Introduction

Thrombocytopenia is a common problem encountered in hospitalized patients. For patients receiving heparin, there are three general explanations for thrombocytope-nia: (1) heparin-induced thrombocytopenia (HIT), (2) non-idiosyncratic heparin-induced platelet activation (see Chapter 4), and—perhaps most often—(3) an unrelated clinical problem, either common (e.g., hemodilution, septicemia) or rare (e.g., posttransfusion purpura, drug-induced immune thrombocytopenic purpura) (see Chapter 11). The availability of sensitive and specific laboratory assays (e.g., enzyme immunoassay [EIA] and serotonin release assay [SRA]) for pathogenic HIT antibodies means that patients with HIT can usually be readily distinguished from those with the other conditions (see Chapter 10). However, the role of heparin in causing thrombocytopenia because of nonimmune platelet activation cannot readily be separated from other common medical problems encountered in hospitalized patients, on either clinical or laboratory criteria. Furthermore, these two conditions can coexist (Chong and Castaldi, 1986). Thus, the term "nonimmune heparin-associated thrombocytopenia" (nonimmune HAT) has been recommended to describe patients who develop thrombocytopenia during heparin treatment in which a role for HIT antibodies cannot be implicated (Warkentin et al., 1998a).

Unfortunately, many early studies of HIT frequency either did not perform laboratory testing or used relatively insensitive or nonspecific assays to diagnose HIT. In contrast, more recent studies have used one, or even two, sensitive and complementary assays. Perhaps for this reason, the understanding of the frequency and clinical impact of HIT has shifted over the years. Formerly, opinions on immune-mediated HIT were divergent: it was considered both nonexistent (Bell, 1988) and common (Kelton, 1986). Nevertheless, both viewpoints acknowledged that thrombosis resulting from HIT was very uncommon. Today's perspective on HIT is very different. The frequency of HIT is now shown to be variable, partly depending on patient population and type of heparin used. For example, the frequency ranges from negligible—e.g., a pregnant patient receiving low molecular weight heparin (LMWH)—to as high as 5% (orthopedic surgical patient receiving unfractionated heparin [UFH] for 2 wk). Anti-platelet factor 4 (PF4)/ heparin antibody formation with use of UFH ranges from 2-5% (cardiac medical patients) to 15-30% (orthopedic surgical patients) to 30-70% (cardiac surgical patients). Most importantly, however, it is now becoming clear that the risk for thrombosis in patients who develop HIT is at least 30-70%, a frequency that is far greater than in control patients who do not develop HIT (see Chapter 2).

The biological basis for this variability in frequency of HIT and antibody formation is now apparent. The HIT antigen is a cryptic autoantigen, or neoanti-gen, on PF4 that is formed when PF4 binds to heparin (see Chapters 5-7). Only stoichiometric concentrations of heparin and PF4 will form the antigen. Thus, it can be hypothesized that the frequency of anti-PF4/heparin antibody formation will be influenced not only by heparin dose and composition, but also by circulating PF4 levels. Conditions associated with fluctuating, but at times high, circulating PF4 and heparin levels (e.g., cardiac surgery) might be ideal for immunization. Thus, real differences in HIT frequency observed among prospective studies can be understood in a biologically plausible context.

The key role of the pathogenic HIT antibodies and the availability of sensitive and specific assays for their detection suggest that HIT should be considered a clinicopathologic syndrome. Consequently, this chapter will focus on studies that have used in vitro testing to evaluate HIT antibodies. However, other features known to be useful to diagnose HIT, such as the timing of the onset of thrombocytopenia in typical HIT, and the rapid platelet count fall on heparin rechallenge, will also be used (see Chapter 2). The importance of confirmatory laboratory testing should not be underestimated: prospective (Greinacher et al., 1994; Lee et al., 1996; Juhl et al., 2006) and retrospective (Look et al., 1997) studies suggest that only 15-55% of sera referred for evaluation test positive for antibodies. Furthermore, systematic analysis of a large clinical trial of heparin treatment (Warkentin et al., 1995, 2003) revealed several patients in whom unusual clinical events subsequently linked to HIT were initially attributed to other problems.

Table 1 lists various biological and technical explanations that underlie the reported variability in frequency of HIT among the prospective studies. We will begin our discussion by summarizing an important technical problem in many studies, namely, the failure to exclude patients with early, nonimmune HAT.

0 0

Post a comment