Hit

Controls

Postorthopedic surgerya

(Warkentin et al., 1995, 2003)

Patients with central lineb (Hong et al., 2003) Medicala (Girolami et al., 2003)

Proximal DVT Bilateral proximal DVT Pulmonary embolism

Any thrombosis 13/18(72.2%) 112/647 (17.3%) 12.4(4.0-45.2) < 0.001 Upper-limb 14/145(9.7%) 3/484 (0.6%) 17.1(4.9-60.5) < 0.001 DVT

Any thrombosis 3/5 (60%)

aHIT defined as >50% platelet count fall.

bHIT defined as any abnormal platelet count fall with positive HIT serology (platelet fall was >50% in 93% of study patients).

Abbreviations: DVT, deep vein thrombosis; HIT, heparin-induced thrombocytopenia; OR, odds ratio.

Regardless of the severity of thrombosis, in any patient who develops a symptomatic venous or arterial thrombosis while receiving heparin, the platelet count should be measured to evaluate whether HIT could be present.

Recently, Levine and colleagues (2006) estimated from published data that HIT could be present in about one in eight patients who develop venous thromboembolism subsequent to UFH treatment or prophylaxis. However, the risk could be substantially higher (~45-75%) among patients who develop symptomatic venous thromboembolism following postoperative UFH thromboprophylaxis (Warkentin, 2006a; Greinacher et al., 2005a).

Lower Limb DVT

Lower limb DVT is the most frequent thrombotic manifestation of HIT. Many venous thrombi are extensive and are often bilateral (Table 4).

Sometimes the DVT is sufficiently severe on clinical grounds as to merit use of the term "phlegmasia cerulea dolens" (i.e., an inflamed, blue, painful limb). However, progression of phlegmasia to venous limb gangrene is rare in the absence of coumarin anticoagulation (discussed subsequently).

There is slight left-sided predominance involving lower limb DVT: we found that 76/137 (56%) of lower limb DVT complicating HIT involved the left lower limb (Hong et al., 2003), a similar proportion as in control patients (57%). A slight left-sided predominance (~55 vs. ~45%) for lower limb DVT has also been noted in non-HIT populations (Kerr et al., 1990; Markel et al., 1992). This is attributed to the left iliac vein crossing the left iliac artery, causing an increase in left-sided lower limb venous pressures. Pregnancy amplifies further this phenomenon, thus explaining the marked predominance (>95%) of left lower limb DVT in pregnancy (Ginsberg et al., 1992; Breinachet et al., 2005a).

Upper Limb DVT

Upper limb DVT is relatively common in HIT, occurring in about 5% of patients with HIT (Hong et al., 2003). Notably, in these patients the upper limb DVT occurred at the site of a current or recent central venous catheter. Most (86%) of the patients therefore had right upper limb DVT complicating HIT, reflecting strong physician preference to using the right neck veins for insertion of central lines. This study suggests that a systemic hypercoagulable state (HIT) interacts with a local factor (location of central lines) to result in clinical events (upper limb DVT).

Recurrence of Venous Thromboembolism

Gallus and colleagues (1987) identified HIT as a significant risk factor for recurrence of venous thromboembolism in a prospective treatment study: three of the nine patients with HIT developed recurrent venous thromboembolism, compared with 12 of the 223 patients in whom HIT was not diagnosed (odds ratio, 8.8; p < 0.01).

Warfarin-Induced Venous Limb Gangrene

Venous limb gangrene is one of two clinical syndromes associated with HIT in which coumarin anticoagulation paradoxically plays an important pathogenic role (Fig. 9). Venous limb gangrene is defined as acral (extremity) necrosis that occurs in a limb affected by DVT. Additional features include (1) absence of large artery occlusion (i.e., there are palpable or doppler identifiable pulses); (2) extensive thrombotic occlusion of large and small veins, as well as venules; and (3) the characteristic hallmark of a supratherapeutic international normalized ratio (INR), generally >4.0 (Fig. 10).

Anticoagulation with warfarin, phenprocoumon, or other coumarins is a crucial factor to explain the progression of DVT to venous limb gangrene (Warkentin, 1996b; Warkentin et al., 1997). A case-control study of eight patients with HIT-associated venous limb gangrene found a higher median INR, compared with 58 control HIT patients treated with warfarin for DVT who did not develop venous gangrene (5.8 vs. 3.1; p < 0.001). Laboratory studies showed a characteristic hemostatic profile for patients with venous gangrene: persisting in vivo thrombin generation (elevated thrombin-antithrombin complex levels), together with reduced protein C activity (Fig. 11). The high INR is a surrogate marker for severely reduced protein C (through parallel coumarin-induced reduction in factor VII). Thus, venous limb gangrene appears to result from a profound disturbance in procoagulant-anticoagulant balance.

The association between venous limb gangrene and HIT was first reported by Towne and colleagues (1979). They noted a prodrome of phlegmasia cerulea dolens before progression to distal gangrene (information on possible coumarin treatment was not given). Other reports of venous limb gangrene complicating HIT, however, do suggest that warfarin had been used during the evolution to necrosis (Thomas and Block, 1992; Hunter et al., 1993; Kaufman et al., 1998).

Patients have also developed venous limb gangrene during combined treatment with both ancrod and warfarin (Warkentin et al., 1997; Gupta et al., 1998); because thrombin generation increases during treatment of HIT with ancrod (Warkentin, 1998b; Fig. 2 in Chapter 12), ancrod could predispose to a greater risk for venous gangrene during warfarin treatment.

Recently, several patients have been reported who developed venous limb gangrene during the transition to coumarin from parenteral anticoagulation with a direct thrombin inhibitor (lepirudin or argatroban) (Smythe et al., 2002; Srinivasan

FIGURE 9 Coumarin-induced necrosis: HIT is associated with two forms of necrosis: (1) venous limb gangrene, affecting extremities with active deep vein thrombosis, and (2) "classic" CISN, which involves central (nonacral) tissues, such as breast, abdomen, thigh, flank, and leg, among other tissue sites. Coumarin-induced necrosis complicating HIT typically manifests as venous limb gangrene (~90%) (Warkentin et al., 1997, 1999), whereas necrosis in other clinical settings most commonly affects central tissues (~90%) (Cole et al., 1988). Abbreviations: DVT, deep vein thrombosis; CISN, coumarin-induced skin necrosis; HIT, heparin-induced thrombocytopenia. Source: From Warkentin, 1996b.

FIGURE 9 Coumarin-induced necrosis: HIT is associated with two forms of necrosis: (1) venous limb gangrene, affecting extremities with active deep vein thrombosis, and (2) "classic" CISN, which involves central (nonacral) tissues, such as breast, abdomen, thigh, flank, and leg, among other tissue sites. Coumarin-induced necrosis complicating HIT typically manifests as venous limb gangrene (~90%) (Warkentin et al., 1997, 1999), whereas necrosis in other clinical settings most commonly affects central tissues (~90%) (Cole et al., 1988). Abbreviations: DVT, deep vein thrombosis; CISN, coumarin-induced skin necrosis; HIT, heparin-induced thrombocytopenia. Source: From Warkentin, 1996b.

et al., 2003; Warkentin, 2006b). Typically, patients had symptomatic DVT in the affected limb and had their direct thrombin inhibitor started and stopped while they remained thrombocytopenic. Additionally, the INR was supratherapeutic at the time that limb ischemia or gangrene occurred after stopping the direct thrombin

FIGURE 10 (See color insert) Warfarin-associated venous limb gangrene. Progression of deep vein thrombosis to acral necrosis (leading to below-the-knee amputation) occurred despite the presence of palpable arterial foot pulses in this 49-yr-old woman with HIT treated with warfarin (international normalized ratio = 7.2 at the onset of limb gangrene). Source: From Warkentin et al., 1997.

0 0

Post a comment