i i i i i i i i i i i i i i <-7 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 >7

Day of platelet count fall by >50% (when day of HIT thrombosis is defined as day 0)

FIGURE 8 Relation of onset of platelet count decrease and onset of HIT-associated thrombosis. The data summarize 209 patients with HIT-associated thrombosis. About one quarter (26.3%) of patients develop thrombosis on the same day that the thrombocytopenia occurs (defined arbitrarily as the day the platelet count has fallen by more than 50%), and in 33.5% the platelet count reached thrombocytopenia levels only after the occurrence of thrombosis. Abbreviation: HIT, heparin-induced thrombocytopenia. Source: From Greinacher et al., 2005b.

bias caused by symptomatic thrombosis. Nevertheless, the 30-day thrombosis event rate was about 50% (Fig. 2 in Chapter 3). This high frequency of thrombosis occurred whether the heparin administration was simply stopped or substituted by warfarin.

In 1999, Wallis and colleagues provided further support for this concept that isolated HIT had an unfavorable natural history. In their retrospective cohort study of 113 patients with serologically confirmed HIT, these workers also found a relatively high risk of thrombosis (23-38% at 30-day follow-up, depending on whether patients who developed thrombosis at the time heparin was stopped are included) in patients with isolated HIT managed by cessation of heparin. Further, early cessation of heparin (within 48 h after a 50% or greater fall in platelet count) did not appear to reduce risk of thrombosis, compared with patients in whom heparin was discontinued later.

More recently, Zwicker and coworkers (2004) performed a retrospective study that evaluated the risk of symptomatic thrombosis among patients with isolated HIT, based upon the magnitude of a positive enzyme-immunoassay (EIA) for anti-PF4/H antibodies. Thrombosis was seen in five (36%) of 14 patients with a strong positive EIA (>1.0 optical density units), but only three (9%) of 34 patients with a weak positive test (p = 0.07). This is consistent with data indicating that the greater the magnitude of a positive EIA, the greater the likelihood that the patient has heparin-dependent platelet-activating antibodies and, hence, clinical HIT (Warkentin et al., 2005a).

Meta-analyses of two prospective cohort studies also found a high initial thrombotic event rate (6.1% per day after stopping heparin therapy and before beginning alternative anticoagulant therapy with lepirudin) (Greinacher et al., 1999, 2000) (Fig. 4 in Chapter 14). Taken together, these large retrospective and prospective cohort studies suggest the following rule:

Rule 5

HIT is associated with a high frequency of thrombosis despite discontinuation of heparin therapy with or without substitution by coumarin: the initial rate of thrombosis is about 5-10% per day over the first 1-2 days; the 30-day cumulative risk is about 50%.

About 5% of patients (3 of 62) in the largest study died suddenly, two with proved or probable pulmonary embolism (Warkentin and Kelton, 1996). This experience supports the recommendation that further anticoagulation be considered for patients in whom isolated HIT has been diagnosed (Hirsh et al., 1998, 2001; Warkentin and Greinacher, 2004; Warkentin et al., 1998) (see Chapters 1 and 12-15).

D. Clinical Factors in the Pathogenesis of HIT-Associated Thrombosis

Clinical factors help determine the location of thrombosis in HIT. For example, Makhoul and colleagues (1986) observed prior vessel injury (e.g., recent angiography) in 19 of 25 patients with lower limb HIT-associated thrombosis. Similarly, central venous catheters are crucial for the occurrence of an upper limb DVT in patients with HIT (Hong et al., 2003).

Prospective studies of HIT in medical patients show that venous and arterial thrombotic events occur in approximately equal numbers; in contrast, there is a marked predominance of venous thrombosis when HIT occurs in surgical patients (Table 6 in Chapter 3). In a retrospective study, Boshkov and colleagues (1993) found that HIT patients with cardiovascular disease were more likely to develop arterial thrombosis, whereas venous thrombosis was strongly associated with the postoperative state.

Rule 6

Localization of thrombosis in patients with HIT is strongly influenced by independent acute and chronic clinical factors, such as the postoperative state, arteriosclerosis, or the location of intravascular catheters in central veins or arteries.

E. Venous Thrombosis

Large case series suggest that venous thrombotic complications predominate in HIT (Warkentin and Kelton, 1996; Nand et al., 1997; Greinacher et al., 2005b) (Tables 6 and 7 in Chapter 3). Indeed, pulmonary embolism occurs more often than all arterial thrombotic events combined. Furthermore, the strength of association between HIT and venous thromboembolism increases in relation to the severity of thrombosis (Table 4). Other unusual venous thrombotic events complicating HIT include cerebral vein (dural sinus) thrombosis (v.i.), adrenal vein thrombosis (v.i.), hepatic vein thrombosis (Theuerkauf et al., 2000), and perhaps retinal vein thrombosis (Nguyen et al., 2003). Thus:

Rule 7

In patients receiving heparin, the more unusual or severe a subsequent thrombotic event, the more likely the thrombosis is caused by HIT.

TABLE 4 Association of HIT and Thrombosis

Thrombosis rate

Patient population (Ref.)


0 0

Post a comment