EHIT Is Less Frequent in Surgical Patients Receiving LMWH Compared with UFH Prophylaxis

Anecdotal reports indicate that HIT can occur during treatment with LMWH (Ball et al., 1989; Tardy et al., 1990; de Raucourt et al., 1996; Plath et al., 1997; Elalamy et al., 1996; Warkentin, 1998; Gruel et al., 2003; Ng and Lee, 2003). Clinical trial data suggest that the frequency is low, however. Using a sensitive definition for HIT (>50% fall on or after day 5 and confirmed by positive HIT antibodies), two studies in Hamilton found an overall frequency of only 4/439 (0.9%; 95% CI 0.252.32%) for HIT complicating use of LMWH given for postoperative orthopedic patients (Warkentin et al., 1995, 1998b, 2003). In contrast, using the same definition of HIT, patients who received UFH had a much higher frequency of HIT, 16/332 (4.8%; 95% CI 2.78-7.71%) (Warkentin et al., 1995, 2003). A similar high frequency of UFH-induced HIT was observed in a German study, 15/307 (4.9%; 95% CI 2.76-7.93) (Ganzer et al., 1997).

The strongest evidence that LMWH is indeed associated with a lower frequency of HIT was provided by an RCT that directly compared the frequency of HIT between the two types of heparin (Warkentin et al., 1995, 2003, 2005a). The frequency of HIT in patients treated with the LMWH preparation, enoxaparin (itself derived from porcine mucosal heparin), was lower than that seen in patients treated with porcine UFH, irrespective of whether a standard definition (platelet fall to <150 x 109/L on or after day 5 of heparin treatment) or a more sensitive definition (>50% platelet count fall from the postoperative peak) of thrombocyto-penia was used. The frequency of HIT antibody formation also differed between the two patient groups, using either the SRA (Warkentin et al., 1995, 2005a) or a PF4/heparin (or PF4/polyanion) EIA (Warkentin et al., 2000, 2005a). Thrombocytopenia also appeared to be infrequent in other trials of LMWH (Leyvraz et al., 1991; Simonneau et al., 1997; ENOXACAN Study Group, 1997).

UFH appeared also to lead to greater frequency of HIT antibody formation than the LMWH reviparin in a randomized trial of post-hip and knee surgery patients (Ahmad et al., 2003a). HIT antibodies occurred somewhat more often in knee surgery patients. These same investigators also examined HIT antibody formation in orthopedic patients immobilized in a plaster cast who were randomized to receive reviparin or placebo (Ahmad et al., 2003b). A surprising finding was that the number of patients who apparently formed anti-PF4/polyanion antibodies (by EIA) was higher in the placebo group (10 cases vs. 6). No patient in either study developed clinical HIT.

A meta-analysis of five surgical thromboprophylaxis prospective cohort studies and RCTs (four postorthopedic, one postcardiac) that defined HIT using both clinical and serologic criteria found a substantial reduction in HIT frequency (odds ratio [OR], 0.10; p < 0.001) with LMWH compared with UFH (Martel et al., 2005) (Fig. 1). The authors found the absolute risk of HIT to be 0.2% with LMWH and 2.6% with UFH. In contrast, when they examined the OR for HIT using a non-serologically defined definition of thrombocytopenia, the difference in risk of apparent HIT was less marked.

Gruel and colleagues (2003) performed a systematic study that identified 11 patients with HIT (three with HIT-associated thrombosis) that had been exclusively treated with a LMWH preparation (dalteparin, nadroparin, or enoxaparin). Clinical and serologic features were similar to patients with HIT developing during UFH, except that there was evidence that thrombocytopenia may begin somewhat later during LMWH therapy. Based upon estimated relative use of UFH and LMWH in France, the authors estimated the frequency of HIT to be 40-fold less with LMWH, compared with UFH.

The data supporting a lower risk of HIT in postcardiac surgery patients receiving LMWH compared with UFH thromboprophylaxis is discussed later in the chapter (see Section V.B).


Leyvraz et al., 1991** Warkentin et al., 1995* Ganzer et al., 1999* Pouplard et al., 1999** Mahlfeld et al., 2002*

0 0

Post a comment