thrombosis - Journal of Clinical Pathology · thrombosis commenced could be gauged with reasonable accuracy; and some estimate of the probable extent of thrombosis could be obtained

J. clin. Path., 1972, 25, 36-44

The significance of variations in immunoreactiveand clottable fibrinogen in health and followingthrombosisP. WOLF', G. W. FARRELL, AND K. W. WALTON

From the Department of Experimental Pathology, University ofBirmingham

SYNOPSIS In individuals in a steady state of fibrinogen metabolism, immunoreactive and clottablefibrinogen estimates of plasma fibrinogen show close agreement. These estimates also detect anyincrease of plasma fibrinogen (due to increased synthesis) following metabolic stresses but in certaincircumstances discrepancies between immunoreactive and clottable fibrinogen values occur which are

of diagnostic assistance.During extensive thrombosis, circulating 'cryoprofibrin' (fibrin intermediates) may be formed.

These fail to give full quantitative immunodiffusion reactions with antifibrinogen. Values forimmunoreactive are therefore lower than for clottable fibrinogen.When intravascular catabolism (due to plasmin action) accompanies increasedsynthesis, since some

of the molecular breakdown products of fibrin or fibrinogen react with antifibrinogen but are notclottable, immunoreactive values exceed those for clottable fibrinogen. This discrepancy is thereforean indicator of thrombolysis.Each discrepancy in turn may be encountered during the alternating predominance of thrombosis

or thrombolysis in vivo: (a) physiologically, in association with the menstrual cycle; (b) pathologically,following surgical operations or extensive intravascular thrombosis.

Fibrinogen is a rapidly metabolized protein with arelatively short biological half-life of about three tosix days (MacFarlane, Todd, and Cromwell, 1964;Takeda, 1966). Like other 'acute phase reactant'proteins, fibrinogen is synthesized rapidly in responseto trauma, burns, acute infections, and conditionsassociated with haemorrhage and tissue breakdown.As a consequence, the plasma concentration shows arapid rise. From the results of metabolic studies ithas been suggested that the stimulus to increasedsynthesis of fibrinogen may be increased catabolismof the molecule (Reeve, Takeda, and Atencio, 1966).Depending upon the clinical circumstances, differentenzymes may be involved in the intravascularbreakdown of fibrinogen either before, during, orafter its conversion to fibrin.For example, during blood coagulation, thrombin

produces a specific partial proteolysis of the fibrino-gen molecule resulting in the splitting off of two pairs'Member of external scientific research staff, Medical ResearchCouncil, on attachment to the Department of Experimental Path-ology, University of Birmingham.

Received for publication 18 July 1971

of peptide units (Bailey and Bettelheim, 1955) nowknown as fibrinopeptides A and B (Blomback,1967). Following their detachment, the residualportions of the molecule polymerize to forminsoluble fibrin (Fig. IA). Fibrinopeptides A and Bevidently do not contribute substantially to theantigenicity of the original molecule since antiseradirected against fibrin are indistinguishable fromantisera to fibrinogen.On the other hand, when either fibrin or fibrinogen

is exposed to plasmin action (Fig. IB) proteolysis ismore extensive and a large number and variety ofpeptide fragments is released. The end products ofcomplete lysis of fibrinogen, although incoagulableby thrombin (Lipinski, Wegrzynowicz, Budzynski,Kope6, Latallo, and Kowalski, 1967), react withantifibrinogen antibody. Certain of the fragmentsgive precipitin lines in agar diffusion reactions (theD and E fragments of Nussenzweig, Seligman, andGrabar, 1967) while others behave as non-precipitat-ing antigens. Some of the breakdown products offibrin are similarly identifiable by immunologicalmethods.

36

on April 8, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.25.1.36 on 1 January 1972. Dow

nloaded from

http://jcp.bmj.com/

Significance of variations in immunoreactive and clottable fibrinogen in health andfollowing thrombosis 37

Antigen

Fibrinogen

A.e.

Fibrinogen

Fibrin

Enzyme

'Thrombia."(low cOmnbet'cnetation)

Plasmin

Plasmin

Reaction of digestion prod

Fibrin(precipitin*)

Cryoprofibrin (precipitint)

fibrinogen degradation proc

pE) (pn + (precipitins n

pepprecipitins n

luct with antibody

fibrinopeptidesfibrinopeptidsFig. I Schematic representation

+ 0 0 of immunological reactivity ofproducts ofproteolysis offibrinogen andfibrin. *Antigenic

+ ® reactivity offibrin detected byforming clot in plasma containing1% thiomersalate (to blockinteraction with factor XIII)and redissolvingfibrin in 5Mducts (Fon. D. P.) urea. f Quantitative immuno-

D Q Q precipitation of cryoprofibrinon precipitin interfered with by cryoprecipitation

and altered diffusioncharacteristics due to

D O 0 polymerization (see text).ion precipitin

fibrin degradation products (F. D. P.)

The detection of these fragments in serum has beenadvocated as a test for the occurrence of fibrinolysis(or fibrinogenolysis) in various clinical conditions(Merskey, Kleiner, and Johnson, 1966). However,this test, in itself, does not establish whether thefragments were circulating in the plasma or weresubsequently formed in vitro as a result of acceleratedclot lysis in the test tube.As an alternative approach, it has been found that

parallel quantitative comparisons of immuno-reactive and clottable fibrinogen provide new infor-mation about the biological behaviour of fibrinogenand of these degradation products, while discrepan-cies between these two kinds of assay may be ofpractical significance in the assessment of thrombosisand thrombolysis in clinical practice, when theseprocesses are substantial in degree but difficultotherwise to detect and appraise.

Materials and Methods

DETERMINATION OF CLOTTABLE FIBRINOGENThis was carried out by a colorimetric modificationof the method of Fearnley and Chakrabarti (1966) asdescribed by Farrell and Wolf (1971).

DETERMINATION OF IMMUNOREACTIVEFIBRINOGENIn normal plasma this was performed by the methodsof Wolf and Walton (1965). For measurement offibrinogen suspected to be in abnormal polymerizedform this method was modified as follows: the sodiumchloride concentration in the agar and diluting fluidwas increased from 0-15 M to 3 M (17% sodiumchloride) but the concentration of dipotassium3*

sequestrene and the pH of these media were un-altered.

PREPARATION OF ANTIHUMAN FIBRINOGENANTISERUMThis was carried out by immunizing rabbits asdescribed by Farrell and Wolf (1970).

MEASUREMENT OF DIFFUSIONCHARACTERISTICS OF FIBRINOGENA modification of the method of Allison andHumphrey (1960) as applied to measuring thediffusion of human plasminogen (Farrell, Stuckey,and Wolf, 1969) was used, substituting rabbitantifibrin for rabbit antiplasmin.

HUMAN THROMBINHuman thrombin was prepared by the method ofBiggs and Macfarlane (1957). The thrombin solutionwas freeze-dried in 0-2 ml aliquots and reconstitutedbefore use to give a solution which clotted an equalvolume of fresh citrated plasma in 10 to 13 seconds.

STREPTOKINASEThis was obtained from Kabi Biochemicals Ltd.Dried streptokinase (marked 'for laboratory useonly') was reconstituted with distilled water to give aconcentration of 200 units/ml. The solutions werestored in 0 5 ml quantities at -30° until required.The solutions were thawed, used immediately, andany unused residue was discarded.

DIGESTION OF FRESH PLASMA CLOTSTo 0-3 ml plasma, in a water bath at 370, was added0-1 ml of 015 M sodium chloride solution and



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

P. Wolf, G. W. Farrell, and K. W. Walton

0-1 ml of streptokinase solution (containing 200units/ml). After exactly one minute 01 ml ofthrombin solution was added. A clot formed whichdissolved within 60 to 120 seconds. The digest wasincubated for a further 10 minutes before serialdilutions were made for immunological determina-tions.

BLOOD COLLECTIONBlood was collected in trisodium citrate or dipotas-sium sequestrene as detailed elsewhere (Farrell andWolf, 1971). In some instances (see text) epsilonamino caproic acid (EACA) was also included togive a final concentration of 2-6 g/100 ml (0-2 M).

DETECTION OF FIBRIN OR FIBRINOGENDEGRADATION PRODUCTSThis was carried out by immunodiffusion and byimmunoelectrophoresis of serum, collected with andwithout the addition of EACA at the concentrationdetailed above, using antifibrinogen antiserum.

500

450 00

400 . . . . ----0

3250 0 t .. -0~~~~~~~~~~~~~~~~~~~0 0~~~~~~~~~~~~~~~~~~.0~~~~~~~~~~~~~~~~~

0.04..00 * 0

0

o~~~~~~~~~~~~~s0 *00o~~~~~~~~~

0*

,-.350 *0

Results

IMMUNOREACTIVE AND CLOTTABLEFIBRINOGEN IN HEALTHY SUBJECTSWhen immunoreactive fibrinogen levels, as deter-mined by the method of Wolf and Walton (1965),were compared with clottable fibrinogen levels,values lying between 250 and 450 mg/100 ml wereobtained in 23 healthy ambulant subjects (11 malesand 12 females). In all but two subjects (both femaleand referred to in more detail below) the concordancebetween the two methods was acceptable when allow-ance was made for the 2+ 1 dilution system adoptedfor the immunological estimate (Fig. 2).

In two females yielding discrepant results (seepoints indicated by arrows in Fig. 2) the immuno-logical estimate exceeded the clottable value by awide margin. These results were associated with otherfindings suggesting that these subjects differed fromthe remainder in showing evidence of proteolysis offibrinogen both in vitro and in vivo. For example, in

0* -- 88 Fig. 2 Comparison of

4, /measurements of_I / immunoreactive and

0 clottablefibrinogen in23 healthy ambulant

1w / subjects (closedsymbols males; open

_I / symbols females).00.. Regression (continuous

* *.. line) and 95%confidence limits (dottedlines) compared with

.. theoretical slope00. a (dashed line). Two

results marked byarrows in top left-hand

. corner discarded in..0 calculating correlation

coefficient, r: 0 752,P >0 001 (see text).

200 250 300 350 400 450

Clottable fibrinogen (mg/I00hl)

38



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


both subjects (though not in the remainder) 'fibrinfragments' were demonstrated immunologically inthe serum of blood collected in EACA to preventsubsequent fibrinolysis (see Methods) suggesting thepresence of circulating fibrin or fibrinogen degrada-tion products. In addition both subjects showedprogressive decrease of fibrin yield from citratedblood (collected without EACA present) afterincubation of the blood at 370 for one and two hours,demonstrating accelerated clot lysis in the test tube.When the results from these two subjects were

discarded, a highly significant correlation co-efficient (r :0752; p>0O001) was obtained with aslope not significantly different from the theoreticalone (Fig. 2). Statistical comparison of means byStudent's t test showed no significant differencebetween the estimates by the two techniques foreither males or females (Table). Similarly, nosignificant difference was found between the overallmean values between the sexes although the standarddeviation was larger for females. This is also seenfrom Fig. 2 where a wider scatter of individual valuesfor females is evident.

Category No. Measure- Mean ±SE ofMean Comparison ofment (mg/100 ml) Means

t df p

Males 11 IF 3007 ± 90 0-12 9 <0-1CF 298 8 ± 160

Females 10 IF 3190 ±253 0-11 8 <0-1CF 315-3 i 217

Total 21 IF 309-4 ± 137 015 19 <0-1CF 306-7 123

Table Immunoreactive (IF) and clottable fibrinogen(CF) in healthy subjects''Overall means + SE: males 299-8 ± 10-8; females 317-6 ± 10-7; total:308-0 ± 10-7

The occurrence of markedly discrepant results intwo females and the general tendency for results infemales to show a wider scatter suggested a greaterlability of plasma fibrinogen concentration in womenwhich might be physiological in origin. Sequentialexamination of immunoreactive and clottablefibrinogen levels in the two subjects showing thediscrepant results and in three additional women wasundertaken at various times in relation to the men-strual cycle to investigate this possibility. Similardiscrepancies between immunoreactive and clottablefibrinogen levels were again observed in the twosubjects in whom these were originally noted. But thediscrepancy varied in its magnitude, being maximalin the premenstrual phase when clottable fibrinogenvalues were often minimal while immunoreactivefibrinogen values rose to a peak (Fig. 3).

Similar fluctuations in immunoreactive andclottable fibrinogen levels were observed in the threeother female subjects tested but were variable indegree between different women and even in thesame woman between different cycles. Values forimmunoreactive tended to exceed those for clottablefibrinogen, particularly markedly when a minorinfective process (eg, a tooth abscess, as shown inFig. 3) coincided with the premenstrual phase of thecycle. On the other hand, little or no fluctuation inimmunoreactive or clottable fibrinogen levels wasobserved by sequential tests in healthy males exceptwhere tests were carried out during coincident minorinfections (Fig. 3) when immunoreactive levelssometimes showed a tendency to exceed clottablefibrinogen values.The most marked elevation of immunoreactive

values above clottable fibrinogen values coincidedwith the appearance of 'fibrin fragments' in the serumin the premenstrual phase. It was concluded that thediscrepancy between immunoreactive and clottablefibrinogen values probably therefore arose becausethe immunological method detected both the in-creased level of circulating fibrinogen and alsocirculating 'fibrin fragments' generated by thebreakdown of endometrium, haemorrhage, and clotlysis which must occur in this phase of the cycle. Onthe other hand, the 'fibrin fragments', being in-coagulable by thrombin, would not similarlycontribute to the clottable fibrinogen value exceptwhen actually incorporated into the clot. This wasfound sometimes to be the case (see below).

IMMUNOREACTIVE AND CLOTTABLEFIBRINOGEN AFTER MAJOR SURGERYSurgical operations are necessarily attended bylocalized thrombosis in and around the operationarea. During repair and regeneration there is lysis ofboth intra- and extravascular fibrin deposits. Sincethe aim was to examine the relation betweenthrombosis and thrombolysis, a preliminary studyin surgical patients offered several advantages.These were that measurements could be made bothbefore and after elective operations, each patientthus serving as his own control; the time at whichthrombosis commenced could be gauged withreasonable accuracy; and some estimate of theprobable extent of thrombosis could be obtainedfrom knowledge of the nature and extent of theoperative procedure.When immunoreactive and clottable fibrinogen

levels were examined in such patients, changessimilar in kind to those already described in somewomen during the menstrual cycle were observedconsistently during the postoperative period (fromabout the second or third day after operation). The



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


.. ..

IA

IA

v

Av

I

A

v

I I I

Fig. 3. Sequential comparisons of immunoreactive fibrinogen (IF) plotted ascontinuous line and confidence limits (shaded area) with clottable fibrinogen (CF)plotted as dashed line, in a healthy male (A) and in two females (B and C). The resultsin B and C relate to the same subjects giving the abnormal values indicated in Fig. Iwhen a single casual estimate was made. Note constancy and agreement of estimatesin the male and effect on values of minor throat infection (T) in Fig 3a. In contrastnote lability of both immunoreactive and clottable fibrinogen in females and extent ofdiscrepancies between estimates in premenstrual phases of cycle (menstrual periodsindicated by triangles). Solid blocks indicate appearance ofFDP in serum. Alsonote additional effects of dental abscess (A) in Fig. 3C.

40



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


*>

.

w

. .. .......... .,...............<...............

. . . . . . . . . ...<

,... .

.

... ... . .. .. .. .,..... _. . . . .._ . .

.

....af:.. , .: :,:,.,~~~~~~~~~4

Fig. 4 Dailycomparison ofimmunoreactivefibrinogen(continuous linewith hatched area

representingconfidence limits)and clottablefibrinogen (dashedline) in man of46before and afterremoval of benignpleural tumour.Operation at pointmarked by arrow.Note that maximaldiscrepancybetween immuno-reactive andclottablefibrinogendoesnotcorrespondwith appearance offibrin degradationproducts in serum(see text).

*...*~~~~~~~~0Ae*00 0

changes were more marked following surgery,however, and lasted for seven to ten days, dependingupon the severity of the operation. This sequence isshown for a typical case (daily estimations beforeand after removal of a benign pleural tumour in aman of 46 years) in Figure 4. It can be seen that bothimmunoreactive and clottable fibrinogen values rose

sharply, reaching a peak between days 5 and 6 afteroperation and then declining. Over the period offormation of this peak, immunoreactive valuesconsiderably exceeded clottable fibrinogen valueswhile fibrin degradation products in the serumshowed a similar rise and fall but one not preciselyrelated to that of immunoreactive values (see below).Two episodes of fluctuation of fibrinogen levels,

sequentially followed, were seen in another patient,illustrated in Figure 5. In this case renal veinthrombosis had been established radiographically(clinical details to be reported separately by Hard-wicke et al) before fibrinogen levels weie measured.

By the seventh day after catheterization of the renalvein there was a sharp increase in immunoreactiveand clottable fibrinogen levels, the former verygreatly exceeding the latter. During this period ofincrease of fibrinogen levels, there was an accom-panying increase of fibrin degradation products inthe serum. However, in this as in the previous case,the period of appearance of these fragments was notclosely related to the period of elevation of immuno-reactive and clottable fibrinogen values. This wasthought to be because simultaneously demonstrableaccelerated fibrinolysis in vitro contributed to thefragments found.

In the late postoperative period a further extensionof renal vein thrombosis was followed by anothersharp rise in fibrinogen levels. But on this occasionthere was no appreciable discrepancy betweenimmunoreactive and clottable fibrinogen levels at thepeak of response and no increase of fibrin fragmentswas noted in serum.

00

-ioo

700

600

503 _ _

400

300

4-4__

-L_I _



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


This second case illustrates clearly another kind ofanomaly which was noted frequently and which maybe of significance. This was a discrepancy in thereverse direction between immunoreactive andclottable fibrinogen levels, ie, an instance in whichimmunoreactive was found to be markedly belowclottable fibrinogen level. This is seen to be the casein this patient before operation (when thrombosis ofthe renal vein was known to be in progress), forabout 36 hours after operation, and again on the18th day when recurrence of renal vein thrombosiswas detected.

In other previously healthy individuals subjectedto elective surgery, immunoreactive and clottablefibrinogen levels showed the expected agreementbefore and immediately after surgery. But within thefirst 24-48 hours of the postoperative period atransient drop of immunoreactive below clottablefibrinogen levels often occurred. Plasma taken at thistime almost always showed the formation ofcryoprecipitates within four hours of storage at 4°.The following observations and experiments showedthese to be sparingly soluble fibrinogen aggregateswhich were coagulable by thrombin but whichbehaved anomalously in immunodiffusion to giverise to underestimates of immunoreactive fibrinogen.

It was found that some plasma samples takenwithin 24 hours of operation showed visible pre-cipitates formed in the agar and gave greatlyreduced immunodiffusion values despite demon-strably high clottable fibrinogen levels.

Using the technique of Allison and Humphrey(1960) in those instances in which the abnormalplasmas did not show precipitate formation in theagar, it was possible to show that although materialreacting with antifibrinogen was present, it showedabnormal diffusion characteristics compatible withan increase of molecular size (presumably polym-erization).When normal plasma or normal fibrinogen was

added to these abnormal plasmas, even the addition-ally added fibrinogen could not be accuratelymeasured by immunodiffusion. But clottable fibrino-gen values in such cases showed the expectedincrement.

Since the fibrinogen molecule in these plasmasseemed altered in physical form so as to be lessimmunoreactive but still clottable, it seemed ofinterest to examine the effect of its complete con-version to fibrin and subsequent digestion. Fibrinclots of this kind were digested by streptokinase-induced plasmin action (see Methods) and the yield

42

,!I



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


of digestion products was compared with that ofnormal plasmas with equal clottable fibrinogenvalues. In many cases the abnormal plasma samplesyielded peptides with similar diffusion properties andat the same concentration as the matched normalplasmas. But in some instances, the pathologicalplasma samples yielded digests which still gaveimmunological underestimates. In these instances,insoluble material precipitated from the digests aftershort storage at 40 suggesting that aggregates werere-forming.On raising the sodium chloride concentration from

0 15 to 2 M in the agar medium (see Methods)aggregates were dispersed and did not reform. Atthis high salt concentration antigen-antibody inter-action is confined to hydrophobic bonds, reducingthe sensitivity of the immunological assay. However,by adjustment of the dilutions of standard and testplasmas, it was found that immunoreactive fibrino-gen values again approximated closely to theexpected value as compared with clottable andimmunoreactive fibrinogen values ofmatched normalplasmas.

Discussion

From these results it will be evident that comparisonof immunoreactive and clottable fibrinogen valuesyields information indirectly, but simply, about thefunctional utilization of fibrinogen. First, themeasurements confirm the marked responsiveness ofplasma fibrinogen levels to metabolic stresses. Moreelaborate studies of fibrinogen turnover, usingradioisotopic tracers, have established that changesin the plasma concentration reflect alterations in thesynthesis rate (cf, Regoeczi, Regoeczi, andMcFarlane1964; Blomback, Carlson, Franzen, and Zetterqvist,1966).In health, some variation has been attributed to

diurnal 'random' fluctuation of fibrinogen metab-olism (Reeve, Takeda, and Atencio, 1966). But inour hands, sequential studies in healthy males, whowere presumably in a metabolic steady state inrelation to synthesis and catabolism, have shownlittle change in level except in response to minorinfections, etc. In healthy women, a wider range offluctuations was encountered which was apparentlyrelated to the menstrual cycle. Studies on physio-logical variation in fibrinogen levels during preg-nancy and childbirth will be reported separately.

Secondly, the increased concentration of fibrino-gen found following surgery resembles that reportedfollowing injuries, eg, burns, as reported by Davies,Ricketts, and Bull (1966) and by Kukral, Zeineh,Dobryszycka, Pollitt, and Stone (1969), and showssimilar time relations, suggesting an increased

synthesis of fibrinogen as found following burns.In this respect, therefore, the response of fibrino-gen, in resembling that of other 'acute phasereactant' serum proteins (cf Werner and Cohnen,1969) in the postoperative period, is confirmed.Additional information about the fate and

behaviour of fibrinogen can be adduced by con-sideration of discrepancies between immunoreactiveand clottable fibrinogen estimates. When thediscrepancy takes the form of lower immuno-reactive than clottable fibrinogen values, it has beenfound that this is associated with the presence ofaggregates of the fibrinogen molecule in the plasmaand with the occurrence of thrombosis. The charac-teristics of the aggregates resemble the 'cryopro-fibrin' shown by Shainoff and Page (1960) to becomposed of heterogeneous fibrin intermediatesformed as a result of thrombin activity in vivo. Theirpresence in the plasma of our patients within thefirst 24 to 48 hours of surgical intervention, ie, whenthrombosis can be envisaged as occurring in theoperative field, or in an individual with thrombosisin progress (see Fig. 5), is compatible with such anorigin. It seems reasonable to assume that tissuedamage at operation gives rise to massive release ofthromboplastin* and thus to considerable thrombinformation. Thrombin surplus to that utilized forlocal haemostasis and in excess of the amountcapable of neutralization by plasma antithrombinwould be considerably diluted in the circulation. Atlow concentration the enzyme might effect onlypartial and incomplete conversion of fibrinogen to'cryoprofibrin' (see Fig. IB).The failure of fibrinogen in this abnormal physical

form to give full quantitative immunodiffusionreactions may be due to one or more ofthe following:(1) the large size of the aggregates giving rise to veryslow diffusion or actual precipitation in the agar; (2)an effective reduction in molar concentration offibrinogen; or (3) aggregate-induced steric hindranceof access by antibody to reactive sites on the antigen.Whatever the mechanism, the occurrence of thisanomaly appears to be an indication of thrombosisin progress in the body. From reference to Fig. 3it can be seen that this anomaly may occasionally beencountered in healthy women in the menstrualphase, presumably then reflecting the occurrence ofthrombosis in the deeper endometrial vessels. Apartfrom these circumstances, the anomaly has only beenencountered in association with 'pathological'thrombosis of some magnitude. Our experience todate, for example, suggests that the anomaly is onlydetected in ischaemic heart disease when massiveinfarction, associated with intracardiac (mural)thrombosis due to thrombosis of a major coronaryartery, occurs. While of limited value as a general



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

44

screening test for coronary thrombosis, therefore,the detection of the anomaly may be of clinicalsignificance in assessing other varieties of intra- orextravascular thrombosis.

Thirdly, when immunoreactive values are foundmarkedly to exceed clottable fibrinogen values thediscrepancy appears to be associated with activethrombolysis. The phenomenon appears to be a more

sensitive indicator of the presence in the circulationof fibrin degradation products than the estimation ofthese products in serum by haemagglutinationinhibition (Merskey et al, 1966) or other immuno-logical methods of testing serum, presumablybecause these latter methods are also influenced byaccelerated fibrinolysis in vitro due to increasedoutput of plasminogen activator (Bennett, Ogston,and Ogston, 1967).

In comparing inmmunoreactive and clottablefibrinogen values statistically for normal subjects,the clottable value was taken to be the independentvariable, free from error. However, our experiencesuggests that the occurrence of the discrepancy as a

test for the elaboration of fibrin degradation pro-ducts during thrombolysis may sometimes beobscured by spuriously elevated clottable fibrinogenvalues due to the occlusion of the degradationproducts in the clot. For example, underestimates ofthese products in serum in the presence of high levelsin plasma (see Fig. 4) may arise for this reason. Otherevidence in support of this conclusion, and themanner in which it can be demonstrated, will bepublished separately.

Despite this reservation, and without more

elaborate measures to establish occlusion in clots,the occurrence of the anomaly of immunoreactivevalues in considerable excess of clottable fibrinogenvalues has nevertheless been demonstrable in about70% of postoperative cases from about the thirdday onwards. If allowance is made for the occasionaloccurrence of the anomaly in the premenstrualphase, its presence otherwise is suggested to be anindicator of active thrombolysis following 'patho-logical' thrombosis.Once again, in our hands the occurrence of this

anomaly has proved to be more sensitive thanimmunoelectrophoresis of plasma as a means ofdetecting fibrin or fibrinogen breakdown.

We acknowledge gratefully the cooperation of thephysicians and surgeons of the United BirminghamHospitals in allowing us access to cases with sus-


pected thrombosis or thrombolysis and, in particularto Professor J. Hardwicke for permission to presentthe results of a patient with renal vein thrombosisunder his care.

ReferencesAllison, A. C., and Humphrey, J. H. (1960). A theoretical and

experimental analysis of double diffusion'precipitin reactions ingels, and its application to characterization of antigens.Immunology, 3, 95-106.

Bailey, K., and Bettelheim, F. R. (1955). The clotting of fibrinogen.I. The liberation of peptide material. Biochem. biophys. Acta(Amst.), 18, 495-503.

Bennett, N. B., Ogston, C. M., and Ogston, D. (1967). Studies on theblood fibrinolytic enzyme system following acute myocardialinfarction. Clin. Sci., 32, 27-37.

Biggs, R., and Macfarlane, R. G. (1957). In Human Blood Coagula-tion and its Disorders, 2nd ed., p. 392. Blackwell, Oxford.

Blomback, B. (1967). In Blood Clotting Enzynmology, edited by W. H.Seegers, p. 201. Academic Press, New York.

Blomback, B., Carlson, L. A., Franzen, S., and Zetterqvist, E. (1966).Turnover of 13'I-labelled fibrinogen in man. Acta med. scand.,179, 557-574.

Davies, J. W. L., Ricketts, C. R., and Bull, J. P. (1966). Studies ofplasma protein metabolism. III. Fibrinogen in burned patients.Clin. Sci., 30, 305-314

Farrell, G. W., and Wolf, P. (1970). A simple method of preparingmonospecific anti-human fibrin serum in rabbits. J. med. Lab.Technol., 27, 312-316.

Farrell, G. W., and Wolf, P. (1971). Studies on thrombosis andthrombolysis. Med. lab. Technol., 28, 310-318, 319-327,328-336.

Fearnley, C. R., and Chakrabarti, R. (1966). Fibrinolytic treatment ofrheumatoid arthritis with phenformin plus ethyloestrenol.Lancet, 2, 757-760.

Kukral, J. C., Zeineh, R., Dobryszycka, W., Pollitt, J., and Stone, N.(1969). Metabolism of plasma proteins in injury states. I.Turnover rates of fibrinogen in burned patients labelledwith (35S) methionine. Clin. Sci., 36, 221-230.

Lipinski, B., Wegrzynowicz, Z., Budzynski, A. Z., Kopec, M.,Latallo, Z. S., and Kowalski, E. (1967). Soluble unclottablecomplexes formed in the presence of fibrinogen degradationproducts (FDP) during the fibrinogen-fibrin conversion andtheir potential significance in pathology. Thro,nbos. diathes.Haemorrh. (Stuttg.), 17, 65-77.

McFarlane, A. S., Todd, D., and Cromwell, S. (1966). Fibrinogencatabolism in humans. Clin. Sci., 26, 415-420.

Merskey, C., Kleiner, G. J., and Johnson, A. S. (1966). Quantitativeestimation of split products of fibrinogen in human serum,relation to diagnosis and treatment. Blood, 28, 1-18.

Nussenzweig, V., Seligmann, M., and Grabar, P. (1961). Les produitsde degredation du fibrinogene humain par la plasmine. 11.ttude immunologique: mise en 6vidence d anticorps anti-fibrinogene natif possedant des specificites differentes. Ann.Inst. Pasteur, 100, 490-508.

Reeve, E. B., Takeda, Y., and Atencio, A. C. (1966). Some observa-tions on the mammalian fibrinogen system in non-steady andsteady states. Protides biol. Fluids, 14, 283-294.

Regoeczi, E., Regoeczi, G. E., and McFarlane, A. S. (1964). Relationbetween rate of catabolism, plasma concentration and poolsize of fibrinogen. Pflugers Arch. ges. Physiol., 279, 17-25.

Shainoff, J. R., and Page, I. H. (1960). Significance of cryoprofibrin infibrinogen-fibrin conversion. J. exp. Med. 116, 687-707.

Takeda, Y. (1966). Studies of the metabolism and distribution offibrinogen in healthy men with autologous 12"I-labeled fibrino-gen. J. clin. Invest., 45, 103-1 1 1.

Werner, M., and Cohnen, G. (1969). Changes in serum proteins in theimmediate postoperative period. Clin. Sci., 36, 173-184.

Wolf, P., and Walton, K. W. (1965). Investigation of a quantitativeanomaly encountered in the assay of fibrinogen by immuno-diffusion. Immunology, 8, 6-24.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Documents

thrombosis - Journal of Clinical Pathology · thrombosis commenced could be gauged with reasonable accuracy; and some estimate of the probable extent of thrombosis could be obtained