Selective Factor Xa Inhibition Improves Efficacy of Venous Thromboembolism Prophylaxis in Orthopedic Surgery

Philip C. Comp, M.D., Ph.D.

Disclosures

Pharmacotherapy. 2003;23(6) 

In This Article

Selective Factor Xa Inhibition

Because factor X is positioned at the start of the common pathway of the coagulation cascade, inhibition of factor Xa activity would be expected to interrupt fibrin formation initiated by either the extrinsic or intrinsic pathway. In theory, inhibition of factor Xa provides a more efficient mechanism for the control of fibrin formation than does inactivation of thrombin; this is suggested by the finding that while inactivation of one molecule of factor Xa by antithrombin III inhibits the generation of 50 thrombin molecules, inactivation of these same 50 thrombin molecules would require 1300 times as much antithrombin III.[71] Since inhibition of factor Xa leads to decreased thrombin generation rather than inactivation of thrombin's catalytic activity, factor Xa inhibition would not be expected to modulate thrombin's regulatory functions in the control of hemostasis. These regulatory functions are independent of thrombin's primary role in catalyzing the fibrinogen-fibrin transformation and include, among others, procoagulant (factors V and VIII activation), anticoagulant (protein C activation), and prothrombotic (platelet and factor XIII activation) activities.[72]

Although preclinical data suggest a theoretic basis for predicting one antithrombotic agent's superiority over another in reducing the high risk of venous thromboembolism associated with major orthopedic surgery, the controlled clinical trial remains the only true measure for evaluating a drug's actual performance. However, because individual trials, for practical reasons, are limited in the number of specific agents and comparators that can be investigated, obtaining direct clinical evidence to support one agent's superiority over many others -- and a possible reason why -- is not, in fact, realistic. To some extent, meta-analysis offers a useful tool for evaluating collective data from many individual trials, but finding disparate conclusions drawn when comparing the results of one meta-analysis with those of another is not uncommon.

In addressing the question of whether increasing factor Xa selectivity among antithrombotic agents might correlate with improved efficacy in prevention of venous thromboembolism in high-risk orthopedic surgery populations, some insight may be gained from simple inspection of collective data drawn from individual treatment arms of appropriate clinical trials in major hip and knee surgery. For this analysis, because oral anticoagulants (e.g., warfarin) interfere with the synthesis and biologic activity of all -carboxyglutamic acid-containing coagulation proteins, including the factor Xa zymogen factor X, they are the least selective of all antithrombotic agents administered for venous thromboembolism prophylaxis in patients undergoing major orthopedic surgery.[73] Among the heparin family of antithrombotic agents, unfractionated heparin is characterized by equipotent antifactor Xa and IIa activities and is thus more selective than warfarin. The LMWHs and heparinoids have even higher selectivity for Xa relative to IIa, although the proportion of antifactor IIa activity varies among different LMWH preparations. The antithrombotic agents with the highest selectivity for factor Xa inhibition belong to the two new classes of drugs that specifically target factor Xa, that is, the direct factor Xa inhibitors and the indirect factor Xa inhibitors as represented by the synthetic pentasaccharide, fondaparinux. Among these newer classes of factor Xa-specific agents, fondaparinux is the only one, to date, that has been evaluated in large phase II[64] and phase III clinical trials involving all three major orthopedic surgery indications (i.e., hip and knee replacement[74,75,76] and hip fracture[77] surgeries) and that has been approved for use in both North America and Europe.

Figure 3 summarizes the collective DVT frequencies reported for different treatment arms of clinical trials of venous thromboembolism prophylaxis in hip replacement,[74,75,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110] knee replacement,[76,78,79,80,86,87,111,112,113,114,115,116,117,118,119] and hip fracture[77,120,121,122,123,124,125,126,127] surgeries. Trials were identified through a MEDLINE search for thromboprophylaxis clinical trials in major hip and knee surgery reported from January 1978-June 2002 and were selected for inclusion in Figure 3 if efficacy data (total DVT incidence) were based on objective DVT detection, in most cases, venography or fibrinogen uptake. Most included trials were published during the last decade, with only one from 1978 and eight from 1983-1989. Data reported only in abstract form were not used in Figure 3. Antithrombotic agents appear on Figure 3 in order of increasing selectivity for factor Xa inhibition, as described above.

Efficacy data (total deep vein thrombosis [DVT] incidence) of the three classes of antithrombotic agents (oral anticoagulants [white bar], heparin family [gray bars], and indirect selective factor Xa inhibitors [black bar]) are presented for thromboprophylaxis trials in hip replacement (A), knee replacement (B), and hip fracture (C) surgeries. The antithrombotic agents appear on the x axis according to increasing factor Xa selectivity based on reported antifactor Xa:antifactor IIa ratios.[25, 43] Data for total DVT incidence (distal plus proximal) are reported as the weighted mean to take into account the relative contribution of efficacy data from each individual trial to efficacy data from all trials considered for a given anticoagulant. The total number of patients with evaluable efficacy data and the number of data sets these represent (number in parentheses) are indicated beneath each agent.

Efficacy data (total deep vein thrombosis [DVT] incidence) of the three classes of antithrombotic agents (oral anticoagulants [white bar], heparin family [gray bars], and indirect selective factor Xa inhibitors [black bar]) are presented for thromboprophylaxis trials in hip replacement (A), knee replacement (B), and hip fracture (C) surgeries. The antithrombotic agents appear on the x axis according to increasing factor Xa selectivity based on reported antifactor Xa:antifactor IIa ratios.[25, 43] Data for total DVT incidence (distal plus proximal) are reported as the weighted mean to take into account the relative contribution of efficacy data from each individual trial to efficacy data from all trials considered for a given anticoagulant. The total number of patients with evaluable efficacy data and the number of data sets these represent (number in parentheses) are indicated beneath each agent.

Efficacy data (total deep vein thrombosis [DVT] incidence) of the three classes of antithrombotic agents (oral anticoagulants [white bar], heparin family [gray bars], and indirect selective factor Xa inhibitors [black bar]) are presented for thromboprophylaxis trials in hip replacement (A), knee replacement (B), and hip fracture (C) surgeries. The antithrombotic agents appear on the x axis according to increasing factor Xa selectivity based on reported antifactor Xa:antifactor IIa ratios.[25, 43] Data for total DVT incidence (distal plus proximal) are reported as the weighted mean to take into account the relative contribution of efficacy data from each individual trial to efficacy data from all trials considered for a given anticoagulant. The total number of patients with evaluable efficacy data and the number of data sets these represent (number in parentheses) are indicated beneath each agent.

Although the data as presented in Figure 3 are limited by the unequal number of controlled trials published for the various antithrombotic agents and also lack the rigor of a formal meta-analysis, some preliminary and general inferences may be drawn. For each indication, the suggested trend is a decrease in DVT frequency as factor Xa selectivity increases. In all three major orthopedic surgery indications, fondaparinux, the antithrombotic agent with the highest factor Xa selectivity, appears to have the lowest DVT incidence of all the antithrombotic agents represented; although, in the case of hip fracture and knee replacement this is based on only one fondaparinux trial each (albeit with very large randomized study populations, 1711[77] and 1049[76] patients, respectively). Although the suggested relationship between increasing selectivity of factor Xa inhibition and improved efficacy in DVT prevention is not conclusive, the superior efficacy of fondaparinux compared with that of the antithrombotic agents of the heparin family (as suggested in Figure 3 and demonstrated for the LMWH enoxaparin in clinical trials discussed in the following paragraphs) raises the interesting question of how important the capacity to inhibit prothrombinase-associated factor Xa really is, since these two distinct classes of antithrombotic agents each favor inhibition of circulating, free factor Xa. In the absence of head-to-head comparisons in clinical trials, the potential superiority of the direct factor Xa inhibitors, which inhibit free as well as prothrombinase-associated factor Xa, is theoretic.

Direct clinical evidence from the recently completed phase III trials comparing fondaparinux with the LMWH enoxaparin for prevention of venous thromboembolism in hip and knee replacement and hip fracture surgeries argues in support of a relationship between increasingly selective factor Xa inhibition and improved efficacy. These trials revealed a significant, overall 55.2% odds reduction in venous thromboembolism risk (p<0.001) in favor of fondaparinux,[128] which is an exclusive factor Xa inhibitor, compared with enoxaparin, which inhibits factors Xa and IIa in a ratio of 4:1.[25] The administration of fondaparinux compared with enoxaparin significantly reduced the frequency of both proximal and distal DVT in major knee surgery[76] and in major hip surgeries.[74,75,77]

Each of the phase III trials compared the same fondaparinux regimen (2.5 mg/day starting 6 ± 2 hrs after surgical closure) with approved regimens of enoxaparin (30 mg twice/day starting 12-24 hrs after surgery[74,76] or 40 mg/day starting 12 hours before surgery[75,77]). Thus, the first postsurgical dose of enoxaparin was given later than the first postsurgical dose of fondaparinux in two of the trials. Some have suggested that this difference in the timing of administration could have affected the efficacy or safety results. However, the phase III trials used approved regimens of the comparator enoxaparin, and the approved postsurgical regimen does not allow the first dose of enoxaparin to be given earlier than 12 hours after major orthopedic surgery, possibly because of the high risk of major bleeding reported when enoxaparin was started within 8 hours after major knee surgery.[111] Related to this, a recent post hoc analysis demonstrated that the timing of the start of fondaparinux prophylaxis, within a range of 3-9 hours after surgery, did not affect efficacy (p=0.67), although it did affect the frequency of bleeding.[128]

A possible correlation between clinical safety (i.e., bleeding complications) and relative factor Xa selectivity of antithrombotic agents is more difficult to evaluate systematically because of the variation in clinical safety end points -- many often subjective -- used in venous thromboembolism prophylaxis trials.[129,130] Moreover, measurement of circulating antifactor IIa and antifactor Xa levels during antithrombotic therapy and analysis of these data for a possible correlation between relative levels and clinical safety end points has not been a priority in clinical trials. This is the case despite the fact that at least one of the rationales for the development of LMWHs was predicated on preclinical evidence suggesting that antifactor IIa activity in unfractionated heparin preparations contributed to the serious medical complications, including bleeding, associated with administration of unfractionated heparin as an antithrombotic agent.

Bleeding complications do, in fact, pose a serious drawback to the use of the heparin family of antithrombotic agents, particularly unfractionated heparin. A 30% increase in the odds of excessive bleeding or the need for transfusion was reported with prophylactic administration of subcutaneous unfractionated heparin,[131] whereas bleeding in one of the first heparin trials for prevention of venous thromboembolism in orthopedic surgery was a major reason for cessation of the study.[132] Frequent laboratory monitoring of aPTT, therefore, is standard with heparin therapy, both to ensure adequate anticoagulation for maximum efficacy and to prevent excessive anticoagulation and its associated bleeding risk ( Table 1 ).

The safety benefit of LMWHs for routine venous thromboembolism prophylaxis in major orthopedic surgery is controversial. Some studies have reported an increase in bleeding complications, relative to warfarin, with frequencies of 1-5%[78,81,107,133] and 3-12%[78,111,134] found for hip and knee replacement surgery populations, respectively. Findings from other trials in major orthopedic surgery indicate generally lower frequencies and equivalent safety when LMWHs are compared with either unfractionated heparin[88,89,96,115] or warfarin.[80]

Further support for the safety of LMWH for prevention of venous thromboembolism after orthopedic surgery comes from findings in clinical trials of extended venous thromboembolism prophylaxis (> 4 wks). Studies comparing enoxaparin with placebo after hip and knee replacement surgeries showed a low rate of occurrence of hemorrhage or minor bleeding complications (< 1-4%), with no significant difference between the two groups.[135,136] These disparate reports only partially reflect the extent of the controversy surrounding the issue of bleeding complications with LMWH therapy and the resultant reluctance on the part of some orthopedic surgeons to employ LMWHs for venous thromboembolism prophylaxis because of the potential risks that even minor bleeding complications pose to successful surgical outcomes.[137] Wound hematomas, for example, are of particular concern in major hip and knee surgery because of infection risk associated with bleeding at the surgical site and subsequent septic loosening of the prosthesis, requiring repeat surgery.[110,138] Finally, a cautionary FDA warning regarding the administration of all LMWHs and heparinoids in patients undergoing neuraxial block before surgery has alerted the medical community to anecdotal reports of increased occurrences of spinal hematomas in patients in whom anticoagulation with enoxaparin was combined with regional anesthesia.[139] This warning was extended with the recent FDA mandate that package inserts for all anticoagulants must contain a cautionary note regarding the potential risk of spinal hematoma in patients undergoing anticoagulation in conjunction with neuraxial anesthesia or indwelling catheters for analgesia delivery.

The safety of fondaparinux compared with that of enoxaparin has been demonstrated in the phase II[64] and phase III[74,75,76,77] trials of venous thromboembolism prevention in major hip and knee surgeries. In the phase III clinical program, fondaparinux was found to be as safe as enoxaparin, based on a similar low occurrence of clinically relevant bleeding and no increase in wound infection.[128] Some concern has been raised in the medical community regarding the safety of fondaparinux in view of the increased frequency of major bleeding events in the knee replacement trial, as defined by one of the safety outcome measures (i.e., bleeding index ≥ 2). However, a recently reported post hoc analysis of the entire fondaparinux clinical database demonstrated a significant relationship between the timing of the first dose of fondaparinux administration and the frequency of major bleeding (3-9 hrs, p=0.008) and a significant relationship between timing and the frequency of major bleeding associated with a bleeding index of 2 or greater (3-9 hrs, p=0.008).[128] As already discussed, no such relationship was found between the timing of the first dose of fondaparinux administration and the drug's efficacy. Thus, when fondaparinux is administered according to the recommended regimen, especially with respect to the timing of the first dose after surgery (i.e., 6-8 hrs after surgery), its superior efficacy relative to that of approved enoxaparin regimens is not compromised by any increase in clinically relevant bleeding.

Of importance, the frequencies of fatal bleeding, critical organ bleeding, and bleeding events leading to repeat surgery did not differ for both fondaparinux and enoxaparin according to age, sex, body mass index, or duration of surgery.[129] Moreover, analysis of 1266 "fragile" patients (those patients aged > 75 yrs or with creatinine clearance 30-50 ml/min or body weight < 50 kg) in the phase III fondaparinux trials revealed that when the first active postoperative injection of fondaparinux was administered 6 hours or more after surgery, the frequency of major bleeding in the fragile patient population was similar to that in the other patients (1.8% in both groups).[140] These data indicate, therefore, that timing of the first postoperative dose of fondaparinux is a critical factor in controlling bleeding risk in special patient populations. Based on these collective findings, in the absence of any increase in clinically relevant bleeding complications, the overall 55% reduction in venous thromboembolism risk with fondaparinux, compared with enoxaparin, suggests a true clinical advantage of selective factor Xa inhibition for venous thromboembolism prophylaxis in major orthopedic surgery.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....