Adequate Anticoagulants for HD in HIT Patients

Patients with renal failure show plasma hypercoagulability as well as uremic platelet defects, both of which can be worsened by HD (Ambuhl et al., 1997; Sreedhara et al., 1995; Vecino et al., 1998). Therefore, selection of an appropriate anticoagulant in HD patients who also suffer from HIT is difficult.

Reports on specific anticoagulant strategies in HIT are anecdotal. Large studies, especially those comparing different anticoagulant regimens, are lacking. Therefore, no treatment recommendations based on level A or B evidence (i.e., randomized trials) can be provided. Furthermore, because UFH is the routine anticoagulant in use for HD, considerable additional time, effort, and costs are usually required to manage a new anticoagulant for HD, especially during initial use. Ideally, therefore, a center should try to gain experience with a single appropriate alternative anticoagulant for management of these difficult patients. Fear of inducing bleeding should not be used to justify under-anticoagulation, with the potential risk for thrombotic complications.

Danaparoid Sodium

Danaparoid sodium (Orgaran, formerly known as Org 10172) is the alternative anticoagulant that has been most widely used for management of HD in patients with HIT (Chong and Magnani, 1992; Greinacher et al., 1992a, 1993; Henny et al., 1983; Magnani, 1993; Magnani and Gallus, 2006; Neuhaus et al., 2000; Ortel et al., 1992; Roe et al., 1998; Tholl et al., 1997; Wilde and Markham, 1997). However, danaparoid is currently not available in the United States (see Chapter 13). Some of its characteristics require specific attention:

1. The anticoagulant activity of danaparoid can be monitored only by measurement of antifactor Xa levels based on a danaparoid calibration curve; however, many laboratories do not routinely perform these assays. Except for an emergency situation, such as when HIT is strongly suspected and danaparoid is the only available alternative, HD should not be performed without monitoring the antifactor Xa activity to evaluate the dose required for adequate anticoagulation. Once the optimal dose is identified, it can often be used without alteration for several subsequent HD sessions, provided no bleeding or inappropriate clotting occurs and no surgical intervention is scheduled. Periodic measurement of antifactor Xa activity thereafter to validate the dosing of danaparoid is recommended. For maintenance HD without complications, single determination of pre-HD antifactor Xa activity probably suffices. If there are concerns about adequate or excess anticoagulation, then monitoring of levels at three time points is appropriate (e.g., 30-60 min pre-HD, 30 min after beginning HD, and just before completion).

2. Regarding the pharmacokinetics of danaparoid, renal excretion accounts for approximately 40-50% of total plasma clearance; accordingly, diminished clearance of antifactor Xa activity occurs in HD patients (Danhof et al., 1992). The elimination half-life of the antifactor Xa activity (about 24 h in healthy individuals) (Danhof et al., 1992) may reach as high as 4 days (unpublished observations of the author). Thus, significant antifactor Xa levels can be detected in patients undergoing HD with danaparoid even during the inter-dialytic interval. Whether this yields clinical benefit, such as decreased risk of thrombosis or greater maintenance of vascular access, is unknown. An increase in interdialytic bleeding episodes has not been reported.

3. Given its pharmacokinetics, danaparoid is given by initial bolus in intermittent HD, which normally is sufficient to prevent clotting within the extracorporeal circuit during the procedure. Danaparoid anticoagulation may also be useful in critically ill patients on continuous renal replacement therapy. In 13 consecutive intensive care unit (ICU) patients clinically suspected to have HIT, danaparoid was administered by initial bolus followed by continuous infusion (LindhoffLast et al., 2001). This regimen was sufficient to prevent clotting within the extracorporeal circuit both in continuous venovenous hemofiltration (eight patients) and in continuous venovenous HD (five patients), respectively (Lindhoff-Last et al., 2001). Thromboembolic complications did not occur. Despite a mean danaparoid infusion rate of approximately 140 U/h, which is markedly reduced compared to the recommendation of the manufacturer, major bleeding was observed in six of 13 patients (which could be explained by disseminated intravascular coagulation in five patients). However, HIT was confirmed by antibody detection in only two patients. Thrombocytopenic patients not having the prothrombotic state of acute HIT likely are at increased bleeding risk. Therefore, dosing of danaparoid in ICU should be based on the individual patient's risk of bleeding versus thrombosis. With regard to invasive procedures, the long half-life of danaparoid should be considered.

4. No antidote to danaparoid exists. Accordingly, we evaluated hemofiltration as a potential means to rapidly reduce danaparoid plasma concentration. Whereas five different high-flux hemodialyzer membranes did not allow for danaparoid filtration, a plasmapheresis membrane was capable of removing danaparoid from the blood compartment (Schneider et al., 2004). Hence, plasmapheresis may be a way to reduce danaparoid levels in situations of overdosing or bleeding. Again, careful dosing of danaparoid is important to avoid bleeding.

5. HIT antibodies potentially cross-react with danaparoid. Although the respective clinical risk has been claimed to be less than 5% (Warkentin et al., 1998; Magnani and Gallus, 2006), individual patients, nevertheless, may be threatened if this condition occurs. As positive in vitro cross-reactivity is of uncertain clinical significance (Warkentin, 1996; Wilde and Markham, 1997; Newman et al., 1998), attention should focus on platelet count monitoring. A further fall in platelet count, or new fibrin deposits and clot formation within the extracorporeal circuit after application of danaparoid, may indicate clinically relevant cross-reactivity. To differentiate in vivo cross-reactivity from "under-anticoagulation" owing to insufficient dosage, determination of antifactor Xa levels and HIT antibody cross-reactivity studies are needed.

Table 2 lists dose recommendations for use of danaparoid for HD as provided by the manufacturer. The recommendations should be considered as guidelines and not followed uncritically in any individual patient. If applied with appropriate care, danaparoid provides adequate anticoagulation for HD of HIT patients with a favorable benefit/risk ratio, even during long-term use.

Recombinant Hirudin

Native hirudin was the first anticoagulant used for HD over 75 yr ago (Haas, 1925). In recent years, interest in its use for HD has redeveloped because of the availability of recombinant preparations, as well as the clinical need for managing patients with HIT. A preparation of recombinant hirudin (r-hirudin), lepirudin (Refludan® or HBW023), has been used successfully in humans for anticoagulation of both intermittent (Bucha et al., 1999a; Nowak et al., 1992, 1997; Steuer et al., 1999; Vanholder et al., 1994; Van Wyk et al., 1995) and continuous HD (Fischer et al., 1999; Schneider et al., 2000; Saner et al., 2001; Vargas Hein et al., 2001).

For use of r-hirudin anticoagulation in HD, some aspects should be specifically addressed. For further information on r-hirudin in renal insufficiency, the reader is referred to a recent review (Fischer, 2002):

1. As there is repetitive exposure to r-hirudin when used for regular, intermittent HD, immunogenicity of r-hirudin is of particular interest. Initially, r-hirudin appeared to be a weak immunogen (Bichler et al., 1991). However, recent studies revealed frequent development of antihirudin antibodies (AHAb) in patients receiving lepirudin for more than 5 days (Huhle et al., 1999, 2001; Song et al., 1999; Eichler et al., 2000). In addition, allergic reactions (including fatal anaphylaxis) to r-hirudin have been reported (Huhle et al., 1998; Eichler et al., 2000; Greinacher et al., 2003).

r-Hirudin is increasingly used for alternative anticoagulation in HD. Here, repetitive application of r-hirudin in patients on an intermittent maintenance HD regimen is likely to favor both induction and boostering of an immune response against the drug. As prospective studies evaluating sufficient numbers of HD patients on r-hirudin anticoagulation for the generation of AHAb are lacking, the incidence of AHAb and related adverse clinical events in this patient population remain to be elucidated.

Studies of HIT patients treated with lepirudin suggest that AHAb sometimes reduce renal lepirudin clearance (Huhle et al., 1999; Eichler et al., 2000). Indeed, marked reduction of renal lepirudin clearance due to monoclonal AHAb has been demonstrated in rats with normal renal function (Fischer et al., 2003). This was accompanied by a significant increase of both maximal plasma concentration and area under the curve of the alternative anticoagulant when compared to non-AHAb-treated animals. In chronic renal failure patients undergoing HD this may not be an issue. However, even small reductions in residual renal function have been shown to account for relevant prolongation of r-hirudin decay in plasma (Bucha et al., 1999a; Vanholder et al., 1997). Further reduction of renal r-hirudin clearance due to AHAb thus may influence r-hirudin dosing in these patients.

In acute renal failure requiring HD treatment for a prolonged period, reduction of renal r-hirudin clearance attributable to AHAb may be more relevant. Here, in patients suffering from multiorgan failure, the r-hirudin dosage required for sufficient anticoagulation was reduced significantly compared with the dosage needed in patients with normal renal function. In addition, r-hirudin dosage varied markedly depending on the residual renal function (Fischer et al., 1999). AHAb are likely to reduce further the amount of r-hirudin required, and thus may complicate anticoagulation in this challenging patient population.

The animal study also showed a significant decrease in the volume of distribution of lepirudin at steady state in the presence of AHAb (Fischer et al., 2003). Hence, even if further reduction of renal r-hirudin clearance owing to AHAb was negligible, major alterations in r-hirudin plasma concentration could still occur.

2. There remains debate as to which laboratory parameter is best suited for monitoring r-hirudin treatment. Initial studies addressing this in HD patients yielded conflicting results (Vanholder et al., 1994, 1997; Van Wyk et al., 1995). However, it now appears that the ecarin clotting time (ECT) (Nowak and Bucha, 1996) and chromogenic substrate assays (Griessbach et al., 1985) measure the r-hirudin plasma concentration with adequate precision over a wide concentration range and correlate well with each other (Hafner et al., 2000, 2002). However, as these tests are often not available, monitoring of r-hirudin anticoagulation is usually performed with the activated partial thromboplastin time (aPTT). A meta-analysis of two lepirudin treatment trials for HIT revealed a suitable aPTT ratio for reducing clinical thromboembolic complications to be

TABLE 2 Anticoagulation in Hemodialysis of HIT Patients—Dosage Examples of Suitable Alternative Anticoagulants


Dialysis procedure


Continuous infusion

Monitoring Target parameter3 range

Danaparoid sodium (0rg10172, Orgaran®)

Lepirudin (HBW023, Refludan®)

Intermittent HD (every second day)

Intermittent HD (daily)

Continuous HD/HF

Intermittent HD (every second day)

Before first 2 HDs Subsequent HD

First HD Second HD Subsequent HD

Predialytic anti-Xa activityd,f <0.3

Initial bolus First 4 h Next 4 h Subsequently

3750 (2500)

3750 (2500)b 2500 (2000) See above 2500 (2000)b

0 0

Post a comment