CASE REPORTS

Platelet Hypersensitivity and Intravascular Coagulation in Paroxysmal Nocturnal Hemoglobinuria

DAVID STEINBERG, M.D. ANGELINA C. CARVALHO. CAROLYN

M.D.

M. CHESNEY, M.D.’

ROBERT W. COLMAN,

fv4.D.T

Boston, Messachusetts

From the Lahey Clinic and Hematology Research Laboratory, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts. This work was supported in part by NIH Grants HL-14209 and HL-16642, Special Fellowship HL-48190 (C.M.C.) and Career Development Award HL-48075 (R.W.C.). Requests for reprints should be addressed to Dr. Robert W. Colman. Manuscript accepted March 10, 1975. Present address: University of Tennessee, Center of Health Sciences, Memphis, Tennessee 38163. t Present address: Coagulation Unit of the Hematology-Oncology Section, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19 104. l

The patient described had paroxysmal nocturnal hemogtoblnurla associated with recurrent arterial as well as venous thrombosis. Study of platelet function revealed hypersensttlvlty to eptnephrine, adenosine J’phosphate (ADP) and collagen as Judged by their abillty to aggregate platelets as well as to release “C serotonln. The release of total nucleotkles was also markedly increased over normal with all aggregatlng agents. The abnormallty was localized to the platelet since aggregatlon occurred when the patient’s platelets were resuspended in normal plasma but not when normal platelets were incubated in the patient’s plasma. Presumptlve evidence for ongolng intravascular coagulation was an increase in fibrinogen derivatives of heavier molecular weight than the native protein presumably a result of thrombln action. However, factor XII was not activated and fibrlnolysls was not increased. Complement component levels and anttthrombin concentrations were also normal. The findings in this case suggest that hypersensltive platelets may contribute to the intravascular coagulation that is manlfested by the Increased Incidence of thrombosis in patients wlth paroxysmal nocturnal hemoglobinuria. Venous thrombosis is a major cause of death in paroxysmal nocturnal hemoglobinuria [ l-61. Arterial thrombosis, although much rarer, has also been described in patients who had no evidence of atherosclerosis [6-91. Initiation of clot formation has been postulated to be mediated by the release of procoaguiants from hemolyzed red cells [lo]. Although consumption coagulopathy has been documented in other cases of rapid destruction of red cells, such as falciparum malaria [ 11,121 and reactions to transfusions [ 13151, conclusive evidence that the red cell stroma initiated or accelerated plasma coagulation has not always been present. In one study, severe intravascular hemolysis of nonimmune origin failed to cause disseminated intravascular coagulation in experimental animals [ 161. Red cell stroma is a weak procoagulant which in experimental animals may fail to produce very significant intravascular clotting in the face of a normal reticuloendothelial system [ 171. Certainly severe intravascular hemolytic anemia occurs frequently without evidence of intravascular coagulation. Because thrombosis is often such a prominent feature in the morbidity and mortality associated with paroxysmal nocturnal hemoglobinuria, it seems rea-

December 1975

The American Journal of YefJklne

Volume 59

845

PAROXYSMAL NOCTURNAL HEMOCI_OSINURIA-STEINBERG

ET AL.

sonable to examine additional possible causes for the thrombotic tendency seen in paroxysmal nocturnal hemoglobinuria. We describe a patient who had both venous and arterial thrombosis with concomitant abnormalities of platelet function and plasma coagulation which may help to explain the “hypercoagulable state.” CASE REPORT A 45 year old white woman was first seen at the Lahey Clinic in May 1969 because of recurrent postprandial attacks of upper abdominal pain, nausea, vomiting and cramps of six months’ duration. Oral contraceptives had not been ingested for ten months. There was no family history of premature vascular disease. Physical examination was within normal limits except for diffuse abdominal tenderness. Laboratory data included a hemoglobin level of 11.2 g/100 ml, an hematocrit value of 34 per cent and a reticulocyte count of 4.0 per cent. A bone marrow aspirate showed erythroid hyperplasia. The glucose tolerance test was abnormal. Roentgenologic examination of the gastrointestinal tract was within normal limits. An exploratory laparotomy was performed and revealed absent pulsations In the superior mesenteric, hepatic and splenic arteries. Subsequently an aortogram demonstrated obstruction of the celiac and superior mesenteric arteries and marked narrowing of the origin of the inferior mesenteric artery (Figure 1). Because vasculiiis was suspected, the patient was treated with prednisone. She was admitted to the New England Baptist Hospital one month later with persistent abdominal pain. Reexploration revealed a cyanotic small bowel and no pulsations in the celiac and superior mesenteric arteries. A superior mesenteric endarterectomy was performed. Pathologic examination showed atheromatous material. Relief of the pain was only transitory and a repeat aortogram showed reocclusion of the same vessel. Two months later a saphenous vein graft anastamosis was performed between the aorta and superior mesenteric arteries. Because of splenic torsion, splenectomy was performed. Satisfactory pulsations returned to the distal mesenteric arcades. Pathologic examination of the superior mesenteric artery showed atherosclerosis with partially organized thrombus. The postprandial abdominal pain disappeared and did not recur. However, the patient continued to have intermittent attacks of diarrhea, lower abdominal cramps and nausea. Two years later she was readmitted with thrombophlebitis in both legs. Hemoglobin was 8.0 g/100 ml, hematocrit 26 per cent and reticulocyte count 2.4 per cent. She improved following the administration of blood transfusions and heparin. One month later she was hospitalized with another severe attack of nausea, vomiting, diarrhea and lower abdominal cramps. Physical examination was unremarkable except for an abdominal bruit and epigastric tenderness. Laboratory data included a hemoglobin level of 11.4 g/ 100 ml, a hematocrit value of 33 per cent, a reticulocyte count of 4.2 per cent, a white blood cell count of 11 ,200/mm3 with a normal differential and a platelet count of 375,000/mm3. Peripheral blood showed occasional tar-

848

December

1875

The American

Journal ol Medklne

Figure 1.

Aortogram. The arrows

the celiac axis, superior

indicate occlusions of and inferior mesenteric arteries.

get cells, macrocytes and Howell-Jolly bodies consistent with splenectomy. Folic acid was 5.8 ng/ml (normal 7 to 15 ng/ml). Serum lactic dehydrogenase (LDH) was 5 12 mu/ml (normal 120 to 220 mu/ml). The patient’s symptoms gradually subsided; however, she was readmittod to the New England Ba@ist Hospital two months later because she noted red urine. Haptoglobin was 28 mg/lOO ml (normal 33 to 195 mg/lOO ml). Serum LDH was 1,060 mu/ml. A direct Coombs’ test was negative as were a sugar water test and a Ham test. Red cell acetylcholinesterase was 0.306 delta pHIhour (normal 0.95 f 0.06 delta pHIhour). Leukocyte alkaline phosphatase was 12 (normal score 117 f 32). Complement lysis sensitivity tests performed in a manner similar to that described by Rosse et al. [30] revealed 9.9 per cent complement sensitive red cells; 70.9 per cent complement insensitive red cells and 19.2 per cent normal red cells. The diagnosis of paroxysmal nocturnal hemoglobinuria was established. MATERIALS AND METHODS Normal Subjects. Twenty normal subjects were selected from healthy laboratory personnel. They had not ingested any medications for 15 days or more.

Volume 59

PAROXYSMAL NOCTURNAL HEMOG4OBINURIA-STEINBERG

TABLE I Platelet function Studies _I ~~-- .____I__

---I_-__.-_ Platelet Platelet Count

Clot Retraction

(per ~1)

I-

Patient Normal f SD -- -_---.--_-_--___

255,000 250,000

Platelet Aggregation ADP (,uM)

f 50,000

-__

Patient Normal Mean 95% range -I___ --

0.5 1.7 1.0 - 4.0 -_I_______

-_-_-____ Epinephrine (PM)

Collagen (mdml)

0.025 3.25 0.25 -

22 89 10

Adhesiveness

(%I

Preparation of Plasma and Platelets. Platelet-rich plasma: Venous blood was collected with a “butterfly” 19 gauge thin wall siliconized needle (Abbott Laboratories, Chicago) attached to a plastic syringe. Nine volumes of blood were added to one volume of aqueous sodium citrate (0.11 M) solution in plastic tubes. The samples were centrifuged at 23OC for 10 minutes at 100 g. Plasma obtained in this way had a platelet count of 200,000 to 400,000/~l. Platelet-poor plasma: Venous blood was collected as described and centrifuged at 3,000 g for 15 minutes at 24’C. The platelet count was below 2O,OOO/~l. Gel-filtered platelets: A modification [ 181 of the method of Tangen et al. [ 191 was used. Platelet aggregometry: Platelet aggregation was studied as previously described [20]. Platelet nucleotlde release: The release of total nucleotides was measured by a modification [21] of the procedure of Wolfe and Shulman using gel-filtered platelets prepared as described. Platelet 14C-serotonln release reaction: The method of Jerushalmy and Zucker [22] was used to measure the platelet release reaction. Platelet antlheparln factor (PF4): The method of Harada and Zucker [23] was followed and the results were expressed in units of antiheparin factor per IO” platelets. Since this parameter appeared to have a log normal distribution, a 95 per cent range is substituted for the standard deviation. Effects of Normal Plasma on Patient Platelets. To ascertain whether platelet functional changes were due to an abnormality of the platelets or to a component in plasma, two procedures were employed: Aggregation studles: Gel-filtered platelets (0.3 ml) from normal subjects or from the patient were placed in the aggregometer cuvette. To each of these preparations, fresh platelet-poor plasma (0.2 ml) from normal subjects or from the patient was added with stirring, and aggregation was measured. 14C-serotonln release stud&: Gel-filtered platelets from a normal subject or the patient were labeled with 14Cserotonin. Each was then incubated with ABO compatible TABLE II

ET AL.

55-

110 -

--_-

33 32+

3

(set)

(%)

60 43 i 6.5

Platelet Factor

10

37 492

10

normal plasma of the patient’s platelet-poor plasma. After incubation of one hour with stirring, the mixture was centrifuged for 15 minutes at 1,000 g, and the percentage radioactivity released into the supernatant was counted. Gel Flttratlon of Plasma. The sample was obtained when the patient had no clinical evidence of venous or arterial thrombosis. Two milliliters of &rated plasma collected in 0.01 M +aminocaproic acid was filtered on a calibrated column of 4 per cent agarose using a modification of the technic of Fletcher et al. [24]. The fibrinogen-reactive material was measured by the staphylococcal clumping test [25]. The amount of fibrinogenjeactive material was estimated by integration of the area under each peak, and the amount which preceded fibrinogen was expressed as per cent of the total. High molecular weight fibrinogen derivatives all contained clottable protein. Assay of Prekallikreln and Kallikrein Inhlbltors. The kallikrein system was evaluated by the method of Colman et al. [26]. Assay of Factor XII. Factor XII (inactive) was measured by a modification of the activated partial thromboplastin time with the use of congenitally deficient plasma [ 271. Assay of Flbrlnogen Degradatlon Products (FDP). The staphylococcal clumping test was used for measuring FDP in the serum of patients and normal subjects and was expressed in pg/ml fibrinogenequivalents [25]. Assay of Antlthrombln Iii. This assay was performed utilizing the radial immunodiffusion method of Mancini et al. [28] with monospecific antiserums. Comp_lement System. Complement components 3 and 4 and Cl inhibitor were determined by radial immunodiffusion

[=I.

Data Analysis. All normal ranges in this report refer to mean f2 standard deviations. In some of the platelet data which have a lognormal distribution, the geometric mean and 95 per cent confkfence limits are indicated. RESULTS

Platelet Function. In the patient the platelet count, clot retraction, platelet adhesiveness and platelet factor 3 availability were within normal range (Table 1). Platelet Aggregation. The minimal concentration to produce full aggregation was determined for each of the three aggregating agents in 26 normal subjects. The patients’ platelets responded to one half the concentration of ADP, 1/lOO that of epinephrine and one quarter that of collagen when compared to the normal mean (Table II). Platelet Release Studies. Total nucleotlde release:

December 1975

The Amerkan Journalof Medkine

Volume 59

847

PAROXYSMAL NOCTURNAL HEMOGLOSINlJRIA-STEINBERG

ET AL.

The release of total nucleotides was significantly increased on three separate occasions with the three aggregating agents (Table Ill). Platelet factor 4 release: PF4 release was increased only with collagen (Table Ill). Release of 14C-serotonln: The patients’ platelets were hyperresponsive with respect to the release of 14C-serotonin. They required one half the concentration of ADP, one fifth that of epinephrine and one quarter that of collagen necessary for normal platelets to produce an equivalent release of 14C-serotonin. Complement System. All components. of the complement system tested were within the normal range. Whole hemolytic complement 160 C’Hso units/ml (normal 150 to 250 C’Hso units/ml); C3, 90 mg/lOO ml (normal 80 to 140 mg/ 100 ml): C4, 30 mg/ 100 ml (normal 20 to 30 mg/lOO ml): Cl inhibitor, 108 mg/ 100 ml (normal 80 to 120 mg/ 100 ml). Evaluation of lntrinslc Pathway Activation. All components measured were within the normal range. Factor XII was 104 per cent (normal 60 to 130 per cent), prekallikrein 145 pmol toluene sulfonyl L-arginine methyl ester (TAMe) hyd/ml/hour (normal 49 to 145 TAMe hyd/ml/hour), and kallikrein inhibitor was 0.85 kallikrein inhibitor units (normal 0.65 to 1.33 units). Heavy Molecular Welght Fibrinogen Derivatives. The concentration of heavy molecular weight fibrinogen derivatives was markedly increased in the plasma-28 per cent (normal mean 2.9 per cent of total fibrinogen) (Figure 2). Platelet Antlthrombln Ill. The levels of this inhibitor were normal, 24 mg/lOO ml (normal 15 to 29 mg/ 100 ml). Fibrinogen Degradation Products. The fibrinogen degradation products were normal, 4.8 pg/ml (normal 0 to 8.75 pg/ml). Effect of Normal Plasma on Patient Platelets. When the patient’s gel-filtered platelets were added to normal plasma, aggregation occurred (56 per cent), whereas no aggregation occurred when normal platelets were added to the patient’s plateletpoor plasma. When the patient’s gel-filtered platelets labeled with “C-serotonin were incubated with normal plasma, 50 per cent was released compared to 4 per cent in normal platelets. When normal gel-filtered platelets were incubated with the patient’s plasma, no release occurred. COMMENTS The extensive mesenteric thrombosis and atherosclerosis seen in this patient have not been previously reported in paroxysmal nocturnal hemoglobinuria.

848

December

1975

The Amerkan

Journal of Madklne

TABLE I I I Platelet ReleaseStudies Epinephrine Collagen (12.5 PM) (110 mglml)

ADP (4jJvI)

Total Nucleotide Release (mol/lO” Patient Normal Mean 95% range Platelet Patient Normal Mean 95% range

11.2 1.46 0.6 - 30

Factor

platelets)

15.1

10.9

2.23 0.8 - 3.8

4 (antiheparin

units/l

48

90

49 22 - 121

92 60-

0”

2.7 1.0 - 4.3 platelets) 245

170

120 82 - 197

In contrast to the frequency of venous occlusion, arterial thrombosis is not a common complication in this disorder. Since none of the other recognized factors predisposing to vascular disease was present in this premenopausal woman, it was postulated that her platelets might be more reactive than normal platelets, as in the case in type II hyperlipoproteinemia [ 181. This hypothesis was confirmed by the finding that the patient’s platelets responded to concentrations of reagents that do not aggregate normal platelets. Furthermore, the patient’s platelets released more nucleotides and heparin-neutralizing proteins (PF4) than those from normal persons. These findings are similar to those in familial hyper-

h

60

70

60

90

NORMAL

$00

410

SUBJECT

120

VO CuMuL A Tl VE VOL UhiE Cm fI Figure 2. Gel filtration patterns of normal and patient’s plasmas. The eluates were assayed by staphykxtxcal clumping test (SCT). The patient p&ma shows high molecular weight fibrinogen derivatives, Fd (dimer), eluted prior to the fibrinogen peak.

Volume

59

PAROXYSMAL

cholesterolemia but the patient had no lipoprotein abnormalities. The abnormality was demonstrated to be intrinsic to the platelets which aggregated spontaneously in normal plasma. Aster and his associates [29] showed that 10 per cent of the platelets in patients with paroxysmal nocturnal hemoglobinuria have a membrane defect similar to that in red blood cells which are known to be highly susceptible to lysis by complement action [301. The alternate pathway of complement activation has recently been shown to be responsible for hemolysis of red cells from patients with paroxysmal nocturnal hemoglobinuria [3 11. It is tempting to speculate that this pathway is also responsible for lysis of platelets in vivo. Released platelet coagulant factors might then accelerate intravascular coagulation accounting for the increased incidence of thrombotic episodes in paroxysmal nocturnal hemoglobinuria. Evidence supporting the occurrence of intravascular coagulation in this patient is the increase in high molecular weight fibrinogen derivatives which in vitro are produced by thrombin action on fibrinogen [32]. The measurement of high molecular weight fibrinogen derivatives by plasma gel filtration has been suggested to be a specific method for detecting intravascular coagulation [24]. The nature of the high molecular weight fibrinogen derivatives is not entirely clear, but they coukl represent fibrin polymers, fibrin monomer cross-linked by factor XIII or fibrin degradation products [ 321. Such products have also been documented in patients with venous thrombosis [24,33]. However, evaluation of the intrinsic pathway using

NOCTURNAL

HEMOGLOEINURIA-STEINBERG

ET AL.

kallikrein activation as an indicator showed no abnormalities, suggesting that coagulation was not triggered by the activation of factor XII. Antithrombin Ill was not decreased, ruling out another known cause of a thrombotic disorder. The premature mesenteric atherosclerosis seen in this patient might also be secondary to the platelet abnormality. Increased platelet aggregation and release documented in this patient have also been found in patients with type II hyperlipoproteinemia, a group prone to early and severe atherosclerosis [ 181. Mustard [34] has postulated that “in viva” platelet aggregation may not only initiate arterial thrombosis, but also may precipitate atherosclerotic changes. Release of elastase [ 351, cathepsins [ 351 and collagenase [36] by the hypersensitive platelets may digest components of the arterial wall, allowing increased permeability to lipoproteins and cholesterol with formation of a plaque. Coumadin and heparin therapy in patients with paroxysmal nocturnal hemoglobinuria are frequently ineffective [6], suggesting that platelets may play a primary role in the pathogenesis of thrombosis in this disease. The association of hypersensitive platelets with arterial occlusion has potential therapeutic implications and suggests that one of the variety of pharmacologic agents such as aspirin or dipyridamole capable of inhibiting platelet function might be useful in preventing the thrombotic complications in this disease. Another attractive possibility is clofibrate, which has been demonstrated to ameliorate platelet hypersensitivity in patients with type II hyperlipoproteinemia [37].

REFERENCES 1.

2. 3. 4. 5. 6.

7.

8.

Crosby WH:Paroxysmal nocturnal hemogtobinurla. A classic description by Paul Striibing in 1882, and a bibliography of the disease. Blood 6: 270, 1951. Crosby WH: Paroxysmal nocturnal hemoglobinuria: Rehtion of clinical manifestattons to underlying pathogenic mechanisms. Blood 8: 769, 1953. Dacie JV: The Haemolytk Anemias. Congenital and Acquired. Vol. IV. 2nd ed, New York, Crune 8 Stratton, 1967, p 1128. Hartmann RC, Jenkins DE Jr: Analytical review. Paroxysmal nocturnal hemoglobin&a: current concepts of certain pathophysiologic features. Bbod 25: 850, 1965. Manchester RC: Chronic hemolytic anemia with paroxysmat nocturnal hemoglobinurla. Ann Intern Med 23: 935, 1945. Peytremann R, Rhodes RS. Hartmann RC: Thrombosis in paroxysmal nocturnal hemoglobinurla (PNH) with partkzular reference to progressive, dtffuse hepatic venous thrombosis. Ser Haematol3; 115. 1972. Ham CD, Horack HM: Chronic hemofytic anemia with paroxysmal nocturnal hemogiobinurta. Report of a case with only occasional hemoglobtnuria and with complete autopsy. Arch Intern Med 87: 735, 1941. Melkikl A. Morstad KS: A case of paroxysmal nocturnal

9.

10.

11.

12. 13.

14.

15.

December 1975

haemoglobinuria splenectomized twice. Acta Med Stand 164: 299, 1959. Perkins HA, Boulware JM, Feichtmeier TV, Thayer WW, Spaet TH: Paroxysmal nocturnal hemoglobinuria: Report of four cases with observations on treatment with 3,3’methylene-bis (4hydroxy-coumarin/Dicumarol). Ann Intern Med 43: 1218, 1955. McKellar M, Dacie JV: Thromboplastic activity of the plasma in paroxysmal nocturnal haemoglobinuria. Br J Haematol 4: 404. 1958. Dennis LH, Eichelberger JW, lnman MM, Conrad ME: Depletion of coagulation factors in drug-resistant Plasmodtum falciparum malaria. Blood 29: 713, 1967. Devakul K, Harinasuta T, Reid HA: i251-labeled fibrinogen in cerebral malaria. Lancet 2: 886, 1966. Hardaway RM, McKay P, Williams JH: Lower nephron nephrosis (ischemic nephrosis): with speoial reference to hemorrhagic diithesis following incomaatible blood transfusion reaction. Am J Surg 87: 41. 1954. Krevans JR, Jackson DP, Conley CL, Hartmann RC: The nature of the hemorrhagic disorder accompanying hemolytic transfusion reactions in man. Blood 12: 834, 1957. Rock RC, Bove JR, Nemerson Y: Heparin treatment of in-

The American Journal of Medlclne

Volume 59

849

PAROXYSMAL NOCTURNAL HEMOGLOSlNURIA-STEINBERG

16.

17.

16. 19.

20.

21.

22.

23.

24.

25.

26.

850

ET AL.

travascular coagulation accompanying hemolytic transfusion reactions. Transfusion 9: 57, 1969. Mannucci PM, Lobina OF, Caocci L. Dioguardi N: Effect on blood coagulation of massive intravascular hemolysis. Blood 33: 207, 1969. Rabiner SF, Friedman LH: The role of intravascular haemolysis and reticulo-endothelial system in the production of a hypercoagulable state. Br J Haematol 14: 105. 1966. Carvalho A, Colman RW, Lees RS: Platelet function in hyperbetalipoproteinemla. N Engl J Med 290: 434, 1974. Tangen 0, Berman HJ, Marfey P: Gel filtration, a new technique for separation of blood platelets from plasma. Thromb Diith Haemorr 25: 266, 197 1. Born GVR: Aggregation of blood platelets by adenosine diphosphate and its reversal. Nature (Lond) 194: 927, 1962. Wolfe MS, Shulman NR: Inhibition of platelet energy production and release reaction by PGEt. theophylirie and c-AMP. Biichem Biophys Res Comm 41: 126.1970. Jerushalmy 2, Zucker MB: Some effects of fibrinogen degradation products (FDP) on blood platelets. Thromb Diath Haemorr 15: 413,1966. Harada K. Zucker MB: Simultaneous development of platelet factor 4 activity and release of 14C serotonin. Thromb Dfth Haemorr 25: 41, 1971. Fletcher AP, Alkjaersig N, O’Brien JR. Tulevski, VG: Blood hypercoagulabilii and thrombosis. Trans Am Assoc Phys 63: 159, 1970. Hawiger J, Niewiarowski S, Gurewich V: Measurement of fibrinogen and fibrin degradation products in serum by staphylococcal clumping test. J Lab Clin Med 75: 93, 1970. Colman RW, Mason JW, Sherry S: The kallikreinogenkallikrein system of human plasma. Ann Intern Med 7: 763, 1969.

December 1875

The American Jbumal of Medklae

27.

26.

29.

30.

31.

32. 33.

34. 35.

36. 37.

Volume 58

Proctor RR, Rapapport SI: The partial thromboplastin time with kaolin. A simple screening test for first-stage plasma clotting factor deficiency. Am J Clin Pathol 36: 212, 1961. Mancini G. Carbonara CO, Heremans JF: lmmunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 2: 235, 1965. Aster RH, Enright SE: A platelet and granulocyte membrane defect in paroxysmal nocturnal hemoglobinuria: Usefulness for the detection of platelet antibodies. J Clin Invest 46: 1199, 1969. Rosse WF, Dacie JV: Immune lysis of normal human and paroxysmal nocturnal hemoglobinuria (PNH) red blood cells. I. The sensitivity of PNH red cells to lysis by complement and specific antibody. J Clin Invest 45: 736, 1966. Kabakci T, Rosse WF. Logue GL: The lysis of paroxysmal nocturnal hemoglobinuria red cells by serum and cobra venom factor. Br J Hematol 23: 693. 1972. Bang NU, Chang ML: Soluble fibrin complexes. Sem Thromb Hemostasis 1: 91, 1974. Fletcher A, Alkjaersig N: Blood hypercoagulability, intravascular coagulation and thrombosis: new diagnostic concepts. Thromb Diath Haemorr suppl 45, p 369. 197 1. Mustard JF: Platelets and thrombosis in acute myocardiil infarction. Hosp Pratt 7: 115, 1972. Robert BY, Legrand G, Pignaud J, Caen J, Robert L: Activite elastinolytique associee aux plaquettes sanguines. Pathol Biol 17: 615, 1969. Chesney C, Harper E, Colman RW: Human platelet collagenase. J Clin Invest 53: 1647, 1974. Carvalho ACA, Colman RW, Lees RS: Clofibrate reversal of platelet hypersensitivity in hyperbetalipoproteinemia. Circulation 50: 570, 1974.