Acute Infectious Purpura Fulminans

CLINICAL REVIEW Pediatric Dermatology Vol. 15 No. 3 169-183, 1998

Acute Infectious Purpura Fulminans: Pathogenesis and Medical Management Gary L. Darmstadt, M.D. Department of Pediatrics, Children’s Hospital and Regional Medical Center, University of Washington School of Medicine, Seattle, Washington

Abstract: Purpura fulminans (PF) is a potentially disabling and lifethreatening disorder characterized by acute onset of progressive cutaneous hemorrhage and necrosis, and disseminated intravascular necrosis. Acute infectious PF occurs most commonly in the setting of meningococcemia due to elaboration of endotoxin. Presence of purpura, particularly when generalized, is an important predictor of a poor outcome following meningococcal infection. Histopathologic hallmarks of acute infectious PF are dermal vascular thrombosis and secondary hemorrhagic necrosis, findings which are identical to those of the Shwartzman reaction. Acute infectious PF and the Shwartzman reaction have a common pathogenesis, involving a disturbance in the balance of anticoagulant and procoagulant activities of endothelial cells. This disturbance, which is triggered by endotoxin, appears to be mediated by cytokines, particularly interleukin-12, interferon-?, tumor necrosis factor-a, and interleukin-1, leading to the consumption of proteins C and S and antithrombin 111. State-of-the-art therapeutic interventions based on recent advances in our understanding of the pathogenesis of acute infectious PF are discussed. Purpura fulminans (PF) is a potentially disabling and life-threatening disorder characterized by acute onset of progressive cutaneous hemorrhage and necrosis due to dermal vascular thrombosis, and disseminated intravascular necrosis (DIC). It occurs predominantly in three clinical settings (Table 1): (1) in the neonatal period as a manifestation of inherited, homozygous protein C or, rarely, protein S deficiency; ( 2 ) approximately 7 to 10 days after a relatively benign antecedent infection, usually involving the skin, such as varicella or scarlet fever (i.e., “idiopathic” PF) (Table 2); and ( 3 ) in conjunction with an acute infectious illness (i.e., acute infectious PF), particularly sepsis with endotoxin [lipopolysaccharide]

(LPS)-producing gram-negative bacteria (e.g., Neisseria meningitidis) (1). All cases involve dysfunction of hemostasis, with a shift from a quiescent state favoring anticoagulation to a disease state of overwhelming procoagulation. Significant strides have been made in the management of hereditary neonatal PF because of advances in our understanding of its pathogenesis (e.g., protein C deficiency) and the availability of new therapeutic products (e.g., protein C concentrate) (2,3). Idiopathic PF is an uncommon entity that has been reported in less than 100 children (4,5). It develops while the patient is not acutely ill, typically during the postinfectious convalescent phase

Address correspondence to Gary L. Darmstadt, M.D., Division of Infectious Disease, Department of Pediatrics CH-32, Children’s Hospital and Regional Medical Center, 4800 Sand Point Way NE,Seattle, WA 98105, or e-mail: [email protected].

169

170 Pediatric Dermatology Vol. 15 No. 3 May/June 1998

TABLE 1. Classification

of Purpura Fulminans

I. Hemostasis initiated A. Protein C anticoagulant system dysfunction 1. Hereditary a. Homozygous protein C deficiency b. Homozygous protein S deficiency 2. Acquired a. Disseminated intravascular coagulation B. Disorder of other hemostasis regulatory systems 1. Antithrombin I11 11. Idiopathic A. Postinfectious B. Unknown etiology 111. Acute infectious

of varicella, group A streptococcal infection, or a nonspecific viral exanthematous illness. Some cases have followed conditions that apparently were not infectious, such as cutaneous hypersensitivity reaction or stomatitis. Idiopathic PF differs from the acute infectious form in that microthrombi and clinically significant thrombohemorrhagic manifestations in organs other than the skin generally are lacking; distal vascular beds (i.e., extremities) typically are spared; and circulatory collapse is not present initially, although hypovolemic shock and tissue hypoperfusion may develop due to extravasation of blood into the skin. These differences are reflected in the lower mortality rate associated with idiopathic PF (approximately 15%) compared to acute infectious PF. The majority of cases of PF develop in childhood during an acute infection, particularly meningococcal sepsis. This review describes our current knowledge of the pathogenesis of acute infectious PF. Based on this TABLE 2. Infectious Causes of Purpura Fulminans I. Idiopathic (postinfectious) purpura fulminans Varicella Scarlet fever Streptococcal tonsillopharyngitis Viral exanthem Rubella Measles Upper respiratory tract infection Gastroenteritis 11. Acute infectious purpura fulminans Neisseria meningitidis Streptococcus pneumoniae Streptococcus agalactaeae (group B streptococcus) Haemophilus influenzae type b Rickettsia rickettsii Streptococcus pyogenes (group A streptococcus) Staphylococcus aureus Klebsiella pneumoniae Escherichia coli Proteus mirabilis Enterobacter spp. Neisseria catarrhalis Haemophilus aegypticus Capnocytophaga canimorsus

knowledge, the application of new, potentially more effective approaches to its medical management will be discussed.

EPIDEMIOLOGY Purpura fulminans develops in 15% to 25% of those with meningococcemia (6-9). It can develop during infection with any of the meningococcal serogroups; endotoxin production, however, appears to be highest among serogroup C isolates (10). Purpura fulminans occurs only rarely in the course of infection with other organisms, even in the setting of sepsis with DIC. In one series of patients with pneumococcal sepsis, however, 6% (101 165) developed symmetrical peripheral gangrene (1 1). In the neonate, acute infectious PF may be due to group B streptococcus. Although several factors have been identified as predictive of a poor outcome from meningococcemia, size of skin hemorrhage increases with disease severity (12), and the presence of purpura, particularly when generalized, is associated with high morbidity and mortality, as it reflects a profound disturbance in hemostatic mechanisms (8,13-21). While the overall fatality rate in children with meningococcemia in the United States is less than 15%, development of PF heralds mortality in 20% to 60% of cases. Of those who survive, the majority have cutaneous and skeletal deformities due to gangrene (22,23). The presence of petechiae signals septicemia, and onset within 12 hours of initial medical evaluation has been associated with poor outcome in some studies (24,25) but not in others (6,8,17). Thus PF, distinct from petechiae or other cutaneous manifestations of meningococcemia such as a maculopapular eruption, is an important marker of a poor outcome.

CLINICAL CHARACTERISTICS Lesions of PF are similar regardless of the precipitating condition. Cutaneous discomfort develops first, followed by development of erythema with or without edema and petechiae. Sites of involvement appear transiently like ecchymoses, and up to this point the pathologic process in the skin is reversible without progression to necrosis. Lesions evolve rapidly into painful, indurated, welldemarcated, irregularly bordered purpuric papules and plaques which are surrounded by a thin, advancing erythematous border (Figs. 1 and 2). Late findings in necrotic areas are the formation of vesicles and bullae (Fig. 3), which mark the development of hemorrhagic necrosis, and finally firm eschar which ultimately sloughs. Gangrenous necrosis often extends into subcutaneous tissue, and occasionally involves muscle and/or bone (Fig. 4); epiphyseal growth plate necrosis in the growing child

Darmstadt: Acute Infectious Purpura Fulminans

Figure 1. Well-defined, irregularly bordered purpuric plaques on the arm of a child with early acute infectious purpura fulminans due to Neisseria meningitidis.

171

Figure 3. Purpuric plaques and hemorrhagic bullae on the leg of a child with severe, advanced purpura fulminans due to Neisseria meningitidis.

Figure 2. Well-defined, purpuric plaques on the arm of a child, late in the course of meningococcemia.

may lead to limb foreshortening. The distal extremities are often most severely involved, usually in a symmetric manner, probably due to the presence of fewer collateral channels for tissue perfusion, and the relatively greater impact of circulatory collapse on perfusion of distal vascular beds. This differs from idiopathic or postinfectious PF, which typically begins in the skin of the thighs, legs, buttocks, and lower trunk. Acute infectious PF frequently progresses proximally and ultimately may form purpuric plaques of various sizes and shapes in a diffuse, patchy distribution. Development of systemic consumptive coagulopathy (i.e., DIC) is a defining feature of PF, which distinguishes it from other forms of skin necrosis due to dermal vascular occlusion such as warfarin- or heparininduced skin necrosis, thrombotic thrombocytopenic purpura, cryoglobulinemia, antiphospholipid syndrome, or paroxysmal noctural hemoglobinuria. Shock may occur in all forms of PF, but is most characteristic of acute infectious PF. Thrombohemorrhagic manifestations may be found in multiple vascular beds and organ systems,

Figure 4. Gangrenous fifth fingertip due to Neisseria meningitidis.

and multiple organ dysfunction syndrome is common. Fibrinogen, coagulation factors (e.g., factor V and factor VIII), and platelets are consumed in ongoing thrombosis and fibrinolysis. Prothrombin time (PT) and partial thromboplastin time (PTT) are prolonged; fibrin degradation products (e.g., D-dimers) are elevated; and protein C, protein S, and antithrombin I11 levels are reduced. The histopathologic hallmarks of PF are dermal vascular thrombosis and secondary hemorrhagic necrosis (Fig. 5a) (26). Vascular changes are widespread, involving multiple organ systems, particularly the adrenal glands (Waterhouse-Friderichsen syndrome), lungs, and kidneys. The cutaneous vessels most affected are the postcapillary venules in the subpapillary plexus of the papillary dermis, where blood flow velocity is slowest.

172 Pediatric Dermatology Vol. I5 No. 3 May/June 1998

ture of acute infectious PF that distinguishes it from other forms of PF (Fig. 5b). Preliminary identification of N. meningitidis as the cause of PF can sometimes be made by finding the Gram-negative diplococci on Gram’s stain of material obtained by needle aspiration of a petechial skin lesion (27j, or by scraping the lesion with a needle and making a smear of blood (28).

PATH0GENESIS

A

The histopathologic changes in acute infectious PF are essentially identical to those of the Shwartzman reaction. Comparison of these entities has provided valuable insight into possible pathogenetic mechanisms of PF. Out of this increased understanding of the pathophysiology of PF has come the design of more directed and potentially more effective therapy for PF.

The Shwartzman Reaction

a Figure 5. Skin biopsy specimens from children with meningococcemia showing (A) dermal vascular thrombosis and (B) vasculitis. Specimens were stained with hemotoxylin and eosin.

Histopathologic changes begin at the junction of terminal precapillary arterioles with capillaries and generally spare arteries and arterioles. Microthrombi are mixed, composed of fibrin, platelets, and leukocytes, particularly neutrophils; Gram-negative diplococci of N. meningitidis occasionally are visible as well. Endothelial cell swelling is prominent and leads initially to capillary dilatation, manifest clinically as erythema. Acute vascular injury progresses to endothelial cell separation and vessel rupture, allowing for extravasation of formed blood elements into the dermal stroma and development of clinically visible purpura. Extensive hemorrhage is followed by coagulative necrosis. Vasculitis, including a perivascular neutrophilic infiltrate, is a characteristic fea-

The local Shwartzman reaction is a hemorrhagic and necrotizing inflammatory lesion provoked by the injection of endotoxin from Gram-negative bacteria (29,30). Endotoxin from N. meningitidis is 5- to 10-fold more effective at eliciting the reaction than endotoxin from other Gram-negative bacteria (31). Furthermore, endotoxin levels in plasma of patients with fulminant meningococcemia are among the highest ever recorded ( 12), perhaps accounting, at least in part, for the increased incidence of PF in patients with meningococcemia compared to sepsis with other organisms. The two-step Shwartzman reaction is initiated by a local, priming intradermal injection of endotoxin. This elicits a transient perivascular neutrophilic and monocytic inflammatory reaction with increased vascular permeability that is maximal approximately 4 hours after injection. Vessel damage, however, is not seen at this stage. Intravenous challenge with the same endotoxin 18 to 24 hours later results in thrombosis with mixed microthrombi, endothelial cell swelling, dilatation of blood vessels, and necrotizing neutrophilic vasculitis, leading to extravasation of formed blood elements and purpura localized to the prepared dermal site. The degree of hemorrhage elicited by the challenge correlates directly with the accumulation of leukocytes after the priming reaction (32). Development of microthrombi only at skin sites prepared previously with endotoxin suggests that affected endothelial cells were made more thrombogenic during the intervening period between the local and the systemic injections. Of note, a generalized Shwartzman reaction can be produced by two intravenous injections of endotoxin spaced 18 to 24 hours apart, producing a

Darmstadt: Acute Infectious Purpura Fulminans

syndrome which models DIC (30). Tissue damage is similar in the local and generalized Shwartzman reactions, suggesting that they share a common pathogenesis. The local Shwartzman reaction is thought to be more like idiopathic/postinfectious PF, with cutaneous lesions limited to sites prepared in the skin by an antecedent infection, whereas the generalized Shwartzman reaction resembles acute infectious PF with formation of microthrombi in multiple organ systems.

Hemostatic Balance Central to the pathogenesis of the Shwartzman reaction is a disturbance in the balance of anticoagulant and procoagulant activities of endothelial cells. Normally, in the absence of disease, the balance of hemostatic activity on the surface of endothelial cells favors anticoagulation. Systems for ensuring this include the protein Cthrombomodulin system, the antithrombin 111-heparin system, profibrinolytic mechanisms involving plasminogen activation, and inhibition of platelet aggregation through release of prostacyclin (33-36). Thrombomodulin is a high-affinity receptor for thrombin. Binding of thrombin to thrombomodulin activates protein C, which in turn, in concert with its cofactor protein S, destroys clotting factors V and VIII of the intrinsic hemostatic cascade. This prevents factor V-mediated binding of prothrombin to the surface of the platelet, thereby suppressing production of thrombin. Concurrently, interaction of thrombin with thrombomodulin makes thrombin unavailable to stimulate a plethora of pathophysiologic proinflammatory and prothrombotic events in sepsis, including the expression of platelet activating factor and the activation, adhesion, and aggregation of platelets; the transformation of fibrinogen into fibrin with release of fibrinopeptides; the activation of clotting factors V and VIII; and induction of expression of granule membrane protein- 140 on endothelial cells (which promotes adhesion of neutrophils). Thrombin attached to thrombomodulin can be inactivated by antithrombin I11 at a faster rate than can free thrombin. Antithrombin 111 is the major inhibitor of the coagulation cascade. Heparin or heparinlike molecules on endothelial cells stimulate formation of a complex between antithrombin Ill and the active serine center of the serine protease thrombin, neutralizing the activity of thrombin. Similarly, heparin also facilitates inactivation of other serine proteases of the coagulation cascade (i.e., factors IXa, Xa, XIa, XIIa) by antithrombin Ill. Lysis of microthrombi is promoted by endothelial cells through synthesis and release of plasminogen activators such as tissue plasminogen activator (t-PA), which in turn initiates activation of plasminogen.

173

Plasminogen is bound specifically to fibrin, becoming intermeshed and concentrated within the fibrin clot as it forms. There it is sequestered away from protease inhibitors (e.g., a,-antiplasmin) normally present in the blood, and can be activated by fibrin-bound t-PA to plasmin. Plasmin remains bound to fibrin where it is protected from a,-antiplasmin, upregulates its own generation, and degrades fibrinogen and fibrin to dissolve the clot. Outside the milieu of the clot, a,-antiplasmin is able to neutralize plasmin, restricting fibrinolytic activity to the region of the clot. During DIC, however, excessive plasmin is generated and the capacity of a,-antiplasmin to neutralize plasmin in the blood is overcome. Levels of a,-antiplasmin fall due to excessive activation of the fibrinolytic system. Plasmin is then bound to a secondary inhibitor, called a,-macroglobulin, but low-level plasmin activity persists within this complex and systemic fibrinolysis is able to continue. Measurement of the cleaved fibrinogen species produced by the action of plasmin forms the basis for a variety of tests utilized in detecting the systemic fibrinolysis seen in DIC. Biological actions of these cleavage products include potentiation of the hypotensive effects of bradykinin and chemotaxis of monocytes and neutrophils. A final endothelial cell-derived regulator of fibrinolysis is plasminogen activator inhibitor-1 (PAI-1). The major physiologic role of PAI-1 is to suppress the function of t-PA and thus to modulate fibrinolysis.

Cytokines When endothelial cells are injured, a shift in the balance of hemostasis toward a procoagulant state occurs. In acute infectious PF, this is triggered by endotoxin, which in turn causes local production and release of proinflammatory molecules. Principle proinflammatory molecules include the cytokines tumor necrosis factor-a (TNF-a), interferon-? (IFN-y), and interleukin- 1 (IL- 1). During the first or priming step of the Shwartzman reaction, it appears that endotoxin first induces the release of IL- 12 from a variety of cell types, including T and B lymphocytes, natural killer (NK) cells, and mononuclear phagocytes (37). IL-12 in turn induces production of IFN-y, principally from NK cells (38). For optimum generation of IFN-y, however, the presence of TNF-a and macrophages is required (39,40). IFN-y, which plays a central role in the priming reaction (37,38), activates macrophages to produce monokines, including TNF-a and IL1; sensitizes endothelial cells; and primes the endothelial cells to produce TNF-a and IL-1 during the intravenous challenge phase of the Shwartzman reaction. A crucial event during the priming process appears to be upregu-

174 Pediatric Dermatology Vol. 15 No. 3 May/June 1998

lation of intracellular adhesion molecule- 1 (ICAM-1) on vascular endothelial cells in dermal blood vessels (41,42). This may be due principally to the action of endotoxin and TNF-a, and perhaps also IL-1 and IFN-y, all of which are capable of upregulating ICAM-1 expression. Injection of IFN-y, TNF-a, or TNF-a and IL-1 acting synergistically can be used in place of endotoxin to prepare the skin for a local Shwartzman reaction (37,43,44). The inflammatory infiltrate generated during the priming phase, as a consequence of upregulated ICAM-1 expression (41), and the plasma leakage observed during the priming phase depend on local generation of TNF-a(45). Generation of an acute inflammatory cell infiltrate is an absolute requirement of the Shwartzman reaction; the reaction does not occur in neutropenic animals (46). The perivascular, interstitial infiltrate generated during the priming phase, however, does not cause vascular damage or hemorrhage. The second or challenge phase of the Shwartzman reaction, which results in the vasculitis and hemorrhage that becomes manifest clinically as purpura, requires the recruitment and aggregation of neutrophils in the dermal vessels of prepared skin sites (47,48). An early, critical event of the challenge phase appears to be the upregulation of CD1 1b/CD18 (Mac-1) on leukocytes, which subsequently allows them to adhere to the ICAM-1 molecules that were expressed on vascular endothelial cells during the priming phase and to aggregate within affected vessels (41,42). Endothelial cells that are injured by endotoxin, or by exposure to TNF-a and IL-I , and/or infiltrating neutrophils synthesize and express tissue factor (i.e., tissue thromboplastin), the most potent procoagulant known, on their surface (49-52). TNF-a also induces the expression of tissue factor on monocytes (53). In children with meningococcal sepsis, tissue factor expression is increased on circulating monocytes (54). Tissue factor activates the extrinsic pathway, leading ultimately to production of fibrin. Deposition of fibrin due to tissue factor-initiated coagulation within the vessels of prepared skin sites is required for pathogenesis of the Shwartzman reaction (47). Exposure of endothelial cells to endotoxin or to TNF-a and IL-1 also suppresses thrombomodulin activity (55,56). This leads to decreased protein C activity and in turn to impaired degradation of activated factors V and VIII, and to increased thrombin procoagulant activity through enhanced availability of thrombin. Secretion of PAI-1 by endothelial cells also is increased, which further promotes thrombosis (57,58). Cytokines, particularly TNF-a, which are produced by injured endothelial cells as well as by injured keratinocytes and infiltrating monocytes and neutrophils, appear to be the principle mediators of the hemodynamic and histopathologic events during the second (i.e., pro-

voking) phase of the Shwartzman reaction. The effects of TNF-a are potentiated by IL-1 and IFN-y (37). Serum levels of TNF-a were elevated in 91% of children with acute infectious PF, and the level of TNF-a correlated negatively with fibrinogen levels and positively with endotoxin levels (59) and with the severity of infection and mortality (60). A central concept in the pathogenesis of the Shwartzman reaction which has emerged recently is that of cytokine balance. Concurrently with the production of proinflammatory cytokines such as TNF-a, endotoxin also induces the production of antiinflammatory molecules such as IL-10. IL-10 is a potent macrophage deactivator which blocks the synthesis of TNF-a, IL-1, and other proinflammatory cytokines such as IL-6, IL-8, and granulocyte-macrophage colony-stimulating factor (GMCSF) by monocytes (61) and macrophages (62). Furthermore, IL-10 suppresses synthesis of IFN-y by helper T cells (63) and NK cells (64). Mice deficient in IL-10 production were extremely vulnerable to generation of a generalized Shwartzman reaction, requiring 100- to 200fold less endotoxin during the priming phase compared to wild-type mice (65). Production of TNF-a in response to endotoxin sensitization also was exaggerated. Moreover, IL-10 infusion blocked the sensitization step. Perhaps endogenous IL- 10 normally inhibits the production of IL-12 and IFN-y during the preparatory phase of the Shwartzman reaction (65).

Acquired Protein C and Protein S Deficiency Protein C naturally attenuates the proinflammatory activity stimulated by endotoxin and cytokines. It decreases endotoxin-induced cytokine production by monocytes and inhibits the downregulation of CD14 and C D l l b on leukocytes by endotoxin (66). Acquired deficiency of protein C has been described in many patients with acute infectious PF (67-74) and is increasingly implicated in the pathogenesis of this disorder. The degree of protein C deficiency can be correlated with size of skin lesions (73), clinical severity of the illness (9), and mortality (9,70,73). It has been suggested that decreased protein C levels indicate that meningococcal endotoxin induces a relatively greater decrease in natural anticoagulant than procoagulant proteins, and that the coagulopathic effects of meningococcal endotoxin are greater than those observed in most other bacterial infections (9). Most patients with reduced protein C levels have a concomitant reduction in the level of protein S (67-74). Protein S is active only in the free, unbound form. Normally, approximately 60% of total protein S is carried by the complement component C4b binding protein, which is an acute-phase reactant. Elevation of C4b binding pro-

Darmstadt: Acute Infectious Purpura Fulminans

tein during acute infection may lead to a reduction in free protein S activity. Relatively low baseline protein C and S level in infants and young children may explain their increased risk of development of PF compared to adults (67). Some patients with postinfectious PF have had profound depression in levels of protein s, with normal to only slightly decreased protein C at presentation and/or relatively rapid recovery of protein C compared to protein S levels, suggesting that a subtle difference exists in the pathophysiology of their disease compared to those with acute infectious PF (75-79). Recently some patients with postinfectious PF following acute chickenpox have been shown to have antiprotein S IgM and IgG autoantibodies (79,80); many of these patients have also had lupus anticoagulant and antiphospholipid antibodies (77,80,81). These autoantibodies appear to bind to and increase the clearance of protein S. Patients with postinfectious (i.e., post-varicella) transient protein S deficiency, including those with autoantibody-mediated disease, manifest a spectrum of thromboembolic complications, ranging from pulmonary emboli and/or thrombotic events affecting other internal organs but sparing the skin, to PF with coexistent major organ thromboembolic disease (78-80,82,83). The presence of autoantibodies directed against protein S has important implications for therapy, since protein S levels in these patients may not respond to administration of fresh frozen plasma (FFP) (79). Furthermore, infusions of FFP may facilitate the formation of immune complexes and induce a serumsickness-like illness (79). Generation of dermal vascular necrosis and hemorrhagic skin necrosis in PF involves a combination of disordered hemostasis and inflammatory-mediated pathologic changes. Events that are common in all forms of PF include increased tissue factor expression and exposure, downregulation of thrombomodulin expression and protein C activity, and impaired fibrinolysis (1). In patients with PF due to hereditary protein C or protein S deficiency, disordered hemostasis predominates, and little or no inflammatory infiltrate is present. The role of cytokines in the pathogenesis of these lesions is thought to be minor (1). In acute infectious PF, however, cytokine-mediated inflammation may play a predominant role, while disordered hemostasis augments the pathologic process. Reduction of protein C and protein S in acute infectious PF most likely stems from consumption during coagulopathy, and their derangement is thought to perpetuate and exacerbate rather than initiate dermal vascular thrombosis (4). Nevertheless, therapy directed at restoring the functional capacity of the protein C system in animal models had led to promising results. Giving rab-

175

bits human activated protein C attenuated the local Shwartzman reaction (84). In a baboon model of Gramnegative sepsis, blocking protein C activation with monoclonal antibodies increased mortality, suggesting that protein C played an important role in attenuating coagulopathy in sepsis (85). When protein C was administered prophylactically, it prevented the coagulopathic and lethal effects of endotoxin-associated sepsis (85), further suggesting that the protein C system may play a role in modulating the effects of inflammatory cytokines (4).

Consumption of Antithrombin I11 A decrease in antithrombin I11 also occurs in patients with acute infectious PF (68,71-73,86,87). In some patients with PF, derangement in antithrombin I11 appeared to more closely parallel the clinical course than did levels of protein C or protein S (72). Interest in the utility of antithrombin 111 concentrate as a therapeutic measure in PF has increased because of studies in animals and humans which have demonstrated its efficacy as a prophylactic and therapeutic agent in DIC (88-93).

Complement Complement, along with specific antibodies, plays a pivotal role in defense against N. meningitidis. Interestingly, while those with defects in the terminal complement cascade (C5 to C9) have increased risk of acquiring meningococcemia, the severity of their infections tends to be relatively mild; rather, they are at increased risk for recurrent episodes. Overall approximately 4% of those who present with acute, first-episode meningococcemia have a terminal complement deficiency and low total hemolytic complement levels (CH,,). More severe episodes may be associated with a rare, properdin deficiency (94). These patients have normal CH,, but low alternative hemolytic complement. In the course of meningococcal infection, significant complement activation occurs, leading to formation of the anaphylatoxins C3a, C4a, and C5a (95). Complement activation appears to play an important role during the provoking phase of the Shwartzman reaction by upregulating expression of CD 18 on neutrophils, activating them, and inducing them to adhere to ICAM-1 on endothelium of prepared skin sites (42). In addition, generation of the membrane attack complex at the bacterial surface induces release of endotoxin. Increasing severity of disease and mortality are associated with higher levels of C3 activation products and terminal complement complex (96).

176 Pediatric Dermatology Vol. 15 No. 3 May/June 1998

Location of Purpura There is no adequate explanation for why lesions of PF develop at particular skin sites. Hypotheses include sitespecific differences in cytokine release by endothelial cells; variability in endothelial cell expression of procoagulant and anticoagulant factors, cytokine receptors, or transduction of signal following binding of cytokine to its receptor; differences in expression of cell adhesion molecules and thus in leukocyte trafficking; physical properties of the skin such as variations in temperature or blood flow; or cutaneous trauma (1).

MEDICAL MANAGEMENT Greater understanding of the pathophysiology of PF has allowed the design of new approaches to the medical management of this disorder during the past decade (Table 3). Before considering these therapies, however, it must first be emphasized that initial management of the patient with acute infectious PF must be focused on preserving life through respiratory and hemodynamic support and prompt intravenous antibiotic coverage (e.g., third-generation cephalosporin such as ceftriaxone). Circulatory collapse with tissue hypoperfusion and ischemia directly damages endothelial cells and predisposes to thrombosis; the propensity to develop microthrombi is exacerbated in the presence of flow stasis. Furthermore, the high levels of catecholamines produced during shock shunt blood away from the liver, decreasing hepatic synthesis of the proteins involved in preserving hemostatic balance and decreasing hepatic clearance of activated clotting factors. The relative contribution of inflammatory cytokines and components of hemostasis in the pathogenesis of acute infectious PF may vary from paTABLE 3. Therapeutic Modalities in the Medical Management of Acute Infectious Purpura Fulminans Fresh frozen plasma Cryoprecipitate Vitamin K Protein C concentrate Antithrombin I11 concentrate Heparin Tissue plasminogen activator Prostacyclin Epidural sympathetic blockade Topical nitroglycerin Plasmapheresis Double-volume exchange transfusion Hyperbaric oxygen Hirudin (leeches) Antiendotoxin antibodies Anti-TNF-a antibody IL-1 receptor antagonist Platelet activating factor receptor antagonist Pentoxify lline

tient to patient. Consequently, therapy must be individualized, and often will involve a combination of modalities. Acute infectious PF typically is a fulminant disorder with rapid progression to irreversible tissue necrosis, emphasizing the importance of prompt therapeutic intervention. Surgical consultation should be sought early in the course to monitor compartment pressures and intervene in compartment syndrome. Nutritional support is also important, and should be continued during the rehabilitative phase. Initial laboratory determinations that will help to guide therapeutic decisions in the management of patients with PF include a complete blood count, platelet count, PT, PTT, fibrinogen, fibrin degradation products, protein C, free protein S, and antithrombin 111. Blood components that may be necessary in individual cases, depending on the results of these tests, include packed red blood cells for anemia due to massive hemorrhage into the skin, platelets to correct thrombocytopenia due to platelet consumption, and/or cryoprecipitate to replace fibrinogen in patients with DIC. Therapeutic interventions that should be initiated for all patients with PF and DIC include vitamin K and FFP (8-12 mgkg every 12 hours), which are given to correct possible deficiencies of vitamin K-dependent coagulation factors (97,98), antithrombin 111, protein C, and protein s. Fresh frozen plasma does not aggravate consumptive coagulopathy, is not associated with bleeding or thrombocytopenia, and thus is preferable to heparin (99). Continued use of FFP should be guided by regular measurements of protein C, protein S, and antithrombin 111 levels. It is recommended that antithrombin I11 be maintained above 50% of normal and protein C and S at at least 25% of normal (4,74), although it is not precisely known to what level factor activities must be raised to be effective. Clinical response must be monitored carefully, since test results may be unavailable to influence management decisions. Although FFP is effective in many patients, the fluid volume that must be administered during repeated infusions may limit its utility. In such instances, particularly when PF continues to advance in a given patient in whom the levels of specific factors such as protein C or antithrombin 111 remain low despite maximal therapy with FFP, one may consider the use of protein C and/or antithrombin I11 concentrates (see below). An additional, albeit low risk of use of FFP is the transmission of blood-borne disease. Heparin may be effective in reversing the development of skin necrosis in some patients with PF, particularly those with thrombotic or prethrombotic states, including those with autoantibodies against protein s. Heparin acts through a combination of inhibition of clotting, interruption of consumption of anticoagulant fac-

Darmstadt: Acute Infectious Purpura Fulminans

tors, and/or inhibition of the procoagulant and antifibrinolytic effects of thrombin. It is capable of preventing the Shwartzman reaction (77,79,100-102). Two prospective, controlled trials of heparin use in acute meningococcemia, however, demonstrated no effect on survival (103,104). While benefit is often not seen in PF, heparin may limit skin, digit, and extremity necrosis in some patients (7,105,106). Concurrent administration of FFP (which contains antithrombin 111) or antithrombin 111 concentrate may be necessary for an effect to be realized, given that heparin acts by facilitating the activity of antithrombin 111. Heparin may exacerbate bleeding or induce thrombocytopenia. consequently, many experts are reluctant to administer this unproven therapy and correctly assert that control of the underlying pathologic process that triggered DIC is the most important step in controlling it.

Experimental Modalities Because of the limitations of treatment with FFP and/or heparin, additional modalities have been attempted in the treatment of PF. None of these, however, have been evaluated in a controlled prospective trial. Their use, therefore, remains experimental, and, in many cases, controversial. Consequently, while we await further data from controlled clinical trials on efficacy and safety of these experimental products, FFP remains central in the treatment of PF. Protein C Concentrate Five children with acute infectious PF and DIC have been treated successfully with protein C concentrate, 70 to 100 U/kg every 6 hours or a continuous infusion of 10 U/kg/hr during the acute phase of the illness (71,74). Rising protein C levels were associated with an increase in fibrinogen, a decrease in D-dimers, and arrest of DIC. Although protein C concentrate appears to hold promise as an adjunctive treatment in PF, its availability may be limited. It is anticipated that protein S concentrates might be efficacious, particularly in patients with anti-protein S autoantibodies (79), but these products are not yet available. Antithrombin III Concentrate Administration of antithrombin I11 concentrate, 60 IU/kg loading dose, then 60 IU/kg/24 hr for 2 days, to two adults with acute infectious PF resulted in normalization of antithrombin I11 levels within 24 hours, concomitant with improved skin perfusion, reversal of DIC, and improvement in skin lesions (68,72). This occurred despite a more marked and prolonged depression of protein C than antithrombin I11 levels.

177

Tissue Plasminogen Activator By the time PF is apparent clinically, significant intravascular microthromboses and impairment of fibrinolysis have developed. Use of recombinant tissue plasminogen activator @-PA) is predicated upon the notion that dissolution of microthromboses could restore organ perfusion and reduce mortality, since levels of PAI-1 correlate with mortality in acute infectious PF (107). While protein C concentrate has little fibrinolytic effect, rt-PA given as an adjunctive therapy appears to have improved the microcirculation and restored peripheral pulses in three patients with acute infectious PF due to N. meningitidis (108-1 10). It may be particularly useful in lysing clots, such as intracardiac or renal vein thromboses. Given that levels of PAI-1 are highest in patients with meningococcemia during the first few hours after hospital admission (107), this therapy theoretically is of greatest benefit when given as soon as possible. It does not affect blood pressure, and its short half-life (5 minutes) allows for precise use and minimization of side effects. Deleterious bleeding is of concern. Approximately 0.5% to 1.0% of adults given rt-PA following myocardial infarction have developed hemorrhagic stroke (1 11). Similarly, in a review of 154 children given fibrinolytic therapy for femoral artery thrombosis after cardiac catheterization, one child (0.6%) suffered significant intracerebral hemorrhage (1 12). It is likely that the incidence of hemorrhagic complications would be higher in those with severe coagulopathy. Until further data on its use are available, rt-PA should be reserved for those patients with fulminant, life-threatening PF with shock and with major vessel thrombosis who have failed conventional management (108). Prostacyclin Prostacyclin is a vasodilator which also prevents thromboxane-induced platelet aggregation. It has been used successfully in patients with pneumococcal and group B streptococcal PF (1 13-1 15). An added benefit of prostacyclin in the treatment of neonates is its action to relieve pulmonary hypertension (1 13). On the other hand, prostacyclin may cause hypotension, and its use is contraindicated in patients with shock. Vasodilatation Epidural sympathetic blockade with bupivicaine has been utilized successfully in a limited number of patients to restore tissue perfusion, alleviate ischemia, and reverse the development of PF (1 16-1 18). Theoretically this technique works primarily by dilating vessels in microvascular beds which have constricted in response to systemic release of vasoactive mediators or afferent nociceptive signals, and presumably it has its greatest effect in those vessels which are only partly occluded by mi-

178 Pediatric Dermatology Vol. 15 No. 3 May/June 1998

crothrombi (1 16). This modality may be particularly advantageous when utilized early in the course (1 1) in patients with limited PF who require vasopressors for hemodynamic instability and in whom systemic vasodilators are contraindicated. Epidural sympathetic blockade should not be employed in those with meningitis or uncontrolled coagulopathy. Topical application of nitroglycerin (e.g., Nitropaste 2%) was used successfully to restore perfusion and arrest or reverse development of cutaneous necrosis on various areas of the body (e.g., forearms, legs, buttocks, penis) after sympathetic caudal blockade had failed (1 19). Plasmapheresis Plasmapheresis and double volume exchange transfusion are aimed at removing circulating cytokines, endotoxin, activated monocytes, and granulocytes (120-123). These forms of therapy appear to be suited for the rare instance of autoantibody-mediated depletion of an essential hemostatic component, as has been reported in children with PF due to an antibody directed against protein S (79). Plasmapheresis has the added utility of preventing fluid overload in patients requiring large volumes of fluid andlor blood products. Hyperbaric Oxygen Hyperbaric oxygen is controversial, but by increasing the partial pressure of oxygen within patient blood vessels, the diffusion distance of oxygen into surrounding tissue may be increased and tissue oxygenation restored. This modality appears to have prevented or halted the progression of tissue necrosis in some patients (124-126). Medicinal Leeches Medicinal leeches applied to the hands, feet, and legs of one patient appeared to produce a rapid improvement in perfusion (127). Hirudin in leech saliva is the most potent inhibitor of thrombin known (128). Furthermore, it can penetrate thrombi to neutralize thrombin bound to fibrin, an effect which cannot be achieved with heparin (129). It is conceivable that inhibition of thrombin, using recombinant hirudin or hirudin analogs, may be possible in the future. Corticosteruids Although corticosteroids were advocated in the past, their use in PF is best restricted to patients with adrenal gland thrombosis and insufficiency (105).

Lessons from Treatment of Sepsis Based on the notion that massive, uncontrolled inflammation, driven initially by the release of proinflammatory cytokines, is the principle cause of sepsis, a host of novel, targeted immunotherapies for sepsis have been developed and tested for efficacy in double-blind,

placebo-controlled, randomized trials. Agents tested have been directed at blocking the activity of endotoxin [e.g., antiendotoxin antibodies HA- IA and ES generated against the conserved endotoxin core from the J5 mutant of Escherichia coli (130-135); neutralizing the effects of the cytokines whose production is stimulated by endotoxin [e.g., recombinant 1L-1 receptor antagonist (136), dimeric tumor necrosis factor receptor (137), and monoclonal antibody to TNF-cr (138); and inhibiting platelet activation [e.g., platelet activating factor receptor antagonist (139,140)]. Pentoxifylline has a variety of actions including reduction of release of TNF-a (141). Among these therapies, only antiendotoxin antibodies have been tested systematically in patients with PF. Rabbits immunized with the JS mutant of E. culi were protected against development of the Shwartzman reaction when injected with purified endotoxin from N. nzerzingitidis (142). In children with acute infectious PF, however, administration of antiserum raised against E. coli J5 did not alter the course of disease or mortality (143). A randomized, controlled, multicenter, prospective trial of antiendotoxin antibody in patients with meningococcemia is now under way. Results from trials of immunotherapeutic agents in the treatment of sepsis have, in general, been disappointing. Improved survival has been achieved only in subgroups of patients identified retrospectively, not in the entire patient population. Substantiation of benefit to those subgroups of patients has been difficult to demonstrate prospectively, and these agents largely have been deemed ineffective in the treatment of patients with sepsis with or without PF (144-146). This has resulted in a reevaluation of our hypotheses and assumptions about the pathogenesis of sepsis. It is now generally accepted that the host response to a severe infectious challenge involves the generation of both proinflammatory (e.g., TNF-a, IL-1, IL-6) and antiinflammatory molecules (e.g., IL-4, IL-10, IL-11, IL13, soluble TNF receptors, IL- 1 receptor antagonists, transforming growth factor-p) (147,148). The latter molecules provide a buffer or regulatory check on the action of the proinflammatory molecules and act to restore homeostasis. It appears that the outcome for the patient may be dependent on the ability of the immune system to reestablish a balanced immunologic response. An overwhelmingly robust proinflammatory response may lead to shock, while an excessive compensatory antiinflammatory reaction may lead to immunosuppression and increased susceptibility to ongoing, overwhelming, or secondary infection. Consequently, it has been hypothesized that in order to effectively manage patients with sepsis, their state of inflammation must be known (144,146,147). For example, in some patients with a predominant compensatory antiinflammatory response, it may be detrimental to administer agents aimed at down-

Darmstadt: Acute Infectious Purpura Fulminans

regulating proinflammatory mediators. Although it seems counterintuitive, it may, in fact, be necessary in these instances to administer proinflammatory mediators in order to restore immunologic balance (149). It appears unlikely that a single therapeutic agent will be capable of reversing the complex pathophysiologic derangements that occur in PF. However, as we gain in our understanding of the pathophysiologic mechanisms of PF and in our ability to rapidly diagnose infection and, perhaps, the inflammatory state of the patient, our ability to preserve life and prevent disfigurement due to acute infectious PF is likely to improve.

REFERENCES 1. Adcock DM, Bronza J, Marlar RA. Proposed classification and pathologic mechanism of purpura fulminans and skin necrosis. Semin Thromb Hemost 1990;16:333-340. 2. Muller FM, Ehrenthal W, Hafner G, Schranz D. Purpura fulminans in severe congenital protein C deficiency: monitoring of treatment with protein C concentrate. Eur J Pediatr 1996;155:20-25. 3. Dreyfus M, Masterson M, David M, et al. Replacement therapy with a monclonal antibody purified protein C concentrate in newborns with severe congenital protein C deficiency. Semin Thromb Hemost 1995;21:371-381. 4. Francis RB. Acquired purpura fulminans. Semin Thromb Hemost 1990;16:3 10-325. 5. Hjort PF, Rapaport SI, Jorgensen L. Purpura fulminans. Report of a case successfully treated with heparin and hydrocortisone. Review of 50 cases from the literature. Scand J Haematol 1964;1:169-192. 6. Wong VK, Hitchcock W, Mason WH. Meningococcal infections in children: a review of 100 cases. Pediatr Infect Dis J 1989;8:224-227. 7. Kuppermann N, Inkelis SH, Saladino R. The role of heparin in the prevention of extremity and digit necrosis in meningococcal purpura fulminans. Pediatr Infect Dis J 1994;13:867-873. 8. Algren JT, La1 S, Cutliff SA, Richman BJ. Predictors of outcome in acute meningococcal infection in children. Crit Care Med 1993;21:447-452. 9. Powars D, Larsen R, Johnson J, et al. Epidemic meningococcemia and purpura fulminans with induced protein C deficiency. Clin Infect Dis 1993;17:254-261. 10. Andersen BM. Endotoxin release from Neisseria meningitidis. Relationship between key bacterial characteristics and meningococcal disease. Scand J Infect Dis Suppl 1989;64: 1 4 3 . 11. Johansen K, Hansen ST. Symmetrical peripheral gangrene (purpura fulminans) complicating pneumococcal sepsis. Am J Surg 1993;165:642-645. 12. Brandtzaeg P. Pathogenesis of meningococcal infections. In: Cartwright K, ed. Meningococcal disease. New York: John Wiley & Sons, 1995:71-114. 13. Edwards MS, Baker CJ. Complications and sequellae of meningococcal infections in children. J Pediatr 1981;99: 540-545. 14. Leclerc F, Buescart R, Guillois B, et al. Prognostic factors of severe infectious purpura in children. Intensive Care Med 1985;11:140-143. 15. Thomson APJ, Sills JA, Hart CA. Validation of the

179

Glasgow meningococcal septicemia prognostic score: a 10-year retrospective survey. Crit Care Med 1991;19: 26-30. 16. Tesoro LJ, Selbst SM. Factors affecting outcome in meningococcal infections. Am J Dis Child 1991;145:218220. 17. Toews WH, Bass JW. Skin manifestations of meningococcal infections. Am J Dis Child 1974;127:173-176. 18. Sinclair JF, Skeoch CH, Hallworth D. Prognosis of meningococcal septicemia. Lancet 1987;ii:38. 19. Spicer TE, Rau JM. Purpura filminans. Am J Med 1976; 61:S66-57 1. 20. Chu DZJ, Blaisdell F W . Purpura fulminans. Am J Surg 1982;143:356-363. 21. Brandtzaeg P, Dahle JS, Hoiby EA. The occurrence and features of hemorrhagic skin lesions in 115 cases of systemic meningococcal disease. NIPH Am 1983;6: 183203. 22. Herrara R, Hobar PC, Ginsburg CM. Surgical intervention for complications of meningococcal-induced purpura fulminans. Pediatr Infect Dis J 1994;13:734-737. 23. Sheridan RL, Briggs SE, Remensnyder JP, Tompkins RG. Management strategy in purpura fulminans with multiple organ failure in children. Burns 1996;22:53-56. 24. Stiehm ER, Damrosch DS. Factors in the prognosis of meningococcal infection. J Pediatr 1966;68:457467. 25. Niklasson P-M, Lundberg P, Strandell T. Prognositc factors in meningococcal disease. Scand J Infect Dis 1971; 3: 17-25 26. Adcock DM, Hicks MJ. Dermatopathology of skin necrosis associated with purpura fulminans. Semin Thromb Hemost 1990;16:283-292. 27. van Deuren M. van Dijke BJ, Koopman RJJ, et al. Rapid diagnosis of acute meningococcal infections by needle aspiration or biopsy of skin lesions. BMJ 1993;306: 12291232. 28. Periappuram M, Taylor MRH, Keane CT. Rapid detection of meningococci from petechiae in acute meningococcal infection. J Infect 1995;31:201-203. 29. Shwartzman G. A new phenomenon of local skin reactivity to B. typhosus culture filtrate. Proc SOCExp Biol Med 1928;560-561. 30. Bronza JP. Shwartzman reaction. Semin Thromb Hemost 1990;16:326-333. 3 1. Davis CE, Arnold K. Role of meningococcal endotoxin in meningococcal purpura. J Exp Med 1974;140:159-171. 32. Movat HZ, Jaynes BJ, Wasi K, Movat W, Kopaniak MM. Quantitation of the development and progression of the local Shwartzman reaction. In: Agarwal MK, ed. Bacterial endotoxins and host response. Amsterdam: Elsevier, 1980: 179-20 1. 33. Esmon CT. Blood coagulation. In: Nathan DG, Orkin SH, eds. Nathan and Oski’s hematology of infancy and childhood, 5th ed. Philadelphia: WB Saunders, 1998:15311556. 34. Hajjar KA. The molecular basis of fibrinolysis. In: Nathan DG, Orkin SH, eds. Nathan and Oski’s hematology of infancy and childhood, 5th ed. Philadelphia: WB Saunders, 1998:1557-1573. 35. Bauer KA, Rosenberg RD. Control of coagulation reac-

180 Pediatric Dermatology Vol. 15 No. 3 May/June 1998

tions. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ, eds. Williams hematology, 5th ed. New York: McGrawHill, 1995:1239-1252. 36. Levi M, Cate H, van der Poll T, van Deventer JH. Pathogenesis of disseminated intravascular coagulation in sepsis. JAMA 1993;270:975-979. 37. Ozmen L, Pericin M, Hakimi J, et al. Interleukin-12, interferon-?, and tumor necrosis factor-cc are the key cytokines of the generalized Shwartzman reaction. J Exp Med 1994;180:907-915. 38. Heremans H, Dillen C, van Damme J, Billiau A. Essential role for natural killer cells in the lethal lipopolysaccharide-induced Shwartzman-like reaction in mice. Eur J Immunol 1994;24:1155-1160. 39. Wherry JC, Schreiber RD, Unanue ER. Regulation of gamma interferon production by natural killer cells in scid mice: roles of tumor necrosis factor and bacterial stimuli. Infect Immun 1991;59:1709-1715. 40. Farrar MA, Schreiber RD. The molecular cell biology of interferon-gamma and its receptor. Annu Rev Immunol 1993;11:571-611. 41. Argenbright LW, Barton RS. Interactions of leukocyte integrins with intercellular adhesion molecule 1 in the production of inflammatory vascular injury in vivo. The Shwartzman reaction revisited. J Clin Invest 1992;89: 259-272. 42. Argenbright LW, Barton RS. The Shwartzman response: a model of ICAM-I dependent vasculitis. Agents Actions 1991;34:208-210. 43. Movat HZ, Burrows CE, Cybulsky MI, Dinarello CA. Acute inflammation and a Shwartzman-like reaction induced by interleukin-1 and tumor necrosis factor. Am J Pathol 1987;129:463476. 44. Beck G, Habicht GS, Benach JL, Miller F. Interleukin-1: a common endogenous mediator of inflammation and the local Shwartzman reaction. J Immunol 1986;1363025303 1. 45. Norman KE, Williams TJ, Feldmann M, Rossi AG. Effect of soluble P55 tumour-necrosis factor binding fusion protein on the local Shwartzman and Arthus reactions. Br J Pharmacol 1996;117:471-478. 46. Stetson CA, Good RA. Shwartzman phenomenonparticipation of polymorphonuclear leukocytes. J Exp Med 1951;93:49-53. 47. Zivelin A, Rao VH, Rapaport SI. Evidence for an essential role of tissue factor dependent blood coagulation in the pathogenesis of the local Shwartzman reaction. Blood Cells Mol Dis 1995;21:9-19. 48. Toh H, Miyata T, Torisu M. Characterization of neutrophi1 activation by repeated injection of endotoxin in rabbits. Role of neutrophils in the generalized Shwartzman reaction. J Leukoc Biol 1993;53:256-263. 49. Muller-Berghaus G. Pathophysiologic and biochemical events in disseminated intravascular coagulation: dysregulation of procoagulant and anticoagulant activities. Semin Thromb Hemost 1989;15:58-87. 50.Nawroth PP, Handley D, Esmon CT, Stern DM, Interleukin-1 induces endothelial cell procoagulant activity while suppressing cell surface anticoagulant activity. Proc Natl Acad Sci USA 1986;83:3460-3464. 51. Nawroth PP, Stern DM. Modulation of endothelial cell hemostatic properties by tumor necrosis factor. J Exp Med 1986;163:740-745.

52. Brox JH, Osterud B, Bjorklid E, Fenton JW. Production and availability of thromboplastin in endothelial cells: the effect of thrombin, endotoxin and platelets. Br J Haematol 198437:239-246. 53. Conkling PR, Greenberg CS, Weinberg JB. Tumor necrosis factor induces tissue factor-like activity in human leukemia cell line U937 and peripheral blood monocytes. Blood 1988;72:128-133. 54. Osterud B, Flaegstad T. Increased tissue thromboplasin activity in monocytes of patients with mingococcal infections related to unfavourable prognosis. Thromb Haemost 1983;49:5-7. 55. Moore KL, Andreoli SP, Esmon NL, Esmon CT, Bang NU. Endotoxin enhances tissue factor and suppresses thrombomodulin expression of human vascular endothelium in vitro. J Clin Invest 1987;79:124-130. 56. Moore KL, Esmon CT, Esmon NL. Tumor necrosis factor leads to the internalization and degradation of thrombomodulin from the surface of bovine aortic endothelial cells in culture. Blood 1989;73:159-165. 57. Suffredini AF, Harpel PC, Parrillo JE. Promotion and subsequent inhibition of plasminogen activator after administration of intravenous endotoxin to normal subjects. N Engl J Med 1989;320:1 165-1 172. 58. Bevilacqua MP, Schleeef RR, Gimbrone MA, Loskutoff DJ. Regulation of the fibrinolytic system of cultured human vascular endothelium by interleukin-1. J Clin Invest 1986;78:587-591. 59. Wage A, Halstensen A, Espevik T. Association between tumor necrosis factor in serum and fatal outcome in patients with meningococcal disease. Lancet 1987;1:355357. 60. Girardin E, Grau GE, Dayer J-M, et al. Tumor necrosis factor and interleukin-1 in the serum of children with severe infectious purpura. N Engl J Med 1988;319:397400. 61. de Waal-Malefyt R, Abrams J, Bennett B, Figdor CG, de Vries J. Interleukin 10 (IL-10) inhibits cytokine synthesis by human monocytes: an autoregulatory role of IL-10 produced by monocytes. J Exp Med 1991;174:12091220. 62. Fiorentino DF, Zlotnik A, Mosmann TR, Howard M, O’Garra A. IL-10 inhibits cytokine production by activated macrophages. J Immunol 1991;147:3815-3822. 63. Fiorentino DF, Bond MW, Mosmann TR. Two types of mouse helper T-cells. IV. Th2 clones secrete a factor that inhibits cytokine production by Thl clones. J Exp Med 1989;170;2081-209S. 64. Hsu DH, Moore KW, Spits H. Differential effects of IL-4 and IL-10 on IL-2 induced IFN-gamma synthesis and lymphokine-activated killer activity. Int Immunol 1992; 4563-569. 65. Berg DJ, Kuhn R, Rajewsky K, et al. Interleukin-10 is a central regulator of the response to LPS in murine models of endotoxic shock and the Shwartzman reaction but not endotoxin tolerance. J Clin Invest 1995;96:2339-2347. 66. Grey S, Hau H, Salem HH, Hancock WW. Selective effects of protein C on activation of human monocytes by lipopolysaccharide, interferon-?, or PMA: modulation of effects on C D l l b and CD14 but not CD25 or CD54 induction. Transplant Proc 1993;2S:2913-29 14. 67. Powars DR, Rogers ZR, Patch MJ, McGehee WG, Francis RB. Purpura fulminans in meningococcemia: associa-

Darmstadt: Acute Infectious Purpura Fulminans

tion with acquired deficiencies of proteins C and S. N Engl J Med 1987;317:571-572. 68. Fourrier F, Lestavel P, Chopin C, et al. Meningococcemia and purpura fulminans in adults: acute deficiencies of proteins C and S and early treatment with antithrombin 111 concentrates. Intensive Care Med 1990;16:121-124. 69. Madden RM, Gill JC, Marlar RA. Protein C and protein S levels in two patients with acquired purpura fulminans. Br J Haematol 1990;75:112-117. 70. Leclerc F, Hazelzet J, Jude B, et al. Protein C and protein S deficiency in severe infectious purpura of children: a collaborative study of 40 cases. Intensive Care Med 1992; 18:1-4. 71. Gerson WT, Dickerman JD, Bovill EG, Golden E. Severe acquired protein C deficiency in purpura fulminans associated with disseminated intravascular coagulation: treatment with protein C concentrate. Pediatrics 1993;91:418422. 72. Cobcroft R, Henderson A. Meningococcal purpura fulminans treated with antithrombin I11 concentrate: what is the optimal replacement therapy? Aust N Z J Med 1994;24: 575-576. 73. Fijnvandraat K, Derkx B, Peters M, et al. Coagulation activation and tissue necrosis in meningococcal septic shock: severely reduced protein C levels predict a high mortality. Thromb Haemost 1995;73:15-20. 74 Rivard GE, David M, Farrell C, Schwarz HP. Treatment of purpura fulminans in meningococcemia with protein C concentrate. J Pediatr 1995;126:646-652. 75. Alessi MC, Aillaud MF, Boyer-Neumann C, Viard L, Camboulives J, Juhan-Vague I. Cutaneous necrosis associated with acquired severe protein S deficiency. Thromb Haemost 1993;69:524-526. 76. Zenz W, Muntean W. Cutaneous necrosis associated with acquired severe protein S deficiency. Thromb Haemost 1994;71 :396. 77. Manco-Johnson MJ, Nuss R, Key N, et al. Lupus anticoagulant and protein S deficiency in children with postvaricella purpura fulminans or thrombosis. J Pediatr 1996; 128~319-323. 78. Nguyen P, Reynaud J, Pouzol P, Munzer M, Richard 0, Francois P. Varicella and thrombotic complications associated with transient protein C and protein S deficiencies in children. Eur J Pediatr 1994;153:646-649. Cohen H, Young L, Heyder79. Levin M, Eley BS, Louis .I, man RS. Postinfectious purpura fulminans caused by an autoantibody directed against protein S. J Pediatr 1995; 127:3.55-363. 80. D’Angelo A, Valle PD, Crippa L, Grimaldi LME, D’Angelo SV. Brief report: autoimmune protein S deficiency in a boy with severe thromboembolic disease. N Engl J Med 1993;328:1753-1757. 81. Manco-Johnson MJ, Lefkowitz J, Nuss R, Hays T. Purpura fulminans (PF) in a child with varicella, a lupus anticoagulant (LA) and severe protein S (PS) deficiency [abstract]. Blood 1992;80:510a. 82. Pashankar D, Robinson A, Tait RC. Protein S deficiency after varicella. J Pediatr 1996;129:315-316. 83. Ganesan V, Firkham FJ. Mechanisms of ischaemic stroke after chickenpox. Arch Dis Child 1997;76:522-52.5. 84. Pichier L, Hruby E, Schwartz HP. Attenuation of the Shwartzman reaction in rabbits by human activated protein C [abstract]. Thromb Haemost 1991;65:1836a.

181

85. Taylor F, Chang A, Esmon CT, et al. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest 1987;79:918-925. 86. Stenbjerg S, Pedersen EG, Lausen H. Coagulation, fibrinolytic and antithrombin I11 profiles in severely infected patients. Thromb Res 1983;31:635-640. 87. Laursen B, Faber V, Brock A, Gormsen J, Sorensen H. Disseminated intravascular coagulation, antithrombin 111, and complement in meningococcal infections. Acta Med Scand 1981;209:22 1-227. 88. Mammen EF, Miyakawa T, Philips TF, Assarian GS, Brown JM, Murano G. Human antithrombin concentrates and experimental disseminated intravascular coagulation. Semin Thromb Hemost 1985;11:373-383. 89. Emerson TE, Fournel MA, Leach WJ, Redens TB. Protection against disseminated intravascular coagulation and death by antithrombin I11 in the E. coli endotoxemic rat. Circ Shock 1987;21:1-13. 90. Gomez C, Paramo JA, Rocha E. Effect of heparin andlor antithrombin I11 on the generation of endotoxin-induced plasminogen activator inhibitor. Thromb Haemost 1989; 62:694-698. 91. Blauhut B, Necek S, Vinazzer H, Bergmann H. Substitution therapy with an antithrombin 111concentrate in shock and DIC. Thromb Res 1982;27:271-278. 92. Hanada T, Abe T, Takita H. Antithrombin I11 concentrates for treatment of disseminated intravascular coagulation in children. Am J Pediatr Hematol Oncol 1985;7: 3-8. 93. Hauptmann JG, Hassouna HI, Bell TG, Penner JA, Emerson TE. Efficacy of antithrombin 111 in endotoxin induced disseminated intravascular coagulation. Circ Shock 1988; 25:111-122. 94. Densen P, Weiler JM, Griffiss JM, Hoffmann LG. Familial properdin deficiency and fatal meningococcemia. N Engl J Med 1987;316:922-926. 95. Osterud B, Eskeland T. The mandatoly role of complement in the endotoxin-induced synthesis of tissue thromboplastin in blood monocytes. FEBS Lett 1982;149: 75-79. 96. Brandtzaeg P, Mollnes TE, Kierulf P. Complement activation and endotoxin levels in systemic meningococcal disease. J Infect Dis 1989;160:58-65. 97. Bouhasin JD. Purpura fulminans. Pediatrics 1964;34: 264-270. 98. Antley RM, McMillan CW. Sequential coagulation studies in purpura fulminans. N Engl J Med 1967;276:12871290. 99. Branson HE, Katz J. A structured approach to the management of purpura fulminans. J Natl Med Assoc 1983; 75:821-825. 100. Hjort PF, Rapaport SI. The Shwartzman reaction: pathogenetic mechanisms and clinical implications. Annu Rev Med 1965;16:135-168. 101. Cluff L, Berthrong M. The inhibition of the local Shwartzman reaction by heparin. Johns Hopkins Med J 1953;92: 353. 102. Good R, Thomas L. Studies on the generalized Shwartzman reaction. Prevention of the local and generalized Shwartzman reactions with heparin. J Exp Med 1955;102: 263. 103. Manios SG, Kanakoudi F, Maniati E. Fulminant menin-

182 Pediatric Dermatology Vol. 15 No. 3 May/June 1998 gococcemia: heparin therapy and survival rate. Scand J Infect Dis 1971;3: 127-1 33. 104. Haneberg B, Gutteberg TJ, Moe PJ, Osterud B, Bjorvatn B, Lehmann EH. Heparin for infants and children with meningococcal septicemia: results of a randomized therapeutic trial. NIPH Ann 1983;6:43-47. 105. Gerard P, Moriau M, Bachy A, Malvaux P, De Meyer R. Meningococcal purpura: report of 19 patients treated with heparin. J Pediatr 1973;82:780-786. I 06. Hattersley PG. Purpura fulminans: complete recovery with intravenously administered heparin. Am J Dis Child 1970;120:467-471. 107. Brandtzaeg P, Joo GB, Bmsletto B, Kierulf P. Plasminogen activator inhibitor 1 and 2, alpha-2-antiplasmin, plasminogen, and endotoxin levels in systemic meningococcal disease. Thromb Res 1990;57:271-278. 108. Zenz W, Muntean W, Gallistl S, Zobel G, Grubbauer HM. Recombinant tissue plasminogen activator treatment in two infants with fulminant meningococcemia. Pediatrics 1995;96:144-148. 109. Zenz W, Muntean W, Gallistl S, Zobel G, Grubbauer HM, Gallistl S. Treatment of fulminant meningococcemia with recombinant tissue plasminogen activator. Thromb Haemost 1995;74:802-803. 110. Keeley SR, Matthews NT, Buist M. Tissue plasminogen activator for gangrene in fulminant meningococcemia. Lancet 1991;337: 1359. 1 1 1 . The GUSTO Investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N Engl J Med 1993;329:673-682. 112. Zenz W, Muntean W. Fibronolytic therapy in children. Fibrinolysis 1995;9(suppl 1):33-36. 113. Winrow AP. Successful treatment of neonatal purpura fulminans with epoprostenol. J R SOCMed 1992;85:245. 114. Stewart FJ, McClure BG, Mayne E. Successful treatment of neonatal purpura fulminans with epoprostenol. J R Soc Med 1991;84:623-624. 115. Denning DW, Gilliland L, Hewlett A, Hughes LO, Reid CD. Peripheral symmetrical gangrene successfully treated with epoprostenol and tissue plasminogen activator. Lancet 1986;i:1401-1402. 116. Tobias JD, Haun SE, Helfaer M, Nichols DG. Use of continuous caudal block to relieve lower-extremity ischemia caused by vasculitis in a child with meningococcemia. J Pediatr 1989;115:1019-1021. 117. Anderson CTM, Berde CB, Sethna NF, Pribaz JJ. Meningococcal purpura fulminans: treatment of vascular insufficiency in a 2-yr-old child with lumbar epidural sympathetic blockade. Anesthesiology 1989;71:463-464. I1 8. Chiafery MC, Stephany RA, Holliday KJ. Epidural sympathetic blockade to relieve vascular insufficiency in an infant with purpura fulminans. Crit Care Nurs 1993;6: 71-76. 119. Irazuzta J, McManus ML. Use of topically applied nitroglycerin in the treatment of purpura fulminans. J Pediatr 1990;117:993-995. 120. Brandtzaeg P, Sirnes K, Folsland B, et al. Plasmapheresis in the treatment of severe meningococcal or pneumococcal septicaemia with DIC and fibrinolysis. Preliminary data on eight patients. Scand J Clin Lab Invest 1985;45: 53-55. 121. Scharfman WB, Tillotson JR, Taft EG, Wright E. Plasmapheresis for meningococcemia with disseminated in-

travascular coagulation. N Engl J Med 1979;300:12771278. 122. Bjorvatn B, Bjertnaes L, Fadnes HO, et al. Meningococcal septicaemia treated with combined plasmapheresis and leucapheresis or with blood exchange. Br Med J (Clin Res Ed) 1984;288:439441. 123. Togari H, Mikawa M, Iwanaga T, et al. Endotoxin clearance by exchange blood transfusion in septic shock neonates. Acta Paediatr Scand 1983;72:87-91. 124. Karz SC, Thompson RE, Depenbusch F. Hyperbaric oxygen treatment of gangrene and pneumococcemia. JAMA 1971;271:962-963. 125. Dollberg S, Nachum Z, Klar A, et a1. Haernophilus in$uenzue type b purpura fulminans treated with hyperbaric oxygen. J Infect 1992;25:197-200. 126. Dudgeon DL, Kellogg DR, Gilchrist GS, Woolley MM. Purpura fulminans. Arch Surg 1971;103:351-358. 127. de Chalain T, Cohen SR, Burstein FD. Successful use of leeches in the treatment of purpura fulminans. Ann Plast Surg 1995;35:300-304. 128. Markwardt F. Past, present, and future of hirudin. Haemostasis 1991;21(suppl):11-26. 129. Zacharski LR. Successful use of leeches in the treatment of purpura fulminans. Invited discussion. Ann Plast Surg 1995;35:304-306. 130. Wortel CH, von der Mohlen MAM, van Deventer SJH, et al. Effectiveness of a human monoclonal anti-endotoxin antibody (HA-I A) in Gram-negative sepsis: relationship to endotoxin and cytokine levels. J Infect Dis 1992;166: 1367-1 374. 131. Fang KC. Monoclonal antibodies to endotoxin in the management of sepsis. West J Med 1993;158:393-399. 132. Bone RC, Balk RA, Fein AM, et al. A second large controlled clinical study of E5, a monoclonal antibody to endotoxin: results of a prospective, multicenter, randomized, controlled trial. Crit Care Med 1995;23:994-1006. 133. Ziegler EJ, Fischer CJ, Sprung C, et al. Treatment of Gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin. A randomized, double-blind, placebo-controlled trial. N Engl J Med 1991;324:429-436. 134. Greenman RI, Schein RM, Martin MA, et al. A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of Gram-negative sepsis. JAMA 1991;266:1097-1 102. 135. McCloskey Rv, Straube RC, Sanders C, et al. Treatment of septic shock with human monoclonal antibody HA-1A: a randomized double-blind, placebo-controlled trial. Ann Intern Med 1994;121:1-5. 136. Fisher CJ, Dhainaut JFA, Opal SM, et al. Recombinant human interleukin-1 receptor antagonist in the treatment of patients with sepsis syndrome. JAMA 1994;271:18361843. 137. Fisher CJJ Jr, Agosti JM, Opal SM, et al. Treatment of septic shock with the tumor necrosis factor receptor:fc fusion protein. N Engl J Ned 1996;334:1697-1702. 138. Abraham E, Wunderink R, Silverman H, et al. Efficacy and safety of monoclonal antibody to human tumor necrosis factor-a in patients with sepsis syndrome. JAMA 1995;273:934-941.

Darmstadt: Acute Infectious Purpura Fulminans

139. Dhainaut JF, Tenaillon A, Letulzo Y, et al. Platelet activating factor receptor antagonist BN52021 in the treatment of severe sepsis: a randomized, double-blind, placebo-controlled, multicenter clinical trial. Crit Care Med 1994;22:1720-1728. 140. Dhainaut JF, Tenaillon A, Hemmer M, et al. Confirming phase I11 clinical trial to study the efficacy of a PAF antagonist, BN 52021, in reducing mortality of patients with severe Gram-negative sepsis [abstract]. Am J Respir Crit Care Med 1995;151:A447. 141. Zeni F, Pain P, Vindimian M, et al. Effects of pentoxifylline on circulating cytokine concentrations and hemodynamics in patients with septic shock: results from a double-blind, randomized, placebo-controlled study. Crit Care Med 1996;24:207-214. 142. Davis CE, Ziegler EJ, Arnold KF. Neutralization of meningococcal endotoxin by antibody to core glycolipid. J Exp Med 1978;147:1007-1017. 143. J5 Study Group. Treatment of severe infectious purpura in children with human plasma from donors immunized

183

with Escherichia coli J5: a prospective double-blind study. J Infect Dis 1992;165:695-701. 144. Bone RC. Why sepsis trials fail. JAMA 1996;276:565566. 145. Abraham E, Raffin TA. Sepsis therapy trials. Continued disappointment or reason for hope? JAMA 1994;271: 1876-1878. 146. Bone RC, Grodzin CJ, Balk RA. Sepsis: a new hypothesis for pathogenesis of the disease process. Chest 1997;112: 235-243. 147. Bone RC. Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 1996;125: 680-687. 148. Bone RC. The pathogenesis of sepsis. Ann Intern Med I99 I ;115:4577469. 149. Kox WJ, Bone RC, Krausch D, et al. Interferon gamma1p in the treatment of compensatory anti-inflammatory response syndrome. Arch Intern Med 1997;157:389-393.