HEPATIC STORAGEANDEXCRETIONOF SULFOBROMO-PHTHALEIN SODIUM IN THE DOG*

By HENRY0. WHEELER,ROBERTM. EPSTEIN,t ROSCOER. ROBINSON:tAND ERIC S. SNELL §

(From the Department of Medicine, Columbia University College of Physicians and Surgeons,and the Presbyterian Hospital, New York, N. Y.)

(Submitted for publication June 22, 1959; accepted October 1, 1959)

The use of a halogenated phthalein dye for theinvestigation of hepatic function was first proposedby Rowntree, Hurwitz and Bloomfield (1) in1913. Their technique required the measurementof the fecal excretion of phenoltetrachlorphthaleinfollowing its parenteral administration. However,it was subsequently shown by Rosenthal and White(2) that intravenously administered phenoltetra-bromophthalein disulfonate (BSP) is removedfrom the blood almost exclusively by the liverand that the function of the liver may therefore beevaluated by measuring the rate of disappearanceof this dye from the plasma.

Various lines of evidence support the view thatthe removal of BSP (and presumably many othersubstances) from the plasma involves two proc-esses: 1) storage of the material in the liver and2) biliary excretion. Following the intravenousadministration of BSP to dogs it was first shownby Wirts and Cantarow (3) that the output ofBSP in the bile continues for over three hours afterits virtual disappearance from the plasma, indi-cating that this material is first taken up by theliver and then gradually excreted. This observa-tion was confirmed by Brauer and Pessotti (4),also in dogs. The existence of a similar hepaticBSP storage mechanism in man is evident fromthe studies of Mendeloff, Kramer, Ingelfinger andBradley (5) in which a retardation of the BSPdisappearance rate from plasma was noted whensuccessive identical intravenous dosts were ad-ministered 30 minutes apart. These authors pos-

* This investigation was supported by a grant fromthe Department of the Army (Contract DA-49-007-MD-205).

tWork done during tenure of a Postdoctoral Fellow-ship, New York Heart Association.

tWork done during tenure of a Postdoctoral Fellow-ship, American Heart Association.

§ Work done while a Rockefeller Traveling Fellow inMedicine.

tulated that "saturation" of the hepatic storagecompartment occurs after a single injection of dyeso that the rate of uptake of subsequent doses islimited by the biliary excretion mechanism. Fluo-resence microscopy of living frog and rat liversby the technique of Grafflin and Bagley (6) hasprovided direct visualization of the storage phe-nomenon. Following intravenous injection bothfluorescein and thioflavine-S accumulate in highconcentration in hepatic parenchymal cells as wellas in the bile canaliculi. Finally, the same processhas been demonstrated after injection of I31-la-beled rose bengal by Taplin, Meredith and Kade(7) using an external counter over the liver.There is a rapid accumulation of isotope in theliver followed by a gradual disappearance as thedye is excreted. The latter phase can be blockedby acute common bile duct ligation in rabbits butthe "storage" uptake is still apparent.

The rate of uptake of BSP from the plasma intohepatic storage may be very rapid immediatelyafter an intravenous injection. In man, for ex-ample, the BSP concentration in plasma falls at therate of 10 to 15 per cent per minute following asingle injection (8) so that the initial rate of re-moval of a 300 mg dose must be of the order of 40mg per minute. However, the liver is incapableof removing BSP this rapidly for an extended pe-riod of time. Constant BSP infusions at ratesmuch over 3 mgper m2 per minute in man (9) or1.4 mg per 10 kg per minute in the dog (10) re-sult in a continuously rising plasma concentration.Mason, Hawley and Smith ( 11 ) reported a higherfigure of 4.4 mg per 10 kg per minute for dogsbut arrived at the same general conclusion. Theseobservations have led to the hypothesis that thebiliary excretion of BSP is a rate-limited processand that the rate of uptake of BSP from plasma isultimately limited by the maximal rate of excretion.

This mechanism was explored in more detail

236

HEPATIC STORAGEANDEXCRETIONOF BSP

by Combes, Wheeler, Childs and Bradley (10) us-ing 1181-labeled rose bengal to measure hepaticblood flow so that actual rates of hepatic BSP re-moval could be calculated from the splanchnicarteriovenous concentration difference. It wasfound that the removal rate of BSP reaches alimiting "maximum" during a BSP infusionwhich is sufficiently rapid to produce a continu-ously rising plasma concentration. This observa-tion was interpreted as further evidence for theexistence of a limited excretory transport mecha-nism, and the maximal rate of transport was desig-nated "Tm." However, when a large "priming"dose of BSP was administered at the time thatconstant infusion was started, hepatic removalrates much higher than the apparent excretory"Tm" were observed transiently, suggesting rapidinitial uptake of BSP into a "storage" compartment.It was postulated that the liver is capable of veryrapid uptake of BSP from the plasma when thedye is entering the storage compartment but thatthe uptake approaches the excretory "Tm" oncethe storage compartment is "saturated."

The high plasma concentrations of BSP re-quired for more detailed exploration of the storagemechanism interfered markedly with the removalof rose bengal making it impossible to obtain ac-curate measurements of hepatic blood flow. In thestudies reported here a mechanical device (rotam-

eter) was therefore employed for the measurementof flow in order to determine hepatic BSP uptakeaccurately over a wide range of plasma levels inthe dog. The results have proved consistent with'the view that biliary excretion of BSP involves arate-limited process but that accumulation (orstorage) of BSP in hepatic parenchymal cells isdetermined by plasma concentration.

METHODS

In 14 mongrel dogs (14 to 26 kg) anesthetized withpentobarbital (30 mg per kg) hepatic venous outflowwas measured using the rotameter and optical recordingsystem described by Shipley and Wilson (12). In thefirst 6 dogs hepatic venous blood was collected from theinferior vena cava below the liver and returned througha chest wound to the thoracic inferior vena cava (Fig-ure 1). Respiration was maintained by means of a Star-ling pump. A femoral-to-jugular shunt was provided forreturn of venous blood from structures below the hepaticcannula. In the remaining 8 dogs the arrangement shownin Figure 2 was employed. Hepatic venous blood wasled to the rotameter through a no. 20 Fr. woven nyloncatheter threaded through a femoral vein and tied inplace by a ligature around the vena cava above the renalveins. After passing through the rotameter the bloodwas delivered to a reservoir beneath the animal boardfrom which it was returned by finger pump to an ex-ternal jugular vein. Venous return from structures be-low the liver was collected from the opposite femoral veinand returned to the heart by way of the same reservoirsystem. A circulating warm water jacket around the

FIG. 1. SYSTEMFOR HEPATIC BLOODFLOWMEASUREMENTUSED IN FIRST SIX STUDIES.Hepatic venous blood was collected through cannula (A), delivered to rotameter (B)and returned to heart by way of cannula in thoracic inferior vena cava (C). Bloodfrom structures below (A) was collected through femoral vein cannulae (D) and re-turned to heart via jugular cannulae (E). Mixed hepatic venous blood samples wereobtained at F.

237

WHEELER,EPSTEIN, ROBINSONAND SNELL

B

FIG. 2. ROTAMETERSYSTEM USED IN LAST EIGHT STUDIES. Hepatic ve-

nous blood was collected through catheter (A), passed through rotameter(B) and delivered to reservoir (C). Blood from structures below the liverwas collected from the opposite femoral vein (D) and delivered to the samereservoir. A short section of Penrose drain (E) acted as a valve to preventemptying of reservoir as blood was pumped (F) through return tubing (G)to the jugular vein. Warmwater flowing through plastic tube jacketing (G)maintained body temperature. Balloon (H) directed hepatic venous bloodthrough the rotameter. Mixed hepatic venous blood samples were obtainedat I.

return tubing was regulated to maintain constant bodytemperature (measured by an esophageal thermistor).A cardiac catheter equipped with a balloon was insertedthrough the other external jugular vein and maneuveredinto the thoracic inferior vena cava. Its position wasdetermined by palpation through the intact diaphragm.Inflation of this balloon directed the entire hepatic ve-nous outflow through the rotameter. Hepatic venouscongestion did not occur, because the resistance of thetubing and rotameter to hepatic venous outflow was bal-anced by the siphoning effect of the blood delivered tothe dependent reservoir.

In all studies heparin was used as the anticoagulant(10 mg per kg initially and 5 mg per kg every hour, intra-venously). The external circuits were filled beforehandwith 6 per cent dextran in saline to prevent sudden ini-tial reduction in blood volume. The rotameter was cali-brated frequently with dog blood at body temperatureand in each of the last 8 studies this calibration was car-ried out immediately after completing the experiment.Sulfobromophthalein sodium (BSP) solutions were ad-ministered intravenously using a Bowman pump.

BSP concentration in arterial and hepatic venous plasma

was measured in a Beckman DU spectrophotometer at

580 m,u after dilution of 1 ml of plasma with 10 ml ofnormal saline and addition of 0.2 ml. of 20 per cent po-tassium hydroxide. The optical density of a similarlytreated alkalinized blank plasma (obtained before BSPadministration) was subtracted. Standard concentra-tions for this determination were prepared in dog plasma.Frequent hematocrit determinations were carried out inWintrobe tubes centrifuged 30 minutes at 2,000 rpm.These were not corrected for trapped plasma. Totalplasma volume was measured after the external circuitswere functioning, using I't-labeled human serum albumin.A blood sample obtained 10 minutes after injection wascompared with an aliquot of the injection solution dilutedto standard volume with carrier plasma to avoid adsorp-tion by glass. In the last 8 dogs this measurement was

repeated near the end of the study and intervening valuesof plasma volume were estimated by interpolation.Splanchnic plasma volume was assumed to be 20 percent of total plasma volume (13). The precision of thisassumption has not been tested under the conditions ofthese experiments. However, as shown in the equa-tion below, the value for splanchnic plasma volume is in-

238

HEPATIC STORAGEAND EXCRETIONOF BSP

volved in correcting for a comparatively small amount ofBSP which is distributed in the splanchnic plasma sothat only a reasonable approximation is required.

The hepatic BSP removal rate was calculated from thesplanchnic arteriovenous concentration difference on theassumption that removal of BSP by splanchnic tissuesother than the liver was practically negligible (14) ex-cept for changes in the BSP content of the splanchnicplasma itself. The following formula was employed:

RH= F(A-V)- AP( X SPV\At/where:

RH = hepatic BSP removal rate, mg per minuteF = hepatic plasma flow, ml per minute = rotameter

flow X (1 - hematocrit)AV

= arterial plasma BSP concentration, mg per ml= mixed hepatic venous BSP concentration, mg

per ml

P= rate of change of arterial plasma concentration,mg per ml per minute

SPV = splanchnic plasma volume, ml = 0.2 X totalplasma volume.

Measurements of flow and collection of blood specimenswere carried out synchronously at intervals of 3 to 10minutes and the values of F, A and V, obtained at thebeginning and end of each interval, were averaged andsubstituted into the above equation to calculate Ri for theinterval. The difference between the arterial concen-trations at the beginning and end of each period was di-vided by the time interval to obtain At During suddenchanges in concentration following large injections ofBSP (see Figures 4 and 5) continuous blood sampleswere drawn at a constant rate for 2 successive periods of2 minutes' duration in order to obtain representative av-erage values for A and V.

RESULTS

The first six studies demonstrate the effect ofsudden alterations in arterial BSP concentrationon the hepatic BSP removal rate. In the study il-lustrated in Figure 3 (Dog Hu) a constant BSPinfusion of 3.8 mg per minute was administered.As anticipated from earlier work (10) the hepatic

FEMORALARrERY

HEPATIC VEIN*.*

w.~~~~~~~~~OO Omgo

DSP 3.82mg/min IV.

t50mg asp Iv.V

U

10 20 30 40 50 60 70 80 90 100 110MINUTES

FIG. 3. EFFECT OF A SUDDENINTRAVENOUSINJECTION OF BSP ON HEPATICREMOVALRATE. A BSP infusion of 3.82 mg per minute was started at zero time

and continued throughout the study. Intravenous injection of 50 mgof BSP re-sulted in transient increase in arterial concentration and hepatic removal rate.

239

MG%Dog Hu24 Kg

8

7

uo

606a.

U)44-

3

MG/MII6

40

2

_ .o

0

N

120 130 1409 I

WHEELER,EPSTEIN, ROBINSONAND SNELL

MG%12

1 1

10

9 I-8

t)z0 7U

36

54-J

2

Dog Si/0.2 Krg

FEMORALARTERY

HEPAr/C VEIN

-SP 2. 333o.g/minmsmroPPEDs

MG/MIN

' °r Iw

-i a4

8

G.U)

a.wz

I

I I I I I

10 20 30 40MINUTES

50 60 70 80

FIG. 4. EFFECT OF SUDDEN INCREMENT AND SUDDEN DECREMENTIN AR-

TERIAL BSP CONCENTRATIONON HEPATIC BSP REMOVALRATE. A BSP in-fusion of 3.89 mg per minute was started 80 minutes before time zero and re-duced to 2.33 mg per minute 10 minutes before time zero. During steadystates hepatic removal rate was constant and independent of plasma concen-tration, but sudden increment and decrement in concentration produced, re-spectively, transient increase and decrease in hepatic removal rate representingnet uptake and release of "stored" BSP.

BSP removal rate achieved a constant value (3.2mgper minute) at about the time the arterial con-centration reached 2.5 mg per 100 ml, while thelatter continued to rise. After 107 minutes ofsteady BSP infusion, a sudden increment in ar-terial concentration was produced by the intra-venous injection of 50 mg of BSP. There was amarked increase in hepatic removal rate to as highas 7 mg per minute followed by a gradual returnduring the next 20 minutes to the original rateof 3.2 mg. per minute. This phenomenon is re-peated in the next illustration and the opposite ef-fect is shown to occur after a sudden decrement inplasma BSP concentration (Figure 4). In this

study a BSP infusion (4 mg per minute) was be-gun 80 minutes beforehand and slowed to 2.3 mgper minute 10 minutes before sampling. Duringthe succeeding 30 minutes a relatively constant ar-terial level of 4 mg per 100 ml was thus obtained.and the hepatic BSP removal rate was constant at2 mg per minute. After this control period aninjection of 50 mg of BSP produced, as in the firststudy, a marked transient increase in hepatic re-moval rate followed by a gradual return to 2 mgper minute as the arterial concentration establisheda new level of 6 mg per 100 ml. The BSP in-fusion was then discontinued for a 10 minute in-terval. During the resulting decrement in arterial

240

HEPATIC STORAGEANDEXCRETIONOF BSP

concentration the hepatic removal rate decreasedto as low as 0.7 mgper minute but returned againto 2 mg per minute after resumption of the infu-sion. This reduction in BSP removal rate oc-curred even though the absolute arterial BSP con-centration was always higher than during the con-trol period.

The other four studies in this series (Figure 5)confirm the observations described above. Suddenincrements in plasma level produced either by in-stantaneous intravenous doses of BSP (Figure5A, B, C) or temporary acceleration of infusionrate (Figure 5D) invariably resulted in signifi-cant transient increase in hepatic removal rate.Moreover, in Dog Sa (Figure 5D) the responseselicited by each of two successive concentration in-crements were identical. On the other hand, con-centration decrements (Figure 5A, B, C) resultedin transient decreases in BSP removal rate in eachcase.

The subsequent eight studies were designed tomeasure the effect of more gradual but continuouschanges in plasma concentration and were con-ducted as illustrated in Figure 6 (Dog Ot). A"priming" dose of BSP, approximately 12 mgper kg was administered, and this was followedby an infusion of approximately 0.15 mg per kgper minute. After a 30 to 40 minute equilibrationperiod this rate of infusion usually produced arelatively constant or slowly changing arterialplasma level above 2.5 mg per 100 ml as in theillustration. After an initial phase of rapid"storage" uptake the BSP removal rate also be-camne constant. Arterial and hepatic venoussampling was continued for an additional 30 min-utes. The infusion rate was then doubled andmeasurements were continued at 10 minute in-tervals for another 50 minutes. At the high in-fusion rate the arterial concentration was observedto rise rapidly at first, but within approximately20 minutes the rate of increase became less rapidand quite constant. Coincident with this the he-patic removal rate increased to a new constantvalue, considerably greater than the rate observedduring the slower infusion. In the study illus-trated the rate of change of arterial concentrationwith respect to time during the last 30 minutes ofthe initial BSP infusion was - 0.0014 mg per 100ml per minute and achieved a new constant value

of + 0.0346 mg per 100 ml per minute during themore rapid infusion. The corresponding hepaticBSPremoval rates were 2.62 and 4.76 mgper min-ute, respectively. In the same manner values ofarterial concentration change and correspondinghepatic removal rates were computed during eachof two BSP infusion rates in the other seven stud-ies (Columns 4 and 5 of Table I).

Extrahepatic removal rate of BSP was cal-culated in the last eight studies by means of the

me%11 - Dog De

6.7-A 177i04 94L7 -

3 _MP5 I.V.I 2 - SP 1V4 MP#/0I V.

mG/MINi

12

to 8

i 0z 0 2 40 60 s0 IO 2

M6%10 Dog MO

9-C mhZ2.. t

064~5BJ4 SSP 75mg I.V.

4

rr &SP ZS2omgp'alf V

MG/MIN

IQ_

w 10

, 6f- Lg4

20 40 60 80 100MINUTES

MG/MIN12 -

W10 -

48a

w 24

t 0 20 40 60 80 100MINUTES

FIG. 5. EFFECT OF SUDDENCHANGESIN ARTERIAL PLASMABSP CONCENTRATIONON HEPATIC REMOVALRATE IN FOURDOGS. A BSP infusion of approximately 6 mg per minutewas started about 70 to 100 minutes before time zero, andthis was reduced to the indicated infusion rate 15 minutesbefore time zero. In each case increments in plasmaconcentration resulted in transient increase in hepaticremoval rate, and decrements caused a transient decrease.In D an identical increase in removal rate accompaniedeach of two successive increments in plasma level indi-cating additional "storage" uptake with each increment.

241

WHEELER,EPSTEIN, ROBINSONAND SNELL

-4 00 In 0% en --4 C4 \0 `d0% \0 1-o o 0 - 0 0o 6o - -

00 tU)O 0O

C)~~~~~U

'0 006 . .

~4 C'- 00

J - ) U)

,-- C\ oo

4t In C% 0 U-) -4

cl .4 cl r -

- N 0% 0 0 o

V4 en4 CN

+ oo +

C. .)(N ( %

' - CU oN \0

(N 4 o 00(N -S N

m c,+ CSo0. H

0 -400

Co C

o %0

C) (

00 0O bi

C)

U) 0%

O U)

U 0E t1

%0 0

(N NC

00 (N c-i+' 00 %00'(N ) (N N'

O 0 d0 %00 00 - +- '0 0 U)O 0 0 0

+ +

U) 00 0 00

(N U) (N U)

4 (N_ (N

en

C-

0-

dz O

(N

IC

ON

00

c-i

c-i

00

oj4 00U) 0

'0 00tO 066C

+

m \0C eq

(N U)

00

3-

(N 0%_ 0

U) 00U) C-i

-d4 \0

6 oE

(N

( 4UN

u) es

U) 0-U) C-

U)

o4 C-

O00 '0 U)dq m

c4 U-) 0% \%

Cdq

Os08 0 -'0 0 0%o

0 0 0 0O 0 0

U) 00 U) cU) '0

0% U)oU) U)o. o

cCo

0\

-0+

1.

4-

CO

"0ceC)

10"CO::0)C

0

U_

0)

C)dX

CO

0.

0

+EC.)0)

C)

CO0)

>00U)

E

C0

H

0)

~0)

4-C=

CO'-

E

CO.

>00

0)0

Uct)

0)

N 0)

..o

COt

C)

Ct

ceO E

Q =

r 0

0C)CIO

" CO

CO5

~CO

CO_

*0

CO

242

-

-0 '0

*)

0*

0,%,

*0

0.4

R

Q4

00.

Q-06

004

04

04C

bo4

00

(0

S-(

0/2

rEX d'0

,i>

Cd0~

06

¢ od

._

bO

00

- 0

HEPATIC STORAGEAND EXCRETIONOF BSP

following formula:

Extrahepatic removal rate

I A -V) + (P) X (ESPV)]where:

I = infusion rate, mg per minuteF = hepatic plasma flow, ml per minute

A - V = splanchnic arteriovenous difference,mg per ml

AP = rate of change of arterial concentra-

tion, mg per ml per minuteESPV = extrasplanchnic plasma volume, ml

(assumed to be 0.8 X total plasmavolume).

The magnitude of the extrahepatic removal rate,both in absolute terms and relative to total infu-sion rate, is shown in Table II. The average ex-trahepatic removal rate under the conditions ofthese studies was 9.2 per cent of the rate of in-fusion over a range of plasma BSP concentrationsof approximatetly 3 to 16 mgper 100 ml.

TABLE II

Extrahepatic removal during estimation of BSP transportmaximum and storage capacity

1 2 3 4 5Arterial Extra- Per centconcen- hepatic extra-

Infusion tration removal hepaticDog rate range rate removal

mg/min mg/OWml mg/min 4/2 X100Ai 2.57 3.1- 3.8 0.14 5.4

4.95 3.1- 5.3 0.18 3.6

Bi 3.12 5.8- 6.3 0.08 2.66.22 6.3-12.2 0.59 9.56.22 12.2-16.4 1.16 18.6

To 2.66 5.5- 5.8 0.36 13.55.31 5.5- 8.1 0.48 9.0

Tr 2.53 7.4- 7.8 0.25 9.95.08 7.4-12.0 0.50 9.8

Ot 2.70 2.7- 2.8 0.10 3.75.38 2.8- 5.1 0.21 3.9

Du 2.63 5.8- 5.9 0.31 11.85.26 5.8-10.2 0.83 15.8

Ia 2.73 5.4 0.29 10.65.28 5.4- 9.2 0.39 7.4

He 3.33 6.4- 6.7 0.29 8.76.62 6.7-13.3 0.80 12.1

Average 9.2

DISCUSSION

A. Hepatic excretory and storage mechanisms.The movement of BSP from the plasma into theliver, as indicated in the introduction, dependsupon both the excretory and the storage functionsof this organ. Thus, in arithmetic terms, the he-patic removal rate of BSP at any instant must beequal to the sum of the hepatic excretory rateplus the rate of accumulation (or minus the rateof depletion) of BSP in the hepatic parenchyma.The various lines of evidence cited in the introduc-tion, although indirect, strongly suggest that theBSP excretory process involves a rate-limitedtransport mechanism and that a constant maximalexcretory rate (Tm) is established when a suffi-cient quantity of BSP is infused to maintain plasmaconcentrations above, approximately, 2 to 3 mg

per 100 ml. This conclusion has been drawn,in part, from repeated observations (10) of thesort illustrated in the first portion of Figure 3.Here, for example, a constant hepatic BSP re-

moval rate was achieved and maintained as the

plasma concentration (and hence the availableload) of BSP increased continuously. If onepostulated a variable excretory rate, then the con-stancy of hepatic removal rate observed underthese circumstances could only be explained by theunlikely assumption that every change in excretoryrate is accompanied by an equal and oppositechange in the rate of accumulation of BSP in thehepatic parenchyma. Thus it is assumed thatBSP excretory rate is constant (and equal, bydefinition, to "Tm") whenever an adequate quan-tity of BSP is available for maximal transport(i.e., plasma concentrations consistently in therange of 2 to 3 mgper 100 ml or above). If theseconditions are satisfied, increments or decrementsin hepatic BSP removal rate can then be takenas an indication of changes in the rate of move-ment of dye into (or out of) hepatic parenchymalstorage.

Application of this interpretation to the firstsix studies (Figures 3, 4, 5) leads to the conclu-sion that increments in plasma concentration re-sult in an increase in hepatic BSP storage whereas

243

\VHEELER, El'STEIN, ROBINSONAND SNELL

decrements in concelntration are followed by a net(lepletion of stored BSP (since, in the latter case,sufficient BSP is certainly available in the liver tomiaintain a maximal excretory rate during the pe-rio(d of markedly reduced hepatic uptake fromplasma). Apparently, then, the quantity of BSPwhich the liver can lhold in storage is determinedb the concentration of the dye in the surroundingplasma. The capacity of the liver to store BSP(loes not appear, within the limited range of plasmacolncentrations explored, to become 'saturated"in an absolute sense, since successive incrementsin plasma concentration result in additional "stor-alge 'uptake of BSP 1h the liv er (Figure 5D).Moreover, in the last eight studies the fact thathepatic remiioval rate is much higher throughout aleriod of rising plasma concentration than duringa relatively steady state indicates that a continu-ously increasing quantity of BSP can be storedin response to a constantly increasing plasma coIn-centration. Under these circumstances the hepaticremovTal rate includes a constant rate of movement

C2C

ii.

I-

Ir-4

I--40.I

of BSP into hepatic parenchymal storage and is,to this extent, greater than the excretory Tm.This also applies to all experiments of the typeillustrated by the first portion of Figure 3, wlicrethe plasmiia BSP concentration is changing. Wh-lileit does not modify the previous argumentS in fa-v-or of a constant excretory Tm, the potential in-e(juality l)etween hepatic BISP remov al rate atndexcretory- Tmii must obviously l)e recognized in anyquantification of the latter v-lhich is basedn(lmeasurements of the former.

B. Qua(niltification of liepatic storage and c'rcC-tory 11u1. Because clhanges in the quantity ofB)SP storedl in the liver depend upon the differencebetween uptake from the plasma and excretion inltothe bile, the quantification of BSP storage requiresa knowledge of both the hepatic removal rate and(the excretory Tm. Theoretically, the Tm couldbe assumed to be equal to the hepatic removalrate observed toward the end of a protracted pe-riod of constant high plasma level, since the inetquantity of stored BSP would ultimately aclhiev-e a

8

6

4

2

o 0

FIG. 6. RELATIONSHIP BETWEENHEPATIC REMOVALRATE AND RATE OF CHANGE

OF PLASMACONCENTRATION. A "priming" BSP injection was followx ed by two

successive constant infusions of BSP. The rising plasma concentration duringthe more rapid infusion was associated with continuous uptake of BSP into"storage" as indicated by the higher hepatic removal rate.

MG oog 0 220 Kg

5- FMORALARrERY

- .0014mg%/mmn

3

2 a HEPArIC VEIN, BSPI255mg IV.

-0- BSP 2.70 mg/rn/n I.t V BSPa 5.3mglmln/ I. VMG/ MIN

l0 r

20 30 40 50 60 70 80 90 100 11O 120Ml NUTES

244

3

HEPATIC STORAGEAND EXCRETION OF BSP

constant value under these circumstances. Inpractice, however, this approach would requirea knowledge of the Tmbefore it could actually beestimated. In other words, after administering asuitable "priming dose" of BSPone would have tofind, by trial and error, the precise BSP infusionrate that would maintain a constant plasma levelfor an indefinite period. This laborious procedurecan be avoided by the application of a generalequation for hepatic removal rate based on theobserved dependence of storage upon plasma con-centration. If it is assumed, over a limited rangeof concentrations, that the quantity of stored BSPis directlv proportional to the plasma concentra-tion, then a proportionality factor can be definedto express the number of milligrams of BSP takenup into storage per mg per 100 ml of plasmaconcentration. This factor is designated "rela-tive storage capacity" (S). The net rate of up-take of BSP into storage (in mg per minute) dur-ing a period of continuously changing plasma con-centration is then equal to the product of S timesthe rate of change of plasma concentration (inmilligrams per 100 ml per minute). The generalequation for hepatic removal rate thus becomes:

Equation 1: RH = Tm+ S X (A)where:

RH = hepatic removal rate, mg per minute

Tm = excretory transport maximum, mg perminute

S = relative storage capacity, mgper mgper100 ml

AP = rate of chanige of plasma concentration,Atmg per 100 ml per minute.

Substitution into this general equation of the val-

ues of RH and AP, observed during two differentAt'rates of infusion, provides a set of simultaneousequations that can be solved for both Tmand S.

This calculation has been carried out using the

values of RH and A shown in Columns 4 and 5Atof Table I, and the resulting estimates of TmandS are shown in Columns 6 and 8 of the same table.These have also been reduced to a standard body

weight of 10 kg in Columns 7 and 9. The Tmvalues calculated in these dogs were not obviouslydifferent, per unit of body weight, from those es-timated in studies reported previously (10). How-ever, it must be emphasized that the present stud-ies were carried out under conditions of anesthesia,severe trauma and often reduced hepatic bloodflows. While these conditions produced no obvi-ous progressive impairment of the mechanismsunder investigation, the precise numerical valuesmay not apply to normal animals.

C. Concentration of BSP in the liver. The rela-tive storage capacity, as defined, has the dimen-sions of a volume, expressed in units of 100 ml.Thus, in a sense, an average "S" of 19.9 mg permg per 100 ml per 10 kg means that the hepaticstorage compartment of a 10 kg dog behaves asif it were an extension of the plasma compartmentwith a volume of 1,990 ml. Theoretically, there-fore, one could define S as the hepatic "volume ofdistribution" of BSP, but compared with theknown size of the liver the numerical values of Sare absurdly large. Actually the dog's liver, ex-clusive of blood content, weighs only 235 g per10 kg (an average figure for 14 dogs in which aprior injection of Cr5l-labeled red cells made itpossible to estimate and subtract the blood con-tent). In order to accommodate 19.9 mg of BSPfor every mg per 100 ml change in plasma con-centration, the liver must be capable, therefore,of storing BSP at more than eight times the con-centration in the surrounding plasma. ( If thematerial were stored at a concentration just equalto that in plasma, a 235 g liver would have a stor-age capacity of only 2.35 mg per mg per 100 ml.)Since part of the weight of the liver is attributableto comparatively inert supporting tissue, and sincethere is no assurance that dye is uniformly distrib-uted through the remaining parenchyma, it is prob-able that concentration of BSPin the actual hepaticstorage sites may achieve an even higher ratio withrespect to plasma concentration.

D. Storage and transport mechanisms. Thepresent studies provide no information regardingthe nature of the mechanisms by which BSP isconcentrated in the liver and transported into thebile nor of the possible relationship between thetwo processes. The schema presented here is un-doubtedly oversimplified. It has been shown byKrebs and Brauer (15), for example, that BSP is

245

WHEELER,EPSTEIN, ROBINSONAND SNELL

partially converted by the liver into a numberof chromatographically separable molecular spe-cies, some of which have definitely lower extinctioncoefficients than the parent compound at 580 m,u.Whatever the behavior of these products withrespect to storage and excretion their formationrepresents, in effect, a "loss" of BSP color fromthe body and would hence be included in the ex-cretory Tm as calculated herein. However, theresulting overestimation of Tmcompared with theactual rate of biliary excretion is probably notvery great, since over 70 per cent of administeredBSE' is ultimately recoverable in the bile by simplecolorimetry (3). Moreover, the major quantityof the derivatives is ultimately excreted in thebile, and to this extent properly belongs in the"Tm" calculation.

E. Simplification of technique. Even at thehigh plasma concentrations required for thesestudies the calculated removal of BSP by struc-tures other than the liver was comparatively small.The figures shown in Table II tend, if anything, tooverestimate the actual removal by extrahepatictissues. They are undoubtedly magnified by theeffect of blood loss and subsequent hemodilution(due to repeated sampling plus wound ooze afteranticoagulation). Moreover, the hematocrits werenot corrected for trapped plasma so that the valuesfor hepatic plasma flow, and hence the hepatic re-

moval rates, were subject to a systematic under-estimation which would also exaggerate apparentextrahepatic removal.

If one ignores the effect of extrahepatic BSPremoval altogether, then the values for Tmand Scan be calculated without measurement of hepaticblood flow and hepatic vein concentration, as fol-lows:

R =I-PVX ()

where:

R = BSP removal rate, mg per minute

I = BSP infusion rate, mg per minute

AP = rate of change of plasma BSP concen-

tration, mg per 100 ml per minute

PV = plasma volume in hundreds of ml.

Substituting this expression into Equation 1:

Equation 2: 1 = Tm+ (S + PV) X At

This equation may be solved for Tmand S by sub-

stitution of two sets of values for I and 'AP This

calculation, in the eight studies shown in Table II,yielded values for Tmand S (Columns 10 and 11)which were quite comparable to the figures calcu-lated from the actual hepatic removal rates. Asshown in Columns 12 and 13, Tmwas thus over-estimated by an average of 9 per cent and S by4 per cent when extrahepatic removal was ignored.

It is apparent from this that reasonable esti-mates of BSP transport maximum and storagecapacity can be obtained by a relatively simpletechnique utilizing constant BSP infusions andmultiple peripheral blood samples. Studies ofthis type are in progress (17); the results will bereported later in detail.

SUMMARY

The hepatic removal rate of sulfobromophthaleinsodium (BSP) has been estimated in 14 dogs bymeans of measurements of hepatic blood flow (ro-tameter) and splanchnic arteriovenous concentra-tion difference. During a constant BSP infusionhepatic BSP removal reaches a constant rate atan arterial concentration of 2 to 3 mg per 100 mleven when the latter continues to rise. Sudden in-crements in arterial concentration produce amarked transient increase in hepatic BSP removalrate and decrements produce a transient reduction.The hepatic removal rate is higher during continu-ously rising than during constant arterial BSPconcentrations.

The data support the view that hepatic excre-tion of BSP depends upon a rate-limited trans-port mechanism and that the quantity of BSPstored in the liver parenchyma depends directlyupon the concentration in the surrounding plasma.A method for calculating the excretory trans-port maximum (Tm) and the relative storagecapacity (S) has been presented, based on ob-

1 It should be noted that this equation is, in fact, theequivalent of one of the graphic methods proposed forestimation of the maximal BSP excretory rate by Lewis(16).

246

HEPATIC STORAGEAND EXCRETIONOF BSP

servations during two different constant rates ofBSP infusion. The calculated values of storagecapacity indicate that BSP in the hepatic storagecompartment is maintained at a concentrationmore than eight times that in the plasma. In therange of plasma concentrations studied there wasno evidence of "saturation" of the storage com-partment. The extrahepatic removal rate of BSPis low enough so that Tmand S may be reasonablyestimated from observations of peripheral plasmaBSP concentrations during different rates of con-stant infusion, without resort to hepatic venoussampling or hepatic blood flow measurements.

ACKNOWLEDGMENT

The authors are indebted to Miss Marion Yulinsky,Mrs. Susan Stassa and Miss Evelyn Audioun for theirvaluable technical assistance.

REFERENCES

1. Rowntree, L. G., Hurwitz, S. H., and Bloomfield,A. L. An experimental and clinical study of thevalue of phenoltetrachlorphthalein as a test forhepatic function. Bull. Johns Hopk. Hosp. 1913,24, 327.

2. Rosenthal, S. M., and White, E. C. Studies in he-patic function. VI. A. The pharmacological be-havior of certain phthalein dyes. B. The value ofselected phthalein compounds in the estimation ofhepatic function. J. Pharmacol. exp. Ther. 1924,24, 265.

3. Wirts, C. W., Jr., and Cantarow, A. A study of theexcretion of bromsulphthalein in the bile. Amer.J. dig. Dis. 1942, 9, 101.

4. Brauer, R. W., and Pessotti, R. L. Hepatic uptakeand biliary excretion of bromsulphthalein in thedog. Amer. J. Physiol. 1950, 162, 565.

5. Mendeloff, A. I., Kramer, P., Ingelfinger, F. J., andBradley, S. E. Studies with bromsulfalein: II.Factors altering its disappearance from the blood

after a single intravenous injection. Gastroenterol-ogy 1949, 13, 222.

6. Grafflin, A. L., and Bagley, E. H. Studies of hepaticstructure and function by fluorescence microscopy.Bull. Johns Hopk. Hosp. 1952, 90, 395.

7. Taplin, G. V., Meredith, 0. M., Jr., and Kade, H.The radioactive (I'-tagged) rose bengal uptake-excretion test for liver function using externalgamma-ray scintillation counting techniques. J.Lab. clin. Med. 1955, 45, 665.

8. Ingelfinger, F. J., Bradley, S. E., Mendeloff, A. I.,and Kramer, P. Studies with bromsulfalein: I. Itsdisappearance from the blood after a single intra-venous injection. Gastroenterology 1948, 11, 646.

9. Bradley, S. E., Ingelfinger, F. J., Bradley, G. P.,and Curry, J. J. The estimation of hepatic bloodflow in man. J. clin. Invest. 1945, 24, 890.

10. Combes, B., Wheeler, H. O., Childs, A. W., andBradley, S. E. The mechanisms of bromsulfaleinremoval from the blood. Trans. Ass. Amer.Phycns 1956, 69, 276.

11. Mason, M. F., Hawley, G., and Smith, A. Applica-tion of the saturation principle to the estimationof functional hepatic mass in normal dogs. Amer.J. Physiol. 1948, 152, 42.

12. Shipley, R. E., and Wilson, C. An improved re-cording rotameter. Proc. Soc. exp. Biol. (N. Y.)1951, 78, 724.

13. Bradley, S. E., Marks, P. A., Reynell, P. C., andMeltzer, J. The circulating splanchnic blood vol-ume in dog and man. Trans. Ass. Amer. Phycns1953, 66, 294.

14. Selkurt, E. E. Validity of the bromsulphalein (BSP)method for estimating hepatic blood flow. Amer.J. Physiol. 1953, 175, 461.

15. Krebs, J. S., and Brauer, R. W. Metabolism ofsulfobromophthalein sodium (BSP) in the rat.Amer. J. Physiol. 1958, 194, 37.

16. Lewis, A. E. Investigation of hepatic function byclearance techniques. Amer. J. Physiol. 1950, 163,54.

17. Wheeler, H. O., Meltzer, J. I., Epstein, R. M., andBradley, S. E. Hepatic storage and biliary trans-port of bromsulfalein in dog and man (abstract).J. clin. Invest. 1958, 37, 942.

SPECIAL NOTICE TO SUBSCRIBERS

Post Offices will no longer forward the Journal when you move.Please notify The Journal of Clinical Investigation, Business

Office, 333 Cedar Street, New Haven 11, Conn., at once when youhave a change of address, and do not omit the zone number ifthere is one.

247