group consisted of 5 white and 3 negro malesand 1 negro female. Four patients were inoculated by thebites of Anopheles quadrimaculatus infected with Plas-modium vivax (McCoy strain), 1 patient by the intra-venous administration of trophozoites of P. vivax (McCoystrain), and 4 patients by the intravenous injection oftrophozoites of P. falciparum (Costa strain). These dataare summarized in Table I.

1 The work described in this paper was done under acontract, recommended by the Committee on Medical Re-search, between the Office of Scientific Research andDevelopment and the University of Tennessee.

2 Major, MC, AUS.8 Captain, MC, AUS.4For the purpose of this report, the period of active

malaria is considered to be the period of visible para-sitemia. Consequently, the days of disease are the daysof parasitemia, while days which precede or follow thedays of parasitemia are identified by minus or plus signsrespectively.

Parasite counts were performed daily on the peripheralblood (1).

Plasma volume was measured (in fasting subjects)with the dye T-1824 (2). However, it was suggested (3)that a single 10-minute sample of blood be used in placeof an extrapolation curve to calculate the degree of dyedilution and this was done, except that we employed asingle 15-minute sample. Patients were given 5 ml. of a0.5 per cent solution of the dye5 for each measurementand the absorption coefficient of the dyed plasma wasthen determined with a Beckman photoelectric spectro-photometer.

TABLE I

Summary of descriptive data concerning 9 malaria patientsstudied by the T-1824 blood volume method

Patient Sex Color Age Species Method of inoculation

1 M W 29 P. vivax Trophozoites2 M C 35 P. falciparum Trophozoites3 M C 43 P. falciparum Trophozoites4 M W 39 P. vivax Sporozoites5 M W 41 P. vivax Sporozoites6 M W 34 P. vivax Sporozoites7 F C 36 P. falciparum Trophozoites8 M W 48 P. vivax Sporozoites9 M C 43 P. fakiparum Trophozoites

All blood samples were heparinized and, in addition to

dye concentration, were analyzed for plasma protein con-centration (4), for plasma bilirubin (5), and for redblood cell concentration by means of the hematocrit index.Hematocrits were done in duplicate and centrifuged for30 minutes at 3,000 r.p.m. in Wintrobe tubes.

White patients who had plasma volumes determined onconsecutive or nearly consecutive days became visibly dis-colored, and this served as a limiting factor for the fre-quency with which determinations could be made in thesame individual. This "blueing" of the skin graduallydisappeared over a period of from several days to a fewweeks.

Each patient had one or more "base-line" plasma vol-ume measurements during the period prior to the clinicalexpression of malaria. Further plasma volume determina-tions were made during both febrile and afebrile phasesof clinical activity and, whenever possible, duringconvalescence.

6 Obtained from the William R. Warner Company ofNew York, N. Y., in ampoules of this size.

780

BLOODVOLUMEALTERATIONS IN HUMANMALARIA

0I

(4

§I Ig O 0 00 I

-4 ~ ~ ~ - .~

-4 +00U) U) 00t-.

1 el 1 -: n%q q0%.t-! 0~1 c ~claq q 0*U) 4U)Vq V*0*04p jU)*000 00000000 U) 0U)'O0%~~'.. t10%if) (' U) U)o0-.i MP) 00~o0eq 0 W) ~%O U)0m %V~

.0- - - -~~~~~~~~~~~~~N-4

- --N94 V- -43l-

0%'m0, -. U)00.- N 0z'.. %O) in) 400f)0 '0"t- '-400%( 0%0%'0'O% "Momif

0%0'0 %if)V U)C'4-4 00%(' o0 V 400.# V)U)OU) o0"

40 ~ , ++++ ,"+ i+- 1'4 + T I+ +

~~ ~ ~ U)C0F~~~~t~r NW0 if 008 MUM0if 808 W)U) 0%4IW~ 9 000 ('N("8('.VsoU)V~-l-Now 1-.t-.t'.00 co V 00 Z- Ul) (- l- U) ) Z- N 0 '00 co U)i - (- It-(-

.0 - -~~~~~~qtCt -o~ mc NO- in - -2W --(1 T)10 ac

o '~~~~~~~~~ -000-) if) ~'-04 . '0q 40 0 ~('4 0. -- 0%.t NO'-4 N000 0%(n'400NOint-V 1~tU)co 4co 0'0t

I ~~~~~~ 40NMN 4 M0 . 4 00Vfol (' 4( N)U) 0%044 4.0(' i)-NN"N

"Oq 0o . o -.0 ~4 o' if0 %(' t- co"0%UV~U ('0%0 Q)t. -('00 0(-l

riK o-.

4 t00

tn

0'000

.~~~ .0 ~~~~~ ++++I +1+ +++ +++ ++ + (+ + +

.~oVV0%.--. 0 ('400'4if) 00 ) Vif U)0%nW )Coi -V NO'4UU))(-( 0mifnr 0% Nt0U) ) ('.00 in)'0

-

0%if 0..'00 if) 0%

m%0 0 - OV0%0it- ('4(4'O -00g%'-4(1 ..of

N ) M %p00 NOk-U e'ot- 0 (-000 a en W) co U) - 000000 in in N k-t-z~~ .~~~~ t ~~+++II II + + + ++I ++I 1+ I1+1 II

0 VVVVU)U) U) U) cau 0~~~~~~~~~~~~VV U)VU) V U)VU) U)U)U)~~~~~~~~~~~~ U)U)U)~~~~~U) U)U)V '4 U) U)W)WV mnml

008000 0oo0 0. 00 0000 0000 0000 00000 01 000 0000

V 0000UWC~k)C0 OU)soO"--C )W000 0000 0C000 00000~0oo O0001 00004*W00nm vC0k C 04Vc)co a %0

('4 -- 04( 'v' 4-- - '" 4N -4'4"- '-4 C4I ++I I+I++I + +1

04

0

0

4-)

0.210)

4-J

0)

4C0

4-.

4U0.

4)

A4

I'n

781

CU

4-

U

&4)

a

44)40.-

IV

le

0

0

04I00t0Z.

a

I t10

HARRYA. FE1LUMANAND FRANKLIN D. MURPHY

64's § K E'p b

QfF1: I g0

782

00

4,

0

In4,

4,

17

00

LD U'

2-

0$ 8

U'

. 0

BLOODVOLUMEALTERATIONS IN HUMANMALARIA

-Io '.0

CD 6is12 s °s v0

F--~~~~~

783

0C

U)

Ez

0

UMN0

CQ

C-l

0 o _

ot -

1

HARRYA. FELDMANAND FRANKLIN D. MURPHY

40 _

+30 _0

+20

* *

1oG 0 e 000~~~

0 0~~-to 0 0 0

0 08-20, ~~~00 0

0 0100-*

3Q ~~~~~~0-30~~~~~~ot 0 ~ ~ 0

40~~~~~~~~~0*-30 ' ESoOK

*a* ASNA VOLUME

_~~~~~~~~~~~~~~~~~~~~~.8 .oHEMTC"^T lNX-SO e A1 C Q If 165 I 110 ese .5- 034 rim ONz- - 4 v IV I It Ib 18ZD Z 4 D UOV Z 4 6 8 1tO 12 14 16 18

IPLUSDAYS OF DISEASE

FIG. 3. RELATION OF PLASMAVOLUMETO HEMATOCRITALTERATIONS

DATA AND DISCUSSION

Plasma volumeErythrocyte destruction is a constant concom-

itant of all active malarial infections. Hence, itwas expected, a priori, that the plasma volumes ofthe patients studied would be increased if a stableblood volume were to be maintained. The datasummarized in Table II tend to support this con-

cept, for the plasma volume is increased almostwithout exception during the active phase of thevivax and falciparum infections studied. How-ever, the plasma volume increase may exceed thecalculated loss in circulating erythrocytes, resultingin hemodilution with a lowering of the venous

hematocrit out of proportion to the actual decreasein total erythrocyte mass (Figures 1 to 3, andTable II, patients 5 and 8). This dilution phe-nomenon, which has also been observed in pneu-mococcal pneumonia (6), tends to occur shortlyafter the end of a paroxysm.6 The plasma volumetends to be lower during the paroxysm than dur-

6 The febrile excursion associated with schizogony,whether or not accompanied by a definite chill.

ing the afebrile periods, probably the result ofexcessive loss of extracellular water by perspira-tion.

Erythrocyte mass

The mass of circulating erythrocytes was de-termined in all instances by subtracting the plasmavolume from the calculated total blood volume.As would be anticipated in a disease in which thereis continuing blood cell destruction, the erythro-cyte mass progressively decreases with the courseof the disease (Table II).

The validity of the value for erythrocyte mass

obtained by the T-1824 method has been chal-lenged by several workers (7, 8), who suggest thatthe calculated value is about 25 per cent higherthan the actual value. However, since our experi-ments are concerned with alterations in totalamounts rather than the total value per se, thiserror, if relatively constant, would appear to playa minor role in the interpretation of these data.The course of a patient who received 600 ml. ofred blood cells immediately following a bloodvolume determination is illustrated in Figure 4.

784

Z.02

CK4z

I

LU

w

I--

:3

Z

2

0LL

40I.-2

0

BLOODVOLUMEALTERATIONS IN HUMANMALARIA 7

4 -.

10 k 10xm<-~~~4

]>L S B

> X t+11 '/s:~~~

o oQ Oo _

I.

.t

> 8

0D IL0

0s I ¢ s ° e i00 >--

'4'- j? F) N2 cun *

A. If85

HARRYA. FELDMANAND FRANKLIN D. MURPHY

84§Q

so

9-4

w~~~~~~

0i IL§0

0 V-~~~~o

i4 |0 ° g g

*5qt A

7186

#.i

toNInNC-i

NC-iN

0N Mz

D `4P4

.- UXe.'40 t

04qt u

0 0

9o

F)

N

zDietIO

08 8-Q O^

4.

VA

us ' t,4 '1-

BLOODVOLUMEALTERATIONS IN HUMANMALARIA

0 0

2&h a6aII

hiF4

.

ON0

Idra

aS

0o

6

-6 12-~~~

787

HARRYA. FELDMANAND FRANKLIN D. MURPHY

On the following day another bloo;d volume de-termination revealed that the red cell mass hadincreased by 530 ml. and the plasma volume by50 ml. Another such patient, in whom519 of 600ml. of transfused erythrocytes were detected onthe day following transfusion, is shown in Figure5. Since the red blood cells were of necessitysuspended in a small amount of plasma, theaccuracy of the above determinations is probablygreater than would appear. In 1 patient (Figure6), however, the red cell mass increased only98 ml. on the day following a transfusion of 500ml. of erythrocytes, while the plasma volume in-creased 330 ml. This could be explained by theoccurrence of a hemolytic transfusion reaction asreflected by the sharp increase in plasma bilirubinfrom 0.99 to 2.45 mgm. per 100 ml.

.V-

'1

-o5

40

0 .0/

oo * o

20 * 0/oo // *

10 / .// .0 0/

0/

0/ 0t 30

f//

/o/o

* OURING ARASITIEMA0 DURING CONVALESCNCI

40 50o SO

PERCENT FALL IN HEMATOCRIT

FIG. 7. CORRELATIONBETWEENDEFICIT IN ERYTHROCYTEMASSAND HEMATOCRITFALL

Total blood volume

The total blood volume, calculated from thehematocrit and the value for plasma volume, tendsto be slightly lower during fever and higher dur-ing the immediate post-paroxysm, afebrile phase.These fluctuations, usually not great, result mainlyfrom changes in the plasma component of theblood. Observations on 4 patients during various

stages of convalescence (Table II, patients 1, 5,6 and 8) suggest that the total blood volume tendsto be decreased in convalescence primarily be-cause the plasma volume does not remain suffi-ciently increased to make up for the reduction inred cell mass. The lowered plasma volume may

result in part from the significant loss of bodyweight which these patients incur during thecourse of the disease.

Venous hematocrit index

A progressive fall in the venous hematocrit in-dex begins on the third or fourth day after theappearance of parasites and continues for severaldays after their disappearance. This hematocritfall reflects the direction of change but does notindicate the quantitative alterations in erythrocyte-mass (Figure 7). The latter inaccuracy results,in many instances, from hemodilution, so that,for example (Figure 2), a drop of 7 points in thehematocrit on day 12 is due almost entirely to thisphenomenon. In general, it may be said that inhuman malaria of the character reported in this

study a single hematocrit index gives only an ap-proximation of the true state of anemia, althougha series of such determinations will point the di-rection of erythrocyte change.

Plasma biirubinThe plasma bilirubin increases shortly after

parasites appear in the blood and returns tonormal following the end of the parasitemia.This increase, which is magnified considerablyby calculation of the total circulating bilirubin(Table II), appears to be a reflection of the de-struction of erythrocytes (Figure 8). As will benoted, this evidence of destruction may reach con-siderable proportions.

AnemiaThese studies infer that at least 2 factors are

operating in the production of the normochromic,normocytic anemia of malaria; i.e., erythrocytedestruction and hemodilution. Bone marrowstudies (9) suggest the possibility of a thirdfactor; namely, maturation arrest of the erythro-cyte -series.

Plasma proteinsThe decrease in the concentration of plasma

proteins during malarial infections has been re-ported (10, 11) to be due principally to the re-duction of the albumin fraction. While our basicdata are in agreement with the belief that therelative concentration of plasma proteins falls,

-788

BLOODVOLUMEALTERATIONS IN HUMANMALARIA 789

0- +50 *12 096zO 0F .40_4

40

.30+30

o '20

z 0- +10 0

c 10000.0 0~~~0 0 00

b .-IIL. 0 0

0

CD -20 0 0z 0

o ~~~~~~~~0-301- ~~~~~~~~~~0

Z 0

- 40 0 0' ERYTHRMYTEMASS

w *. *PLASMA SLIRUSIN(REDUCEO 10 XI

- 2 4 6 8 10 12 14 16 18 20 22 24 26 28 1 3 5 7 9 11 13 15 17 19PUS'

DAYS OF DISEASEFIG. 8. RELATION OF PLASMABILIRUBIN TO ALTERATION IN THE ERYTHROCYTEMASS

) +25 *29z0

4'20z

tr t

LU 010I *

co 0Z iOO _ @ @ -'.° *.;0

O0 0 0 0

c.5

0 0 0

-200 0 0

0~~~~~~~~~~~~0~~~~~~

2100 so~~ tGM, 00Mo 2 00

a ~~~~~~~~~~~~~~~~~~~~** TOT" CWCULATNG MM"

lx 0~~~~~

-25 0_ 4 6 8 10 12 14.16 18 202224 26 28 113- 5 7 9 11 13 15 A7 '19

PLUS0

D)AYS OF DISEASEFIG. 9. RELATION OF CHANGESIN TOTAL CIRCULATING PROTEIN TO CHANGESIN PROTEIN CONCENTRATION

(GRAMSPER 100 ML.)

HARRYA. FEIDMAN AND FRANKLIN D. MURPHY

the picture is altered when the total amount ofcirculating proteins is calculated (Figure 9), forit then becomes apparent that instead of a fall,there is more often no change or even a slightincrease. While plasma dilution accounts for someof the fall in the concentration of plasma proteinsduring malaria, it fails to explain the reported(10, 11) selective plasma albumin decrease. Astudy now in progress, which combines bloodvolume studies with electrophoretic analysis of theplasma, may serve to clarify the question of quan-titative alterations in the plasma protein fractions.Other factors which may play a r6le in the proteinalterations, and which still remain to be appraisedcritically, are nitrogen imbalance, hepatic dysfunc-tion, and the effect of the increased metabolic ac-tivity which accompanies fever.

Shock

During the course of these studies, 1 patientwas encountered (Figure 10; Table II, Patient2) who presented findings of sufficient interest towarrant a detailed discussion. This patient, onthe tenth day of a severe falciparum infection,suddenly developed clinical signs of shock, in-cluding a rapid, almost imperceptible pulse, hypo-tension, moist, cooling skin, and semiconscious-ness. Blood volume and other studies wereperformed while he was still in shock and priorto the administration of any intravenous fluids ormedication. Instead of the anticipated elevationof the venous hematocrit, which is usually associ-ated with the hemoconcentration of shock, thehematocrit index in this patient was 30, 32 percent lower than the control value and 19 per centless than the index obtained 2 days previously.The plasma protein concentration also fell. Theplasma volume had fallen 46 per cent below thecontrol value to 1628 ml., erythrocyte mass 71per cent to 697 ml., and the total blood volume 57per cent to 2325 ml. In other words, althoughthere had been a drastic reduction in plasma vol-ume, this was not indicated by the hematocritbecause of marked and almost concomitant de-struction of eryjthrocytes. The latter was reflectedin the enormous increase in plasma bilirubin,which had first been noted 2 days before. Itseems reasonable to assume that the anoxemiaresulting from erythrocyte destruction initiated

the chain of events which terminated in the plasmadecrease and shock. The marked reduction intotal proteins, of the same general degree as theplasma volume, may be partly explained by in-creased capillary leakage.

The- patient responded promptly to infusions of500 ml. of ery-throcytes and 500 ml. of plasma.Although he also received intramuscularly admin-istered atabrine dihydrochloride, this could nothave altered the immediate picture. Three dayslater the total blood volume had returned tonormal with increases of 100 per cent above theshock levels in both plasma and erythrocyte vol-umes. The hematocrit, as expected, was still low.In spite of the slight increase in the concentrationof plasma proteins, the calculated total amount ofprotein had increased greatly to a level slightlylarger than the control value.

We have not been able to confirm these dataon other patients as this is the only instance inwhich we have encountered this relatively rarecomplication of therapeutic malaria. The tech-niques employed were those used throughout thisstudy; normal values were obtained on a controlpatient on the same day. Since the error whichhas been stated to occur in the study of shock bythe use of T-1824 is thought to be due to in-creased loss of dye from the vascular bed andthus is responsible for plasma volume valueswhich are inordinately high (12), the significanceof the low values obtained in our patient wouldnot be diminished by this factor. The correlationbetween the biochemical and volume studies doneduring this shock state, the studies done on priorand subsequent days, the clinical findings, andthe theoretical concepts lead us to accept thequalitative accuracy of the direction and degreeof the above described alterations.

The therapeutic implications of this case areclear, for had treatment been directed toward theimprovement of shock by plasma transfusionsalone, the clinical state undoubtedly would havebeen worsened. Although it was necessary tobolster the blood volume, to have done so withoutincluding erythrocytes might have been disastrousin view of this patient's marked anemia. Sincethe destruction of erythrocytes is such an integralpart of malarial infections, it should be emphasizedthat transfusions of whole blood or erythrocytesare vital in the treatment of the shock which is not

790

BLOODVOLUMEALTERATIONS IN HUMANMALARIA

008 8^o o -

0)Id-

II!o _

w31

04 2

791

a

3rI-,

0W

to

:10

2 -

OD _e4

4

82 §8 0

I ID PI ° a

72ARRYA. FELDMANAND FRANKLIN D. MURPHY

uncommonly encountered in neglected falciparuminfections. Plasma infusions alone may reducethe ratio of circulating red cells below the criticallevel and result in a fatality.

SUMMARYAND CONCLUSIONS

1. The alterations in blood volume during thecourse of induced malaria in 9 patients have beenstudied by means of the dye T-1824.

2. (a) Plasma volume is almost uniformly in-creased during the active phase of vivax andfalciparum infections. This increase reflects theability of the body to maintain a fairly constantblood volume in the presence of erythrocytedestruction.

(b) The plasma volume increase may be dis-proportionately great following a paroxysm andthus the degree of anemia as measured by thehematocrit may be exaggerated.

(c) The plasma volume tends to be lowerduring the paroxysm than during the afebrileintervals.

3. -The calculated mass of circulating erythro-cytes progressively decreases with the course ofvivax and falciparum infections.

4. (a) The calculated total blood volume tendsto be slightly lower during the malaria paroxysmand higher during the afebrile intervals. Thesefluctuations are due mainly to alterations in theplasma volumes.

(b) The total blood volume tends to be re-duced during convalescence.

5. The concentration, as well as total amountof plasma bilirubin, increases during the periodof visible parasitemia and reflects erythrocytedestruction.

6. The normocytic anemia of clinically activemalaria may be due to at least two factors; namely,erythrocyte destruction and hemodilution.

7. The concentration of plasma proteins fallsduring the clinical phase of vivax and falciparuminfections. However, the total amount of circu-lating plasma proteins remains fairly constantand may in fact be increased.

8. (a) In one case of shock due to severefalciparum malaria, a drastic decrease in plasmavolume and eyrthrocyte mass was observed.

(b) These studies suggest that in the treatmentof the shock which may be encountered in malarialinfections, the administration of whole blood orerythrocytes is mandatory while plasma trans-fusions alone may be harmful.

It is a pleasure to acknowledge the constant advice andcooperation of Doctors Robert Briggs Watson, HenryPacker and William B. Wendel.

BIBLIOGRAPHY

1. Watson, R. B., The enumeration of malaria parasites.Bull. U. S. Army M. Dept., 1944, 81, 99.

2. Gibson, J. G., 2nd, and Evans, W. A., Jr., Clinicalstudies of the blood volume. I. Clinical applicationof a method employing the azo dye "Evans blue"and the spectrophotometer. J. Clin. Invest., 1937,16, 301.

3. Hopper, J., Jr., Tabor, H., and Winkler, A. W.,Simultaneous measurements of the blood volume inman and dog by means of Evans blue dye, T1824,and by means of carbon monoxide. I. Normal sub-jects. J. Clin. Invest., 1944, 23, 628.

4. Phillips, R. A., Van Slyke, D. D., Dole, V. P., Emer-son, K., Jr., Hamilton, P. B., and Archibald, R. M.,Copper sulfate method for measuring specific gravi-ties of whole blood and plasma Bull. U. S. ArmyM. Dept., 1943, 71, 66.

5. Malloy, H. T., and Evelyn, K A., The determinationof bilirubin with the photoelectric colorimeter. J.Biol. Chem., 1937, 119, 481.

6. Rutstein, D. D., Thomson, K. J., Tolmach, D. M.,Walker, W. H., and Floody, R. J., Plasma volumeand "extravascular thiocyanate space" in pneumo-coccus pneumonia. J. Clin. Invest., 1945, 24, 11.

7. Hahn, P. F., Ross, J. F., Bale, W. F., Balfour, W. M.,and Whipple, G. H., Red cell and plasma volumes(circulating and total) as determined by radio ironand by dye. J. Exper. Med., 1942, 75, 221.

8. Stead, E. A., Jr., and Ebert, R. V., Relationship ofplasma volume and cell plasma ratio to total redcell volume. Am. J. Physiol., 1941, 132, 411.

9. Diggs, L. W., Personal communication..10. Kopp, I., and Solomon, H. C., The relationship of

hypoalbuminemia to the edema of malaria. Am. J.M. Sc., 1941, 202, 861.

11. Boyd, M. F., and Proske, H. O., Observations onthe blood proteins during malaria infections. Am.J. Trop. Med., 1941, 21, 245.

12. Hopper, J., Jr., Winkler, A.. W., and Elkinton, J. R.,Simultaneous measurements of the blood volumein man and dog by means of Evans blue dye, T1824,and by means of carbon monoxide. II. Under ab-normal conditions including secondary shock. J.Clin. Invest., 1944, 23, 636.

792