Septicemia Associated Purpura Fulminans

Septicemia complicated by DIC occasionally results in progressive ischemia and necrosis of fingers or hands and toes or feet, producing a syndrome of symmetrical peripheral gangrene also known as purpura fulminans (Knight et al., 2000). The association with DIC suggests that increased thrombin generation in vivo, together with severe consumption and depletion of natural anticoagulant factors (e.g., protein C, protein S, antithrombin), leads to dysregulated fibrin deposition in the microvasculature. Other contributing factors can include hypotension or shock, pharmacological vasoconstriction (e.g., dopamine, epinephrine, norepinephrine) (Winkler and Trunkey, 1981; Hayes et al., 1992), vessel injury from invasive catheters, impaired hepatic synthesis of natural anticoagulants (e.g., vitamin K deficiency, postoperative hepatic dysfunction or failure), postsplenectomy status, or congenital deficiency of natural anticoagulants. Rarely, purpura fulminans occurs several weeks after varicella infection, usually because of autoantibodies reactive against protein S (Smith and White, 1999).

Meningococcemia in particular is often complicated by peripheral tissue necrosis that seems to parallel the severity of protein C depletion (Fijnvandraat et al., 1995). Recent trials suggest that protein C replacement therapy improves the natural history of this infection (Smith and White, 1999; White et al., 2000). Other infections that sometimes are complicated by symmetrical peripheral gangrene include septicemia secondary to pneumococcus (Johansen and Hansen Jr., 1993), Escherichia coli (Rinaldo and Perez, 1982), Haemophilus influenzae type b (Hayes et al., 1992), and Capnocytophaga canimorsus (Kullberg et al., 1991), among others. Sometimes severe systemic inflammatory response syndromes, such as ARDS, in the absence of demonstrable infection, can be complicated by limb necrosis (Bone et al., 1976). Acquired antithrombin deficiency in such patients with ARDS could be associated with thrombosis (Owings et al., 1996).

The development of acral tissue ischemia or necrosis in a thrombocytopenic, septic patient receiving heparin may suggest HIT. Although a common therapeutic response to such a diagnostic dilemma might be to stop heparin pending results of diagnostic testing for HIT antibodies, this could result in further ischemic injury, because anticoagulants might help prevent microvascular thrombosis (White et al., 2000). Furthermore, alternative non-heparin anticoagulants could be relatively contraindicated in a patient with significant renal or hepatic dysfunction. Thus, a reasonable treatment approach might well include continued heparin if clinical judgment posited a higher likelihood of septicemia, rather than HIT, as the cause of the microvascular thrombosis.

Only a small minority of septic patients develop acral limb ischemia or necrosis. Many, however, develop thrombocytopenia, with or without laboratory evidence for DIC. The predominant explanation for increased platelet destruction in sepsis is uncertain, but appears to involve the underlying inflammatory host response (Aird, 2003a,b). Since hospitalized septic patients frequently are exposed to heparin, diagnostic confusion with HIT can result. Low protein C levels correlate with poor outcomes in sepsis (Yan et al., 2001), and recombinant human activated protein C (drotrecogin, Xigris) has been shown to reduce mortality in septic patients (Bernard et al., 2001). It is possible that this therapy might reduce risk of limb ischemia from microvascular thrombosis in this patient population. A potential dilemma is that septic patients with severe thrombocytopenia (<30 X 109/L) were excluded in the clinical trials because of the bleeding potential of drotrecogin; however, as relative and absolute efficacy was greatest in the patients with the most severe sepsis, it has been suggested that otherwise eligible patients with such severe thrombocytopenia be considered as candidates for drotrecogin following platelet transfusion (Warkentin et al., 2003b).

0 0

Post a comment