Comparison with Other Treatments for HIT

Mortality rates in patients with HIT remained at approximately 20-30% for more than a decade (King and Kelton, 1984; AbuRahma et al., 1991; Warkentin and Kelton, 1996; Nand et al., 1997). Notably, these rates are two to three times higher than those observed in the HAT studies. In addition to lepirudin, other drugs with antithrombin activity (e.g., argatroban) or antifactor Xa activity (e.g., danaparoid) may be appropriate for management of HIT (see Chapters 13 and 15-17).

Comparisons of the various clinical trials of agents used to treat HIT need to be interpreted with caution, since there have been no direct comparative trials and the studies employed different designs. Trials of lepirudin and argatroban, however, utilized similar clinical endpoints and historical controls for comparison. The most obvious differences between the lepirudin and the argatroban trials are (1) the need for laboratory confirmation of HIT in the lepirudin trials; (2) treatment duration, which was consistently longer than 10 days in the lepirudin-treated patients but less than 7 days in the argatroban-treated patients (potential to increase apparent efficacy and also bleeding with lepirudin); (3) the observation period, which started at the time of diagnosis in lepirudin-treated patients compared with the time of treatment initiation in the argatroban trials (potential to underestimate the efficacy of lepirudin); and (4) a considerable proportion of patients in the historical control group of the HAT trials had been treated with danaparoid (potential to underestimate the efficacy of lepirudin). To allow a more direct comparison, we reanalyzed the data of the HAT trials as per the argatroban trials, i.e., analyzing those events occurring from start of active treatment only (Tables 4b and 5).

The rates for the combined endpoint were consistently lower in the lepirudin trials than in the two argatroban studies. For patients with isolated HIT, the composite event rate was 19.8% in the HAT-1, -2, and -3 meta-analysis (Lubenow et al., 2004), compared with 25.6% and 28.0% in the two argatroban trials. For patients with HIT and thrombosis, the combined endpoint occurred in 19.1% (meta-analysis) with lepirudin (Lubenow et al., 2005), and in 43.8% and 41.5% of patients treated in the two argatroban trials.

Because of the often critical condition of the patient population under study, the rate of deaths observed for both DTIs is not likely to be solely attributable to treatment failure and may vary considerably with different patient populations. As the argatroban trials included many patients who most likely did not have HIT, the death rate associated with HIT might have been overestimated, as non-HIT patients with a decrease of platelet count are often very sick (e.g., septicemia, DIC). Death rates were 14.3% (meta-analysis) and 12.3% (DMP) for patients with isolated HIT treated with lepirudin, but 18.1% and 23.1% in those treated in the two argatroban trials. In patients with HIT complicated by thrombosis, death rates with lepirudin were 11.9% (meta-analysis) and 10.9% (DMP), compared with 18.0% and 23.1% in the argatroban trials.

Among patients with isolated HIT, limb amputation occurred in 3.3% (metaanalysis) and 1.3% (DMP) when treated with lepirudin, and in 1.9% and 4.2% of those treated with argatroban. The amputation rates were higher in HIT with thrombosis; 5.5% (meta-analysis) and 5.8% (DMP) in those treated with lepirudin, versus 11.1% and 14.8% in those treated with argatroban.

There was also a difference in the incidences of new thrombosis, which is arguably the most important parameter for assessing efficacy of an alternative anticoagulant in HIT. In those with isolated HIT, it was 4.4% (meta-analysis) and

2.1% (DMP) with lepirudin, and 6.9% and 5.8% in those treated with argatroban. In HIT with thrombosis it was 6.8% (meta-analysis) and 5.2% (DMP) for lepirudin, and 14.6% and 13.1% for argatroban-treated patients.

The risk of major bleeding between lepirudin and argatroban appears similar if treatment duration is taken into account. In the lepirudin trials, 17.6% of patients experienced major bleeding (Table 4b) over a mean treatment period of 15 days, i.e., a risk for major bleeding of 1.17% per treatment day. In the argatroban trials, major bleeding occurred in 6.9% (Arg 911) and 5.7% (Arg 915) with a mean treatment period of 6 days in both trials, i.e., a risk for major bleeding of 1.05% per treatment day.

Similar rates of efficacy and safety for the two drugs were also observed in a retrospective chart analysis enrolling 61 lepirudin-treated and 29 argatroban-treated HIT patients in a single center (Smythe et al., 2005). Effective anticoagulation was achieved in 77.8% of argatroban patients and 69.5% of lepirudin patients (p = 0.61). Major bleeding occurred in 10.3% and 11.5% of argatroban and lepirudin patients, respectively (p = 1.0).

Because there are no prospective data comparing lepirudin and danaparoid for treatment of HIT, we retrospectively compared 126 danaparoid-treated patients with 175 lepirudin-treated patients who fulfilled the same inclusion and exclusion criteria (Farner et al., 2001). In the patients with HIT without TECs at baseline, a time-to-event analysis showed that the cumulative risk of the combined endpoint was higher in danaparoid-treated patients than in the lepirudin-treated patients (p = 0.02 by log rank test; hazard ratio [HR] = 2.9 [95% CI 1.1-7.6]; p = 0.027). This was due primarily to an increased incidence of new TECs (20% [95% CI 8.436.9] for danaparoid vs. 6.3% [95% CI 1.3-17.2] for lepirudin; p = 0.087). Of note, patients with isolated HIT usually received only prophylactic-dose danaparoid. In contrast, HIT patients with thrombosis at baseline, and who were therefore treated with a therapeutic-dose regimen of danaparoid, had a similar outcome as patients receiving lepirudin (p = 0.913). However, the risk for major bleeding was lower in the danaparoid-treated group (p = 0.012) (Farner et al., 2001).

The major conclusions of these comparisons are: (1) HIT patients seem to benefit from a longer treatment period with an alternative anticoagulant, with 10 days better than 5 days; (2) the prophylactic-dose regimen of danaparoid (750 U sc two or three times daily) approved in the European Union for HIT with isolated thrombocytopenia appears to be suboptimal. Thus, patients with acute HIT require anticoagulation generally in therapeutic doses (see also Chapter 12).

0 0

Post a comment