- Email: [email protected]
Purpura Fulminans in Sepsis
Purpura Fulminans in Sepsis ALEX P. BETROSIAN, MD; TOM BERLET, DEAA, EDIC; BANWARI AGARWAL, MBBS, MD, MRCP, FRCA, EDIC
ABSTRACT: Sepsis-induced purpura fulminans is a rare but life-threatening disorder, characterized by hemorrhagic infarction of the skin caused by disseminated intravascular coagulation and dermal vascular thrombosis. The pathogenesis is linked to enhanced expression of the natural procoagulants and depletion of the natural anticoagulant proteins particularly protein C. Meningococcal sepsis is the most common cause, followed by pneumococcal sepsis in adults. The syndrome is associated with more than 50% mortality secondary to multiple organ dysfunction syndrome and is accompanied by long-term morbidity. Necrotic lesions usually progress
P
urpura fulminans is an uncommon disorder first described by Guelliot in 1884. Initially noted in children in association with bacterial or viral infection, it was later seen in neonates with a family history of homozygous protein C or protein S deficiency without infection.1 Purpura fulminans secondary to sepsis is a lifethreatening condition characterized by hemorrhagic infarction of the skin caused by disseminated intravascular coagulation and dermal vascular thrombosis.1,2 It is associated with high mortality rate (⬎50%) owing to multiple organ dysfunction and is also accompanied by considerable long-term morbidity. Meningococcal sepsis is the most common form, followed by pneumococcal sepsis in adults. Necrotic lesions usually progress to distal ischemia, often necessitating skin grafting or limb amputation. Early antibiotic administration and intensive care management according to the recommendations of severe sepsis and shock is crucial for patients’ survival. Adjuvant therapies against inflammatory and coagulation cascades and augmenting fibrinolysis are controversial and need further assessment.
From the Intensive Therapy Unit, Royal Free Hospital, London, United Kingdom. Submitted May 23, 2006; accepted in revised form July 19, 2006. Correspondence: Alex Betrosian, MD, Hippokration General Hospital, Intensive Care Unit, Vas. Sofias 114, 11527 Athens, Greece (E-mail: abetrosian@ gmail.com). THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
to distal ischemia, and skin grafting and extremities or limb amputation are often required. Early antibiotic administration and intensive care management according to the recommendations of severe sepsis and shock is crucial for patients’ survival. Adjuvant therapies against inflammatory and coagulation cascades and augmenting fibrinolysis are still controversial and need further assessment. Among them activated protein C and supplementation therapy have given promising results. KEY INDEXING TERMS: Purpura fulminans; Disseminated intravascular coagulation; Sepsis. [Am J Med Sci 2006; 332(6):339–345.]
Etiology Purpura fulminans can be identified in three clinical conditions: 1) in patients with preexistent inherited or acquired abnormality of the protein C or protein S anticoagulant pathway; 2) in patients with acute severe infection, predominantly gram-negative bacterial infections, called acute infectious purpura fulminans; and 3) in patients without known abnormalities of the protein C pathway or acute infections, termed idiopathic purpura fulminans. The last subtype occurs primarily in children, usually after benign infections.3 A multitude of infections can cause this syndrome, Neisseria meningitidis being the foremost among the microorganisms that have been implicated in the development of purpura fulminans, followed by streptococcal infections.6 Pneumococcal purpura fulminans (Streptococcus pneumoniae) is reported with increased frequency (⬎60%) in the setting of asplenia or functional hyposplenia.4 Other encapsulated bacteria such as Haemophilus influenzae and group A and B streptococci have been implicated in infants and adults. Staphylococcus aureus as a cause of purpura fulminans is considered rare, probably because these cases are reported as toxic shock syndrome, but may be as common as other nonmeningococcal organisms. Other organisms associated with purpura fulminans are listed in the Table 1.5–13 Time Course of Cutaneous Manifestations of Skin Necrosis The cutaneous lesions of sepsis-induced purpura fulminans are similar regardless of the causative microorganism. Clinically cutaneous necrosis begins 339
Purpura Fulminans in Sepsis
Table 1. Infective Causes of Purpura Fulminans Bacterial: Neisseria meningitides Streptococcus pneumoniae Group A and B streptococci Enterococcus faecalis Haemophilus influenza, Haemophilus aegyptius Staphylococcus aureus Enterobacteriacae: Klebsiella pneumoniae Escherichia coli Proteus spp (mirabilis, vulgaris) Pseudomonas spp Acinetobacter calcoanitratus Xanthomonas maltophilia Vibrio spp Salmonella paratyphi Mycobacterium tuberculosis Capnocytophaga canimorsus Rickettsia rickettsii Viral: Varicella zoster Protozoal: Plasmodium falciparum
with a region of dermal discomfort that quickly progresses within hours to petechiae, which coalesce to form purple ecchymoses.14 At this stage, the cutaneous lesions can be completely reversible without progressing to skin necrosis. The hallmark of hemorrhagic infarction is the development of hemorrhagic bullae, which can progress to frank skin necrosis and gangrene (Figure 1). The stage of necrosis is associated with considerable morbidity and mortality (⬎50%). The pathologic feature of skin necrosis in purpura fulminans is characterized by occlusion of dermal venules and capillaries by microthrombi (Figure 2), resulting in hemorrhagic infarction.1,3 Soon after the vascular occlusion, the endothelial cells become edematous, resulting in capillary dilation with congestion of the vessels by erythrocytes. Clinically, this stage is manifested as erythema and some edema at the site of skin injury.14 Progression of ischemia and capillary
Figure 2. Fibrin thrombi occlusion in the vessels of the dermis (arrows). (H&E stain; original magnification ⫻40.)
dilatation results in extravasation of blood elements into dermis. Clinically this stage is characterized by petechiae. A minimal to mild inflammatory neutrophil infiltrate may be found in the perivascular region. As ischemia and vascular damage progresses, there is further hemorrhage into dermis and ecchymoses develop. Ecchymoses formation is followed by the development of hemorrhagic bullae in the subepidermal area.15 Affected areas are painful and indurated. The end result is coagulative necrosis of the dermal and subcutaneous tissues with extensive dermal hemorrhage manifested as gangrenous necrosis. Although initially sterile, secondary infection of gangrenous tissue may occur with time, contributing to late morbidity and mortality. Occasionally gram-stained microorganisms are present within the microthrombi, the vessel walls, and the perivascular region. In cases of sepsis-induced purpura fulminans, these vascular changes are not limited to the skin but are widespread, involving multiple organs such as lungs, kidneys, and adrenal glands (WaterhouseFriderichsen syndrome when associated with adrenal hemorrhage), leading to multiple systems organ failure.16 Pathogenesis
Figure 1. Cutaneous lesions in purpura fulminans with gangrene of the tip of the little finger.
340
The pathogenesis of petechiae in sepsis-induced purpura fulminans is strongly suggestive of the localized Shwartzman phenomenon. Shwartzman reaction is initiated by a local, priming intradermal injection of endotoxin, which elicits a transient perivascular inflammatory reaction and increased vascular permeability. Intravenous challenge with the same endotoxin 12 to 24 hours later results in vascular thrombosis and necrotizing vasculitis.17 Sepsis-induced purpura fulminans and the Shwartzman reaction share a pathogenesis involving a disturbance in the balance of anticoagulant and procoagulant activities of endothelial cells. This disturbance, which is triggered by endotoxin in gramDecember 2006 Volume 332 Number 6
Betrosian et al
negative and exotoxin in gram-positive sepsis, appears to be mediated by cytokines, particularly tumor necrosis factor-alpha (TNF-␣), and interleukin-1 (IL-1), leading to the consumption of the natural anticoagulants, proteins C and S and antithrombin III.3,17 Virtually all patients with sepsis have coagulation abnormalities that may range from a mild drop in platelet count and subclinical prolongation of the clotting parameters without clinical signs of bleeding or thrombosis (nonovert disseminated intravascular coagulation [DIC]) to fulminant DIC, characterized by widespread microvascular thrombosis and bleeding from various sites.18 DIC has been shown to be present in more than 90% of these cases, supporting the notion that purpura fulminans is a cutaneous marker of DIC.1,4,5 The development of DIC is mediated by several simultaneously occurring mechanisms initiated by the proinflammatory cytokines TNF-␣, IL-1, and IL-6. The mechanisms include ongoing generation of thrombin, the inhibition of the natural anticoagulant mechanisms, and the impairment of fibrinolysis caused by the plasminogen activator inhibitor-1.18,19 Thrombin Generation In sepsis, the activation of the coagulation cascade originates from the extrinsic pathway with the expression of tissue factor (TF) from cytokine-activated mononuclear cells, which have a pivotal role in the host response to infection.20 Data from animal models and from healthy volunteers have shown that even minimal amounts of circulating endotoxin or TNF-␣ can cause activation of coagulation cascade within few minutes.21 TF activates factor VII and the TF-VII complex cleaves factor X to its active form, Xa. Factor Xa, in conjunction with cofactor Va, cleaves prothrombin to form thrombin, which at the same time amplifies coagulation by activating factors VIII and XI of the intrinsic pathway. Thrombin converts fibrinogen to fibrin and is a potent platelet activator. Activated platelets can accelerate coagulation activation through several pathways.18 –21 Inhibition of the Natural Anticoagulant Mechanisms Thrombin generation, however, is limited by the natural anticoagulant mechanisms involving antithrombin III, protein C/activated protein C system, and tissue factor pathway inhibitor (TFPI). But in cases of sepsis-induced DIC, these regulatory systems are defective and are degraded as a result of endothelial dysfunction and release of acute-phase proteins, respectively.22 Antithrombin III is the major inhibitor of thrombin and factors Xa, IXa, VIIa, and XIa. The anticoagulant action of antithrombin III is enhanced by glycosaminoglycans present on endothelial membranes. During severe infection, antithrombin III THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
levels are very low due to consumption, impaired synthesis, and degradation by elastase from activated neutrophils. Also, a reduction in glycosaminoglycans activity due to endothelial dysfunction further contributes to reduced antithrombin III function in sepsis.23,24 The protein C/activated protein C system pathway is also significantly impaired during sepsis. Protein C is activated when thrombin binds to thrombomodulin on vascular endothelium surfaces. The activated protein C interacts with protein S, which then catalyzes the inactivation of factors Va and VIIa, thus stopping thrombin formation. Down-regulation of thrombmodulin on endothelial cells by proinflammatory cytokines, enhanced consumption, impaired synthesis, and vascular leakage are the main causes of inadequacy of protein C in sepsis.23,25 Tissue factor pathway inhibitor is an endogenous serine protease inhibitor, synthesized and secreted by the endothelial cells, which inhibits factor Xa directly and the factor VIIa/TF catalytic complex in a factor Xa– dependent fashion.26 It contains three Kunitz-type domains. The first domain binds factor VIIa, the second binds factor Xa, and the third domain has a binding site for heparin. Free TFPI binds factor VIIa very slowly in comparison with its binding of factor Xa, whereas factor Xa–TFPI complex is a potent inhibitor of factor VIIa–TF. Their interaction results in the formation of a quaternary Xa-TFPI-VIIa-TF inhibitory complex.23,26 Both experimental and clinical evidence, however, indicates that functionally active TFPI levels are inadequate within the microcirculation to prevent ongoing coagulation and organ dysfunction in sepsis and DIC.26 Inhibition of Fibrinolysis The fibrinolytic system is activated in sepsis and has a major role in the control of fibrin deposition in the microcirculation.27 Major fibrinolytic activators and inhibitors are synthesized and stored in endothelial cells. There is strong evidence of increase in tissue type plasminogen activator, followed by an increase in plasminogen activator inhibitor (PAI-1), which inhibits plasmin generation and contributes to further fibrin deposition in the microcirculation and subsequent organ failure.28 The proinflammatory cytokines TNF-␣ and IL-1 increase PAI-1 synthesis or its release from endothelial cells and decrease the plasminogen activator.18,28 Inhibition of fibrinolysis in cases of fulminant DIC associated with sepsis-induced purpura fulminans is exaggerated by thrombin-activatable fibrinolysis inhibitor, a zymogen that is activated by the thrombin-thrombomodulin complex28,29. Thromin-activatable fibrinolysis inhibitor down-regulates fibrinolysis by removing the carboxy-terminal lysine and arginine residues from fibrin and contributes to the ongoing inhibition of fibrinolysis in the later stages of sepsis.29 341
Purpura Fulminans in Sepsis
Clinical Features
Management
Along with purpuric rash, the clinical features of sepsis-induced purpura fulminans are mainly those of the underlying sepsis and its complications. It is usually preceded by a flu-like illness, with fever or chills, sore throat, malaise, and occasionally gastroenteric symptoms occurring 12 to 24 hours before the development of the spreading petechial rash. The majority of affected patients develop septic shock and DIC. In one study among 70 patients with purpura fulminans of various origins, 67% appeared with septic shock and 78% were associated with DIC.30 Nearly all patients experienced deterioration of their condition and required aggressive fluid resuscitation and inotropic and mechanical ventilatory support. Of interest is the absence of shock in 21 of 43 cases (51%) of pneumococcal sepsis—induced purpura fulminans, suggesting a difference in the rate and severity of activation of the cytokine cascade between gramnegative and gram-positive strains (endotoxin versus exotoxins). However, DIC was present in the majority of the reported cases.4
Purpura fulminans is a syndrome, not a specific disease; therefore, treatment must be aimed toward the underlying infection along with on-going supportive management. Successful management is facilitated by high index of suspicion, early diagnosis, aggressive antibiotic treatment, systemic organ support, and prompt surgical consultation.29 Early administration of antibiotics as soon as the diagnosis of purpura fulminans has been evoked is a significant therapeutic advantage. Since the majority of the reported cases are due to Neisseria meningitidis, empiric antibiotic treatment should consist of a third-generation cephalosporin, which also provides good cover against streptococcus.6,29 Vancomycin or teicoplanin should be added if MRSA is suspected. The above regimen should be altered once the causative microorganism is identified and sensitivities obtained. However, despite early antibiotic intervention and intensive care treatment, the mortality rate remains high and is reported to average 40% (range, 20 –70%).32
Differential Diagnosis
All patients with sepsis-induced purpura fulminans should be treated in an intensive care setting for optimal organ support management. Hemodynamic monitoring for continuous measurement of blood pressure and arterial blood gases are necessary. Aggressive fluid resuscitation to restore intravascular volume, inotropic and ventilatory support, and continuous renal replacement therapy when required are major therapeutic interventions.
In patients presenting with purpura fulminans and hypotension, the diagnosis of infectious purpura fulminans is obvious. In more complex cases, however, the differential diagnoses should include other causes of purpura such as Henoch-Scho¨nlein purpura, postinfectious thrombocytopenic purpura, thrombotic thrombocytopenic purpura, and idiopathic purpura fulminans. The skin lesions of Henoch-Scho¨nlein purpura are typically small and rarely lead to extensive necrosis.1 Thrombotic thrombocytopenic purpura is associated with hemolysis, thrombocytopenia, and hyaline thrombosis of small vessels but never cause massive skin necrosis. However, thrombotic thrombocytopenic pupura occasionally is associated with DIC.30 Sepsis-induced purpura fulminans differs from acute infectious or idiopathic purpura fulminans in that it is associated with an overwhelming acute infection and shock. The idiopathic variety is commonly associated with normal blood pressure and a well-preserved peripheral pulse.1 The skin involvement also differs in these two conditions. In cases of sepsis-induced purpura fulminans, skin necrosis often begins in the distal extremities affecting feet, toes, tip of the nose, earlobes, hands, and fingertips, while progressing gradually to the entire body in a patchy distribution. In contrast, idiopathic purpura fulminans skin necrosis usually involves the skin of the thighs, the buttocks, and the lower trunk and less commonly the extremities.1,31 342
Supportive Treatment
Surgical Treatment Purpura fulminans almost invariably leads to some full-thickness skin loss. The similarity between skin necrosis secondary to DIC and full-thickness cutaneous burns forms the rationale for treating them as if they were full-thickness burns.33 Surgical treatment includes aggressive debridement of necrotic tissues to decrease the risk of secondary contamination and sepsis. Areas of dry gangrene of the digits should be left to demarcate, since early surgery might be unnecessarily extensive.34 Monitoring of compartment pressure and early fasciotomy before the development of demarcation may improve the prognosis of limb preservation. The majority of cases with skin necrosis would need skin grafting.35 Often tissue necrosis is so severe that limp amputations are performed. In a retrospective analysis of patients with purpura fulminans in a burn center, 87% of them required skin grafting and 90% limb amputations. In one-fourth of them, amputation was performed in all extremities.29 Streptococcal-induced purpura fulminans resulted in December 2006 Volume 332 Number 6
Betrosian et al
more extensive amputations when compared with other infections with a mean of three amputations per person. Adjuvant Therapy Many unconventional treatment strategies have been used for the treatment of sepsis-induced purpura fulminans, meningococcal in particular.36 However, these could be applied to any case of sepsis-induced purpura fulminans irrespective of the causative microorganism. The therapies have focused on approaches that could modulate the inflammatory and the coagulation cascade. Therapies Against Inflammatory Cascade Corticosteroids inhibit many components of inflammatory cascade, including complement activation, cytokine production, and neutrophil function and migration. The use of steroids in treating sepsis and purpura fulminans remains controversial. Although there is a survival benefit with low-dose, short-term hydrocortisone in septic patients with relative adrenal insufficiency, evidence does not conclusively support routine use of corticosteroids in sepsis-induced purpura fulminans of any cause. Other anti-inflammatory treatment strategies that had been used for meningococcal sepsis include the use of recombinant bactericidal permeability-increasing protein and human monoclonal antibodybinding endotoxin, which have failed to show a survival benefit for this disease.37–39 Several case reports and a few small randomized trials used plasmapheresis and hemofiltration as methods to remove proinflammatory cytokines in meningococcal purpura fulminans. However, no effect on survival has been reported. These techniques in theory could remove cytokines from circulation, but whether they can remove cytokines from tissues remains questionable.39 Plasma exchange has been used in pediatric patients with sepsis-induced purpura fulminans and was reported to improve prognosis by virtue of a significant increase in the level of coagulation factors II, V, VII, and VIII.40 Therapies Against Coagulation Cascade Since microvascular thrombosis appears to be the predominant pathophysiologic feature of purpura fulminans, the majority of therapeutic approaches have focused on reduction of coagulation activation and improvement of fibrinolysis.36,37 Heparin The use of heparin in cases of sepsis-induced purpura fulminans may appear counterintuitive given the hemorrhagic nature of the disease. However, there is some evidence that heparin could be of THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
benefit in reducing the severity of distal necrosis by inhibiting thrombus formation and consumption of coagulation factors in sepsis-induced purpura fulminans. Arbitrary schemes of unfractionated or low molecular weight heparin have been used, but until now, no effect on survival has been demonstrated in animal models or small clinical trials.41 Antithrombin III Sepsis-induced purpura fulminans is often associated with markedly reduced levels of antithrombin III due to consumptive coagulopathy. Antithrombin III is not only an important physiologic inhibitor of coagulation but also possesses anti-inflammatory properties through release of prostaglandin I2 from endothelial cells.24,25 Several authors have shown that antithrombin III supplementation therapy in patients with severe sepsis and shock is beneficial in restoring organ function and reversing DIC. In adults with purpura fulminans, a loading dose of 100 IU/kg followed by 100 IU/kg/day was able to correct the levels of antithrombin activity, which was subsequently maintained by continuous infusion.42 In contrast to the initial positive results, a large-scale multicenter randomized controlled trial investigating the use of antithrombin III in more than 2300 patients with severe sepsis and shock failed to demonstrate any significant reduction in mortality.43 However, the trial did not investigate DIC, but rather severe sepsis in general. Protein C/Activated Protein C Dysfunction of the protein C pathway during sepsis is one of the major factors contributing to the development of purpura fulminans. Severe depletion of protein C leads to uncontrolled hemorrhage and thrombosis and is associated with poor outcome.22 Treatment with protein C concentrate or recombinant activated protein C (drotrecogin alfa, activated) has been shown to be beneficial in cases of sepsis-induced purpura fulminans.25,29,36,44 Several case reports and open-label studies on protein C (unactivated) substitution therapy have shown encouraging results.45,46 Whether this therapy is effective in achieving adequate levels of activated protein C remains questionable. However, a phase II dose-finding study in children with meningococcal sepsis and purpura fulminans showed that a dose of 150 IU/kg or more is required to increase the levels of activated protein C and a cumulative dose of 600 IU/kg/day is necessary for sustained protein C activation.47 Drotrecogin alpha activated has been approved for the treatment of severe sepsis and organ dysfunction (APACHE II ⬎24). Patients with sepsis-induced purpura fulminans fulfill the criteria for adjuvant therapy with this drug at the dose of 24 g/kg per hour for 96 hours.48 A few case 343
Purpura Fulminans in Sepsis
reports with encouraging results have been published on the use of this drug in patients with purpura fulminans.7,49 However, no randomized trial is yet available.50 Therapies Augmenting Fibrinolysis Recombinant Tissue Plasminogen Activator Administration of a fibrinolytic agent such as recombinant tissue plasminogen activator (t-PA) is an alternative for limiting ongoing thrombin generation. Concentrations of PAI-1 are increased in purpura fulminans and correlate with risk for mortality. Although a few case reports have shown promising results, a retrospective analysis on the use of recombinant t-PA in children with meningococcal purpura fulminans showed a high incidence of intracerebral hemorrhage.51,52 Since there has not been a randomized trial, no definite answers on the use of recombinant t-PA in sepsis-induced purpura fulminans could be drawn. Other Strategies Fresh frozen plasma infusions containing anticoagulant factors have been suggested as an adjuvant therapy in purpura fulminans, providing both fluid resuscitation and immunomodulation. Cryoprecipitate, fibrinogen concentrate, or recombinant factor VIIa can be used in patients with severe bleeding not responding in other treatment options.36,37,53 Plasma exchange using fresh frozen plasma has been used successfully in patients with sepsis-induced purpura fulminans by restoring low levels of protein C and antithrombin III.54 A few anecdotal reports have suggested improved distal perfusion and pain relief with the use of vasodilating agents such as prostacyclin or epoprostenol, sodium nitroprusside, or topical nitroglycerin.53 Hyperbaric oxygenation has been used, and one study demonstrated improvement of limb perfusion in 12 of 14 patients with sepsis-induced purpura fulminans caused by various pathogens.37 Because this treatment needs to be instituted early in the course of the disease, its usefulness is limited for logistical reasons. Extracorporeal membrane oxygenation has been used to support patients with intractable meningococcal sepsis and purpura fulminans. However, there is no exact answer to whether this technique offers any advantage over conventional therapy.53,55 The efficacy and benefits of the above-mentioned treatment strategies are difficult to interpret, since they are often used in combination. Despite the positive preclinical data and promising case reports, no adjuvant therapy has been shown to reduce the rates of morbidity and mortality associated with sepsis-induced purpura fulminans.36,37 Although treatment with recombinant human activated protein C has been approved for the treatment of severe 344
sepsis, its use in patients with purpura fulminans needs further validation. Until then, the management of sepsis-induced purpura fulminans must stand upon early recognition, prompt antibiotic treatment, and intensive care management, according to the recommendations for severe sepsis and shock.56,57 References 1. Marlar RA, Neumann A. Neonatal purpura fulminans due to homozygous protein C or protein S deficiencies. Semin Thromb Hemost 1990;16:299–309. 2. Francis RB Jr. Acquired purpura fulminans. Semin Thromb Hemost 1990;16:310–25. 3. Adcock DM, Brozna J, Marlar RA. Proposed classification and pathologic mechanisms of purpura fulminans and skin necrosis. Semin Thromb Hemost 1990;16:333–40. 4. Pathan N, Faust SN, Levin M. Pathophysiology of meningococcal meningitis and septicaemia. Arch Dis Child 2003; 88:601–7. 5. Molos MA, Hall JC. Symmetrical peripheral gangrene and disseminated intravascular coagulation. Arch Dermatol 1985;121:1057–61. 6. Carpenter CT, Kaiser AB. Purpura fulminans in pneumococcal sepsis: case report and review. Scand J Infect Dis 1997;29:479–83. 7. Kravitz GR, Dries DJ, Peterson ML, et al. Purpura fulminans due to Staphylococcus aureus. Clin Infect Dis 2005;40:941–7. 8. Childers BJ, Cobanov B. Acute infectious purpura fulminans: a 15-year retrospective review of 28 consecutive cases. Am Surg 2003;69:86–90. 9. Olowu WA. Klebsiella-induced purpura fulminans in a Nigerian child: case report and a review of literature. West Afr J Med 2002;21:252–5. 10. Huemer GM, Bonatti H, Dunst KM. Purpura fulminans due to E. coli septicemia. Wien Klin Wochenschr 2004;116:82. 11. Gurses N, Ozkan A. Neonatal and childhood purpura fulminans: review of seven cases. Cutis 1988;41:361–3. 12. Deshmukh PM, Camp CJ, Rose FB, et al. Capnocytophaga canimorsus sepsis with purpura fulminans and symmetrical gangrene following a dog bite in a shelter employee. Am J Med Sci 2004;327:369–72. 13. Keri JE, Thomas K, Berman B, et al. Purpura fulminans in a patient with malaria. Eur J Dermatol 2000;10:617–9. 14. Adcock DM, Hicks MJ. Dermatopathology of skin necrosis associated with purpura fulminans. Semin Thromb Hemost 1990;16:283–92. 15. Cole MS, Minifee PK, Wolma FJ. Coumarin necrosis: a review of the literature. Surgery 1988;103:271–7. 16. Varon J, Chen K, Sternbach GL. Rupert Waterhouse and Carl Friderichsen: adrenal apoplexy. J Emerg Med 1998;16: 643–7. 17. Brozna JP. Shwartzman reaction. Semin Thromb Hemost 1990;16:326–32. 18. Dempfle CE. Coagulopathy of sepsis. Thromb Haemost 2004;91:213–24. 19. Levi M, de Jonge E, van der Poll T. Sepsis and disseminated intravascular coagulation. J Thombosis Thrombolysis 2003;16:43–7. 20. Levi M. Current understanding of disseminated intravascular coagulation. Br J Haematol 2004;124:567–76. 21. Burrell R. Human responses to bacterial endotoxin. Circ Shock 1994;43:137–53.
December 2006 Volume 332 Number 6
Betrosian et al
22. Jagneaux T, Taylor DE, Kantrow SP. Coagulation in sepsis. Am J Med Sci 2004;328:196–204. 23. Okajima K. Regulation of inflammatory responses by natural anticoagulants. Immunol Rev 2001;184:258–74. 24. Roemisch J, Gray E, Hoffmann JN, et al. Antithrombin: a new look at the actions of a serine protease inhibitor. Blood Coagul Fibrinolysis 2002;13:657–70. 25. Aird WC. Natural anticoagulant inhibitors: activated Protein C. Best Pract Res Clin Haematol 2004;17:161–82. 26. Doshi SN, Marmur JD. Evolving role of tissue factor and its pathway inhibitor. Crit Care Med 2002;30(Suppl):241–50. 27. Mavrommatis AC, Theodoridis T, Economou M, et al. Activation of the fibrinolytic system and utilization of the coagulation inhibitors in sepsis: comparison with severe sepsis and septic shock. Intensive Care Med 2001;27:1853–9. 28. Hack CE. Derangements of coagulation and fibrinolysis in infectious diseases. Contrib Microbiol 2003;10:18–37. 29. Zeerleder S, Schroeder V, Hack E, et al. TAFI and PAI-1 levels in human sepsis. Thrombosis Res 2006;118:205–12. 30. Warner PM, Kagan RJ, Yakuboff KP, et al. Current management of purpura fulminans: a multicenter study. J Burn Care Rehabil 2003;24:119–26. 31. Alexander G, Basheer HM, Ebrahim MK, et al. Idiopathic purpura fulminans and varicella gangrenosa of both hands, toes and integument in a child. Br J Plast Surg 2003;56:194–5. 32. Rosenstein NE, Perkins BA, Stephens DS, et al. Meningococcal disease. N Engl J Med 2001;344:1378–88. 33. Dinh TA, Friedman J, Higuera S. Plastic surgery management in pediatric meningococcal-induced purpura fulminans. Clin Plast Surg 2005;32:117–21. 34. Potokar TS, Oliver DW, Russel RR, et al. Meningococcal septicaemia and plastic surgery: a strategy for management. Br J Plast Surg 2000;53:142–8. 35. Edlich RF, Winters KL, Woodard CR, et al. Massive soft tissue infections: necrotizing fasciitis and purpura fulminans. J Long Term Eff Med Implants 2005;15:57–65. 36. Fischer M, Hilinski J, Stephens DS. Adjuvant therapy for meningococcal sepsis. Pediatr Infect Dis J 2005;24:177–8. 37. Leclerc F, Leteurtre S, Cremer R, et al. Do new strategies in meningococcemia produce better outcomes? Crit Care Med 2000;28(Suppl):600–33. 38. Levin M, Quint PA, Goldstein B, et al. Recombinant bactericidal/permeability-increasing protein (rBPI21) as adjunctive treatment for children with severe meningococcal sepsis: a randomised trial. rBPI21 Meningococcal Sepsis Study Group. Lancet 2000;356:961–7. 39. van Deuren M, Brandtzaeg P, van der Meer JW. Update on meningococcal disease with emphasis on pathogenesis and clinical management. Clin Microbiol Rev 2000;13:144–66. 40. Churchwell KB, McManus ML, Kent P, et al. Intensive blood and plasma exchange for treatment of coagulopathy in meningococcemia. J Clin Apheresis 1995;10:171–7. 41. Kupperman N, Inkelis S, Saladino R. The role of heparin in the prevention of extremity and digit necrosis in meningococcal purpura fulminans. Pediatr Infect Dis J 1994;13:867–87. 42. Cobcroft R, Henderson A, Solano C, et al. Meningococcal purpura fulminans treated with antithrombin III concen-
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53. 54.
55.
56.
57.
trate: what is the optimal replacement therapy? Aust N Z J Med 1994;24:575–6. Warren BL, Eid A, Singer P, et al. KyberSept Trial Study Group. Caring for the critically ill patient: high-dose antithrombin III in severe sepsis: a randomized controlled trial. JAMA 2001;286:1869–78. Dhainaut JF, Yan SB, Claessens YE. Protein C/activated protein C pathway: overview of clinical trial results in severe sepsis. Crit Care Med 2004;32(Suppl):194–201. Rintala E, Kauppila M, Seppala OP, et al. Protein C substitution in sepsis-associated purpura fulminans. Crit Care Med 2000;28:2373–8. Betrosian AP, Balla M, Kofinas G, et al. Protein C in the treatment of coagulopathy in meningococcal sepsis. Crit Care Med 1999;27:2849–50. de Kleijn ED, de Groot R, Hack CE, et al. Activation of protein C following infusion of protein C concentrate in children with severe meningococcal sepsis and purpura fulminans: a randomized, double-blinded, placebo-controlled, dosefinding study. Crit Care Med 2003;3:1839–47. Bernard GR, Vincent JL, Laterre PF, et al. Recombinant human protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study group. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001;344:699–709. Cone LA B, Waterbor R, Sofonio MV. Purpura fulminans due to Streptococcus pneumoniae sepsis following gastric bypass. Obes Surg 2004;14:690–4. Vincent JL, Nadel S, Kutsogiannis DJ, et al. Drotrecogin alfa (activated) in patients with severe sepsis presenting with purpura fulminans, meningitis, or meningococcal disease: a retrospective analysis of patients enrolled in recent clinical studies. Crit Care 2005;9:331–43. Akol H, Boon ES, van Haren FMP, et al. Successful treatment of fulminant pneumococcal sepsis with recombinant tissue plasminogen activator. Eur J Intern Med 2002; 13:389–91. Zenz W, Zoehrer B, Levin M, et al. International Paediatric Meningococcal Thrombolysis Study Group. Use of recombinant tissue plasminogen activator in children with meningococcal purpura fulminans: a retrospective study. Crit Care Med 2004;32:1777–80. Nolan J, Sinclair R. Review of management of purpura fulminans and two case reports. Br J Anaesth 2001;86:581–6. Munteanu C, Bloodworth LLO, Korn HE. Antithrombin concentrate with plasma exchange in purpura fulminans. Pediatr Crit Care Med 2000;1:84–7. Luyt DK, Pridgeon J, Brown J, et al. Extracorporeal life support for children with meningococcal septicaemia. Acta Paediatr 2004;93:1608–11. Rice TW, Bernard GR. Drotrecogin alfa (activated) for the treatment of severe sepsis and septic shock. Am J Med Sci 2004;328:205–14. Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005 365:63–78.
345
ABSTRACT: Sepsis-induced purpura fulminans is a rare but life-threatening disorder, characterized by hemorrhagic infarction of the skin caused by disseminated intravascular coagulation and dermal vascular thrombosis. The pathogenesis is linked to enhanced expression of the natural procoagulants and depletion of the natural anticoagulant proteins particularly protein C. Meningococcal sepsis is the most common cause, followed by pneumococcal sepsis in adults. The syndrome is associated with more than 50% mortality secondary to multiple organ dysfunction syndrome and is accompanied by long-term morbidity. Necrotic lesions usually progress
P
urpura fulminans is an uncommon disorder first described by Guelliot in 1884. Initially noted in children in association with bacterial or viral infection, it was later seen in neonates with a family history of homozygous protein C or protein S deficiency without infection.1 Purpura fulminans secondary to sepsis is a lifethreatening condition characterized by hemorrhagic infarction of the skin caused by disseminated intravascular coagulation and dermal vascular thrombosis.1,2 It is associated with high mortality rate (⬎50%) owing to multiple organ dysfunction and is also accompanied by considerable long-term morbidity. Meningococcal sepsis is the most common form, followed by pneumococcal sepsis in adults. Necrotic lesions usually progress to distal ischemia, often necessitating skin grafting or limb amputation. Early antibiotic administration and intensive care management according to the recommendations of severe sepsis and shock is crucial for patients’ survival. Adjuvant therapies against inflammatory and coagulation cascades and augmenting fibrinolysis are controversial and need further assessment.
From the Intensive Therapy Unit, Royal Free Hospital, London, United Kingdom. Submitted May 23, 2006; accepted in revised form July 19, 2006. Correspondence: Alex Betrosian, MD, Hippokration General Hospital, Intensive Care Unit, Vas. Sofias 114, 11527 Athens, Greece (E-mail: abetrosian@ gmail.com). THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
to distal ischemia, and skin grafting and extremities or limb amputation are often required. Early antibiotic administration and intensive care management according to the recommendations of severe sepsis and shock is crucial for patients’ survival. Adjuvant therapies against inflammatory and coagulation cascades and augmenting fibrinolysis are still controversial and need further assessment. Among them activated protein C and supplementation therapy have given promising results. KEY INDEXING TERMS: Purpura fulminans; Disseminated intravascular coagulation; Sepsis. [Am J Med Sci 2006; 332(6):339–345.]
Etiology Purpura fulminans can be identified in three clinical conditions: 1) in patients with preexistent inherited or acquired abnormality of the protein C or protein S anticoagulant pathway; 2) in patients with acute severe infection, predominantly gram-negative bacterial infections, called acute infectious purpura fulminans; and 3) in patients without known abnormalities of the protein C pathway or acute infections, termed idiopathic purpura fulminans. The last subtype occurs primarily in children, usually after benign infections.3 A multitude of infections can cause this syndrome, Neisseria meningitidis being the foremost among the microorganisms that have been implicated in the development of purpura fulminans, followed by streptococcal infections.6 Pneumococcal purpura fulminans (Streptococcus pneumoniae) is reported with increased frequency (⬎60%) in the setting of asplenia or functional hyposplenia.4 Other encapsulated bacteria such as Haemophilus influenzae and group A and B streptococci have been implicated in infants and adults. Staphylococcus aureus as a cause of purpura fulminans is considered rare, probably because these cases are reported as toxic shock syndrome, but may be as common as other nonmeningococcal organisms. Other organisms associated with purpura fulminans are listed in the Table 1.5–13 Time Course of Cutaneous Manifestations of Skin Necrosis The cutaneous lesions of sepsis-induced purpura fulminans are similar regardless of the causative microorganism. Clinically cutaneous necrosis begins 339
Purpura Fulminans in Sepsis
Table 1. Infective Causes of Purpura Fulminans Bacterial: Neisseria meningitides Streptococcus pneumoniae Group A and B streptococci Enterococcus faecalis Haemophilus influenza, Haemophilus aegyptius Staphylococcus aureus Enterobacteriacae: Klebsiella pneumoniae Escherichia coli Proteus spp (mirabilis, vulgaris) Pseudomonas spp Acinetobacter calcoanitratus Xanthomonas maltophilia Vibrio spp Salmonella paratyphi Mycobacterium tuberculosis Capnocytophaga canimorsus Rickettsia rickettsii Viral: Varicella zoster Protozoal: Plasmodium falciparum
with a region of dermal discomfort that quickly progresses within hours to petechiae, which coalesce to form purple ecchymoses.14 At this stage, the cutaneous lesions can be completely reversible without progressing to skin necrosis. The hallmark of hemorrhagic infarction is the development of hemorrhagic bullae, which can progress to frank skin necrosis and gangrene (Figure 1). The stage of necrosis is associated with considerable morbidity and mortality (⬎50%). The pathologic feature of skin necrosis in purpura fulminans is characterized by occlusion of dermal venules and capillaries by microthrombi (Figure 2), resulting in hemorrhagic infarction.1,3 Soon after the vascular occlusion, the endothelial cells become edematous, resulting in capillary dilation with congestion of the vessels by erythrocytes. Clinically, this stage is manifested as erythema and some edema at the site of skin injury.14 Progression of ischemia and capillary
Figure 2. Fibrin thrombi occlusion in the vessels of the dermis (arrows). (H&E stain; original magnification ⫻40.)
dilatation results in extravasation of blood elements into dermis. Clinically this stage is characterized by petechiae. A minimal to mild inflammatory neutrophil infiltrate may be found in the perivascular region. As ischemia and vascular damage progresses, there is further hemorrhage into dermis and ecchymoses develop. Ecchymoses formation is followed by the development of hemorrhagic bullae in the subepidermal area.15 Affected areas are painful and indurated. The end result is coagulative necrosis of the dermal and subcutaneous tissues with extensive dermal hemorrhage manifested as gangrenous necrosis. Although initially sterile, secondary infection of gangrenous tissue may occur with time, contributing to late morbidity and mortality. Occasionally gram-stained microorganisms are present within the microthrombi, the vessel walls, and the perivascular region. In cases of sepsis-induced purpura fulminans, these vascular changes are not limited to the skin but are widespread, involving multiple organs such as lungs, kidneys, and adrenal glands (WaterhouseFriderichsen syndrome when associated with adrenal hemorrhage), leading to multiple systems organ failure.16 Pathogenesis
Figure 1. Cutaneous lesions in purpura fulminans with gangrene of the tip of the little finger.
340
The pathogenesis of petechiae in sepsis-induced purpura fulminans is strongly suggestive of the localized Shwartzman phenomenon. Shwartzman reaction is initiated by a local, priming intradermal injection of endotoxin, which elicits a transient perivascular inflammatory reaction and increased vascular permeability. Intravenous challenge with the same endotoxin 12 to 24 hours later results in vascular thrombosis and necrotizing vasculitis.17 Sepsis-induced purpura fulminans and the Shwartzman reaction share a pathogenesis involving a disturbance in the balance of anticoagulant and procoagulant activities of endothelial cells. This disturbance, which is triggered by endotoxin in gramDecember 2006 Volume 332 Number 6
Betrosian et al
negative and exotoxin in gram-positive sepsis, appears to be mediated by cytokines, particularly tumor necrosis factor-alpha (TNF-␣), and interleukin-1 (IL-1), leading to the consumption of the natural anticoagulants, proteins C and S and antithrombin III.3,17 Virtually all patients with sepsis have coagulation abnormalities that may range from a mild drop in platelet count and subclinical prolongation of the clotting parameters without clinical signs of bleeding or thrombosis (nonovert disseminated intravascular coagulation [DIC]) to fulminant DIC, characterized by widespread microvascular thrombosis and bleeding from various sites.18 DIC has been shown to be present in more than 90% of these cases, supporting the notion that purpura fulminans is a cutaneous marker of DIC.1,4,5 The development of DIC is mediated by several simultaneously occurring mechanisms initiated by the proinflammatory cytokines TNF-␣, IL-1, and IL-6. The mechanisms include ongoing generation of thrombin, the inhibition of the natural anticoagulant mechanisms, and the impairment of fibrinolysis caused by the plasminogen activator inhibitor-1.18,19 Thrombin Generation In sepsis, the activation of the coagulation cascade originates from the extrinsic pathway with the expression of tissue factor (TF) from cytokine-activated mononuclear cells, which have a pivotal role in the host response to infection.20 Data from animal models and from healthy volunteers have shown that even minimal amounts of circulating endotoxin or TNF-␣ can cause activation of coagulation cascade within few minutes.21 TF activates factor VII and the TF-VII complex cleaves factor X to its active form, Xa. Factor Xa, in conjunction with cofactor Va, cleaves prothrombin to form thrombin, which at the same time amplifies coagulation by activating factors VIII and XI of the intrinsic pathway. Thrombin converts fibrinogen to fibrin and is a potent platelet activator. Activated platelets can accelerate coagulation activation through several pathways.18 –21 Inhibition of the Natural Anticoagulant Mechanisms Thrombin generation, however, is limited by the natural anticoagulant mechanisms involving antithrombin III, protein C/activated protein C system, and tissue factor pathway inhibitor (TFPI). But in cases of sepsis-induced DIC, these regulatory systems are defective and are degraded as a result of endothelial dysfunction and release of acute-phase proteins, respectively.22 Antithrombin III is the major inhibitor of thrombin and factors Xa, IXa, VIIa, and XIa. The anticoagulant action of antithrombin III is enhanced by glycosaminoglycans present on endothelial membranes. During severe infection, antithrombin III THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
levels are very low due to consumption, impaired synthesis, and degradation by elastase from activated neutrophils. Also, a reduction in glycosaminoglycans activity due to endothelial dysfunction further contributes to reduced antithrombin III function in sepsis.23,24 The protein C/activated protein C system pathway is also significantly impaired during sepsis. Protein C is activated when thrombin binds to thrombomodulin on vascular endothelium surfaces. The activated protein C interacts with protein S, which then catalyzes the inactivation of factors Va and VIIa, thus stopping thrombin formation. Down-regulation of thrombmodulin on endothelial cells by proinflammatory cytokines, enhanced consumption, impaired synthesis, and vascular leakage are the main causes of inadequacy of protein C in sepsis.23,25 Tissue factor pathway inhibitor is an endogenous serine protease inhibitor, synthesized and secreted by the endothelial cells, which inhibits factor Xa directly and the factor VIIa/TF catalytic complex in a factor Xa– dependent fashion.26 It contains three Kunitz-type domains. The first domain binds factor VIIa, the second binds factor Xa, and the third domain has a binding site for heparin. Free TFPI binds factor VIIa very slowly in comparison with its binding of factor Xa, whereas factor Xa–TFPI complex is a potent inhibitor of factor VIIa–TF. Their interaction results in the formation of a quaternary Xa-TFPI-VIIa-TF inhibitory complex.23,26 Both experimental and clinical evidence, however, indicates that functionally active TFPI levels are inadequate within the microcirculation to prevent ongoing coagulation and organ dysfunction in sepsis and DIC.26 Inhibition of Fibrinolysis The fibrinolytic system is activated in sepsis and has a major role in the control of fibrin deposition in the microcirculation.27 Major fibrinolytic activators and inhibitors are synthesized and stored in endothelial cells. There is strong evidence of increase in tissue type plasminogen activator, followed by an increase in plasminogen activator inhibitor (PAI-1), which inhibits plasmin generation and contributes to further fibrin deposition in the microcirculation and subsequent organ failure.28 The proinflammatory cytokines TNF-␣ and IL-1 increase PAI-1 synthesis or its release from endothelial cells and decrease the plasminogen activator.18,28 Inhibition of fibrinolysis in cases of fulminant DIC associated with sepsis-induced purpura fulminans is exaggerated by thrombin-activatable fibrinolysis inhibitor, a zymogen that is activated by the thrombin-thrombomodulin complex28,29. Thromin-activatable fibrinolysis inhibitor down-regulates fibrinolysis by removing the carboxy-terminal lysine and arginine residues from fibrin and contributes to the ongoing inhibition of fibrinolysis in the later stages of sepsis.29 341
Purpura Fulminans in Sepsis
Clinical Features
Management
Along with purpuric rash, the clinical features of sepsis-induced purpura fulminans are mainly those of the underlying sepsis and its complications. It is usually preceded by a flu-like illness, with fever or chills, sore throat, malaise, and occasionally gastroenteric symptoms occurring 12 to 24 hours before the development of the spreading petechial rash. The majority of affected patients develop septic shock and DIC. In one study among 70 patients with purpura fulminans of various origins, 67% appeared with septic shock and 78% were associated with DIC.30 Nearly all patients experienced deterioration of their condition and required aggressive fluid resuscitation and inotropic and mechanical ventilatory support. Of interest is the absence of shock in 21 of 43 cases (51%) of pneumococcal sepsis—induced purpura fulminans, suggesting a difference in the rate and severity of activation of the cytokine cascade between gramnegative and gram-positive strains (endotoxin versus exotoxins). However, DIC was present in the majority of the reported cases.4
Purpura fulminans is a syndrome, not a specific disease; therefore, treatment must be aimed toward the underlying infection along with on-going supportive management. Successful management is facilitated by high index of suspicion, early diagnosis, aggressive antibiotic treatment, systemic organ support, and prompt surgical consultation.29 Early administration of antibiotics as soon as the diagnosis of purpura fulminans has been evoked is a significant therapeutic advantage. Since the majority of the reported cases are due to Neisseria meningitidis, empiric antibiotic treatment should consist of a third-generation cephalosporin, which also provides good cover against streptococcus.6,29 Vancomycin or teicoplanin should be added if MRSA is suspected. The above regimen should be altered once the causative microorganism is identified and sensitivities obtained. However, despite early antibiotic intervention and intensive care treatment, the mortality rate remains high and is reported to average 40% (range, 20 –70%).32
Differential Diagnosis
All patients with sepsis-induced purpura fulminans should be treated in an intensive care setting for optimal organ support management. Hemodynamic monitoring for continuous measurement of blood pressure and arterial blood gases are necessary. Aggressive fluid resuscitation to restore intravascular volume, inotropic and ventilatory support, and continuous renal replacement therapy when required are major therapeutic interventions.
In patients presenting with purpura fulminans and hypotension, the diagnosis of infectious purpura fulminans is obvious. In more complex cases, however, the differential diagnoses should include other causes of purpura such as Henoch-Scho¨nlein purpura, postinfectious thrombocytopenic purpura, thrombotic thrombocytopenic purpura, and idiopathic purpura fulminans. The skin lesions of Henoch-Scho¨nlein purpura are typically small and rarely lead to extensive necrosis.1 Thrombotic thrombocytopenic purpura is associated with hemolysis, thrombocytopenia, and hyaline thrombosis of small vessels but never cause massive skin necrosis. However, thrombotic thrombocytopenic pupura occasionally is associated with DIC.30 Sepsis-induced purpura fulminans differs from acute infectious or idiopathic purpura fulminans in that it is associated with an overwhelming acute infection and shock. The idiopathic variety is commonly associated with normal blood pressure and a well-preserved peripheral pulse.1 The skin involvement also differs in these two conditions. In cases of sepsis-induced purpura fulminans, skin necrosis often begins in the distal extremities affecting feet, toes, tip of the nose, earlobes, hands, and fingertips, while progressing gradually to the entire body in a patchy distribution. In contrast, idiopathic purpura fulminans skin necrosis usually involves the skin of the thighs, the buttocks, and the lower trunk and less commonly the extremities.1,31 342
Supportive Treatment
Surgical Treatment Purpura fulminans almost invariably leads to some full-thickness skin loss. The similarity between skin necrosis secondary to DIC and full-thickness cutaneous burns forms the rationale for treating them as if they were full-thickness burns.33 Surgical treatment includes aggressive debridement of necrotic tissues to decrease the risk of secondary contamination and sepsis. Areas of dry gangrene of the digits should be left to demarcate, since early surgery might be unnecessarily extensive.34 Monitoring of compartment pressure and early fasciotomy before the development of demarcation may improve the prognosis of limb preservation. The majority of cases with skin necrosis would need skin grafting.35 Often tissue necrosis is so severe that limp amputations are performed. In a retrospective analysis of patients with purpura fulminans in a burn center, 87% of them required skin grafting and 90% limb amputations. In one-fourth of them, amputation was performed in all extremities.29 Streptococcal-induced purpura fulminans resulted in December 2006 Volume 332 Number 6
Betrosian et al
more extensive amputations when compared with other infections with a mean of three amputations per person. Adjuvant Therapy Many unconventional treatment strategies have been used for the treatment of sepsis-induced purpura fulminans, meningococcal in particular.36 However, these could be applied to any case of sepsis-induced purpura fulminans irrespective of the causative microorganism. The therapies have focused on approaches that could modulate the inflammatory and the coagulation cascade. Therapies Against Inflammatory Cascade Corticosteroids inhibit many components of inflammatory cascade, including complement activation, cytokine production, and neutrophil function and migration. The use of steroids in treating sepsis and purpura fulminans remains controversial. Although there is a survival benefit with low-dose, short-term hydrocortisone in septic patients with relative adrenal insufficiency, evidence does not conclusively support routine use of corticosteroids in sepsis-induced purpura fulminans of any cause. Other anti-inflammatory treatment strategies that had been used for meningococcal sepsis include the use of recombinant bactericidal permeability-increasing protein and human monoclonal antibodybinding endotoxin, which have failed to show a survival benefit for this disease.37–39 Several case reports and a few small randomized trials used plasmapheresis and hemofiltration as methods to remove proinflammatory cytokines in meningococcal purpura fulminans. However, no effect on survival has been reported. These techniques in theory could remove cytokines from circulation, but whether they can remove cytokines from tissues remains questionable.39 Plasma exchange has been used in pediatric patients with sepsis-induced purpura fulminans and was reported to improve prognosis by virtue of a significant increase in the level of coagulation factors II, V, VII, and VIII.40 Therapies Against Coagulation Cascade Since microvascular thrombosis appears to be the predominant pathophysiologic feature of purpura fulminans, the majority of therapeutic approaches have focused on reduction of coagulation activation and improvement of fibrinolysis.36,37 Heparin The use of heparin in cases of sepsis-induced purpura fulminans may appear counterintuitive given the hemorrhagic nature of the disease. However, there is some evidence that heparin could be of THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
benefit in reducing the severity of distal necrosis by inhibiting thrombus formation and consumption of coagulation factors in sepsis-induced purpura fulminans. Arbitrary schemes of unfractionated or low molecular weight heparin have been used, but until now, no effect on survival has been demonstrated in animal models or small clinical trials.41 Antithrombin III Sepsis-induced purpura fulminans is often associated with markedly reduced levels of antithrombin III due to consumptive coagulopathy. Antithrombin III is not only an important physiologic inhibitor of coagulation but also possesses anti-inflammatory properties through release of prostaglandin I2 from endothelial cells.24,25 Several authors have shown that antithrombin III supplementation therapy in patients with severe sepsis and shock is beneficial in restoring organ function and reversing DIC. In adults with purpura fulminans, a loading dose of 100 IU/kg followed by 100 IU/kg/day was able to correct the levels of antithrombin activity, which was subsequently maintained by continuous infusion.42 In contrast to the initial positive results, a large-scale multicenter randomized controlled trial investigating the use of antithrombin III in more than 2300 patients with severe sepsis and shock failed to demonstrate any significant reduction in mortality.43 However, the trial did not investigate DIC, but rather severe sepsis in general. Protein C/Activated Protein C Dysfunction of the protein C pathway during sepsis is one of the major factors contributing to the development of purpura fulminans. Severe depletion of protein C leads to uncontrolled hemorrhage and thrombosis and is associated with poor outcome.22 Treatment with protein C concentrate or recombinant activated protein C (drotrecogin alfa, activated) has been shown to be beneficial in cases of sepsis-induced purpura fulminans.25,29,36,44 Several case reports and open-label studies on protein C (unactivated) substitution therapy have shown encouraging results.45,46 Whether this therapy is effective in achieving adequate levels of activated protein C remains questionable. However, a phase II dose-finding study in children with meningococcal sepsis and purpura fulminans showed that a dose of 150 IU/kg or more is required to increase the levels of activated protein C and a cumulative dose of 600 IU/kg/day is necessary for sustained protein C activation.47 Drotrecogin alpha activated has been approved for the treatment of severe sepsis and organ dysfunction (APACHE II ⬎24). Patients with sepsis-induced purpura fulminans fulfill the criteria for adjuvant therapy with this drug at the dose of 24 g/kg per hour for 96 hours.48 A few case 343
Purpura Fulminans in Sepsis
reports with encouraging results have been published on the use of this drug in patients with purpura fulminans.7,49 However, no randomized trial is yet available.50 Therapies Augmenting Fibrinolysis Recombinant Tissue Plasminogen Activator Administration of a fibrinolytic agent such as recombinant tissue plasminogen activator (t-PA) is an alternative for limiting ongoing thrombin generation. Concentrations of PAI-1 are increased in purpura fulminans and correlate with risk for mortality. Although a few case reports have shown promising results, a retrospective analysis on the use of recombinant t-PA in children with meningococcal purpura fulminans showed a high incidence of intracerebral hemorrhage.51,52 Since there has not been a randomized trial, no definite answers on the use of recombinant t-PA in sepsis-induced purpura fulminans could be drawn. Other Strategies Fresh frozen plasma infusions containing anticoagulant factors have been suggested as an adjuvant therapy in purpura fulminans, providing both fluid resuscitation and immunomodulation. Cryoprecipitate, fibrinogen concentrate, or recombinant factor VIIa can be used in patients with severe bleeding not responding in other treatment options.36,37,53 Plasma exchange using fresh frozen plasma has been used successfully in patients with sepsis-induced purpura fulminans by restoring low levels of protein C and antithrombin III.54 A few anecdotal reports have suggested improved distal perfusion and pain relief with the use of vasodilating agents such as prostacyclin or epoprostenol, sodium nitroprusside, or topical nitroglycerin.53 Hyperbaric oxygenation has been used, and one study demonstrated improvement of limb perfusion in 12 of 14 patients with sepsis-induced purpura fulminans caused by various pathogens.37 Because this treatment needs to be instituted early in the course of the disease, its usefulness is limited for logistical reasons. Extracorporeal membrane oxygenation has been used to support patients with intractable meningococcal sepsis and purpura fulminans. However, there is no exact answer to whether this technique offers any advantage over conventional therapy.53,55 The efficacy and benefits of the above-mentioned treatment strategies are difficult to interpret, since they are often used in combination. Despite the positive preclinical data and promising case reports, no adjuvant therapy has been shown to reduce the rates of morbidity and mortality associated with sepsis-induced purpura fulminans.36,37 Although treatment with recombinant human activated protein C has been approved for the treatment of severe 344
sepsis, its use in patients with purpura fulminans needs further validation. Until then, the management of sepsis-induced purpura fulminans must stand upon early recognition, prompt antibiotic treatment, and intensive care management, according to the recommendations for severe sepsis and shock.56,57 References 1. Marlar RA, Neumann A. Neonatal purpura fulminans due to homozygous protein C or protein S deficiencies. Semin Thromb Hemost 1990;16:299–309. 2. Francis RB Jr. Acquired purpura fulminans. Semin Thromb Hemost 1990;16:310–25. 3. Adcock DM, Brozna J, Marlar RA. Proposed classification and pathologic mechanisms of purpura fulminans and skin necrosis. Semin Thromb Hemost 1990;16:333–40. 4. Pathan N, Faust SN, Levin M. Pathophysiology of meningococcal meningitis and septicaemia. Arch Dis Child 2003; 88:601–7. 5. Molos MA, Hall JC. Symmetrical peripheral gangrene and disseminated intravascular coagulation. Arch Dermatol 1985;121:1057–61. 6. Carpenter CT, Kaiser AB. Purpura fulminans in pneumococcal sepsis: case report and review. Scand J Infect Dis 1997;29:479–83. 7. Kravitz GR, Dries DJ, Peterson ML, et al. Purpura fulminans due to Staphylococcus aureus. Clin Infect Dis 2005;40:941–7. 8. Childers BJ, Cobanov B. Acute infectious purpura fulminans: a 15-year retrospective review of 28 consecutive cases. Am Surg 2003;69:86–90. 9. Olowu WA. Klebsiella-induced purpura fulminans in a Nigerian child: case report and a review of literature. West Afr J Med 2002;21:252–5. 10. Huemer GM, Bonatti H, Dunst KM. Purpura fulminans due to E. coli septicemia. Wien Klin Wochenschr 2004;116:82. 11. Gurses N, Ozkan A. Neonatal and childhood purpura fulminans: review of seven cases. Cutis 1988;41:361–3. 12. Deshmukh PM, Camp CJ, Rose FB, et al. Capnocytophaga canimorsus sepsis with purpura fulminans and symmetrical gangrene following a dog bite in a shelter employee. Am J Med Sci 2004;327:369–72. 13. Keri JE, Thomas K, Berman B, et al. Purpura fulminans in a patient with malaria. Eur J Dermatol 2000;10:617–9. 14. Adcock DM, Hicks MJ. Dermatopathology of skin necrosis associated with purpura fulminans. Semin Thromb Hemost 1990;16:283–92. 15. Cole MS, Minifee PK, Wolma FJ. Coumarin necrosis: a review of the literature. Surgery 1988;103:271–7. 16. Varon J, Chen K, Sternbach GL. Rupert Waterhouse and Carl Friderichsen: adrenal apoplexy. J Emerg Med 1998;16: 643–7. 17. Brozna JP. Shwartzman reaction. Semin Thromb Hemost 1990;16:326–32. 18. Dempfle CE. Coagulopathy of sepsis. Thromb Haemost 2004;91:213–24. 19. Levi M, de Jonge E, van der Poll T. Sepsis and disseminated intravascular coagulation. J Thombosis Thrombolysis 2003;16:43–7. 20. Levi M. Current understanding of disseminated intravascular coagulation. Br J Haematol 2004;124:567–76. 21. Burrell R. Human responses to bacterial endotoxin. Circ Shock 1994;43:137–53.
December 2006 Volume 332 Number 6
Betrosian et al
22. Jagneaux T, Taylor DE, Kantrow SP. Coagulation in sepsis. Am J Med Sci 2004;328:196–204. 23. Okajima K. Regulation of inflammatory responses by natural anticoagulants. Immunol Rev 2001;184:258–74. 24. Roemisch J, Gray E, Hoffmann JN, et al. Antithrombin: a new look at the actions of a serine protease inhibitor. Blood Coagul Fibrinolysis 2002;13:657–70. 25. Aird WC. Natural anticoagulant inhibitors: activated Protein C. Best Pract Res Clin Haematol 2004;17:161–82. 26. Doshi SN, Marmur JD. Evolving role of tissue factor and its pathway inhibitor. Crit Care Med 2002;30(Suppl):241–50. 27. Mavrommatis AC, Theodoridis T, Economou M, et al. Activation of the fibrinolytic system and utilization of the coagulation inhibitors in sepsis: comparison with severe sepsis and septic shock. Intensive Care Med 2001;27:1853–9. 28. Hack CE. Derangements of coagulation and fibrinolysis in infectious diseases. Contrib Microbiol 2003;10:18–37. 29. Zeerleder S, Schroeder V, Hack E, et al. TAFI and PAI-1 levels in human sepsis. Thrombosis Res 2006;118:205–12. 30. Warner PM, Kagan RJ, Yakuboff KP, et al. Current management of purpura fulminans: a multicenter study. J Burn Care Rehabil 2003;24:119–26. 31. Alexander G, Basheer HM, Ebrahim MK, et al. Idiopathic purpura fulminans and varicella gangrenosa of both hands, toes and integument in a child. Br J Plast Surg 2003;56:194–5. 32. Rosenstein NE, Perkins BA, Stephens DS, et al. Meningococcal disease. N Engl J Med 2001;344:1378–88. 33. Dinh TA, Friedman J, Higuera S. Plastic surgery management in pediatric meningococcal-induced purpura fulminans. Clin Plast Surg 2005;32:117–21. 34. Potokar TS, Oliver DW, Russel RR, et al. Meningococcal septicaemia and plastic surgery: a strategy for management. Br J Plast Surg 2000;53:142–8. 35. Edlich RF, Winters KL, Woodard CR, et al. Massive soft tissue infections: necrotizing fasciitis and purpura fulminans. J Long Term Eff Med Implants 2005;15:57–65. 36. Fischer M, Hilinski J, Stephens DS. Adjuvant therapy for meningococcal sepsis. Pediatr Infect Dis J 2005;24:177–8. 37. Leclerc F, Leteurtre S, Cremer R, et al. Do new strategies in meningococcemia produce better outcomes? Crit Care Med 2000;28(Suppl):600–33. 38. Levin M, Quint PA, Goldstein B, et al. Recombinant bactericidal/permeability-increasing protein (rBPI21) as adjunctive treatment for children with severe meningococcal sepsis: a randomised trial. rBPI21 Meningococcal Sepsis Study Group. Lancet 2000;356:961–7. 39. van Deuren M, Brandtzaeg P, van der Meer JW. Update on meningococcal disease with emphasis on pathogenesis and clinical management. Clin Microbiol Rev 2000;13:144–66. 40. Churchwell KB, McManus ML, Kent P, et al. Intensive blood and plasma exchange for treatment of coagulopathy in meningococcemia. J Clin Apheresis 1995;10:171–7. 41. Kupperman N, Inkelis S, Saladino R. The role of heparin in the prevention of extremity and digit necrosis in meningococcal purpura fulminans. Pediatr Infect Dis J 1994;13:867–87. 42. Cobcroft R, Henderson A, Solano C, et al. Meningococcal purpura fulminans treated with antithrombin III concen-
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53. 54.
55.
56.
57.
trate: what is the optimal replacement therapy? Aust N Z J Med 1994;24:575–6. Warren BL, Eid A, Singer P, et al. KyberSept Trial Study Group. Caring for the critically ill patient: high-dose antithrombin III in severe sepsis: a randomized controlled trial. JAMA 2001;286:1869–78. Dhainaut JF, Yan SB, Claessens YE. Protein C/activated protein C pathway: overview of clinical trial results in severe sepsis. Crit Care Med 2004;32(Suppl):194–201. Rintala E, Kauppila M, Seppala OP, et al. Protein C substitution in sepsis-associated purpura fulminans. Crit Care Med 2000;28:2373–8. Betrosian AP, Balla M, Kofinas G, et al. Protein C in the treatment of coagulopathy in meningococcal sepsis. Crit Care Med 1999;27:2849–50. de Kleijn ED, de Groot R, Hack CE, et al. Activation of protein C following infusion of protein C concentrate in children with severe meningococcal sepsis and purpura fulminans: a randomized, double-blinded, placebo-controlled, dosefinding study. Crit Care Med 2003;3:1839–47. Bernard GR, Vincent JL, Laterre PF, et al. Recombinant human protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study group. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001;344:699–709. Cone LA B, Waterbor R, Sofonio MV. Purpura fulminans due to Streptococcus pneumoniae sepsis following gastric bypass. Obes Surg 2004;14:690–4. Vincent JL, Nadel S, Kutsogiannis DJ, et al. Drotrecogin alfa (activated) in patients with severe sepsis presenting with purpura fulminans, meningitis, or meningococcal disease: a retrospective analysis of patients enrolled in recent clinical studies. Crit Care 2005;9:331–43. Akol H, Boon ES, van Haren FMP, et al. Successful treatment of fulminant pneumococcal sepsis with recombinant tissue plasminogen activator. Eur J Intern Med 2002; 13:389–91. Zenz W, Zoehrer B, Levin M, et al. International Paediatric Meningococcal Thrombolysis Study Group. Use of recombinant tissue plasminogen activator in children with meningococcal purpura fulminans: a retrospective study. Crit Care Med 2004;32:1777–80. Nolan J, Sinclair R. Review of management of purpura fulminans and two case reports. Br J Anaesth 2001;86:581–6. Munteanu C, Bloodworth LLO, Korn HE. Antithrombin concentrate with plasma exchange in purpura fulminans. Pediatr Crit Care Med 2000;1:84–7. Luyt DK, Pridgeon J, Brown J, et al. Extracorporeal life support for children with meningococcal septicaemia. Acta Paediatr 2004;93:1608–11. Rice TW, Bernard GR. Drotrecogin alfa (activated) for the treatment of severe sepsis and septic shock. Am J Med Sci 2004;328:205–14. Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005 365:63–78.
345