COAGULATION HISTORY, OXFORD 1951–53

Abstract

The purpose of this review is to record the history of coagulation studies in the Department of Clinical Pathology in Oxford, 1951–53, under the direction of R. G. Macfarlane and Rosemary Biggs. During that time the efficiency of the intrinsic pathway for prothrombin conversion was established, Christmas disease was characterized, and the thromboplastin generation test was developed. In order to understand the development of knowledge of coagulation prior to 1951 an introduction is given, based mainly on previous historical accounts. R. G. Macfarlane in 1967 said 'Other people's reminiscences can be pretty tedious' and 'One has to remember what it was like not to know things which now seem self-evident'. This was a personal reminiscence of events in development of knowledge of blood coagulation in Oxford, 1951–53. The work was done in the Department of Clinical Pathology, at the Radcliffe Infirmary, Oxford, under the leadership of Gwyn Macfarlane (1907–1987) (later Professor R. G. Macfarlane, CBE, FRS) and Dr Rosemary Biggs. I was privileged to be in their Department and to make a contribution to its work during the tenure of an MRC Research Fellowship; moreover I had the great good fortune to arrive when earlier work was about to pay dividends. In Oxford I worked on concepts provided by my supervisors; these came as a distillate of their scientific wisdom. If I had not gone to Oxford I would not have done the experiments described in our published work ( Biggs et al, 1952 ; Biggs & Douglas, 1953a, b; Biggset al, 1953a , b, c). After this academic training in blood coagulation in Oxford, I did experiments which followed my own ideas. Professor Macfarlane died in 1987 and Dr Biggs lives in retirement near Oxford. In the later part of her career, Dr Biggs was Director of the Oxford Haemophilia Centre and made an enormous contribution to the field, bringing much of Professor Macfarlane's vision to fruition. To refer to surnames nowadays seems too formal, so I shall use Gwyn and Rosemary. When I arrived in Oxford in October 1951 my technical skills were limited to performing the whole blood clotting time, the plasma recalcification time, and Quick's one-stage prothrombin time; I learned not only from my two senior supervisors but also from Dr Ethel Bidwell, who with great patience taught me the necessary biochemical methods. My interest first in haemostasis and then in thrombosis was triggered by two patients. On my first day as house physician to the late Professor J. W. McNee in the Western Infirmary, Glasgow, a teenage haemophiliac had a tooth extracted and I held a swab to the socket intermittently for 18 days. On discharge from the Army I held a second house-physician's post in Inverness, and one of the patients was a middle-aged severe haemophiliac who later died from myocardial infarction with a coronary thrombus found at autopsy. To put the 1951–53 advances in perspective, I shall trace briefly the development of knowledge on blood clotting during the previous century and a half. Most of this work came from investigators on mainland Europe. These slow advances concerned the role of tissue in the mechanism. It was known that blood collected by clean venepuncture coagulated without tissue contamination, but this phenomenon had been much neglected. There have been well-researched and detailed accounts on the history of blood coagulation. I have not re-researched this but have worked from these accounts (1909–97) and given them recognition ( Mellanby, 1909; Howell, 1935; Owren, 1947; Biggs & Macfarlane, 1953; Douglas, 1962; Quick, 1966; Tullis, 1976; Macfarlane, 1976; Ratnoff, 1980, 1996; Forbes, 1997). The two reviews by Ratnoff (1980, 1996) are commended for extended reading. I have given Buchanan's work special mention because I had access to the Proceedings of the Glasgow Philosophical Society, and because none of the previous historical accounts have given enough detail on it ( Buchanan, 1836, 1845). Although some observations were made in ancient times, I choose to start in the second half of the seventeenth century, when Malpighi (1686) separated fibres from clotted blood free of red cells and serum, and identified these using the single-lens microscope. When William Hewson, a young surgeon, was working in London in 1771 ( Hewson, 1772, 1846), there was much inflammation and other gross pathology amongst his patients, and therefore red cell sedimentation would have been rapid. He observed that the coagulum of blood arose from the upper liquid part of the blood, above the red cells, which had previously been thought to be the source of the coagulum. The term fibrin was introduced by Fourcroy (1801) working in France. He demonstrated that plasma contained soluble substances, albumin and globulin, and that the precursor of fibrin was a soluble substance present in plasma but not in serum. The term globulin was used to describe the part of plasma precipitated when diluted in water, albumins remaining in solution. These observations led to the conclusions that there was a precursor of Malpighi's fibres and the term fibrinogen was applied to it ( Babington, 1830). It was much later that crude preparations of fibrinogen were made. Denis (1859) made a preparation of fibrinogen by salt precipitation and later, using the same starting technique, Hammersten (1879/80) duplicated this, providing a purer preparation by repeated precipitation. Mellanby (1909) made an important precipitation containing fibrinogen by diluting plasma and acidifying with acetin acid following directions given by Schmidt (1892, 1895). We too used this technique in some of our work, because this precipitate also contains prothrombin freed from antithrombin. Andrew Buchanan (1798–1882), a junior surgeon in Glasgow Royal Infirmary in 1836, reported to the Glasgow Philosophical Society on experiments he had conducted earlier in 1830–31 ( Buchanan, 1836). He had taken a freshly formed blood clot and placed it in a linen cloth, squeezing the liquid into a separate container. He added this strained liquid to hydrocoele fluid and obtained a coagulum. Later in 1841, by then Professor of Physiology at the Institute of Medical Sciences in the University of Glasgow, he repeated those experiments and reported further. This more definitive paper was published in the London Medical Gazette ( Buchanan, 1845) and because of its importance was reprinted in the Journal of Physiology in 1879. In these later experiments he diluted blood with water and allowed it to clot overnight. He then removed the coagulum, washed it in water, wrapped it in a linen cloth, and squeezed out the fluid. This extract also coagulated hydrocoele fluid. In the earlier experiments Buchanan had observed that fresh serum would coagulate hydrocoele fluid, which he believed was the same as the fluid part of the blood. Buchanan (like Hewson) found that, when sedimented blood was allowed to clot, the upper part of the clot — the part with the 'buffy coat' of white corpuscles — produced most of his coagulant material, but that it was also present to some extent in the lower red-cell part. Buchanan made the additional observation that the coagulum in hydrocoele fluid did not retract in the same way as did whole blood clot (he did not know about platelets). In the Proceedings of the Glasgow Philosophical Society of February 1845, he wrote 'It is scarcely necessary for me to add that I am now satisfied that the fibrin of animal fluids exists in them in solution'. Buchanan also conducted experiments adding washed tissues to hydrocoele fluid and obtained a coagulum. These results were less convincing as the clotting time was very long, and despite the washing the tissue may have contained some of his coagulant material from serum; the role of tissue required subsequent work to substantiate its importance. Blood coagulation research in the second half of the nineteenth century was dominated by Alexander Schmidt (1831–1894) ( Schmidt, 1872, 1892, 1895; Ratnoff, 1980, 1996). Schmidt worked in Germany and later at the University of Dorpat in Estonia from 1862 onwards. Schmidt repeated the experiments described by Buchanan and called the activity 'fibrin ferment' and this term was still in use in Mellanby's papers ( Mellanby, 1909); Schmidt found that 'fibrin ferment' could be precipitated by addition of alcohol to fresh serum. Virchow was the first to use the term thrombin ( Tullis, 1976). When publishing previously, Schmidt wrote two books (1892, 1895) recounting his observations. He had the rare gift of changing his hypotheses in the light of advancing experimental information. Since the presence of thrombin in the circulation did not allow blood to remain fluid, Schmidt realized that it could not exist as such in blood but must have a precursor, prothrombin. At first, therefore, prothrombin was hypothetical. Thackrah (1819, 1834) found that blood collected after flowing over tissues — say in an open wound — coagulated more quickly than if the blood had been drawn directly from the blood vessel. De Blainville (1834) demonstrated that if macerated brain tissue was injected into an animal's vein the animal died rapidly, due to widespread coagulation. Following the lead given by Buchanan, the action of tissue was demonstrated more positively by Schmidt (1892, 1895), when he mixed blood and tissue in a container and found immediate coagulation; the role of tissue in converting the hypothetical prothrombin to thrombin was called zymoplastic. It was found that tissue which had been well washed and then macerated would accelerate whole-blood clotting, but would not coagulate Hammersten's fibrinogen. Arthus & Pagès (1890) and Sabbatini (1902) observed that calcium precipitants inhibited coagulation, and that this effect was reversed when sufficient calcium was re-added. Calcium was not required for the reaction of thrombin with fibrinogen, but was required for the conversion of the hypothetical prothrombin to thrombin. Mellanby (1909) described the inhibition of coagulation by oxalate and citrate, these being immensely valuable for subsequent studies. Citrate was used for preservation of blood for transfusion. To recapitulate (see also Howell, 1935), the central findings by the end of the nineteenth century were: (a) Alexander Schmidt (1872): Coagulation in shed blood is due to a ferment-like substance (thrombin) which exists in circulating blood as an inactive precursor (prothrombin). (b) Hammersten (1879/80): Fibrin is the consequence of a reaction between thrombin and fibrinogen. (c) Arthus & Pagès (1890): Calcium is essential in the earlier stages of the clotting process, but not in the clotting of fibrinogen by thrombin. (d) Alexander Schmidt (1892, 1895): Formation of thrombin involved an essential constituent from tissue cells. These findings were used by Morawitz (1940) and Fuld & Spiro (1904) in the formulation of the 'classical theory', which survived relatively unchallenged for about 40 years (see Fig 1). 686). Paul Morawitz (1879–1936), a 26-year-old Strasbourg physician, confirmed that tissue extracts did not coagulate fibrinogen ( Morawitz, 1904; Ratnoff, 1980). Morawitz renamed Schmidt's zymoplastic substance thrombokinase. Later, Nolf (1908) in France called this activity thromboplastic and Howell (1914) used the term 'tissue factor', both of which terms are still in use. Thus at the beginning of the twentieth century the theory was that prothrombin was converted to thrombin under the influence of thrombokinase (thromboplastin) and calcium, whereas thrombin converted fibrinogen to fibrin (see Fig 1). The mechanism for converting prothrombin using tissue was called the extrinsic pathway. It was known (see above) that blood clotted without tissue, albeit rather less quickly. This was the intrinsic pathway and was much neglected for almost half a century. After the activity which culminated in the classical theory at the end of the nineteenth century there was a relative lull until the early 1940s. However, during that period there were some important events, the significance of which were not recognized until later, and an account of these events now follows. Whipple (1912) reported the absence of prothrombin in a case of melaena neonatorum, and later ( Whipple, 1913) he found similar features in a case of jaundice. In the earlier paper he described a 3-day-old infant who developed melena and other haemorrhagic symptoms. The blood remained fluid after prolonged contact with tissue juice. Using Schmidt's alcohol-precipitation technique (as described above), no thrombin could be extracted from the infant's blood, and a deficiency of prothrombin was postulated. In his later paper, Whipple (1913) described abnormal haemorrhage in a patient with terminal obstructive jaundice due to carcinoma of the head of the pancreas. He collected oxalated plasma, and to this he added tissue (spleen extract) and calcium. The clotting-time was slower than that obtained from his preferred normal (canine) plasma. In using this technique, Whipple was almost quarter of a century ahead of Quick, who developed the one-stage prothrombin-time more definitively, using brain extract ( Quick, 1935). In 1916 the story of heparin began when Jay McLean ( McLean, 1916), a medical student, was working at Johns Hopkins in the laboratory of the great coagulationist William H. Howell (1860–1945). Again, the finding was a chance one. He studied ether/alcohol extracts of brain, heart and liver. These extracts accelerated blood clotting, but after storage for some months they had lost their coagulant activity, and indeed were now anticoagulant. Howell & Holt (1918) studied this phenomenon further, and coined the term heparin for the anticoagulant. 'Sweet clover disease' in cattle was first noted in Canada (1921–23) when excessive or fatal haemorrhage occurred after castration or dehorning in cattle. Later Roderick (1931) traced the disease to improperly cured hay made from common types of sweet clover. The disease was reversible when cattle ceased being fed on spoiled hay. Roderick found slow clotting of the blood, but measurement of prothrombin had not been adequately developed at the time. Howell (1914) had shown that a crude preparation of prothrombin could be made by acetone-precipitation of bovine plasma. Such solutions from normal bovine plasma corrected the deficit in the blood of cattle suffering 'sweet clover disease', but not when the same solutions were made from blood of animals with the disease. In the early 1940s Link isolated and synthesized dicoumarol from 'spoiled sweet clover'. The synthesis of other coumarins such as warfarin followed; these important historical events have been described by Douglas (1962). A new burst of activity was induced by the discovery of vitamin K, an event which highlighted the need for an improved method of measuring prothrombin. Henrik Dam in 1935 fed an ether-extracted diet to chickens, which developed a haemorrhagic diathesis ( Dam, 1935). None of the vitamins known at the time would correct this nutritional deficiency. The missing fat-soluble vitamin K was found in the photosynthetic portion of plants, e.g. alfalfa grass and horse chestnut leaves ( Douglas, 1962). Following the discovery of prothrombin deficiency in obstructive jaundice ( Whipple, 1913), it was established that bile salts were needed to absorb vitamin K. Appropriate preparations of vitamin K came into use in the prophylaxis and treatment of bleeding in haemorrhagic disease of the newborn, obstructive jaundice and excessive coumarin therapy. (i) One-stage prothrombin-time. This technique was described by Quick (1935) and depended on the classical theory; this suggested that, given enough tissue, calcium and fibrinogen, there is only one factor limiting the time of clotting: prothrombin. It is now recognized that this result is limited by deficiencies of factors additional to prothrombin, but it has been of the greatest importance in advancement of knowledge, and it is used world-wide in the control of coumarin therapy. Armand J. Quick (1894–1977) made immense contributions to our understanding ( Ratnoff, 1980). Of the many distinguished coagulationists I visited later in my life, he was the only one to set aside a whole day taking the history from a patient with an obscure bleeding disorder and, with the help of one technician, to work at the bench doing the experiments he thought were relevant. I understand that Oscar Ratnoff, when possible, also treated his visitors to this privilege; Oscar's experiments and thinking dominated the field for much of the second half of this century. (ii) Two-stage prothrombin time. At Iowa University, Warner, Brinkhous & Smith (1936) developed a method of measuring prothrombin which became known as the two-stage technique, in contrast to the one-stage method of Quick. Plasma (test or normal), tissue and, subsequently, calcium were mixed, and the technique measured formation of thrombin, by transferring an aliquot to fibrinogen, the coagulation time being a measure of thrombin. In the thrombin–fibrinogen reaction the shorter the clotting-time the higher the concentration of thrombin, and using this reciprocal relationship thrombin was measured in units. Ken Brinkhous (1939) reported experiments on the rate of utilization of prothrombin in whole blood delivered to a glass tube. He found that prothrombin consumption in haemophilic blood was defective; even after the start of clotting, prothrombin was utilized slowly, and the fibrin formation in haemophilic blood, once started, very The time from the formation of the first fibrin to the was In normal blood, in the fibrinogen was coagulated rapidly, and the prothrombin was converted to thrombin after clotting was Brinkhous used the two-stage technique to measure the of prothrombin in serum. Quick reported similar at to after blood had an aliquot of the serum was added to brain plasma (see and calcium, and the clotting time In addition to finding in he also that the same result could be obtained in earlier by the great & had reported this in plasma. In and the one-stage prothrombin time was but the prothrombin consumption test was (i) The one-stage prothrombin-time as by Quick (1935) made the immediate between the coagulation in and that in obstructive jaundice. a Quick had realized that there be an additional needed to a normal result in his When he oxalated not plasma, the clotting-time was prolonged and he called this phenomenon During the in in the Paul a patient with a in plasma there was a prolonged one-stage and found that this result could be corrected by adding plasma the prothrombin had been removed by an precipitate ( & for a of the The missing coagulation factor was called factor (see Fig Owren, found that factor was not present in serum, it like prothrombin, used during coagulation. Fig conversion to thrombin. The classical theory had been by the discovery of factor and factor of the observations to the discovery of factor was by & who that the addition of normal serum prothrombin factor the one-stage prothrombin-time of coumarin plasma. years other had described to which the factor was given term first used by et al, 1952 (see Fig In Oxford, in the work during 1951–53, crude by were most The of the hypothetical prothrombin was by experiments in France ( & added an precipitate to plasma, and the came to be called plasma. This would not coagulate following the addition of tissue and calcium. The of prothrombin from the was by a containing prothrombin from the calcium and by adding this to the plasma, which then coagulated In our work used as the precipitation can be used to make some of of in whole or normal plasma ( Quick, The globulin precipitate made by adding contains fibrinogen and the substance missing in globulin, The after this first precipitation can be treated by with to a factor ( Quick, These preparations can only be as

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.