Antihirudin antibodies following low-dose subcutaneous treatment with desirudin for thrombosis prophylaxis after hip-replacement surgery: incidence and clinical relevance

There are 2 recombinant hirudins approved in the European Union: desirudin (Revasc; Aventis, Frankfurt, Germany) and lepirudin (Refludan; Schering, Berlin, Germany). Desirudin is approved for thrombosis prophylaxis following orthopedic hip-replacement surgery and lepirudin for anticoagulation in patients with heparin-induced thrombocytopenia (HIT) and thromboembolic complications. Both compounds differ from each other in their N-termini: desirudin possesses a valine-valine and lepirudin a leucine-threonine structure. Otherwise the desirudin and lepirudin are very similar, with molecular weights of 6.96 kDa and 6.98 kDa, respectively.

Rhesus monkeys did not show a specific immune response following intravenous boluses of desirudin.1 In contrast, balb/c mice showed a 100% antihirudin antibody response after immunization by albumin-conjugated hirudin or hirudin fragments.2,3 Also, dogs generated antihirudin antibodies following immunization with desirudin.4 

In healthy volunteers, a few allergic reactions were observed following intravenous as well as subcutaneous applications of desirudin.5 Allergic reactions occurred in 3 volunteers, in another 4 the skin test became positive. In one volunteer antihirudin antibodies were detected.

In none of the major clinical studies of patients with acute coronary syndromes was an immune reaction toward 1 of the 2 hirudins reported.6-10 However, these patients were usually treated for 2 to 3 days only, which is too short of a period for a B-cell response.

In 1996, we reported that a high proportion of patients receiving lepirudin for HIT developed antihirudin antibodies,11,12 a finding confirmed by others.13,14 During a mean duration of treatment of 14.4 days, 44% of 198 patients treated with lepirudin developed antihirudin antibodies of the immunoglobulin G (IgG) class, which occurred as early as day 4 (peak at days 8-9). Of 87 patients with detectable antihirudin antibodies, 61 (70.1%) developed during treatment, 23 (26.4%) occurred after cessation of lepirudin, and 3 (3.5%) became positive during a second treatment course with lepirudin only.12 

In another study with lepirudin given to patients with HIT (up to 26 days), up to 74% of patients developed antihirudin antibodies of the IgG, IgA, and/or IgM class.13,14 None of these patients developed allergic reactions.

Desirudin is approved for thrombosis prophylaxis following orthopedic hip-replacement surgery based on 3 prospective trials.15-17 These patients usually require treatment for 7 to 10 days, which is within the time frame for induction of antihirudin antibodies. The purpose of the present study was to assess whether desirudin given subcutaneously in prophylactic dose in patients following orthopedic hip-replacement surgery is immunogenic, and to compare the outcomes of patients testing positive and negative for antihirudin antibodies.

Patients were enrolled in a prospective dose-finding study, as described elsewhere.15 Approval was obtained from the Göteborg University and the Greifswald University institutional ethics boards for these studies. Informed consent was provided according to the Declaration of Helsinki. Serum was obtained between days 6 and 11 after start of desirudin (mean day, 8 ± 1). Samples were frozen at −30°C or below.

Desirudin was provided by Novartis Pharma (Nürnberg, Germany) and lepirudin, by Aventis.

Antihirudin antibody enzyme-linked immunosorbent assay (ELISA) was performed as described.12 Microtiter wells (Covalink, Nunc, Germany) coated with desirudin or lepirudin were incubated with patient plasma (1:50 for 60 minutes at 37°C), and IgG binding assessed by anti-human IgG-peroxidase labeled antibodies. Sera of HIT patients with known antihirudin antibodies served as positive controls, plasma of healthy blood donors served as negative controls. The assay was performed by personnel blinded for the dose of desirudin given to the patients.

Concentrations of free desirudin in patient plasma samples were assessed by a chromogenic assay using the thrombin substrate S2238, as described previously.18 

Incidences of deep vein thrombosis and pulmonary embolism, allergic reactions, and hemorrhages were compared by Fisher exact test (2-tailed). Occurrence of antihirudin antibodies over time was calculated by Kaplan-Meier statistics. Plasma concentrations of antihirudin antibody–positive and –negative patients were compared by Wilcoxon rank-sum test.

Samples from 112 patients were studied of 20 patients who received a dosage of 20 mg desirudin subcutaneously twice a day, of 75 patients who received a dosage of 15 mg subcutaneously twice a day, and of 17 patients who received a dosage of 10 mg subcutaneously twice a day.

In 11 (9.8%) of 112 patients, antihirudin antibodies were detectable. Results were identical whether desirudin or lepirudin was used in the ELISA, indicating a 100% cross-reactivity. Seroconversion rates were similar among the 3 dosing groups: 2 (10%) of 20 patients received 20 mg desirudin twice a day, 6 (8.0%) of 75 patients received 15 mg twice a day, and 3 (17.6%) of 17 patients received 10 mg twice a day. Compared with the occurrence of antihirudin antibodies in HIT patients treated with lepirudin, no differences were observed (P = .113, log-rank test, Figure1).

Desirudin plasma concentrations of antihirudin antibody-positive patients did not differ from those of antihirudin antibody-negative patients within the 3 dosing groups: (1) 10 mg, 0.1 μg/mL versus 0.094 μg/mL, respectively (P = 1.0); (2) 15 mg, 0.17 μg/mL versus 0.28 μg/mL, respectively (P = .33) and (3) 20 mg, 0.1 μg/mL versus 0.17 μg/mL, respectively (P = .63).

Clinical end points could be evaluated in 109 patients. Outcomes of antihirudin antibody–positive patients did not differ compared with antihirudin antibody–negative patients: deep vein thrombosis and/or pulmonary embolism, 2 (18.2%) of 11 vs 11 (10.9%) of 101,P = .61; allergic reactions, 0 of 11 vs 1 of 101,P = 1.0; hemorrhage, 1 (9.1%) of 11 vs 6 (5.9%) of 101,P = .53.

The present study shows that subcutaneous desirudin is immunogenic in patients after major orthopedic surgery. At day 8 of treatment, approximately 10% of patients are immunized. Numerically this rate is much lower than the rate observed in patients with HIT receiving lepirudin.12-14 However, using Kaplan-Meier calculation, the risk of immunization per time exposed seems to be identical in both patient populations (Figure 1). Thus the total rate of immunization depends on treatment duration and time point of blood sampling.

The 2 hirudins differ in the structure of the N-terminal part. However, antibodies of patients treated with desirudin showed a 100% cross-reactivity to lepirudin and vice versa, and therefore must recognize structures common to both hirudins with the same immunogenicity.

Also, different routes of application (intravenous versus subcutaneous) are of no relevance for immunization; in the present study, patients received hirudin only subcutaneously, whereas lepirudin was given only intravenously in HIT patients.

As the dose was higher in lepirudin-treated patients (approximately 250 mg per day in a 70-kg patient) compared with the present study, immunization is not dose related, at least if 10 mg twice a day is given.

The result is surprising as the 2 patient groups are very different. We suspected HIT patients to be more likely to form antihirudin antibodies, because they have already formed antibodies toward platelet factor 4/heparin complexes. This indicates that HIT patients do not have an increased tendency to form antibodies to other drugs.

Antihirudin antibody desirudin complexes cannot be filtered by the glomeruli. However, neither a difference in desirudin plasma levels, nor in bleeding complications, could be observed between antihirudin antibody–positive and –negative patients. Also we found no evidence for inhibitory antibodies, which have been described in HIT patients receiving lepirudin.12,19 Because of the small numbers of immunized patients, this observation must be considered preliminary.

In conclusion, desirudin given in prophylactic dose in patients after orthopedic hip-replacement surgery is immunogenic. The overall risk of immunization seems to be the same as in HIT patients receiving lepirudin in therapeutic dose. As the treatment and observation period was shorter in the present study, the percentage of immunized patients was lower. The overall number of immunized patients, however, is still small, and thus less frequent severe adverse effects might have been overlooked until now. Patients under treatment should be carefully observed for adverse events.