Abnormal Plasminogen-Plasmin System Activity (Fibrinoly ... · Journal ofClinical Investigation Vol. 43,oNo. 4, 1964 AbnormalPlasminogen-Plasmin System Activity (Fibrinoly-sis) in

Journal ofClinical InvestigationVol. 43,oNo. 4, 1964

Abnormal Plasminogen-Plasmin System Activity (Fibrinoly-sis) in Patients with Hepatic Cirrhosis: Its Cause

and Consequences *

ANTHONYP. FLETCHER, OLIVER BIEDERMAN, DONMOORE,NORMAALKJAERSIG, AND SOL SHERRY

(From the Department of Internal Medicine, Washington University School of Medicine, andthe Jewish Hospital, St. Louis, Mo.)

Patients suffering from hepatic cirrhosis ofmoderate or advanced degree frequently have anill-defined coagulation disorder and may, espe-cially if subjected to operative stress, develop asevere hemorrhagic diathesis, often associatedwith pathological plasma proteolysis (pathologicalfibrinolysis). Moreover, many such patients mayexhibit, constantly or intermittently, evidence ofabnormal plasminogen-plasmin system activity(fibrinolysis), documented by rapid lysis of theirclotted blood and other tests.

Interest in this latter phenomenon stems fromGoodpasture's paper (1), published in 1914, inwhich he reported that blood, drawn from fourcirrhotic patients, when clotted, underwent spon-taneous lysis more rapidly than did that from con-trol subjects. Later Ratnoff (2) confirmed thisfinding and substantially extended study of thisphenomenon. He noted that rapid plasma clotlysis was a frequent accompaniment of Laennec'scirrhosis, was sometimes observed in patients whoexperienced hepatic damage during the course ofsome other illness, but was not observed in eitheracute hepatitis or obstructive jaundice withoutobvious associated hepatic damage. Subsequently,other authors (3-8) have reported that a highproportion of patients, suffering from advancedhepatic cirrhosis, exhibited rapid spontaneouswhole blood clot lysis, shortened euglobulin lysistimes, and other evidence of enhanced plasmino-gen-plasmin system activity. Moreover it hasbeen reported (7, 9, 10) that the administration

* Submitted for publication September 6, 1963; acceptedDecember 4, 1963.

Presented in part at the Seventy-sixth Annual Meetingof the American Association of Physicians, Atlantic City,N. J., 1963. This work was supported by a grant fromthe National Heart Institute (1H-3745), U. S. PublicHealth Service.

of drugs capable of eliciting enhanced thrombo-lytic activity in the normal subject produced ananomalous and exaggerated response in the cir-rhotic patient.

Hitherto, despite the apparent clear-cut natureof these phenomena, their interpretation and sig-nificance have remained obscure owing to de-ficiencies in understanding of the function of theplasminogen-plasmin system in vivo. Though re-cent studies (11-14) have delineated the crucialrole of plasminogen activator in the induction ofphysiological thrombolytic phenomena and in thegenesis of pathological plasma proteolytic (fibrino-lytic) states, neither the mechanisms underlyingin vivo release of plasminogen activator nor thoseresponsible for the rapid disappearance of thismoiety from the circulation have been fully elu-cidated.

The aim of the present study has been to definethe anomaly of plasminogen system function re-sponsible for the enhanced fibrinolytic activity inthe cirrhotic patient and to assess the significanceof this enhanced activity in the genesis of the co-agulation disorder that may complicate this con-dition. The first problem has been investigatedprimarily by determination of the respectiveplasma clearance rates for plasminogen activatorin normal subjects and in cirrhotic patients andthe second by studies designed to relate changesin the activity of the plasminogen system to thedevelopment of the recently described and dis-tinctive coagulation disorder, defective fibrinpolymerization (15-18). In this latter disorder,arising secondarily to pathological plasma pro-teolysis, high molecular weight fibrinogen frag-ments (19, 20) circulate in plasma and, by in-hibiting the polymerization of fibrin monomer,produce disturbances in hemostatic function.

681

FLETCHER, BIEDERMAN, MOORE,ALKJAERSIG, AND SHERRY

The investigative protocol involved the induc-tion of enhanced plasma thrombolytic activity inpatients suffering from moderately advanced oradvanced hepatic cirrhosis and in other subjectswithout liver disease by one of two procedures.The majority of studies were performed with in-travenously administered nicotinic acid used toinduce enhanced plasma thrombolytic activity, butin some patients, suffering from mental disease,studies were performed after therapeutic electro-shock. With both stimuli the response of thecirrhotic patients, both with respect to its degreeand duration, greatly exceeded that of subjectswithout liver disease. The degree and persistenceof the rise in plasma thrombolytic activity in thecirrhotic patients was of such magnitude as toinduce coagulation alterations of the type observedin patients suffering from pathological plasmaproteolysis. Moreover, the data on the clearanceof plasminogen activator indicated that these find-iligs were attributable to gross deficiency in theability of cirrhotic patients to clear plasminogenactivator from the circulation. Consequently, itwas concluded that the normal liver exerts im-portant control functions over plasma thrombo-lytic activity and that loss of this hepatic clearancefunction, as a consequence of certain hepatic dis-eases, increases susceptibility to the developmentof pathological plasma proteolysis and to the coag-ulation disorder of defective fibrin polymerization.

MethodsPlasma thrombolytic activity was assayed by the eu-

globulin lysis time as previously modified to permit ex-pression of the results in units (11) and by the IP`-labeled clot lysis assay (12, 13). The plasma euglobulinswere precipitated from 0.5 ml plasma according to themethod of Milstone (19 vol water containing 0.32 vol1% acetic acid added to 1 vol plasma) (21). After cen-trifugation the precipitate was resuspended in 0.5 ml 0.01MVeronal buffer, pH 7.4, and clotted with 0.5 ml throm-bin (2 U per ml). Clots were incubated at 370 C, andthe time for complete lysis was recorded. For the I...assay, clots were prepared from 0.2 ml bank bloodplasma to which 0.5 casein U of plasminogen and suffi-cient I'31-labeled fibrinogen were added to yield a finalcount for the fully lysed clot of 25,000 cpm per clot.0.2 ml of the patient plasma to be tested (refrigeratedon drawing and assayed within a 2-hour period) was in-cubated with a single clot for 2 hours at 370 C. De-termination of supernatant radioactivity and plasmathrombolytic activity was carried out as previously de-scribed (13).

Plasma plasminogen, antiplasmin, fibrinogen, the plasmathrombin clotting time (by a procedure utilizing throm-bin titration mixture), and theplasma urokinase lysistime (a method for detecting the presence of inhibitorsto plasminogen activator) were determined as previouslydescribed (12, 22).

Patient groups: electroshock series. Observationswere made on 12 patients suffering from mental depres-sion, before and after unmodified electroshock therapy.The group contained three patients, in whom the diag-nosis of hepatic cirrhosis had been confirmed by hepaticneedle biopsy, and nine patients without clinical or labora-tory evidence, including normal bromsulfophthalein(B.S.P.) excretion tests, of hepatic disease. Venoussamples, refrigerated on drawing, were collected beforeand 1, 5, 15, 30, 60, 120, and 180 minutes after electro-shock, during which each patient experienced a typicalconvulsive seizure.

Nicotinic acid series. There were 32 patients in thisgroup: eight patients with moderately advanced or ad-vanced hepatic cirrhosis; three patients in whom hepaticcirrhosis was suspected clinically and on biochemicalfindings but in whom repeated hepatic biopsy failed toconfirm the diagnosis; four patients with carcinomatousinvolvement of the liver (one primary and three meta-static from breast, colon, and stomach); and 17 subjectswithout clinical or biochemical evidence of hepatic dis-ease. Each received 100 mg nicotinic acid intravenouslyover a 3-minute period and developed a typical flush.Though differences in intensity and duration of the flushwere noted, this was not correlated with the subsequentresponse in plasma thrombolytic activity. One patient,suffering from hepatic cirrhosis, experienced, immediatelyafter the injection, a brief episode of syncope and hypo-tension, probably of vasovagal origin. Blood samples,immediately refrigerated, were withdrawn before and10, 15, 30, 60, 120, 180, and 240 minutes after completionof the injection.

The existence of hepatic cirrhosis was documented inall patients with this diagnosis by appropriate historical,physical, and laboratory findings, and in all patients ex-cept one, where the diagnosis was confirmed at autopsy,by positive needle biopsy histology. The initial bio-chemical findings in the cirrhotic group, in whom the ma-jority of studies with nicotinic acid was performed, areshown in Table I; the findings attest to the generallyadvanced nature of the disorder in this patient group.

Control observations were made on convalescentmiddle-aged hospitalized patients without fever, hemato-logical disorders, or other clinical or biochemical evi-dence of hepatic disease. Control series patients hadusually been admitted for the study of chronic ratherthan acute medical disorders. The data shown in Fig-ures 5 and 6 were obtained from apparently healthylaboratory personnel.

Statistical testing was by the t test, using the method ofpaired comparison (23) for determining the significanceof experimental deviation from control values. Stand-ard terminology has been used to designate componentsof the plasminogen-plasmin system (24).

682

FIBRINOLYSIS IN HEPATIC CIRRHOSIS

TABLE I

Diagnostic and biochemical findings in the group of cirrhotic patients who received nicotinic acidand in whomthe majority of observations was made*

Alka-line

phos- BSPBili- phat- Albu- Glob- reten-

Age Sex Remarks rubin ase min ulin S-GOT S-GPT tion C-CF

mg/ Bodan- g/100 g/100 U/ml U/ml %100 ml sky ml ml

U/ml72 F Biopsy-Laeinec's cirrhosis, 4.5 6.4 2.2 3.2 116 39 33 4+

ascites, esophagealvarices; died

19 F Biopsy-postnecrotic 2.6 9.0 3.5 4.4 150 100 4+cirrhosis (pregnancy)

70 M Biops)-Laennec's cirrhosis, 14.7 4.8 2.0 4.5 610 390 29 3+Prothrombin time, 53%

54 F Biopsy-Laennec's cirrhosis, 21.1 10.6 2.8 3.2 80 51 28esophageal varices

65 M Biopsy-Laeinec's cirrhosis, 2.6 2.4 2.4 3.7 46 6 40 4+ascites

59 F Biopsy-Lainnec's cirrhosis, 6.6 3.5 2.5 2.3 65 52 70 4+esophageal varices

45 xi Biopsy-Laennec's cirrhosis 4.9 5.2 3.0 3.0 140 62 41

* Abbreviations: S-GOT, serum glutamic oxaloacetic transaminase; S-GPT, serum glutamic pyruvic transaminase;BSP, bromsulfophthalein; C-CF, cephalin-cholesterol flocculation.

Results

Enhlanced plaswa throtbolytic activity afternicotinic acid or electroshock. Figure 1 illustratesserial determinations of plasma thrombolytic ac-tivity in cirrhotic patients (the two left-hand por-tionls of Figure 1) and in normal subjects with-out hepatic disease (the two right-hand portions)after the iv injection of 100 mg nicotinic acid.Plasma thrombolytic activity was determined bythe euglobulin lysis time (upper portion of Fig-ure 1 ) and by the 1131-labeled fibrin clot method(lower portion). Several patients were studiedby both methods, but some were studied by onlyone; consequently, the two series, illustrated re-spectively at the top and bottom of Figure 1,though comparable, are not identical.

Nicotinic acid injection produced, in both nor-mal subjects and cirrhotic patients, a sharp in-crease in plasma thrombolytic activity. This re-sponse was maximal at approximately 15 minutes,but both the degree of response and its durationin the cirrhotic subject greatly exceeded that ofthe normal subjects. At 15 minutes, the averageincrease of plasma thrombolytic activity in thecirrhotic group was 2 to 3 times that of the

normal subjects, and whereas plasma thrombo-lytic activity in the normal subjects had returnedto base-line values within 1 hour, enhanced plasmathrombolytic activity persisted for 2 to 4 hours inthe cirrhotic patients.

Statistical testing showed that the difference inresponse between the two groups was highly sig-nificant, p < 0.001 for each of the two methods(euglobulin lysis and isotopic assay) tested inde-pendently. The ratio of response, with respectto plasma thrombolytic activity, between the twogroups was determined by measurement, usingplanimetry of the areas under the response curves.Comparison, in this manner, revealed that theaverage over-all increase of plasma thrombolyticactivity in the cirrhotic patients was 5 times thatof the controls, the euglobulin method showing a6-fold difference and the isotopic method a 4-foldone. This discrepancy between the results ofthe two methods probably was attributable to thefact that the patient groups tested by each method,although overlapping, were not identical. Figure1 illustrates that the results with the two methodswere well correlated and exhibited no divergencebetween the general shape of the response.

683


EUGLOBULIN LYSIS AFTER NICOTINIC ACID

IxI\%I \

\\I'

I'' Cirrhofics

\ Mean and Range of\% 7

Normalof

//

1131 CLOT LYSIS AFTER NICOTINIC ACID

III I C'irrhoticsI Mmean and Range of

I 5Ip1

NNormalI IIMean and Range ofI I /3III'II II I

0 1 2 3 4 0TIME-HOURS POST INJECTION

1 2 3

FIG. 1. PLASMATHROMBOLYTICACTIVITY (MEAN AND RANGEOF OBSERVA-TIONS, ORDINATE) AFTER NICOTINIC ACID INJECTION IN CIRRHOTIC PATIENTSAND IN NORMALSUBJECTS AS A FUNCTION OF TIME (ABSCISSA). Plasmathrombolytic activity was determined by euglobulin lysis and by the I' clotlysis technique.

The striking differences demonstrated betweennormal subjects and cirrhotic patients with re-spect to the pattern of enhanced plasma thrombo-lytic activity developing after nicotinic acid wasalso apparent after the use of a quite dissimilarstimulus, therapeutic electroshock. These data,plasma thrombolytic activity being assayed by theeuglobulin lysis technique, are shown in Figure 2.Whereas, in the normal subjects, electroshock pro-duced a mean peak elevation of plasma thrombo-lytic activity of 0.4 U at 1 minute with decline tovirtually base-line values within 30 minutes, thecorresponding observation for the three cirrhoticpatients was a level of 1.6 U at 1 minute withmaintenance of sustained levels for 2 to 3 hours.The difference in response between the two

, 2-ii

U)

I-z

0

I-

0z

ID-

0 60 120 180TIME (MINUTES)

FIG. 2. PLASMA THROMBOLYTICACTIVITY (MEAN AND

RANGE OF OBSERVATIONS) IN EUGLOBULIN UNITS (ORDI-NATE) AS A FUNCTION OF TIME (ABSCISSA) AFTER ELEC-

TROSHOCKIN THREE CIRRHOTIC PATIENTS AND NINE NOR-

MAL SUBJECTS.

684

2

CD0-j

-j

0zFE

.E

z-JI-0-J

-(

3 CIRRHOTIC PTS. 9 NORMALPTS.

2-

I~~~~~I

L60I


z

° 2-j

32L&J

- I

co0z

co

OLA0

N.

cn 14

(n

L-z 12im,2

c-

a-

2

z

~iI(n

Ci

Thrombin Clotting Time

0 30' 2 3 4TIME- HOURSPOST INJECTION

FIG. 3. OBSERVATIONS MADE IN A SINGLE CIRRHOTIC

PATIENT AFTER THE IV INJECTION OF 100 MG NICOTINIC

ACID. The ordinates are, respectively, from top to bot-tom: plasma thrombolytic activity determined by botheuglobulin lysis and I13. lysis techniques, the thrombinclotting time, plasma fibrinogen, and plasminogen. Theabscissa is time after injection.

groups, determined by planimetry of the areas

under the respective response curves, was 16times greater for the cirrhotic as compared to thenormal group. Testing of this difference, by themethod of paired comparison, indicated a statis-tical significance at a level of p < 0.001.

Secondary effects of plasminogen activator ex-

cess. The exaggerated and long-sustained in-crease in plasma plasminogen activator concentra-tion that occurred in cirrhotic patients after nico-tinic acid injection was associated with alterationsin other plasminogen system components and inplasma substrates susceptible to plasmin digestion.Figure 3 illustrates such findings in a patient suf-fering from advanced hepatic cirrhosis who re-

ceived 100 mg nicotinic acid at zero time. Theupper portion of Figure 3 documents, by the useof both the euglobulin lysis and I131-labeled clotassays, the substantial long-sustained rise ofplasma thrombolytic activity that followed nico-tinic acid injection and again illustrates the ex-cellent correlation observed between these twomethods. Concomitantly with the early maximalincrease of plasma plasminogen activator, whichoccurred at 15 minutes, there was a sharp fall ofplasma plasminogen (bottom section of Figure 3)from 2.6 to 0.9 casein U per ml plasma; the plas-minogen concentration remained depressed dur-ing the period of maximal thrombolytic response,subsequently rising to 1.9 casein U per ml plasmaat the end of the observation period (4 hours).The initial rate of plasmin formation, consequentto the activation of 1.7 casein U of plasminogenper ml plasma in 15 minutes, was so rapid thatplasmin inhibition by antiplasmin was incomplete;as a consequence, plasma fibrinogen fell 28%(bottom of Figure 3), although by the end of theobservation period it was returning towards itscontrol value. The thrombin clotting time (mid-dle of Figure 3) rose from 10 to 13.8 seconds, afinding attributable to the presence of fibrinogenproteolysis products in plasma, as demonstratedby the use of a specific immunochemical assay.'

Plasma plasminogen, fibrinogen, and thrombinclotting time determinations for all subjects re-ceiving nicotinic acid are presented in Table II.The top line of each table section displays zerotime determinations for the normal and cirrhoticgroups. Plasma plasminogen and fibrinogen weresignificantly higher in the normal group (plas-minogen, 2.71 versus 1.86 casein U per ml plasmaand fibrinogen, 3.18 versus 2.58 mg per mlplasma), but the thrombin time differences, 10.0seconds as compared to 10.7 seconds, were notstatistically different.

After nicotinic acid, the normal group exhibited1 During in vivo states resulting in fibrinogen proteo-

lysis (pathological plasma proteolytic states) high mo-lecular weight fibrinogen fragments may be readilydemonstrated in plasma by immunoelectrophoretic tech-niques. Such specific fibrinogen fragments were demon-strated, during the present investigation by using pre-liminary electrophoresis of plasma in acrylamide gel forproteolysis product separation followed by immunodif-fusion against specific antisera for identification (25).The full findings will be reported separately.

30' 1 2 3 40

0

( brinogen0-~

PlIsmnogeI

f i

685


only minor decreases in plasma plasminogen(maximum, 0.33 casein U per ml at 15 minutes)and in plasma fibrinogen (maximum, 0.46 mg

per ml at 15 minutes) with a small increase inthe plasma thrombin clotting time (1.1 secondsat 30 minutes); although the fall of plasma fib-rinogen reached the customary boundary of sta-tistical significance (p = 0.05), all changes were

transitory, and testing of the combined determina-tions, obtained over the 1-hour observation period,against the control values showed only nonsig-nificant decreases for plasma plasminogen andfibrinogen with an increase of doubtful signifi-cance (0.05 < p < 0.1) for the plasma thrombinclotting time. In contrast, effects of much greatermagnitude were observed in cirrhotic subjects re-

ceiving nicotinic acid. At 15 minutes plasminogendecrease averaged 0.92 casein U per ml plasma,fibrinogen fall averaged 0.83 mg per ml plasma,and the increase in the thrombin clotting time was

3.2 seconds. All these changes were statisticallysignificant. Moreover the table shows that thesechanges were persistent, and testing of all valuesobtained during the 1-hour observation periodagainst the control determinations revealed thatthe cirrhotic group developed, in response to the

stimulus, highly significant decreases of plasmino-gen and fibrinogen with a similarly significantincrease of the thrombin clotting time.

Plasma inhibitors. In vivo, the biological ac-

tivities of the plasminogen-plasmin system are inpart controlled by and dependent upon the ratioof plasma inhibitors to active enzymatic moieties.Antiplasmin concentration was assayed in theplasma of five cirrhotic patients and in 57 normalsubjects. Antiplasmin assays, expressed as per-

centage inhibition of a standard plasmin solution,were 68% (range, 55 to 80%) for 0.1 ml cir-rhotic plasma and 69.1%o (range, 57 to 84%) fornormal plasma, whereas for 0.05 ml cirrhotic plasmathe assays were 42% (range, 25 to 55%o) and41.6% (range, 16 to 66%) for normal plasma;neither of these differences was statistically sig-nificant, (p > 0.1). Thus, despite the abnormalplasma protein patterns of the cirrhotic group

(Table I), plasma antiplasmin in cirrhotic pa-

tients was within the normal range.

Plasma activator inhibitor was assayed in nor-

mal and cirrhotic plasma by determining plasmainhibitory action on the plasminogen activator,urokinase. Figure 4 illustrates the mean lysistimes and their standard deviations for four cir-

TABLE 1I

Plasma plasminogen, fibrinogen, and thrombin clotting time determinations in normaland cirrhotic patients receiving nicotinic acid at zero time*

p Values,normal versus

Significance of Significance of cirrhotic sub-Normal deviation from Cirrhotic deviation from jects (zero

Assay Time after injection subjects control values patients control values time)

min p pPlasminogen, casein 0 2.71 4 0.72 1.86 4 0.45 0.02 0.1 0.94 ± 0.17 0.001 0.1 1.00 ± 0.27 0.001 0.1 1.21 ±t0.40 0.02 0.1 1.05 ± 0.32 <0.00160 values combined

Fibrinogen, ng/tml 0 3.18 4 0.31 2.58 4 0.67 0.02 0.1 2.05 4 0.36 >0.160 3.18 4 0.39 >0.1 2.30 I 0.55 >0.115, 30, and 2.94 4 0.40 >0.1 2.03 ± 0.45 0.02 0.1clotting time, 15 10.7 ± 1.15 >0.1 13.9 ± 1.8 0.001 <p<0.01seconds 30 11.09 ± 1.33 >0.1 14.1 ± 1.6 0.001 <p<0.01

60 10.9 ± 1.44 >0.1 12.3 i: 1.7 0.02 <p<0.0515, 30, and 10.9 4 1.22 0.05 <p <0.1 13.4 4 1.8 0.001 <p <0.0160 values combined

* Means and standard deviations are given for group determinations at 15, 30, and 60 minutes and for the combined readings obtained duringthe observation period (15, 30, and 60 minutes totaled). Significance values for each patient group were calculated by the method of paired com-parison with zero time values as control figures. The right-hand column shows significance for the differences between groups at zero time.

686


(n .08

z2 .06

F .04UnFn

i .02N.z

0 90 180 240 360UROKINASE UNITS/ML PLASMA

FIG. 4. MEANAND RANGEOF UROKINASE LYSIS TIME

DETERMINATIONS ON PLASMA FRO-M FIVE NORMAL AND

FOUR CIRRHOTIC PATIENTS. The ordinate is reciprocalof lysis time, and the abscissa, urokinase units per mil-liliter plasma. Significant differences between the twosets of plasmas were not detected (p > 0.1).

rhotic and five normal sera; fibrinogen was addedand the mixture clotted by thrombin after the ad-dition of four different concentrations of urokin-ase. For each individual urokinase concentrationthe cirrhotic plasmas exhibited somewhat slowerlysis times than did the normal plasmas, althoughthe differences were not statistically significant(p> 0.1). The absence of a statistically signi-ficant difference between the two sets of deter-minations indicates that the abnormal response

of the cirrhotic to nicotinic acid and electroshockcould not be explained by a significant lack ofactivator inhibitor.

Studies in patients with he patic disorders otherthan cirrhosis. Three patients, all alcoholics, inwhom the diagnosis of hepatic cirrhosis had beensuspected on clinical and laboratory grounds butnot confirmed on biopsy, were also studied afterthe injection of 100 mg nicotinic acid. In thesecases, hepatic biopsy, repeated in two patients,showed only fatty infiltration with mild fibrosisand no evidence of regenerative phenomena.

These patients responded to nicotinic acid in a

manner indistinguishable from that of the normalsubjects and unlike that of the cirrhotic group.

Determinations for plasma activator, plasma plas-minogen, and plasma fibrinogen and the change inthrombin clotting time all lay within the range ofthe normal, or the low normal, response.

Four patients with carcinoma of the liver (onewith primary hepatoma and three with metastatic

hepatic involvement) were studied after nicotinicacid injection. In this patient group there was avirtual failure of plasma thrombolytic activity toincrease in response to nicotinic acid injection.In striking contrast to the behavior of both nor-mal and cirrhotic groups, only one of the serialspecimen euglobulin lysis tests (totaling 16)showed any alteration from control values andthat only to 0.82 U. Test by the chi square pro-cedure (i.e., detectable versus no detectable al-teration of serial sample from control samplevalue) revealed that this failure of the neoplasticgroup to respond to nicotinic acid injection washighly significant when compared with the be-havior of the normal group (p < 0.001). Thetechnical validity of this finding was confirmed bythe results of the isotopic assay method run ontwo lots of these serial samples; this method alsoshowed no deviation from base-line values with

Uf)U1)

0

-j

_

IN VIVO

I \I'\

15 -,1

,05'10,_ I 0\

TIM \

Ho 60

TI ME OF SAMPLING

15

10l

5

IN VITRO

o0 0

"\NO

N

0N

1%0

N-

I'l

A

A\A''\ "~-.' 'a

---

0w0 30 60 90

INCUBATION TIME

FIG. 5. RESPONSEOF PLASMA THROMBOLYTICACTIVITY

TO THE INTRAVENOUSINJECTION OF NICOTINIC ACID (left-hand section) AND THE in vitro ACTIVITY DECAY CURVES,

AT 370 C, OF THESE PLASMA SPECIMENS (right-handsection). The ordinates are plasma thrombolytic ac-

tivity; the left-hand abscissa, time after nicotinic acidinj ection; and the right-hand abscissa, time of plasmaspecimen incubation, in vitro, at 370 C. Three patients,their specimens marked, respectively, A and B, C and D,and E and F, were studied (left-hand section), and theactivator decay curve, in vitro, of each similarly labeledplasma sample is shown in the right-hand section.

* Normal.0 Cirrhotic

I I I I-~~~~/HZ"//

/ - I-

687

I

I


-_

a.~~~~~~~~~~~~~~

D~~~ ~ ~ ~ ~ ~ ~ -0

0

(I)~~~~~~~~~~~~~~~~~~~~~3

10-

o 50-

30 0 30 60 90 120TIME- MINUTES

FIG. 6. THE DATA FROMFIGURE 5 IN EXPONENTIAL FORM. The ordinateis log activator concentration, and the abscissas are time. On the left, peakin vivo activity is defined as zero time, and on the right, zero time marksthe commencement of in vitro plasma incubation. Both the in vivo disap-pearance rate and the in vitro activator decay rates at 37° C are an expo-nential function of activator concentration, although the former is muchmore rapid.

these sample sets. Moreover assays for plasmino-gen, on three of the sample sets, failed to revealany alteration in this component as a result ofnicotinic acid injection. One patient, who re-

ceived a subcutaneous dose of 0.5 ml 1/1,000epinephrine also failed to show any plasma throm-bolytic activity response, a highly unusual find-ing (7, 26).

In vivo clearance of plasminogen activator.Figure 5 illustrates the response of plasma throm-bolytic activity (measured by the Il1l radiochemi-cal assay) in three normal subjects, after theintravenous injection of 100 mg nicotinic acid.The left-hand portion of the diagram showsthat, like the normal plasma thrombolytic response

illustrated in Figure 1, plasma thrombolytic ac-

tivity reached its peak at 10 to 15 minutes and

then declined rapidly, so that by 60 minutes ithad fallen to near control or control levels.

The right-hand panel of Figure 5 displays datadesigned to determine the in vitro decay rate ofplasminogen activator in plasma. Plasma samplesfrom the subjects studied in the left-hand panel,drawn during the period of enhanced thrombolyticactivity, were incubated at 370 C for varying timeintervals before reassay for thrombolytic activity.These decay curves, shown in the right-handpanel, are marked to indicate their identity withrespect to the left-hand panel samples, i.e., thedecay curve marked A on the right was derivedfrom plasma sample A shown on the left. Thedecay curves for plasminogen activator in plasmaat 37° C, shown on the right, indicate that decayrates are at least partially a function of plasmino-

688


gen activator concentration, being more rapidwhen it is high and lessening as concentrationfalls. Figure 6, an exponential plot of the datacontained in Figure 5, demonstrates that both thein vivo plasma "clearance" data and the in vitrodecay data for plasminogen activator, incubated inplasma at 370 C, are exponential functions of ac-tivator concentration 2 with respect to time. Con-sequently the data for each of these individualpatient studies may be properly expressed as 50%plasma clearance or 50% plasma decay times.Moreover, the data of Figures 5 and 6 show thatthe in vivo disappearance rate of plasminogen ac-tivator (left-hand sections) is much faster thancan be accounted for on the basis of in vitro

2It has been determined by in vitro investigation thatI'1-labeled clot assay values are linearly related to un-bound plasma plasminogen activator concentration withstreptokinase (12), with urokinase (27), and with plas-minogen activator extracted from tissues (unpublisheddata).

- IN VIVO6-

4

0.6

IC)OA 'aI I

activator decay rates at 37° C (right-hand por-tions).

Plasminogen activator clearance data in normaland cirrhotic patients. Analysis of the data ob-tained from the cirrhotic patients indicated, in linewith the findings from the normal patients dis-cussed above, that both the in vivo clearance ofplasminogen activator and the in vitro decay rateof plasminogen activator in cirrhotic plasma at370 C were exponential with time. Consequentlythe total averaged data for the normal and cir-rhotic patients are shown in Figure 7, plotted inexponential form as log activator concentrationversus time. Each individual point shown inFigure 7 represents the averaged value for allrelevant determinations, and a close fit of thedata is demonstrated. In vivo clearance data areshown on the left of the figure and in vitro decaydata on the right, the interrupted lines represent-ing the I131 isotopic assay determinations and the

TIME (MINUTES)

FIG. 7. TOTAL AVERAGEDDATA FOR ALL in vSvo AND in vitro PLASMAACTIVATOR DETERMINA-TIONS. The figure is similar in design to Figure 6. On the left, all determinations from eachpatient class have been averaged, i.e., CIRRH-Nic.A. represents the mean data for all cirrhoticpatients receiving nicotinic acid; NORMAL-ESTrepresents all normal subjects after elec-troshock. The right-hand section represents the total in vitro decay data plotted similarly.Both ordinates are log activator concentration, and the abscissas are time. The left-handdata demonstrate that for both cirrhotic and normal patients stimulated by either nico-tinic acid injection or electroshock in zivo activator disappearance is an exponential function ofactivator concentration. Similarly, the right-hand data demonstrate that, in vitro, plasmaactivator decay is also exponential for all classes of patient plasma. Solid lines designate thedata obtained with the I' assay, and the broken lines, the euglobulin data.

689


TABLE III

In vivo 50% activator clearance rate for each class of patient and stimulus distinguished by the assaymethod employed and in vitro plasma activator decay data*

Half lite

No. Assay Stimulus Group Range Mean SD p

"'in8 Normal 2-8.5 5.0 1: 2.4

Euglobulin Electroshock <0.0013 Cirrhotic 40-50 45 : 5

In vivo 9 Normal 6-26.5 12.3 :1: 6.4Euglobulin Nicotinic acid <0.001

5 Cirrhotic 30-85 59 4 21

9 Normal 6-22 13.2 i 4.51131 assay Nicotinic acid <0.001

5 Cirrhotic 32-104 46.7 :1: 17

3 Normal 45-105 81 i 13Euglobulin Nicotinic acid >0.1

3 Cirrhotic 52-123 744 14

In vitro 3 Normal 53-95 74 4 8I13s assay Nicotinic acid >0.1

3 Cirrhotic 60-100 82 i: 9

* In all instances differences between cirrhotic patients and normal subjects were statistically highly significant(p < 0.001). In vitro plasma activator decay data did not reveal significant differences between cirrhotic and normalplasma (p > 0.1). Differences between the in vivo data and the in vitro decay data were highly significant (p <0.001)

solid lines those obtained by the euglobulinmethod.

The left-hand portion of Figure 7 demonstratesthat striking differences were evident betweenplasminogen activator clearance from the plasmain normal subjects and in cirrhotic patients; more-over, these differences were of the same order ofmagnitude for both stimuli, nicotinic acid andelectroshock. The right-hand portion of Figure7 reveals that the averaged lines for activator de-cay in plasma incubated at 370 C were approxi-mately parallel, indicating that decay of plasmino-gen activator in vitro proceeded at a rate inde-pendent of the type of plasma in which it wascontained. Individual plots of this exponentialtype were used to calculate the 50% plasma clear-ance for each patient and the 50%7o plasma decayrate for each such study. The results are shownin Table III.

The upper portion of Table III lists the cal-culated in vivo half life of plasminogen activatorobtained from observations employing the euglob-ulin lysis method and the radiochemical assay;the lower portion of Table III cites the calculatedin vitro 50%o plasma decay rates. The resultsobtained with the two different assay methods

demonstrate very satisfactory agreement, bearingin mind that the patient groups studied with eachmethod were overlapping rather than identical.For instance the in vivo 50% plasma clearanceof activator in normal subjects, after nicotinicacid, was 12.3 + 6.4 minutes determined by theeuglobulin lysis method and 13.2 + 4.5 minuteswith the radiochemical assay. The findings inTable III indicate that the plasma clearance ofplasminogen activator in the cirrhotic patient wasgreatly slowed as compared to that of the normalsubject, the 50% plasma clearance time beingapproximately 9 times slower after electroshocktherapy (euglobulin method) and 5 times (euglob-ulin method) or 4 times slower (radiochemicalmethod) after nicotinic acid stimulation. Allthese differences were significant at p < 0.001.In contrast, the results with both assay methodsindicated that in vitro decay rates for plasminogenactivator in plasma at 37° C were similar in eithercirrhotic or normal plasma (p > 0.1).

Discussion

Both the enhanced plasma thrombolytic activity,developing after nicotinic acid injection, electro-shock, and other stimuli (9-11, 26), and the very

690


low level of thrombolytic activity detectable inplasma from resting man (13) have been shownto be due to the presence of plasminogen activatorin plasma. The present work offers interestingconfirmation of this conclusion, for whereas theprevious evidence was necessarily indirect [studyof the action of specific activators in a plasmamilieu (27). the action of such plasma on differentsubstrates (1 1, 12, 27), and the demonstrationof plasminogen activator activity in the euglobu-lin fraction of plasma ( 11 ) ], the present evidenceis direct, resting upon the in vivo demonstration ofplasminogen activation. In both normal and cir-rhotic patients, the injection of nicotinic acid wasshown to induce a fall of plasma plasminogen,although the decrease in the cirrhotic patients wasof much greater magnitude (p < 0.001).

Since most of the conclusions derived duringthis study rest upon the accuracy of plasminogenactivator assay in plasma, some considerationmust be given to the validity of the methodologi-cal approach. Theoretically, the euglobulin lysistime assay method for activator might haveyielded artifactual results in plasma from cir-rhotic patients because of variation in plasmacomponents other than plasminogen activator, i.e.,the euglobulin clot formed from such plasmawould have been of abnormal composition. Al-though differences were detected between thetwo subject groups with respect to the pretreat-ment values for plasma fibrinogen and plasmaplasminogen (two moieties influencing this test),the radiochemical assay would have been unaf-fected by changes in plasma composition, sincea standardized substrate with a defined plasmino-gen and fibrin concentration is employed for thisassay procedure. In fact both assay proceduresyielded comparable results that were excellentlycorrelated (Table III). Calculation from thetotal experimental data yielded (regression ofeuglobulin lysis on I131 lysis) r = 0.61 (p <0.001 ) for normal subjects, the line equationbeing y = 2.85 + 5.09x, and r = 0.87 (p < 0.001)for the cirrhotic patients, the line equation beingv = 0.32 + 5.45x. Since the calculated regressionslopes lay within 10% of each other, the influenceof technical artifact in the use of the euglobulinactivator assay with plasma from cirrhotic pa-tients can be confidently excluded.

The respective responses of normal and cir-rhotic subjects to a stimulus capable of elicitingenhanced plasma activator activity has been pre-viously investigated by Kwaan, MlacFadzean, andCook (7). They studied a disease entity, com-mon in Hong Kong, termed cryptogenic spleno-megaly (28), in which idiopathic splenomegalycommonly antedates the onset of hepatic cirrhosis,the condition not infrequently terminating as pri-mary hepatoma. They demonstrated, usingsingle plasma samples taken 20 minutes after thesubcutaneous injection of epinephrine, exercise,or other suitable stimuli that patients, in whomthe manifestations of hepatic cirrhosis were ofpredominant importance, developed significantlygreater enhancement of plasma thrombolytic ac-tivity than did patients with cryptogenic spleno-megaly without cirrhosis or normal subjects.Moreover in four patients, first observed duringthe cirrhotic phase of the disorder (at which timeenhanced thrombolytic activity occurred in re-sponse to the stimulus), the development of hepa-toma resulted in virtual abolition of the patients'response to the stimulus.

The present studies, while confirming the ob-servation that patients with hepatic carcinomamay have a decreased response to a standardthrombolytic stimulus, were primarily designedto investigate other facets of the relationship be-tween hepatic function and plasma thrombolyticactivity. The striking finding3 that the ini vivo50% plasma clearance in normal subjects afternicotinic acid averaged 12.7 + 5.4 minutes andafter electroshock therapy, 5.0 ± 2.4 minutes,whereas the in vitro 50% decay time for plas-minogen activator in either cirrhotic or normalplasma averaged 78 + 9.2 minutes indicates that,in vivo, plasma clearance mechanisms for acti-vator play, apart from the spontaneous decay ordestruction of activator in plasma, a crucial rolein the regulation of plasma thrombolytic activity.This rapid in vivo clearance, or inactivation ofplasma activator, or both, explain the many dif-

3 References here and subsequently are to the averagedresults from Table III without distinction as to the as-say method employed. This simplification of the data isjustified by the fact that the data obtained by the sepa-rate assay procedures are virtually identical (Table III);consequently, the averaged figures used hereafter, al-though distinct from those of the table, are little different.

691


ficulties encountered during attempts to inducea therapeutic thrombolytic state by the adminis-tration of pharmacological agents that induceactivator release-such states have invariably beentransient and of insignificant degree. Moreoverit has been reported (29) that the 50% in vivoplasma clearance rate of the plasminogen activatorstreptokinase was 83 minutes, a rate so unex-pectedly rapid that reticuloendothelial systemclearance was suggested as its cause.

The present data provide relatively precise evi-dence as to the nature of the functional anomalyresponsible for the strikingly different activatorclearance rates determined for normal and cir-rhotic patients. The finding that both stimuli,nicotinic acid injection and electroshock, inducedsimilar effects excludes the possibility of eitherabnormal nicotinic acid metabolism or aberrantresponse to the electroshock as being the uniquecause of the anomaly. Indeed in conjunction withthe previously cited data concerning the responseof the cirrhotic patient to other stimuli (7), theresults indicate that cirrhotic patients respond inan exaggerated fashion to all stimuli capable ofinducing thrombolytic activity in the normal sub-ject. This abnormal response of the cirrhoticpatient is not related to deficient plasma inhibitorycharacteristics as a) assays for plasma antiplasminrevealed no significant differences between thenormal and cirrhotic groups, and determinationof the urokinase lysis time showed only a non-significant decrease of inhibitory activity in cir-rhotic plasma (Figure 4), and significantly b)plasminogen activator decay rates in plasma fromthe two different classes of subjects were identical(Table III).

Consequently the finding of a 4-fold differencebetween the 50% plasma activator clearance ratesfor normal and cirrhotic patients after nicotinicacid stimulation (12.7 + 5.4 versus 53 + 19 min-utes) and a 9-fold difference between these groupsafter electroshock therapy (5.0 + 2.4 versus 45+ 5.0 minutes) assumes crucial etiological sig-

nificance. Whereas the 50% in vivo plasma clear-ance of activator in the normal subject was 7 to 16times as rapid as the 50% in vitro plasma decayrate (i.e., 78 + 9 minutes), the corresponding invivo plasma clearance rate of activator in thecirrhotic patient was less than twice the in vitrodecay rate of plasminogen activator in plasma

(Table III). Thus while the normal subjectclears plasminogen activator from the plasmawith remarkable speed, this ability is almost ab-sent in the cirrhotic patient. Since there is noreason to believe that hepatic cirrhosis inducesextraneous changes in intravascular-extravascularequilibria that could retard loss of activator fromthe plasma, the results suggest that the almosttotal failure of the cirrhotic subject to activelyclear this moiety from the circulation is a directconsequence of the structural or functional hepaticchanges consequent upon cirrhosis. The normalliver must play an important role in the activeclearance of plasminogen activator from the circu-lation and, in this manner, exert important regu-latory functions over physiological and pathologi-cal activity of the plasminogen-plasmin system.

This hypothesis is supported by a previous re-port (7) that hepatic vein blood exhibited asomewhat lower level of thrombolytic activitythan did that from the inferior vena cava andby the significant observations of Grossi, Rous-selot, and Panke (5) made on the portal andsystemic vein blood of cirrhotic patients under-going portacaval shunt procedures. They ob-served, using the euglobulin lysis time assay, thatalthough the thrombolytic activity of portal andsystemic venous blood was similar in 18 patients,in 15 others (mostly those patients showing agrossly shortened euglobulin lysis time) activatoractivity in the portal vein blood exceeded, some-times substantially, that found in the systemicvein blood. All the experimental evidence, in-cluding that of the present study, indicating thatthe liver actually clears plasminogen activatorfrom the plasma, has been obtained during statesof enhanced plasma thrombolytic activity. Never-theless there is little reason to doubt that hepaticclearance of plasminogen activator is of signifi-cance under strictly physiological circumstances,for many of the clearance studies performed dur-ing the present investigation were made duringstates of only mild enhancement of plasma throm-bolytic activity, and the evidence in the cirrhoticpatient, in whomhepatic clearance may be grosslydeficient, indicates that virtual absence of thishepatic function predisposes to the developmentof pathological plasma proteolysis.

The precise functional defect responsible forfailure of the cirrhotic liver to clear plasminogen

692


activator must, at present, remain conjectural,although the evidence is sufficient to permit use-ful speculation. Ratnoff's evidence (2) thatrapid whole plasma clot lysis, commonly detectedin the cirrhotic patient, was seldom if ever ob-served in patients suffering from hepatitis (evenof such severity to cause deep jaundice, coma, andultimately death), and our evidence that, in threepatients with hepatic failure but without biopsyevidence of cirrhosis, there was no apparent fail-ure to clear plasminogen activator, would suggestthat hepatic cellular injury or death per se wasnot the responsible cause. Nevertheless the ob-servation that four patients, with carcinomatousinvolvement of the liver but without marked signsof hepatic failure, failed to exhibit enhancedplasma thrombolytic activity after nicotinic acidinjection (p < 0.001) coupled with the interest-ing work of Kwaan, Lo, and McFadzean (30),suggesting that hepatic neoplasm might producean inhibitor of plasminogen activation, indicatesthat multiple factors complicate interpretation ofthis problem. Moreover there is the possibilitythat blood shunt formation, secondary to cirrhosis,might interfere with hepatic clearance of acti-vator. However, shunt formation is unlikely toexplain completely the findings, as recent inves-tigation (31) of extrahepatic shunt flow in pa-tients with moderately advanced or advancedhepatic cirrhosis indicated that it averaged only14% of total hepatic blood flow, although the cal-culated intrahepatic shunt flow was somewhathigher. Tentatively we suggest that clearance ofplasminogen activator is probably a function ofhepatic cell mass and that reduction of this func-tion may occur during various disease states, itseffects being particularly evident as a result ofhepatic cirrhosis. Such hepatic functional defectmight be masked in vivo by a capacity of hepaticcells, as a result of injury or neoplastic involve-ment, to produce an inhibitor, probably non-specific, to plasminogen activation.

Patients suffering from hepatic cirrhosis andother chronic forms of nonobstructive hepaticdisease frequently exhibit a coagulation disorderof usually ill-defined etiology, and the present ob-servations may bear upon its pathogenesis. De-spite the attention of many investigators (3, 5,8, 32-35), these disorders are relatively poorlyunderstood. The patients usually display throm-

bocytopenia of mild or moderate degree, and co-agulation assays usually indicate the presence ofa multifactorial defect involving chiefly Factors II(prothrombin), V (accelerator globulin), VII(proconvertin), IX (Christmas factor), and pos-sibly other thromboplastic factors. Conventionalcoagulation assays tend to correlate poorly withhemorrhagic phenomena and sometimes fail toprovide an adequate explanation for the clinicalfindings.

Although many investigators (1-8, 33-35)have reported the presence of rapid whole bloodclot lysis, an accelerated euglobulin lysis time, orother evidence of abnormal plasminogen-plasminsystem activity in from 20 to 60% of patientswith advanced cirrhosis, these phenomena werenot considered relevant to the hemorrhagic dia-thesis (except in rare instances where they wereextreme in nature), because plasma fibrinogenlevels usually remained within the low normalrange. The recent description of a unique co-agulation defect, termed defective fibrin poly-merization (15-18), in which the polymerizationof fibrin monomer is inhibited through the pres-ence in plasma of fibrinogen proteolysis productssuggests that the high incidence of abnormalfibrinolytic activity in cirrhotic patients may playan important role in the genesis of the coagulationabnormalities and accompanying hemorrhagicdiathesis.

In this connection the present demonstration ofthe exquisite sensitivity of cirrhotic patients tostimuli, capable of inducing release of activatorin the normal subject, assumes unusual impor-tance. Whereas the injection of nicotinic acid inthe normal subject caused only enhanced plasmathrombolytic activity of transient degree withoutother significant findings, the same procedure inthe cirrhotic patient caused greatly increasedplasma activator activity of prolonged degree, suf-ficient to induce a substantial fall in plasma plas-minogen (p < 0.001), a substantial fall of plasmafibrinogen (0.001 < p < 0.01), and a significantincrease in plasma thrombin clotting time (0.001< p < 0.01). These findings in the cirrhotic pa-tients, although in this instance clinically inconse-quential, differ only in degree from those ob-served in patients suffering from defective fibrinpolymerization of clinical significance. IndeedRatnoff (36) suggested that the raised plasma

693


thrombin clotting time, commonly found in thecirrhotic patient, might reflect a defect in thecoagulation of fibrinogen by thrombin, and others(37-38) have reported a depression of plasmaplasminogen and fibrinogen in the cirrhotic pa-

tient of comparable degree to our own findings.The present data suggest that the coagulation

disorder, defective fibrin polymerization, result-ing from the failure of the cirrhotic patient to

clear plasminogen activator, may play a hithertounsuspected role in the coagulation disorders thatcomplicate this condition. Confirmation of thishypothesis would have important therapeutic im-plications, since treatment with specific inhibitorsmight prove advantageous (14).

Summary

1) Enhanced plasma thrombolytic activity ofshort duration developed in control subjects re-

ceiving 100 mg nicotinic acid intravenously, butpatients with hepatic cirrhosis exhibited a grosslyexaggerated response with greatly enhanced peakactivity of much prolonged duration. Similardifferences were observed after therapeutic elec-troshock in normal and cirrhotic patients. Nico-tinic acid injection induced in cirrhotic, but notin normal subjects, significant decreases in plasmaplasminogen and fibrinogen and a rise in thethrombin clotting time, although plasma inhibi-tors did not differ significantly between the two

groups.

2) The in vivo 50% plasma clearance rate forplasminogen activator in normal subjects afternicotinic acid injection was 13 5 minutes com-

pared to 53 + 19 minutes in the cirrhotics. Dif-ferences were not detected in the in vitro decayrates of plasminogen activator in normal and cir-rhotic plasmas, the mean 50% time being 78 9minutes for both. Essentially similar findingswere obtained with normal and cirrhotic patientswho received therapeutic electroshock.

3) Abnormal fibrinolysis in the cirrhotic pa-

tient may b)e due to a failure of hepatic clearancemechanisms for plasminogen activator. The evi-dence also suggested that the normal liver exerts

an important control function over plasminogen-plasmin system activity.

Acknowledgments

We are greatly indebted to the staff of the JewishHospital, particularly to Dr. Albert Eisenstein, for theirinterest and for the opportunity to study patients withcirrhosis and a normal population. Dr. Walter Mooreat St. Mary's Hospital and Dr. Franz Hornung at theMalcolm Bliss Hospital kindly provided facilities for thestudy of patients undergoing electroshock therapy.

References

1. Goodpasture, E. W. Fibrinolysis in chronic hepaticinsufficiency. Bull. Johns Hopk. Hosp. 1914, 25,330.

2. Ratnoff, 0. D. Studies on a proteolytic enzyme inhuman plasma IV. The rate of lysis of plasmaclots in normal and diseased individuals, withparticular reference to hepatic disease. Bull.Johns Hopk. Hosp. 1949, 84, 29.

3. Finkbiner, R. B., J. J. McGovern, R. Goldstein, andJ. P. Bunker. Coagulation defects in liver dis-ease and response to transfusion during surgery.Amer. J. Med. 1959, 26, 199.

4. Grossi, C. E., A. H. Moreno, and L. M. Rousselot.Studies on spontaneous fibrinolytic activity in pa-tients with cirrhosis of the liver and its inhibitionby epsilon amino caproic acid. Ann. Surg. 1961,153, 383.

5. Grossi, C. E., L. M. Rousselot, and W. F. Panke.Coagulation defects in patients with cirrhosis ofthe liver undergoing portasystemic shunts. Amer.J. Surg. 1962, 104, 512.

6. Kwaan, H. C., A. J. S. McFadzean, and J. Cook.Plasma fibrinolytic activity in cirrhosis of theliver. Lancet 1956, 1, 132.

7. Kwaan, H. C., A. J. S. McFadzean, and J. Cook.On plasma fibrinolytic activity in cryptogenicsplenomegaly. Scot. med. J. 1957, 2, 137.

8. Bergstr6m, K., B. Blombdck, and G. Kleen. Studieson the plasma fibrinolytic activity in a case of livercirrhosis. Acta med. scand. 1960, 168, 291.

9. Weiner, M., K. de Crinis, W. Redisch, and J. M.Steele. Influence of some vasoactive drugs onfibrinolytic activity. Circulation 1959, 19, 845.

10. Weiner, M. The fibrinolytic response to nicotinicacid in abnormal liver states. Amer. J. med. Sci.1963, 246, 294.

11. Sherry, S., R. I. Lindemeyer, A. P. Fletcher, andN. Alkjaersig. Studies on enhanced fibrinolyticactivity in man. J. clin. Invest. 1959, 38, 810.

12. Alkjaersig, N., A. P. Fletcher, and S. Sherry. Themechanism of clot dissolution by plasmin. J. clin.Invest. 1959, 38, 1086.

13. Sawyer, W. D., A. P. Fletcher, N. Alkjaersig, andS. Sherry. Studies on the thrombolytic activityof human plasma. J. clin. Invest. 1960, 39, 426.

14. Fletcher, A. P., N. Alkjaersig, and S. Sherry. Fi-brinolytic mechanisms and the development of

694


thrombolytic therapy. Amer. J. Med. 1962, 33,738.

15. Fletcher, A. P., N. Alkjaersig, and S. Sherry.Pathogenesis of the coagulation defect developingduring pathological plasma proteolytic ("fibrino-lytic") states. I. The significance of fibrinogenproteolysis and circulating fibrinogen breakdownproducts. J. clin. Invest. 1962, 41, 896.

16. Alkjaersig, N., A. P. Fletcher, and S. Sherry.Pathogenesis of the coagulation defect developingduring pathological plasma proteolytic ("fibrino-lytic") states. II. The significance, mechanismand consequences of defective fibrin polymeriza-tion. J. clin. Invest. 1962, 41, 917.

17. Bang, N. U., A. P. Fletcher, N. Alkjaersig, and S.Sherry. Pathogenesis of the coagulation defectdeveloping during pathological plasma proteolytic("fibrinolytic") states. III. Demonstration ofabnormal clot structure by electron microscopy.J. clin. Invest. 1962, 41, 935.

18. Latallo, Z. S., A. P. Fletcher, N. Alkjaersig, and S.Sherry. Inhibition of fibrin polymerization byfibrinogen proteolysis products. Amer. J. Physiol.1962, 202, 681.

19. Niewiarowski, S., and E. Kowalski. Un nouvel anti-coagulant derive du fibrinogene. Rev. Hemat.1958, 13, 320.

20. Triantaphyllopoulos, D. C. Effects of intravenousinj ections of the anticoagulant fraction of incu-bated fibrinogen on blood coagulation. Canad. J.Biochem. 1960, 38, 909.

21. Milstone, H. A factor in normal human blood whichparticipates in streptococcal fibrinolysis. J. Im-munol. 1941, 42, 109.

22. Fletcher, A. P., N. Alkjaersig, and S. Sherry. Themaintenance of a sustained thrombolytic state inman. I. Induction and effects. J. clin. Invest.1959, 38, 1096.

23. Ostle, B. Statistics in Research. Ames, Iowa, IowaState College Press, 1954.

24. Sherry, S., A. P. Fletcher, and N. Alkjaersig. Fi-brinolysis and fibrinolytic activity in man. Physiol.Rev. 1959, 39, 343.

25. Fischer, S., A. P. Fletcher, N. Alkjaersig, and S.Sherry. Fibrinogenolysis in vivo: identification,occurrence, and characterization in man (abstract).J. clin. Invest. 1963, 42, 931.

26. Genton, E., F. Kern, and K. von Kaulla. Fibrinoly-sis induced by pressor amines. Amer. J. Med.1961, 31, 564.

27. Sawyer, W. D., N. Alkjaersig, A. P. Fletcher, andS. Sherry. A comparison of the fibrinolytic andfibrinogenolytic effects of plasminogen activatorsand proteolytic enzymes in plasma. Thrombos.Diathes. haemorrh. (Stuttg.) 1960, 5, 149.

28. McIntosh, J. F. Cryptogenic splenomegaly (Banti'sdisease). China med. J. 1932, 46. 992.

29. Fletcher, A. P., N. Alkjaersig, and S. Sherry. Theclearance of heterologous protein from the circu-lation of normal and immunized man. J. clin. In-vest. 1958, 37, 1306.

30. Kwaan, H. C., R. Lo, and A. J. S. 'McFadzean. Anti-fibrinolytic activity in primary carcinoma of theliver. Clin. Sci. 1959, 18, 251.

31. Nakamura, T., S. Nakamura, T. Kaneko, T. Suzuki,K. Tokita, and S. Abe. Measurement of extra-hepatic shunted blood flow in liver cirrhosis. J.Lab. clin. Med. 1962, 60, 889.

32. Rapaport, S. I., S. B. Ames, S. Mikkelsen, and J. R.Goodman. Plasma clotting factors in chronichepatocellular disease. New Engl. J. Med. 1960,263, 278.

33. Zucker, M. B., M. Siegel, E. E. Cliffton, J. W. Bell-ville, W. S. Howland, and C. E. Grossi. Theeffect of hepatic lobectomy on some blood clottingfactors and on fibrinolysis. Ann. Surg. 1957, 146,772.

34. Zucker, M. B., M. Siegel, E. E. Cliffton, J. WV. Bell-ville, W. S. Howland, and C. E. Grossi. General-ized excessive oozing in patients undergoing ma-

jor surgery and receiving multiple blood trans-fusions. J. Lab. clin. Med. 1957, 50, 849.

35. Zetterqvist, E., and I. von Francken. Coagulationdisturbances with manifest bleeding in extra-hepatic portal hypertension and liver cirrhosis.Acta med. scand. 1963, 173, 753.

36. Ratnoff, 0. D. An accelerating property of plasmafor coagulation of fibrinogen by thrombin. J.clin. Invest. 1954, 33, 1175.

37. Blix, S. The proactivator of the fibrinolytic sys-

tem in human plasma. Acta med. scand. 1962, 171,83.

38. Purcell, G., and L. L. Phillips. Fibrinolytic activityin cirrhosis of the liver. Surg. Gynec. Obstet.1963, 117, 139.

695

Documents

Abnormal Plasminogen-Plasmin System Activity (Fibrinoly ... · Journal ofClinical Investigation Vol. 43,oNo. 4, 1964 AbnormalPlasminogen-Plasmin System Activity (Fibrinoly-sis) in