THE COAGULATION of HUMAN BLOOD by J

62 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from

THE COAGULATION OF HUMAN BLOOD By J. F. ACKROYD, M.B., M.R.C.P. The Medical Unit, St. Mary's Hospital, London, W.2

The classical theory of blood coagulation ad- ProthrombJn LlThromboplastin C vanced by Schmidt (I892) and Morawitz (1905) is represented schematically in Fig. i. This theory has stood the test of time and repeated experimental investigation, and it is on this foundation that all further hypotheses must be built. Conclusive proof that blood coagulation involved other factors in addition to those included Fibr±n in the classical theory was first provided by

Owren's (I944, I947) discovery of factor V, FIG. i.-The classical theory of blood coagulation. Protected by copyright. although it is clear that Nolf's (I908) work, which had largely been overlooked, had anticipated outline some of the more important observations Owren's discovery by many years. Owren's that led to their discovery. Finally an attempt will factor V, which was soon to be rediscovered in- be made to show how they, and the older estab- dependently by Fantl and Nance (I946) and by lished factors, interact to cause coagulation of the Ware, Guest and Seegers (I947), stimulated wide- blood. As many of the coagulation factors have spread interest in the factors concerned in blood been given different names by different workers, a coagulation with the result that numerous new list of some of the more commonly used synonyms coagulation factors, some hypothetical and in- is given in Table i. adequately supported by experimental evidence, have been postulated. It is the purpose of this The Factors Recognized in the Classical short article to describe very briefly the more im- Theory portant properties of the factors included in the Tissue Thromboplastin classical theory of blood coagulation; to describe Tissue extracts have been known for many years http://pmj.bmj.com/ the action pf those newly discovered factors whose to hasten the coagulation of blood or plasma. They existence seems adequately established; and to act by accelerating the conversion of prothrombin

TABLE I COMMONLY USED SYNONYMS FOR COAGULATION FACTORS FACTOR V FACTOR VII (Koller) (Owren) on September 30, 2021 by guest. Thrombogen (Nolf) Stable component (Stefanini) Labile factor (Quick) Co-factor V (Owren) Prothrombin accelerator (Fantl and Nance) Preconvertin-->convertin (Owren) Proaccelerin- >accelerin (Owren) Serum prothrombin conversion accelerator (S.P.C.A.) Plasma Ac. globulin.->Serum Ac. globulin (Seegers) (Alexander and de Vries) Prothrombin conversion factor (Owen and Bollman) Co-thromboplastin (Mann and Hum) ANTIHAEMOPHILIc GLOBULIN PLASMA THROMBOPLASTIN COMPONENT (P.T.C.) (Lewis, Tagnon, Davidson, Minot and Taylor) (Aggeler, White, Glendenning, Page, Leake and Bates) Thromboplastinogen (Quick) Christmas factor (Biggs, Douglas, Macfarlane, Dacie, Thrombocytolysin (Brinkhous) Pitney, Merskey and O'Brien) Plasma thromboplastic factor (Stefanini) Factor IX (Koller) Factor VIII (Koller) February 1954 ACKROYD: The Coagulation of Human Blood 63 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from to thrombin. This activity depends to a con- abnormality of blood coagulation in a patient with siderable extent upon the concentration of calcium a hitherto undescribed haemorrhagic disease which in the mixture (Biggs and Macfarlane, 1953a). Owren called parahaemophilia. This patient had The active principle of the extracts, thrombo- a history of abnormal bleeding since the age of 31 plastin, is particulate, and can be removed by years. The whole blood clotting time was pro- centrifugation at relatively low speeds (31,000 g.) longed, on one occasion having been as long as 70 (Chargaff et al., I944). It can be obtained from minutes. The prothombin time by Quick's (I935) almost any tissue in the body, brain and lung being one-stage method was also prolonged. the most commonly used. Thromboplastin will ac- In this test the coagulation time of oxalated celerate clotting even when diluted as much as plasma is measured after the addition of optimal I:100,000. It seems probable, therefore, that it concentrations of calcium chloride and brain acts as a catalyst. Thromboplastin will not clot thromboplastin. On the basis of the classical fibrinogen solutions or preparations of plasma theory, if the plasma contains an adequate amount from which the prothrombin has been removed. of fibrinogen, and calcium and thromboplastin are present in optimal concentrations, then the co- Plasma Thromboplastin agulation time should be inversely proportional to Blood will clot in glass vessels without admixture the concentration of prothrombin. The pro- with tissue extracts and must therefore, according longed coagulation time with this test should to the classical theory, contain thromboplastin. It therefore have indicated that the prothrombin con- has generally been believed that this came from tent of this patient's plasma was reduced. Owren the platelets. These, however, contain relatively found, however, that the prothrombin time could be little, but as will be shown later, they are essential greatly shortened by the addition of normal plasma for normal thromboplastin formation. from which the prothrombin had been removed by adsorption on to aluminium hydroxide, or by Prothrombin and Thrombin Seitz filtration. It seemed clear, therefore, that Protected by copyright. These have been prepared in a highly purified this patient's plasma lacked a substance which is state by Seegers, McClaughry and Fahey (I950). present in normal plasma and which is left in the They are proteins. go per cent. of the pro- plasma after removal of the prothrombin. Owren thrombin moves electrophoretically with a mobility called this substance factor V. He described a of a, globulin. The mobility of thrombin is- method for its partial purification and showed that slightly less. Purified prothrombinwis converted it was necessary for the rapid conversion of pro- into thrombin when dissolved in 25 per cent. thrombin to thrombin, and that it disappeared sodium citrate. It therefore seem&;reasonabre to rapidly from the blood after clotting. The pro- conclude that prothrombin contains all the struc- longed prothrombin time observed with the blood tural material needed for the formation of thrombin of his patient was clearly due to the slow and in- (Seegers et al., 1950). The action of thrombin in complete formation of thrombin resulting from the converting fibrinogen to fibrin appears to be absence of factor V from the plasma.

enzymatic, for thrombin will convert many times Several other cases of factor V deficiency have http://pmj.bmj.com/ its weight of fibrinogen. This actiorn is accelerated now been described (Frank, Bilhan and Ekren, by calcium (Biggs and Macfarlane, I953a), and by 1950; de Vries, Matoth and Shamir, '95'; platelet extracts (Ware, Fahey and Seegers, 1948). Stohlman, Harrington and Moloney, I95I; Brink and Kingsley, I952; Cosgriff and Leifer, 1952; Fibrinogen and Fibrin Owren, 1953). In some of these cases there was a Fibrinogen is a protein with an elongated mole- familial incidence of the disease. cule about 8oo A in length and 40 A in width. Shortly after Owren had announced his dis- Fibrin appears to be formed by the polymerization covery, Fantl and Nance (1946) and Ware, Guest on September 30, 2021 by guest. of individual fibrinogen molecules. Electron and Seegers (1947), working independently with microscopic studies suggest that an end to end purified preparations of prothrombin, described linkage of fibrinogen molecules occurs forming experiments which demonstrated the existence of a fibrils which later aggregate into bundles which factor required for the rapid conversion of pro- constitute the individual fibrin strands. thrombin to thrombin. These workers found that although prothrombin prepared by salt precipita- Recently Discovered Coagulation Factors tion rapidly formed thrombin in the presence of calcium and thromboplastin, yet prothrombin pre- Coagulation Factors in Plasma and Serum pared by adsorption on to aluminium hydroxide Factor V and subsequent elution, was only converted to The existence of this factor was first established thrombin very slowly. The rate of conversion by Owren's (I944, 1947) brilliant analysis of the could be accelerated by the addition of a small 64 POSTGRADUATE MEDICAL JOURNAL February 1954 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from quantity of the plasma from which the pro- which they have partially purified (Alexander. thrombin had been adsorbed. The factor re- Goldstein and Landwehr, 1950), and which they sponsible was partially purified by both groups of call serum prothrombin conversion accelerator workers. Fantl and Nance (I948) called the factor (S.P.C.A.). They showed that dicoumarol causes prothrombin accelerator, while Ware and Seegers a reduction in both plasma prothrombin and the (I948) gave it the name Ac globulin. There seems serum prothrombin conversion accelerator (Alex- little doubt that these factors are identical with ander, de Vries and Goldstein, 1949). Owren's factor V, and Owren's term will therefore In 1950 Owren (195oa) showed that thrombin be used hereafter. The labile factor which dis- was formed when human serum was added to appears from oxalated plasma on storage, and Seitz filtered ox plasma in the presence of calcium which was described by Quick in 1943, is probably and thromboplastin. He thought at first that the also identical with factor V, although it must be prothrombin must be in the serum but later stated that the clotting defect produced by storage showed that it was in the ox plasma which had been is not a clearly defined entity, and that storage inadequately filtered (Owren, 195oa; 195I). This almost certainly alters factors other than factor V remaining prothrombin was not converted to in plasma. Ware, Fahey and Seegers (1948) have thrombin by calcium, thromboplastin and factor V shown that platelets, as well as plasma, have alone. Owren therefore concluded that the serum factor V activity. must contain a further factor necessary for prothrombin conversion, and that this factor had been Factor VII adsorbed from the plasma by Seitz filtration. It That there is another factor in addition to factor is also adsorbed by aluminium hydroxide (Biggs, V which is responsible for accelerating the con- Douglas and Macfarlane, 1953a). Owren called version of prothrombin to thrombin was first this factor Convertin. It seems almost certain that suggested Owen and by Bollman (1948). They it is identical with the factor described by Owen Protected by copyright. were investigating Schofield's (I924) observation and Bollman (1948) and the serum prothrombin that spoiled sweet clover disease in cattle (which conversion accelerator of Alexander and his co- is due to the presence of dicoumarol in the fodder) workers (1949, 1950). is improved by the injection of normal serum. Koller and his colleagues (1951, 1952) have re- This condition was believed to be due solely to a peated Owren's experiments and have shown con- reduction in plasma prothrombin which was un- clusively that this factor accelerates the conversion likely to have been significantly increased by the of prothrombin to thrombin. They call it factor administration of serum which normally contains VII, factor VI being the name given by Owr rn little or no prothrombin. These authors therefore to a hypothetical substance derived from factor V. investigated the blood of dogs fed on, or injected The importance of this factor, which will here- intravenously with, dicoumarol. They found that after be referred to as factor VII, was finally estab- the plasma prothrombin, as estimated by the one- lished by Alexander and his co-workers (I95I) stage method, was greatly reduced, but that the ad- who described a patient with a haemorrhagic syn-

dition of normal prothrombin free serum caused a drome apparently due to a congenital deficiency of http://pmj.bmj.com/ considerable rise in the apparent prothrombin the factor. The disease was characterized by a level as indicated by this test. They concluded normal whole blood clotting time and a prolonged from this and other experiments that normal serum one-stage prothrombin time which was corrected contains a factor necessary for prothrombin con- in vitro by the addition of crudely purified factor version. This factor is unlikely to have been VII and temporarily, in vivo, by the intravenous factor V because factor V disappears rapidly from administration of human serum. The factor V serum (Owren, 1947) and also because factor V is content of the patient's plasma was shown to be not significantly reduced in the plasma of patients normal. Similar cases have recently been reported on September 30, 2021 by guest. receiving dicoumarol (Owren, I95ob). The factor by Beaumont and Bernard (I953) and by Jiirgens described by Owen and Bollman appears, there- (I953); and Owren (I953) has described a family fore, to be a further prothrombin conversion factor. of which several members were affected by this In 1949, de Vries, Alexander and Goldstein condition. observed that oxalated serum, in the presence of Further studies have shown that deficiency of calcium and thromboplastin, accelerates the co- this factor is the main abnormality caused by tro- agulation of oxalated plasma. This occurred if the mexan therapy (Douglas, I953). It is also reduced oxalated plasma was diluted with plasma which in advanced liver disease and in the newborn (de had been adsorbed with barium sulphate to re- Nicola, 1953). move most of its prothrombin whilst leaving an excess of factor V. They therefore considered that Factor X the serum contained a new coagulation factor The existence of this factor has recently been February 1954 ACKROYD: The Coagulation of Human Blood 65 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from postulated by Koller (I953) who found that if Aggeler and his co-workers (1952), Schulman and factor VII was prepared in a sufficiently pure state Smith (I95z), Biggs, Macfarlane and Dacie and it lost its power completely to correct the coagula- their colleagues (Biggs et al., 1952), Poole (1953), tion defect of the blood of patients receiving tro- Lewis and Ferguson (1953) and by Cramer, mexan. He suggested that crude factor VII con- Matter and Loeliger (i953). tained two factors: factor VII and a new factor The clinical picture in all these cases was in- which he called factor X, factors VIII and IX distinguishable from that of haemophilia. The being antihaemophilic globulin and P.T.C. (Christ- condition is inherited as a sex-linked recessive mas factor) respectively. Koller found that factor character, although sporadic cases occur. As in X, like factor VII, is present in high concentration haemophilia, the coagulation time is usually pro- in serum, and is reduced in advanced liver disease, longed. The one-stage prothrombin time is in the newborn, and in the blood of patients re- normal, although the conversion of prothrombin ceiving dicoumarol and tromexan. He believes to thrombin during coagulation is reduced. The that he has seen -a case of congenital factor X condition is not due to, a deficiency of anti- deficiency. haemophilic globulin, for the plasma of these patients reduces the clotting time of haemophilic Antihaemophilic Globulin plasma. The clotting time of the plasma of these Haemophilia is inherited as a sex-linked re- patients is reduced by normal and by haemophilic cessive character, although sporadic cases, result- plasma, but not by antihaemophilic globulin (Biggs ing from spontaneous mutation, frequently occur. et al., 1952). The blood characteristically shows a prolonged Investigation of the factor concerned shows that, clotting time although this is not an essential unlike anti-haemophilic globulin, it is present in feature of the disease. The one-stage prothrombin large amounts in normal serum. It is removed

time is normal and there is an impaired conversion from normal serum or plasma by adsorption on to Protected by copyright. of prothrombin to thrombin during coagulation. aluminium hydroxide, or by Seitz filtration (Biggs It has been known for many years that the and Macfarlane, 1953a), or by adsorption on to clotting time could be reduced to normal by the barium sulphate (Rosenthal et al., 1953). In these addition of thromboplastin or of small quantities of respects it resembles factor VII. That it is dis- normal plasma (Weil, I906o; Addis, i9ii; tinct from factor VII is shown by the following Govaerts and Gratia, I93I). These findings observations: (i) Its deficiency does not cause strongly suggest that haemophilia is caused by prolongation of the one-stage prothrombin time. deficient formation of thromboplastin, and that (2) Serum from patients deficient in the factor this results from the absence of some factor present corrects the coagulation defect of the plasma of in normal plasma. This factor appears to be patients receiving tromexan. (3) The coagulation utilised during coagulation and is almost absent defect in patients deficient in the factor is corrected from normal serum. It is not adsorbed by by some serum samples from patients receiving aluminium hydroxide (Biggs et al., I952) or barium tromexan (Biggs and Macfarlane, 1953a). sulphate (Rosenthal et al., 1953). It occurs mainly As the one-stage prothrombin time of the plasma http://pmj.bmj.com/ in Cohn's plasma fraction I, which also contains of patients deficient in this factor is normal, it the fibrinogen (Cohn, I946). It will be referred to appears that brain thromboplastin rectifies the below as the antihaemophilic globulin. Factor defect, and therefore that this factor, like anti- V (Owren, 1947) and factor VII (Owren, I95oa) haemophilic globulin, must be concerned in the are present in normal concentration in haemo- formation of thromboplastin. The factor has been philia. partially purified by Aggeler and his colleagues (1952), who have called it the plasma thrombo- Plasma Thromboplastin Component (P. T.C.) or' plastin component (P.T.C.). Macfarlane and on September 30, 2021 by guest. Christmas Factor Dacie and their co-workers (Biggs et al., I952) Pavlovsky, in 1947, observed the surprising fact have called it the Christmas factor after the first that occasionally a patient is found suffering from patient in whom they found it to be deficient. what is apparently 'classical haemophilia, whose Both terms will be used when the factor is referred blood or plasma has the power to reduce the to below. clotting time of the blood of another haemophilic patient. Plasma Thromboplastin Antecedent (P.T.A.) Three years later Koller and his colleagues Rosenthal, Dreskin and Rosenthal (I953) have (I950) reported the results of their investigations recently reported a family of which three members on a patient who seems to have been a further had a mild haemorrhagic disease which resembled example of the condition described by Pavlovsky. haemophilia except for the fact that two of those Since then further cases have been reported by affected were females. The clotting time was pro- 66 POSTGRADUATE MEDICAL JOURNAL February 1954 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from longed and there was an abnormal utilization of 120 prothrombin. The one-stage prothrombin time was normal. The abnormality was corrected by the 110 addition of blood from patients with either haemo- 100 philia or with P.T.C. deficiency (Christmas 90 disease). It appears therefore that this condition 80 may be due to deficiency of a third factor which 8 70 . these authors have called plasma thromboplastin antecedent (P.T.A.). According to Rosenthal and A 60 . his colleagues, the three conditions can readily be 50\ distinguished as follows: The clotting defect in V 40 ' haemophilia is corrected by normal plasma ad- 30- sorbed with barium sulphate, but not by normal serum; P.T.C. deficiency is corrected by normal 20 serum, but not by barium sulphate adsorbed 10 plasma; and P.T.A. deficiency is corrected by both 10 20 3040 50 60 70 80 90 100 110 120 140 normal serum and normal plasma which has been T6nbmoplestin % adsorbed with barium sulphate. FIG. 2.-Thromboplastin dilution curve. The curve shows the relation between clotting time and Plasma Thromboplastin thromboplastin concentration. Reproduced from Biggs and Douglas, Y7. Clin. Path., 1953, 6, One of the main problems in understanding 23, by courtesy of the authors and publishers. blood coagulation on the basis of the classical theory is the origin of thromboplastin. As blood intervals by observing the coagulation time of taken by vein puncture, without contamination Protected by copyright. with tissue juices, clots readily when placed in a samples of normal oxalated plasma to which cal- glass vessel, it is clear that blood contains all the cium chloride and aliquot quantities of the above- factors necessary for coagulation. The platelets mentioned mixture are added. As the mixture have usually been assumed to be the origin of the contains no significant amount of prothrombin, thromboplastin under these conditions, but plate- the coagulation of the normal plasma will be due lets, in fact, contain very little (Ware, Fahey and almost entirely to the thromboplastin developed Seegers, 1948). The addition of platelets to re- in the mixture, and the coagulation times will be calcified oxalated plasma does not reduce its inversely proportional to the concentration of clotting time to less than about 6o to 8o seconds thromboplastin. Under these conditions thrombo- (Biggs and Macfarlane, 1953a), whereas with brain plastin forms gradually over a period of 4 to 5 thromboplastin it is easily possible to reduce the minutes, but when fully developed has a potency clotting time to about 12 seconds. Is this weak at least as great as that of a potent brain extract. thromboplastic action sufficient to promote normal By observing the coagulation times with different a coagulation or is there some other source of dilutions of this thromboplastin, thromboplastin http://pmj.bmj.com/ thromboplastin in normal plasma? dilution curve can be drawn (Fig. 2) relating The most convincing demonstration that plasma clotting time to thromboplastin concentration. contains a potent thromboplastin has been pro- The diagram (Fig. 3) shows the thromboplastin vided by the thromboplastin generation test of concentrations generated in 40 different tests. Biggs and Douglas (I953). In this test plasma is After five minutes, when the thromboplastin adsorbed with aluminium hydroxide which re- generated in the mixtures was maximal, the mean moves prothrombin, factor VII and P.T.C. clotting time of normal plasma to which the (Christmas factor) and, as shown by Koller (I953), mixture was added was less than io seconds. This on September 30, 2021 by guest. probably also factor X, but leaves antihaemophilic is as short, if not shorter, than that obtained with globulin and factor V in the plasma which, in the most potent tissue extracts. It is clear, there- addition, probably contains plasma thrombo- fore, that normal plasma contains all the factors plastin antecedent (P.T.A.), the factor recently required for producing a potent thromboplastin. described by Rosenthal and his colleagues (1953). Moreover, using purified reagents, Biggs, Douglas The deficiencies. in the factors other than pro- and Macfarlane (I953b) have shown that all the thrombin, removed by adsorption with aluminium following components: platelts, antihaemophilic hydroxide, are made good by the addition of serum globulin, P.T.C. (Christmas factor), factor V and freed of both prothrombin and thrombin. Plate- factor VII, are required, and that if any one of lets are then added to the mixture. Finally calcium these factors is missing, abnormal thromboplastin chloride is added and the thromboplastic activity formation will result. The reagents used in these that develops in the mixture is estimated at investigations almost certainly contained the re- February 1954 ACKROYD: The Coagulation of Human Blood 67 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from

140.----- So far no attempt appears to have been made to discover which of these factors is utilized in the generation of plasma thromboplastin. The pro- 120- perties of factors I and III suggest that these must be required, but it is not known whether these _100 alone are sufficient or whether the platelets have still other properties which are also involved. In thrombocytopenic purpura, presumably be- - I0 -80. I cause there are insufficient platelets to permit o E g0 normal thromboplastin formation, prothrombin conversion is incomplete. Apart from this, the 60 I association of a haemorrhagic disease with an abnormality of the platelets has been reported by two groups of workers. van Creveld and Paulssen 40 I (I95I, 1952) have described two cases of a haemorrhagic disorder which they attributed to a partial 20 - or complete absence of factor III from the platelets, and Biggs and Douglas (953), although they made no attempt to assay the platelets for their content of the different platelet factors, were able to show Incubation in minutes that the platelets in a case of thromboasthenia time (Glanzmann) were grossly deficient in their power FIG. 3.-The normal range of the thromboplastin generation test. The central curve represents the to form normal plasma thromboplastin when in- average of 40 different tests. The shaded area cubated with normal serum and aluminium Protected by copyright. shows the range of variation obtained in 38 of 40 hydroxide adsorbed plasma. observations in which the adsorbed plasma was obtained from different normal subjects, but serum The Initiation of Blood Coagulation and platelets were from one subject. The extreme limits show the range of variation when all the If blood is taken from a vein into a silicone lined reagents are varied simultaneously. syringe and is transferred to a silicone lined tube, Reproduced from Biggs and Douglas, J. Clin. Path., 19S3, 6, coagulation may not take place for several hours. 23, by courtesy of the authors and publishers. If the platelets from such a preparation are removed by centrifugation, the plasma may not clot cently described factors X and plasma thrombo- even if it is transferred to a glass container (Patton, plastin antecedent (P.T.A.), and it seems probable Ware and Seegers, 1948). It seems clear, there- that these are also necessary for thromboplastin fore, that both platelets and contact with a suitable formation. surface are necessary, and it may be surmised that contact precipitates platelet disintegration and Coagulation Factors in the Platelets so initiates the formation of plasma thromboplastin. http://pmj.bmj.com/ The thromboplastin generation test has shown In haemophilia, in which the plasma is d2ficient that platelets are necessary for the formation of in antihaemophilic globulin, the blood may not plasma thromboplastin. Three coagulation factors clot for an hour or more even when it is in contact have been extracted from platelets: with glass. Moreover, the platelets in haemo- Platelet factor I. This was first described by philic blood disintegrate much less rapidly during Ware, Fahey and Seegers in 1948. It is found, coagulation than do normal platelets in normal together with platelet factor II, in watery extracts blood (Howell and Cekada, 1926). As plat 1ats of platelets. These two factors have since been from haemophilic blood, when suspended in on September 30, 2021 by guest. separated by van Creveld and Paulssen (I95I). normal plasma, in glass vesszls, disrupt normally Platelet factor I has an activity resembling that of during coagulation (Patek and Stetson, I936), it factor V of normal plasma, but unlike factor V it would seem that the platelets in haemophilia are is particulate and can be made to sediment by probably normal, and that antihaemophilic globu- centrifuging at about 32,000 g. lin may be concerned with the normal disruption Platelet factor II. This accelerates the clotting of platelets when these come into contact with a of fibrinogen by thrombin. suitable surface. Platelet factor III. This is insoluble in water. Further evidence of the importance of contact is When suspended in plasma it antagonizes the provided by the observation that the capacity of action of heparin and has thromboplastic activity. serum to accelerate prothrombin conversion is re- This factor was first described by van Creveld and duced if coagulation is delayed by the use of Paulssen (I95I, 1952). silicone lined apparatus (Alexander, de Vries and 68 POSTGRADUATE MEDICAL JOURNAL February 1954 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from Goldstein, 1949; Biggs and Macfarlane, 1953a). Heparin appears to have two main actions: This observation was originally interpreted as i. It increases the adsorption of thrombin by indicating that factor VII must exist in plasma as fibrin (Klein and Seegers, I950). a relatively inactive precursor which was activated 2. In association with antithrombin it combines by contact, for at this time factor VII was the only with thrombin, greatly increasing the rate of conversion factor known which existed in equal neutralization of thrombin by antithrombin, but concentration in serum and plasma (the concentra- apparently not increasing the amount of thrombin tions of both factor V (Owren, i947) and anti- neutralized (Biggs and Macfarlane, I953a). haemophilic globulin (Graham et al., I951) fall According to Biggs and Macfarlane (1953a), rapidly after coagulation). Subsequently it wasr when heparin is added to whole blood it delays realized that P.T.C. (Christmas factor) and prob- or prevents the conversion of prothrombin to ably the recently postulated factors X and plasma thrombin. They consider that this effect is thromboplastin antecedent (P.T.A.) also exist in probably due to the rapid neutralization of throm- high concentration in serum. More recent work bin, this neutralization preventing the auto- has indicated that contact probably has no effect catalytic effect of thrombin on the development of on factor VII (Biggs and Macfarlane, I953b), and plasma thromboplastic activity (see below). that it activates one or more of the other factors found in serum. Antithromboplastin This substance has been postulated by Tocantins Naturally Occurring Inhibitors ofCoagulation but its existence has not yet been finally proved. Antithrombin Normal blood, in addition to its complex Clot Retraction mechanisms for producing thrombin, also contains When blood clots, the clot occupies a volume

powerful mechanisms for neutralizing it. It has equal to that of the blood from which it wasProtected by copyright. been calculated, from the results of experiments on formed. After a variable period of minutes, the the addition of purified thrombin to plasma, that exact time depending mainly on the temperature, X ml. of normal plasma is capable of neutralizing the clot begins to retract, extruding serum and a rapidly the amount of thrombin that would be small number of red cells. Clot retraction de- required to clot 2,000 ml. of blood in I5 seconds pends upon the presence in the blood of an (Biggs and Macfarlane, 1953a). adequate number of normal platelets. If the Thrombin inactivation depends upon two platelets are removed from plasma, either by mechanisms: centrifugation (Arthus and Chapiro, I908; Le I. Adsorption on to fibrinogen. This is the Sourd and Pagniez, I9iI) or by lysis (Tocantins, cause of the immediate inactivation of thrombin. 1934; Ackroyd, I949a, I949b), clot retraction is The thrombin is not destroyed and can be re- abolished. Tocantins (I934) and Budtz-Olsen covered after lysis of the fibrin (Klein and Seegers, (095i) have produced a considerable amount of 1950). evidence to show that platelets cause clot retraction

2. Neutralization by a constituent of normal by settling on the fibrin threads and later fusing http://pmj.bmj.com/ plasma, antithrombin, which, except with low into small masses, so drawing the fibrin strands concentrations of thrombin, appears to react together. It is not clear what part clot retraction stoichiometrically with thrombin rendering it plays in the haemostatic mechanism. inert (Klein and Seegers, 1950). Fibrinolysis Heparin Normal human blood contains enzymes which, Heparin is a powerful anticoagulant, and under certain circumstances, become activated and although it is not present in the blood in measur- digest fibrin, and sometimes also fibrinogen. The on September 30, 2021 by guest. able quantity, it can be extracted from most of the terminology used below is that of Biggs and tissues of the body, particularly the liver and Macfarlane (1953a). lungs. There is evidence that it arises from the mast cells. These cells are commonly found Plasmin along the capillaries suggesting perhaps that they This enzyme is found in the globulin fraction may have some action in preventing local capillary of normal human plasma. Its activity is inhibited thromboses. by the albumin fraction which has therefore been Heparin is inactive in the absence of the considered to contain a specific inhibitor ' anti- albumin fraction of the plasma which contains a plasmin.' Plasmin normally exists in the blood in substance essential for its action (Quick, 1938). It the form of an inactive precursor ' plasminogen.' seems probable that this substance is antithrombin It can be activated by separation from the albumin (Lyttleton, 1950). fraction of plasma or by the addition to plasma of February 954 ACKROYD: The Coagulation of Human Blood 69 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from chloroform or bacterial filtrates, the most com- These ' serum ' factors then react with anti- monly used being streptokinase which is obtained haemophilic globulin and the products of platelet from certain strains of p-haemolytic streptococci. disintegration (probably platelet factors I and III) Plasmin causes lysis of fibrin and also destroys and with factors V and VII, to produce a highly fibrinogen. active thromboplastin. This develops slowly at first and does not appear in maximal concentration Fibrinolysin for 4 to 5 minutes. The mode of its formation is Human and animal tissues contain an activator entirely obscure. 'fibrinokinase' which activates a fibrinolysin which causes lysis of both fibrin and fibrinogen. ,Conversion of Prothrombin to Thrombin This enzyme is also found in the globulin fraction Thromboplastin is essential for the conversion of of the plasma, but there are reasons for believing that it is not identical with plasmin (Astrup and Permin, 1948). THROMBIN UNITS Fibrinolysis Post Mortem and in Vivo Post mortem blood in cases of sudden death is often found to be fluid, the fibrin in the post- mortem clots having been liquefied by a fibrinolytic enzyme which attacks only fibrin and not fibrinogen (Mole, 1948). Fibrinolytic activity is induced in vivo in man by acute anxiety, severe exerciso or by injections of adrenalin (Macfarlane and Biggs, 1946; Biggs, Macfarlane and Pilling, I947). The enzyme con- Protected by copyright. cerned differs from plasmin and fibrinolysin in that it does not digest fibrinogen (Bidwell and Macfar- 0 2 4 6 S 10 12 14 16 lane, 1951). It closely resembles the lysin found TIUE IN MINUTES post mortem in cases of sudden death. Nothing is FIG. 4.-Thrombin generation in normal whole blood known of the mode of activation of either of these Curve i: Whole blood alone. Curve 2: Blood to enzymes. which a small amount of thrombin was added at the beginning of the test. The Interaction of the Factors Concerned in Reproduced from Macfarlane, Lecture 13 in ' Lectures on the Scientific Basis of Medicine, 195I-52,' by courtesy of the author Blood Coagulation and publishers. The Formation of Thromboplastin THROMBOPLASTIN The first step in the initiation of blood coagula- CONC. tion under normal circumstances is the formation 70 of thromboplastin. Where there is much con- http://pmj.bmj.com/ tamination with tissuejuices these supply thrombo- 60 plastic activity, although there is evidence that the tissues supply a thromboplastin which is biologic- ally incomplete, and that its activity is increased by 6501 the action of factors V and VII (Biggs, Douglas 40 and Macfarlane, I953b). In the absence of tissue thromboplastin, blood 30. coagulation depends upon the production of in- on September 30, 2021 by guest. trinsic plasma thromboplastin, the development of which appears to be dependent upon contact with a water-wettable surface. This seems to be necessary for the disintegration of platelets. Platelet disintegration is probably accelerated by antihaemophilic globulin. 2 4 6 8 10 12 14 16 Contact is apparently also necessary for the TIME IN MINUTES activation of one or more of the factors, other than FIG. S.-The formation of thromboplastin from plate- factor VII, found in high concentration in serum lets, aluminium hydroxide adsorbed plasma, normal (plasma thromboplastin component (Christmas serum and calcium chloride in the presence (x-x) factor) and probably the recently postulated factors and absence (o-o) of thrombin. Reproduced from Biggs and Macfarlane, 'Human Blood Co- plasma thromboplastin antecedent and factor X). agulatiOn,' 1953, by courtesy of the authors and publishers 70 POSTGRADUATE MEDICAL JOURNAL February 1954 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from

CONTACT with water-wettable surface and possibly with fibrin generated during coagulation

Platelets

Products of platelet Anti- P.T.C. ? Plasma thrombo- disintegration (probably haemophilic Factor V Factor VII (Christma3 plastin ante- FFactor X Calcium Thrombin Platelet factors I &III) globulin factor) cedent (P.T.A.)

Calcium Plasma thromboplasti Prothrosmbin

factorI

pFibrinol tc |H F eparin l enzymes / Antithrombin

Lyed Adscorbed | nacltLvated firn \ t / thrombin | thrombin

Fi1 Protected by copyright. FIG. 6.-The factors concerned in normal blood coagulation. prothrombin to thrombin during coagulation. It quantity of thrombin required to produce these appears to act as a catalyst in this reaction. changes is very much greater than that found by Thrombin formation can readily be studied by Biggs, Douglas and Macfarlane (1953a) to be re- aspirating aliquot quantities of blood during co- quired to accelerate thromboplastin generation. It agulation, and transferring them to solutions of seems doubtful therefore whether the action of fibrinogen. The thrombin content will be in- thrombin on the platelets is an important cause of versely proportional to the clotting time of the the autocatalytic action of thrombin. fibrinogen. Fig. 4 curve i shows thrombin production during the clotting of normal whole blood. The Conversion of Fibrinogen to Fibrin There is a lag period of about 3 minutes before any Fibrin begins to appear as soon as thrombin is thrombin appears. It then rapidly increasas in detectable in the blood. The action of thrombin http://pmj.bmj.com/ amount for a further 3 minutes and then rapidly in converting fibrinogen to fibrin is probably disappears, the disappearance being due to the enzymatic. Although not essential, both calcium action of antithrombin. That thrombin can and the products of platelet disintegration (platelet accelerate its own production is shown by the factor II) appear to enhance the activity of throm- observation (Fig. 4 curve 2) that if a very small bin. It has been suggested (Tocantins, I952; amount of thrombin is added to the blood as soon Owren, I952) that fibrin may increase thrombo- as it is taken, thrombin begins to appear after plastin production by forming an extensive water- on September 30, 2021 by guest. about i minute although its subsequent rate of wettable surface. Fibrin on the other hand production is not increased. It seems probable adsorbs large quantities of thrombin, thus acting that this autocatalytic property of thrombin is due as a potent antithrombin. to its capacity to accelerate the production of In conclusion, it will be seen that blood coagula- thromboplastin, an activity that can readily be tion involves several complex mechanisms. These demonstrated by means of the thromboplastin are summarized in Fig. 6. Some of these mechan- generation test (see Fig. 5). Quick (i95i) and isms tend to increase clotting and others to hold it Stefanini (I95i) believe that thrombin accelerates in check. The factors which determine the re- its own formation by causing platelet disruption, sultant of these opposing forces are ill understood, so increasing thromboplastin production. The and it is clear that many of the views expressed rate at which thrombin produces detectable here may have to be modified as further discoveries changes in the platelets is relatively slow and the are made in this rapidly advancing field. February I954 ACKROYD: The Coagulation of Human Blood 71 Postgrad Med J: first published as 10.1136/pgmj.30.340.62 on 1 February 1954. Downloaded from The author is indebted to Dr. Rosemary Biggs and KOLLER, F., LOELIGER, A., and DUCKERT, F. (I95x), Acta Dr. R. G. Macfarlane for their criticisms of the manu- Haemat., 6, I. script and to the Medical Research Council for a grant KOLLER, F., LOELIGER, A., and DUCKERT, F. (19S2), Rev. for expenses. Hematol., 7, I56. LEWIS, J. H., and FERGUSON, J. H. (I9S3), Proc. Soc. Exp. BIBLIOGRAPHY Biol. and Med., N. Y., 82, 445. ACKROYD, J. F. (I949a), Clin. Sd., 7, 249. LYTTLETON, J. W. (I95o), Thesis for degree of Ph.D., Uni- ACKROYD, J. F. (1949b), Ibid., 8, 235. versity of London. ADDIS, T. (i9II), Y. Path. Bact., 15, 427. MACFARLANE, R. G. ('953), Lecture 13 in 'Lectures on the AGGELER, P. M., WHITE, S. G., GLENDENNING, M. B., scientific basis of medicine, 195I-52,' London. PAGE, E. W., LEAKE, T. B., and BATES, G. (1952), Proc. MACFARLANE, R. G., and BIGGS, R. (1946), Lancet, il, 862 Soc. Exper. Biol. and Med., N.Y., 79, 692. MOLE, R. H. (1948), J. Path. Bact., 6o, 413. ALEXANDER, B., GOLDSTEIN, R., and LANDWEHR, G. (I950), Y. Clin. Invest., 29, 88I . MORAWITZ, P. (1905), Ergebn. Physiol., 4, 307. ALEXANDER, B., GOLDSTEIN, R., LANDWEHR, G., and DE NICOLA, P. (953), Blood, 8, 947. COOK, C. D. (I95I), Ibid., 30, 596. NOLF, P. (1908), Arch. Int. Physiol., 6, x. -ALEXANDER, B., DE VRIES, A., and GOLDSTEIN, R. (I949), Blood, 4, 739. OWEN, C. A., and BOLLMAN, J. L. (1948), Proc. Soc. Exper. ARTHUS, M., and CHAPIRO, T. (igo8), Arch. int. Physiol., 6, Biol. and Med., N. Y., 67, 231 298. OWREN, P. A-. (I944), Det Norske videnskaps-akademi i Oslo, ASTRUP, T., and PERMIN, P. M. (1948), Nature, I6I, 689. Arbok, 2I. BEAUMONT, J. L., and BERNARD, J. (I953), Acta med. scand., OWREN, P. A. (I947), 'The coagulation of blood. Investigations 145, 200. on a new clotting factor,' Gundersen, Oslo. BIDWELL, E., and MACFARLANE, R. G. (i9SI), Biochem. 7., OWREN, P. A. (ig5oa), ' International Society of Haematology 49, xlii. Third International Congress, Cambridge,' Grune & Stratton, BIGGS, R., and DOUGLAS, A. S. (I953),A. Clin. Path., 6, 23. New York, p. 379. BIGGS, R., DOUGLAS, A. S., and MACFARLANE, R. G- OWREN, P. A. (I19ob), Ibid., p. 475. (1953a),.7. Physiol., II9, 89. OWREN, P. A. Scand. .7. Clin. Lab. Invest., 3, I68. BIGGS, R., DOUGLAS, A. S., and MACFARLANE, R. G (i9SI), (1953b). Unpublished observations quoted by Biggs and OWREN, P. A. (1952), 'Handbook of biological data,' Saunders, Douglas, 1953. Philadelphia. BIGGS, R., DOUGLAS, A. S., MACFARLANE, R. G., DACIE, OWREN, P. A. (I953), Am..7. Med., 14, 20I. J. V., PITNEY, W. R., MERSKEY, C., and O'BRIEN, J. R. (1952), Brit. med. Y., ii, 1378. PATEK, A. J., and STETSON, R. P. (1936),.7. Clin. Invest., IS, BIGGS, R., and MACFARLANE, R. G. (1953a), 'Human blood 531. Protected by copyright. coagulation and its disorders,' Blackwell, Oxford. PATTON, T. B., WARE, A. G., and SEEGERS, W. H. ('948) BIGGS, R., and MACFARLANE, R. G. (s953b), Personal com- Blood, 3, 656. munication. PAVLOVSKY, A. (I947), Ibid., 2, x85. BIGGS, R., MACFARLANE, R. G., and PILLING, J. (I947), POOLE, J. C. F. (I9S3), Lancet, i, 122. Lancet, i, 402. QUICK, A. (.Q935), 7. Biol. Chem., I"9, ' Scientific proceedings BRINK, A. J., and KINGSLEY, C. S. (1952), Quart. Y. Med., xxix, p. lxXiii. 6s, I9. QUICK, A. J. (1938), Amer. 7. Physiol., 123, 712. BUDTZ-OLSEN, 0. E. (I9sI), 'Clot retraction,' Blackwell, QUICK, A. J. (I943), Ibid., 140, 212. Oxford. QUICK, A. J. (i9si), 'The physiology and pathology of haemo- CHARGAFF, E., BENDICH, A., and COHEN, S S.. (i944),7. stasis,' London. Biol. Chem., I56, I6I. ROSENTHAL, R. L., DRESKIN, 0. H., and ROSENTHAL, N. COHN, E. J. (1946), Blood, 1, 3. (I953), Proc. Soc. Exp. Biol. and Med., N.Y., 82, 171. COSGRIFF, S. W., and LEIFER, E. (1952), 7. Amer. Med. Assoc., SCHMIDT, A. (X892), Quoted by Morawitz (I905). 148, 462. SCHOFIELD, F. W. (1924),.7. Amer. Vet. Med. Ass., 64, 553. CRAMER, R., MATTER, M., and LOELIGER, A. (I953), Helv. SCHULMAN, I., and SMITH, C. H. (1952), Blood 7, 794. Paediat. Acta, 8, x85. SEEGERS, W. H., McCLAUGHRY, R. I., and FAHEY, J. L. vAN CREVELD, S., and PAULSSEN, M. M. P. (ixsi), Lancet, (1950), Ibid., 5, 421. ii, 242. LE SOURD, L., and PAGNIEZ, P. (i9i I), 7. Physiol. Path. Gin., VAN CREVELD, S., and PAULSSEN, M. M. P. (1952), Ibid., i, 23. 13, 56. DOUGLAS, A. S. (1953), Unpublished observations quoted by STEFANINI, M. (i95I), Acta med. Scand., 140, 290. http://pmj.bmj.com/ Biggs and Douglas, 1953. STOHLMAN, F., HARRINGTON, W. J., and MOLONEY, FANTL, P., and NANCE, M. H. (1946), Nature, x58, 708. W. C. (i95I),_7. Lab. Clin. Med., 38, 842. FANTL, P., and NANCE, M. H. (1948), Med. Y. Australia, I, 128. TOCANTINS, L. M. (I934), Amer. 7. Physiol., 110, 278. FRANK, E., BILHAN, VON N., and EKREN, H. (I950), Acta TOCANTINS, L. M. (1952), 'Handbook of biological data,' Haemat., 3, 70. Saunders, Philadelphia. GOVAERTS, P., and GRATIA, A. (193i), Rev. Beige Sci. Med., DE VRIES, A., ALEXANDER, B., and GOLDSTEIN, R. (I949), 3, 689. Blood, 4, 247. GRAHAM, J. B., PENICK, G. D., and BRINKHOUS, K. M. DE VRIES, A., MATOTH, Y., and SHAMIR, Z. (u95i), Acta (1951), Am. Y. Physiol., I64, 710. Haemat., 5, 129. HOWELL, W. H., and CEKADA, E. B. (1926), Ibid., 78, 500. WARE, A. G., FAHEY, J. L., and SEEGERS, W. H. (x948), JURGENS, J. (I953), Communication to the Fourth Congress of Am.Y. Physiol., 154, 140. on September 30, 2021 by guest. the European Society of Haematology. WARE, A. G., GUEST, M. M., and SEEGERS, W. H. (x947), KLEIN, P. D., and SEEGERS, W. H. (I950), Blood, 5, 742. Y. Biol. Chem., I69, 23I. KOLLER, F. (I9S3), Communication to the Fourth Congress of WARE, A. G., and SEEGERS, W. H. (I948), Ibid., x72, 699. the European Society of Haematology. WARNER, E. D., BRINKHOUS, K. M., and SMITH, H. P. KOLLER, F., KRUSI, G., and LUCHSINGER, P. (I950), (1936), Am.Y7. Physiol., 114, 667. Schweiz. med. Wschr., 8o, 110. WEIL, P. P. (I906), BEuI. Soc. Med. Hop., Paris, 23, 1001.

Copies of Title Page and Index for Vol. 29 of The Postgraduate Medical Journal are now available on request. See page I04 for binding particulars.