Rationale for Development of Low-Molecular-Weight Heparins and Their Clinical Potential in the Prevention of Postoperative Venous Thrombosis Jack Hirsh,
MD, Hamilton, Ontario, Canada
Interest
in low-molecular-weight heparins antithrombotic agents was stimulated by two observations in the mid-1970s and early 1980s. The first was the finding that LMWH fractions prepared from unfractionated heparin (UFH) progressively lost their ability to prolong the activated partial thromboplastin time (APTT) while retaining their ability to inhibit Factor Xa. The second was the observation that LMWHs prepared by chemical depolarization of UFH are antithrombotic in experimental animal models but produce less microvascular bleeding in experimental models for an equivalent antithrombotic effect than the UFH from which they are derived. Subsequently, it was shown that LMWHs inhibit platelet function and impair vascular permeability less than standard heparin and that LMWHs have a longer biological half-life than standard heparin. A number of LMWHs have been evaluated in clinical trials in general and orthopedic surgery and in the treatment of venous thrombosis. LMWHs are highly effective in orthopedic surgery, where they appear to be more effective than standard heparin. LMWHs have also been shown to be either as effective or more effective than UFH in preventing postoperative thrombosis following general surgery. In preliminary studies, LMWHs appear to be as effective as standard heparin in the treatment of venous thrombosis, but larger studies are required using clinically relevant outcome measures.
( LMWHs) as potential
I
nterest in low-molecular-weight heparins (LMWHs) as potential antithrombotic agents was stimulated by two observations in the mid-1970s and early 1980s. The first was the finding by Johnson and associates [I] and Andersson and associates [2] that LMWH fractions prepared from standard unfractionated heparin (UFH) progressively lose their ability to prolong the activated partial thromboplastin time (APTT) while retaining their ability to inhibit Factor Xa. The second was the observation that LMWHs produce less bleeding in experimental models for an equivalent antithrombotic effect than the UFH from which they are derived [J]. UFH is a heterogeneous mixture of sulfated polysaccharide chains ranging in molecular weight from 5,000 to 35,000 with a mean molecular weight of between 12,000 and 15,000. The UFH chains consist of alternating residues of D-glucosamine and an uranic acid, either glucouronic acid or iduronic acid. Most of the anticoagulant activity of heparin is accounted for by a unique pentasaccharide with a high-affinity binding sequence to antithrombin III (AT III). Only about one third of the commercial heparin molecules contain the unique pentasaccharide, and its distribution along the heparin chain appears to be random [4-91. PRODUCTION
OF LMWHs
The LMWHs in clinical use are produced from UFH by depolymerization by one of four methods. These are: by treatment with nitrous acid, which causes deamination of N-sulfate glucosamine residues; by treatment with the enzyme heparinase; by hydrolytic cleavage with hydrogen peroxide; or by @-elimination. These LMWHs contain the unique pentasaccharide required for binding to AT III but in lower proportions than the standard heparin from which they have been derived. Two heparinoids are also being evaluated clinically. These are dermatan sulfate and ORG 10172, which is a mixture of heparin sulfate, dermatan sulfate, a LMWH fraction, and chondroitin sulfate. ANTICOAGULANT AND LMWH
EFFECTS
OF HEPARIN
From Hamilton Civic Hospitals Research Centre, Henderson General Division, Hamilton, Ontario, Canada. Requests for reprints should be addressed to Jack Hirsh, MD, Hamilton Civic Hospitals Research Centre, Henderson General Division, 711 Concession Street, Hamilton, Ontario, Canada L8V lC3.
The major effect of heparin as an anticoagulant is due to the heparin-antithrombin interaction [7,IO] which induces a conformational change in antithrombin [11-131 which is responsible for acceleration of the inactivation of clotting enzymes. Heparin increases the rate at which AT III forms inactive complexes with coagulation enzymes [7]. After complexing with these coagulation enzymes, AT III is cleaved at the reactive bond [14-171.
5 12
16 1
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
APRIL
1991
DEVELOPMENT
Heparin serves as a template to which both AT III and thrombin bind [18]. Ternary complexes in which antithrombin and proteinase bind to the same heparin molecule are involved in the mechanism of the reaction with thrombin but have only a minor role in the Factor Xa reaction. Heparin molecules with fewer than 18 saccharides are unable to bind to thrombin and AT III simultaneously, and, therefore, are unable to form ternary complexes necessary for the catalysis of thrombin inhibition by AT III. However, heparin fragments that contain the high-affinity pentasaccharide sequence and that contain fewer than 18 saccharides are able to catalyze the inhibition of Factor Xa by AT III because binding to Factor Xa by heparin is not required to achieve catalysis of inhibition of Factor Xa by heparin [5,29-231. When AT III binds to thrombin or Factor Xa, it acquires a reduced affinity to heparin [21], allowing heparin to dissociate from the complex and become available to catalyze the inhibition of other coagulation enzymes. Thrombin is inhibited more rapidly than Factor Xa by AT III either in the absence or presence of UFH. LMWHs catalyze thrombin inhibition much less efficiently than UFH; the difference between the catalytic action of heparin on Factor Xa and thrombin inhibition is inversely proportional to the molecular weight of the glycosaminoglycans. LMWH also catalyzes Factor Xa inhibition slightly less efficiently than UFH [8,24]. The requirement for an additional sequence of 13 saccharide units over and above the unique pentasaccharide sequence to obtain acceleration of the antithrombinthrombin reaction reflects the need for the glycosaminoglycan to bind to the thrombin molecule by non-specific electrostatic interaction while the inhibition of the enzyme by AT III is facilitated by the conformational change induced in antithrombin by binding to the highaffinity pentasaccharide [4,1 l-l 31. A chemically synthesized high-affinity pentasaccharide containing only the antithrombin-binding sequence of heparin accelerates the inhibition of Factor Xa by antithrombin to nearly the same extent as standard heparin [6l. LMWHs also prolong the APTT to a lesser extent than UFH [25-271. Some studies suggest that LMWHs catalyze Factor Xa inhibition within the prothrombinase complex less efficiently than UFH [27], while others have reported that some LMWHs are more effective than UFH in inhibiting thrombin formation in platelet rich plasma [28]. LMWHs of a chain length less than 18 saccharide units are resistant to neutralization with platelet factor 4, becoming progressively more resistant to neutralization with decreasing molecular size [22]. Similarly, the neutralizing activity of protamine sulfate is influenced by the heparin chain length since anti-Factor Xa activity of LMWH is not completely neutralized by protamine [ 141. ANTITHROMBOTIC AND HEPARINOIDS ANIMAL MODELS
EFFECTS OF LMWHs IN EXPERIMENTAL
The antithrombotic effects of LMWHs, heparin, and heparinoids have been evaluated in experimental models THE AMERICAN
OF LOW-MOLECULAR-WEIGHT
HEPARINS
of venous thrombosis by a number of groups. In these models, activation of blood coagulation is induced by the injection of serum, Factor Xa, thrombin, or tissue factor [3,29-311. Heparin, LMWHs, and dermatan sulfate are all effective in inhibiting thrombosis. On a weight-forweight basis, UFH is the most effective glycosaminoglycan, whereas LMWHs with a mean molecular weight of approximately 5,000 are more effective than dermatan sulfate or heparin sulfate. When compared on a weight basis, unfractionated heparin is approximately twice as effective in preventing experimental venous thrombosis as the LMWHs in current clinical use. The effectiveness of these glycosaminoglycans as antithrombotic agents has also been studied in experiments designed to simulate a therapeutic model. The ability of the various glycosaminoglycans to inhibit accretion of radioactive fibrin onto experimental jugular vein thrombi has also been compared. On a weight-for-weight basis, UFH is the most effective glycosaminoglycan. LMWHs are more effective than dermatan sulfate or the heparinoid ORG 10172 [32,33]. HEMORRHAGIC LMWHs
EFFECTS OF HEPARIN AND
A number of investigators have reported that LMWH fractions are less hemorrhagic than UFH for an equivalent antithrombotic activity [3,29,34,35], while other investigators [36,37] have found that equivalent antithrombotic doses of a particular LMWH fraction have similar hemorrhagic effects as UFH. Differences in the degree of sulfation of the various LMWHs may have contributed to the variations in hemorrhagic effects reported [31]. There is evidence that differences between the effects of UFH and LMWHs on platelet aggregation may be responsible for the differences in the hemorrhagic properties of these glycosaminoglycans. Salzman and associates [38] were the first to report that LMWH gave less enhancement of ADP-induced platelet aggregation than UFH, and similar results have been reported by others. Our group reported marked differences on the effects of UFH and a variety of LMWHs on collagen-induced platelet aggregation ex vim at concentrations of the glycosaminoglycans that induced hemorrhagic effects. A good correlation was found between inhibition of platelet aggregation and bleeding induced by these glycosaminoglycans [31,39&l, suggesting that the observed association is causal. Another potential mechanism for the differences in hemorrhagic effects between UFH and LMWH has been suggested by Blajchman and associates [41]. These investigators reported that doses of UFH that increase the bleeding time in rabbits also increase vessel permeability. In contrast to UFH, similar doses of LMWHs had a markedly lesser effect on vascular permeability [41]. CLINICAL POTENTIAL
OF LMWHs
A number of different LMWHs have been approved for use in Europe, and some are undergoing clinical evaluation in North America. It is now clear that LMWHs have a longer half-life than standard heparin and are JOURNAL
OF SURGERY
VOLUME
161
APRIL
1991
513
TABLE I Randomized Comparlsons Between Standard Heparin and Low-Molecular-Weight Heparlns or Heparln Analogues In General Surgery Dose Anti-Xa Units
First Author Fraxiparin (CY 216: Choav/Sanofi) Kakkar (1965) European (1966)
7,500 7,500
Dose Interval
u u
Daily Daily
Total
7,500 2,500 5,000 2,500 5,000 5,000
u U u U u u
Daily Daily Daily Daily Daily Daily
Total (PK 10169: (1988)
14/199 42/936
(0.027)
56/1,135
(0.070) (0.043)
(“/)
MD, Hamilton, Ontario, Canada
Interest
in low-molecular-weight heparins antithrombotic agents was stimulated by two observations in the mid-1970s and early 1980s. The first was the finding that LMWH fractions prepared from unfractionated heparin (UFH) progressively lost their ability to prolong the activated partial thromboplastin time (APTT) while retaining their ability to inhibit Factor Xa. The second was the observation that LMWHs prepared by chemical depolarization of UFH are antithrombotic in experimental animal models but produce less microvascular bleeding in experimental models for an equivalent antithrombotic effect than the UFH from which they are derived. Subsequently, it was shown that LMWHs inhibit platelet function and impair vascular permeability less than standard heparin and that LMWHs have a longer biological half-life than standard heparin. A number of LMWHs have been evaluated in clinical trials in general and orthopedic surgery and in the treatment of venous thrombosis. LMWHs are highly effective in orthopedic surgery, where they appear to be more effective than standard heparin. LMWHs have also been shown to be either as effective or more effective than UFH in preventing postoperative thrombosis following general surgery. In preliminary studies, LMWHs appear to be as effective as standard heparin in the treatment of venous thrombosis, but larger studies are required using clinically relevant outcome measures.
( LMWHs) as potential
I
nterest in low-molecular-weight heparins (LMWHs) as potential antithrombotic agents was stimulated by two observations in the mid-1970s and early 1980s. The first was the finding by Johnson and associates [I] and Andersson and associates [2] that LMWH fractions prepared from standard unfractionated heparin (UFH) progressively lose their ability to prolong the activated partial thromboplastin time (APTT) while retaining their ability to inhibit Factor Xa. The second was the observation that LMWHs produce less bleeding in experimental models for an equivalent antithrombotic effect than the UFH from which they are derived [J]. UFH is a heterogeneous mixture of sulfated polysaccharide chains ranging in molecular weight from 5,000 to 35,000 with a mean molecular weight of between 12,000 and 15,000. The UFH chains consist of alternating residues of D-glucosamine and an uranic acid, either glucouronic acid or iduronic acid. Most of the anticoagulant activity of heparin is accounted for by a unique pentasaccharide with a high-affinity binding sequence to antithrombin III (AT III). Only about one third of the commercial heparin molecules contain the unique pentasaccharide, and its distribution along the heparin chain appears to be random [4-91. PRODUCTION
OF LMWHs
The LMWHs in clinical use are produced from UFH by depolymerization by one of four methods. These are: by treatment with nitrous acid, which causes deamination of N-sulfate glucosamine residues; by treatment with the enzyme heparinase; by hydrolytic cleavage with hydrogen peroxide; or by @-elimination. These LMWHs contain the unique pentasaccharide required for binding to AT III but in lower proportions than the standard heparin from which they have been derived. Two heparinoids are also being evaluated clinically. These are dermatan sulfate and ORG 10172, which is a mixture of heparin sulfate, dermatan sulfate, a LMWH fraction, and chondroitin sulfate. ANTICOAGULANT AND LMWH
EFFECTS
OF HEPARIN
From Hamilton Civic Hospitals Research Centre, Henderson General Division, Hamilton, Ontario, Canada. Requests for reprints should be addressed to Jack Hirsh, MD, Hamilton Civic Hospitals Research Centre, Henderson General Division, 711 Concession Street, Hamilton, Ontario, Canada L8V lC3.
The major effect of heparin as an anticoagulant is due to the heparin-antithrombin interaction [7,IO] which induces a conformational change in antithrombin [11-131 which is responsible for acceleration of the inactivation of clotting enzymes. Heparin increases the rate at which AT III forms inactive complexes with coagulation enzymes [7]. After complexing with these coagulation enzymes, AT III is cleaved at the reactive bond [14-171.
5 12
16 1
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
APRIL
1991
DEVELOPMENT
Heparin serves as a template to which both AT III and thrombin bind [18]. Ternary complexes in which antithrombin and proteinase bind to the same heparin molecule are involved in the mechanism of the reaction with thrombin but have only a minor role in the Factor Xa reaction. Heparin molecules with fewer than 18 saccharides are unable to bind to thrombin and AT III simultaneously, and, therefore, are unable to form ternary complexes necessary for the catalysis of thrombin inhibition by AT III. However, heparin fragments that contain the high-affinity pentasaccharide sequence and that contain fewer than 18 saccharides are able to catalyze the inhibition of Factor Xa by AT III because binding to Factor Xa by heparin is not required to achieve catalysis of inhibition of Factor Xa by heparin [5,29-231. When AT III binds to thrombin or Factor Xa, it acquires a reduced affinity to heparin [21], allowing heparin to dissociate from the complex and become available to catalyze the inhibition of other coagulation enzymes. Thrombin is inhibited more rapidly than Factor Xa by AT III either in the absence or presence of UFH. LMWHs catalyze thrombin inhibition much less efficiently than UFH; the difference between the catalytic action of heparin on Factor Xa and thrombin inhibition is inversely proportional to the molecular weight of the glycosaminoglycans. LMWH also catalyzes Factor Xa inhibition slightly less efficiently than UFH [8,24]. The requirement for an additional sequence of 13 saccharide units over and above the unique pentasaccharide sequence to obtain acceleration of the antithrombinthrombin reaction reflects the need for the glycosaminoglycan to bind to the thrombin molecule by non-specific electrostatic interaction while the inhibition of the enzyme by AT III is facilitated by the conformational change induced in antithrombin by binding to the highaffinity pentasaccharide [4,1 l-l 31. A chemically synthesized high-affinity pentasaccharide containing only the antithrombin-binding sequence of heparin accelerates the inhibition of Factor Xa by antithrombin to nearly the same extent as standard heparin [6l. LMWHs also prolong the APTT to a lesser extent than UFH [25-271. Some studies suggest that LMWHs catalyze Factor Xa inhibition within the prothrombinase complex less efficiently than UFH [27], while others have reported that some LMWHs are more effective than UFH in inhibiting thrombin formation in platelet rich plasma [28]. LMWHs of a chain length less than 18 saccharide units are resistant to neutralization with platelet factor 4, becoming progressively more resistant to neutralization with decreasing molecular size [22]. Similarly, the neutralizing activity of protamine sulfate is influenced by the heparin chain length since anti-Factor Xa activity of LMWH is not completely neutralized by protamine [ 141. ANTITHROMBOTIC AND HEPARINOIDS ANIMAL MODELS
EFFECTS OF LMWHs IN EXPERIMENTAL
The antithrombotic effects of LMWHs, heparin, and heparinoids have been evaluated in experimental models THE AMERICAN
OF LOW-MOLECULAR-WEIGHT
HEPARINS
of venous thrombosis by a number of groups. In these models, activation of blood coagulation is induced by the injection of serum, Factor Xa, thrombin, or tissue factor [3,29-311. Heparin, LMWHs, and dermatan sulfate are all effective in inhibiting thrombosis. On a weight-forweight basis, UFH is the most effective glycosaminoglycan, whereas LMWHs with a mean molecular weight of approximately 5,000 are more effective than dermatan sulfate or heparin sulfate. When compared on a weight basis, unfractionated heparin is approximately twice as effective in preventing experimental venous thrombosis as the LMWHs in current clinical use. The effectiveness of these glycosaminoglycans as antithrombotic agents has also been studied in experiments designed to simulate a therapeutic model. The ability of the various glycosaminoglycans to inhibit accretion of radioactive fibrin onto experimental jugular vein thrombi has also been compared. On a weight-for-weight basis, UFH is the most effective glycosaminoglycan. LMWHs are more effective than dermatan sulfate or the heparinoid ORG 10172 [32,33]. HEMORRHAGIC LMWHs
EFFECTS OF HEPARIN AND
A number of investigators have reported that LMWH fractions are less hemorrhagic than UFH for an equivalent antithrombotic activity [3,29,34,35], while other investigators [36,37] have found that equivalent antithrombotic doses of a particular LMWH fraction have similar hemorrhagic effects as UFH. Differences in the degree of sulfation of the various LMWHs may have contributed to the variations in hemorrhagic effects reported [31]. There is evidence that differences between the effects of UFH and LMWHs on platelet aggregation may be responsible for the differences in the hemorrhagic properties of these glycosaminoglycans. Salzman and associates [38] were the first to report that LMWH gave less enhancement of ADP-induced platelet aggregation than UFH, and similar results have been reported by others. Our group reported marked differences on the effects of UFH and a variety of LMWHs on collagen-induced platelet aggregation ex vim at concentrations of the glycosaminoglycans that induced hemorrhagic effects. A good correlation was found between inhibition of platelet aggregation and bleeding induced by these glycosaminoglycans [31,39&l, suggesting that the observed association is causal. Another potential mechanism for the differences in hemorrhagic effects between UFH and LMWH has been suggested by Blajchman and associates [41]. These investigators reported that doses of UFH that increase the bleeding time in rabbits also increase vessel permeability. In contrast to UFH, similar doses of LMWHs had a markedly lesser effect on vascular permeability [41]. CLINICAL POTENTIAL
OF LMWHs
A number of different LMWHs have been approved for use in Europe, and some are undergoing clinical evaluation in North America. It is now clear that LMWHs have a longer half-life than standard heparin and are JOURNAL
OF SURGERY
VOLUME
161
APRIL
1991
513
TABLE I Randomized Comparlsons Between Standard Heparin and Low-Molecular-Weight Heparlns or Heparln Analogues In General Surgery Dose Anti-Xa Units
First Author Fraxiparin (CY 216: Choav/Sanofi) Kakkar (1965) European (1966)
7,500 7,500
Dose Interval
u u
Daily Daily
Total
7,500 2,500 5,000 2,500 5,000 5,000
u U u U u u
Daily Daily Daily Daily Daily Daily
Total (PK 10169: (1988)
14/199 42/936
(0.027)
56/1,135
(0.070) (0.043)
(“/)
