Untying the Knot of Thrombotic Thrombocytopenic Purpura and Atypical Hemolytic Uremic Syndrome

REVIEW Untying the Knot of Thrombotic Thrombocytopenic Purpura and Atypical Hemolytic Uremic Syndrome Han-Mou Tsai, MD imah Hematology Associates, New Hyde Park, NY. ABSTRACT Patients presenting with microangiopathic hemolysis and thrombocytopenia are often given the diagnosis of thrombotic thrombocytopenic purpura and treated with plasma exchange until the acute episode is over. Recent findings have shown that acquired thrombotic thrombocytopenic purpura is a chronic autoimmune disease with inhibitory antibodies of a disintegrin and metalloprotease with thrombospondin repeat, member 13 and are at risk of relapses that may be preventable. Furthermore, many of the patients given the diagnosis of thrombotic thrombocytopenic purpura really have atypical hemolytic uremic syndrome due to defective complement regulation that can be more effectively treated to prevent death and end-stage renal failure with eculizumab, a humanized monoclonal antibody of complement C5. These advances indicate that an accurate differential diagnosis of microangiopathic hemolysis is essential for optimal patient management. 2013 Elsevier Inc. All rights reserved. The American Journal of Medicine (2013) 126, 200-209 KEYWORDS: ADAMTS13; Atypical hemolytic uremic syndrome; Complement regulation; Microangiopathic hemolytic anemia; Thrombotic thrombocytopenic purpura First recognized in the 1970s as a disorder distinct from thrombotic thrombocytopenic purpura and the typical shiga toxin associated hemolytic uremic syndrome, atypical hemolytic uremic syndrome refers to the constellation of acute renal failure, thrombocytopenia, and microangiopathic hemolysis without antecedent shiga toxin induced hemorrhagic diarrhea. Subsequent studies show that atypical hemolytic uremic syndrome also occurs sporadically and affects adults. The severity of renal failure is variable, ranging from severe, irreversible renal failure to mild reversible azotemia. Because of its overlap with thrombotic thrombocytopenic purpura in the features of microangiopathic hemolytic anemia and thrombocytopenia, atypical hemolytic uremic Funding: Some of the author s works discussed in the article were supported by Grant R01 HL62136 from the National Heart, Lung, and Blood Institute of the National Institutes of Health. Conflict of Interest: The sole author has been a speaker and a member of the advisory boards of Alexion, Inc, but did not receive support for his research work or preparing this manuscript. Authorship: The author is solely responsible for the content of this manuscript. Requests for reprints should be addressed to Han-Mou Tsai, MD, imah Hematology, 105 Robby Lane, New Hyde Park, NY 11040. E-mail address: hmtsai@gmail.com syndrome has often been considered merely a variant of thrombotic thrombocytopenic purpura. Consequently, the diagnosis of atypical hemolytic uremic syndrome is unrecognized or viewed as a form of thrombotic thrombocytopenic purpura, and treated as such with plasma exchange. Without a pathogenetic basis, the original view of thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome as 2 distinct disorders fell to disfavor and has been held only by a minority of physicians. Advances in the last 15 years have shown that thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome not only differ in pathogenesis but also require different management. The similarity between the 2 disorders in causing thrombocytopenia and microangiopathic hemolysis is an epiphenomenon of microvascular thrombosis. With eculizumab newly approved for the treatment of atypical hemolytic uremic syndrome, it is more critical than ever to distinguish the disease from thrombotic thrombocytopenic purpura. PATHOGENESIS OF MICROANGIOPATHIC HEMOLYSIS Fragmentation of the red blood cells occurs in 2 types of clinical conditions that share the common feature of abnor- 0002-9343/$ -see front matter 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjmed.2012.09.006

Tsai Untying the Atypical Hemolytic Uremic Syndrome Knot 201 mal intravascular shear stress: vascular devices, such as prosthetic heart valves, ventricular assist devices and extracorporeal oxygenators; and microvascular stenosis (Figure 1). In the absence of vascular devices, fragmentation of the red blood cells signifies stenosis in the arterioles and capillaries. At least 5 different types of arteriolar stenosis are observed to be associated pathologically with microangiopathic hemolysis (Table 1): (1) von Willebrand factor platelet thrombosis, as typically observed in patients with thrombotic thrombocytopenic purpura (Figure 2A-D); (2) platelet-fibrin thrombosis, as exemplified in patients with disseminated intravascular coagulopathy; (3) tumor cell invasion of the microvasculature in patients with metastatic neoplasm; (4) microvascular vasculitis complicating autoimmune disorders such as systemic lupus erythematosus or certain infections such as Rocky Mountain spotted fever; and (5) thrombotic microangiopathy, as observed in patients with shiga toxin associated hemolytic uremic syndrome CLINICAL SIGNIFICANCE or atypical hemolytic uremic syndrome (Figure 2E-H). In thrombotic microangiopathy, endothelial changes such as endothelial cell swelling or disruption, accompanied with intimal expansion and cellular proliferation, are prominent. With or without thrombosis, these changes may cause microvascular stenosis or occlusion. In addition, abnormal vascular permeability may cause interstitial edema of the Thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome are major disorders causing microangiopathic hemolysis in patients without vascular devices or other diseases. Severe ADAMTS13 deficiency defines the diagnosis of thrombotic thrombocytopenic purpura, which is treated with plasma exchange or infusion and rituximab. Advanced renal failure, hypertension, or abnormal vascular permeability favor atypical hemolytic uremic syndrome due to complement dysregulation, which is treated with eculizumab. brain and other organs, as well as fluid accumulation in cavitary spaces, contributing to organ dysfunction in patients with thrombotic microangiopathy. In contrast, in thrombotic thrombocytopenic purpura, organ dysfunction results primarily from ischemia; the endothelium and vessel wall structures are intact, and no abnormal permeability or inflammatory cell infiltration is evident. It has been a common practice to equate the clinical constellation of thrombocytopenia and microangiopathic hemolysis with thrombotic microangiopathy (TMA); and the syndrome of thrombocytopenia, microangiopathic hemolysis and renal failure with thrombotic thrombocytopenic purpura (TTP) or thrombotic thrombocytopenic purpura/ hemolytic uremic syndrome (TTP/HUS). Either practice obscures the important difference among the various causes of microangiopathic hemolysis. Two types of molecular mechanisms have been identified to cause idiopathic microangiopathic hemolysis: 1-6 Defective regulation of von Willebrand factor activity due to deficiency in a disintegrin and metalloprotease with thrombospondin repeat, member 13 (ADAMTS13), causing von Willebrand factor platelet aggregation in thrombotic thrombocytopenic purpura; and Defective regulation of the complement system due to mutations or autoantibodies of complement activators or regulators, causing thrombotic microangiopathy in patients with atypical hemolytic uremic syndrome. Whereas ADAMTS13 deficiency is the only known cause of von Willebrand factor platelet thrombosis in thrombotic thrombocytopenic purpura, thrombotic microangiopathy is not specific for defective complement regulation because it may result from other causes of endothelial injury. Figure 1 Pathogenesis of microangiopathic hemolytic anemia. Abnormally high levels of shear stress, created by vascular devices (eg, left ventricular assist device, extracorporeal membrane oxygenator, or prosthetic heart valves) or arteriolar stenosis, may lead to fragmentation of the red blood cells. In the absence of vascular devices, microangiopathic hemolysis signifies arteriolar stenosis, which is accompanied by thrombocytopenia if the stenosis is due to thrombosis. RBC red blood cell. VON WILLEBRAND FACTOR, ADAMTS13, AND THROMBOTIC THROMBOCYTOPENIC PURPURA von Willebrand factor is a glycoprotein secreted from endothelial cells as a large polymeric form but exists in normal plasma as a series of multimers with progressively smaller sizes. The primary function of von Willebrand factor is to support platelet adhesion at sites of microvascular injury, where it binds to type VI collagen and other components of the exposed vessel wall, becoming rapidly unfolded and activated by the high levels of shear stress at the blood-vessel wall boundary, thereby providing the substrate to support platelet adhesion and aggregation. Thanks to its responsiveness to shear stress, von Willebrand

202 The American Journal of Medicine, Vol 126, No 3, March 2013 Table 1 Different Types of Microvascular Pathology Associated With Microangiopathic Hemolytic Anemia Pathology Key Features Examples VWF-platelet thrombosis Fibrin-platelet thrombosis Thrombus with VWF and platelets Endothelium and vessel wall are intact Thrombus with fibrin and platelets Endothelium and vessel wall are intact Thrombotic thrombocytopenic purpura Disseminated intravascular coagulopathy, heparin-induced thrombocytopenia, paroxysmal nocturnal hemoglobinuria, catastrophic antiphospholipid antibody syndrome, HELLP syndrome Microvascular cancer cells Intravascular clusters of cancer cells Metastatic neoplasm Vasculitis Inflammatory cell infiltration Fibrinoid necrosis Fibrous proliferation Disruption of internal elastic lamina Autoimmune diseases, certain infections Thrombotic microangiopathy Endothelial swelling or disruption Subendothelial expansion/cell proliferation Thrombosis present or absent Internal elastic membrane is intact Interstitial edema Cavitary fluid accumulation HELLP hemolysis, elevated liver enzymes and low platelet counts; VWF von Willebrand factor. *Gemcitabine, mitomycin, calcineurin inhibitors, quinine, cocaine, bevacizumab, and others. Atypical hemolytic uremic syndrome, hemolytic uremic syndrome due to shiga toxins or microbial neuraminidases, catastrophic antiphospholipid antibody syndrome, drugs* factor is uniquely capable of supporting platelet adhesion under high shear stress conditions. Because large forms are more responsive to shear stress, this scheme also explains why large von Willebrand factor multimers are hemostatically more effective than small von Willebrand factor multimers. Conformational unfolding and activation of von Willebrand factor by shear stress also occurs in the circulation, Figure 2 Pathology of thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome. A-D, Microvascular von Willebrand factor thrombosis in a patient who died of thrombotic thrombocytopenic purpura. A, Arteriolar thrombosis in the heart with intact endothelial cells and vessel wall. Immunochemical stain shows von Willebrand factor rich thrombi in the arterioles of the brain (B, C) and kidney (D). von Willebrand factor remains detectable in the endothelial cells. E-H, Thrombotic microangiopathy in the kidney of a patient who died of atypical hemolytic uremic syndrome. E, Endothelial disruption and thrombosis at a juxtaglomerular arteriole. F, A reorganizing arteriolar thrombosis with loss of the endothelial cells. The glomerulus is contracted with thickened subendothelial and mesangial matrix. G, Arteriolar fibrosis and stenosis without thrombosis. H, Marked thickening of the intima, due to matrix expansion and cellular proliferation, causes arteriolar stenosis. The internal elastic lamina is intact (Silver stain).

Tsai Untying the Atypical Hemolytic Uremic Syndrome Knot 203 Figure 3 Pathogenesis of microvascular thrombosis in thrombotic thrombocytopenic purpurea. A, At sites of vessel injury, von Willebrand factor binds to the exposed subendothelial collagen and other components, becoming rapidly unfolded and activated by the shear stress at the blood vessel wall boundary. The unfolded von Willebrand factor provides the substrate for platelet adhesion, which is the initial step of normal hemostasis. B, In the circulation, von Willebrand factor is intermittently and partially unfolded by shear stress. ADAMTS13, a plasma metalloprotease, cleaves von Willebrand factor whenever its scissile bonds are exposed. This shear stress-dependent proteolysis maintains the von Willebrand factor in compact, inactive configuration while it becomes progressively smaller multimers. C, Without proteolysis by ADAMTS13, von Willebrand factor is gradually unfolded and activated by shear stress, leading to von Willebrand factor platelet aggregation and intravascular thrombosis characteristic of thrombotic thrombocytopenic purpura. ADAMTS13 a disintegrin and metalloprotease with thrombospondin repeat, member 13; EC endothelial cell; VWF von Willebrand factor. albeit intermittently and gradually. ADAMTS13, a recently discovered plasma metalloprotease, cleaves von Willebrand factor at a scissile bond in the central A2 domain whenever the protein molecule is beginning to undergo shear-induced conformational change. With this process of shear stressdependent proteolysis, von Willebrand factor becomes progressively smaller multimers that maintains their compact, inactive conformations. By this scheme, large von Willebrand factor multimers are available for hemostasis, yet intravascular von Willebrand factor platelet aggregation is prevented. In the absence of adequate ADAMTS13, von Willebrand factor may be gradually activated by shear stress, leading to intravascular von Willebrand factor platelet aggregation and microvascular thrombosis of thrombotic thrombocytopenic purpura (Figure 3). Two causes of severe von Willebrand factor deficiency have been identified: genetic deficiency due to mutations of the ADAMTS13 gene and autoimmune inhibitors of ADAMTS13 activity. 1,2 Clinical observation indicates that platelet consumption does not occur when AD- AMTS13 is greater than 10%. On the other hand, when the ADAMTS13 level is less than 10%, von Willebrand factor platelet thrombosis may but does not invariably occur. This is because the process of von Willebrand factor platelet aggregation is affected by modifiers such as shear stress profile in the microcirculation and other less well understood factors. A variety of pathologic conditions, such as infection, disseminated intravascular coagulopathy, pancreatitis, and pregnancy, may decrease the plasma ADAMTS13

204 The American Journal of Medicine, Vol 126, No 3, March 2013 Figure 4 Activation and regulation of the complement system. The complement system is activated by immune complexes via the classic pathway or microbial mannose via the lectin pathway. The alternative pathway includes a loop of amplification that is essential for effective innate defense against infection and is regulated by several plasma and membrane proteins (in red) whose mutations have been detected in patients with atypical hemolytic uremic syndrome. In some patients, autoantibodies of complement factor H or gain-of-function mutations involving complement factor B or C3 (in blue) are detected. CFB complement factor B; CFH complement factor H; CFHR1 complement factor H related protein 1; CFI complement factor I; CFP complement factor P; MASP-1/2 mannan-binding lectin serine protease 1/2; MCP membrane cofactor protein; THBD thrombomodulin. levels. Although the magnitude of decrease in association with these pathologic conditions is inadequate to cause von Willebrand factor platelet aggregation, it may trigger disease presentation in patients with preexisting ADAMTS13 mutations or inhibitors, giving rise to the appearance of secondary thrombotic thrombocytopenic purpura. UNCONTROLLED COMPLEMENT ACTIVATION AND THE ATYPICAL HEMOLYTIC UREMIC SYNDROME The complement activation pathways and their regulators are illustrated in Figure 4. As an integral part of the innate defense mechanism for controlling infection, the complement system encompasses a loop of amplification via the alternative pathway to ensure rapid generation of membrane attack complex (C5b-9) and anaphylatoxins C3a and C5a. The self-perpetuating alternative pathway activation is controlled via several regulators, including complement factor H (CFH), which acts as a cofactor of complement factor I (CFI), a plasma serine protease that cleaves and inactivates C3b. Complement factor H also regulates the other steps generating the alternative pathway C3 and C5 convertases. Membrane cofactor protein (MCP), thrombomodulin (THBD), and complement factor H related proteins 1 to 5 (CFHR1-5) are other proteins that augment the regulation of the alternative pathway activity. Three types of molecular defects have been detected in patients with atypical hemolytic uremic syndrome: inactivating mutations of the regulators (CFH, MCP, CFI, or THBD), gain-of-function mutations of the activators (CFB or C3), and autoantibodies of CFH. Mutations of CFH are the most common, detected in 20% to 25% of patients with atypical hemolytic uremic syndrome. Together, molecular defects are detected in 40% of sporadic and 75% of familial patients with atypical hemolytic uremic syndrome. Combinations of polymorphisms or haplotypes (complotypes) of CFH, CFHR2, and other complement regulators also are associated with higher risks of atypical hemolytic uremic syndrome. 7 Overall, the complexity of its molecular pathogenesis suggests that protein replacement therapy is unlikely to be a viable strategy for the management of atypical hemolytic uremic syndrome.

Tsai Untying the Atypical Hemolytic Uremic Syndrome Knot 205 Figure 5 Pathophysiology of atypical hemolytic uremic syndrome. In atypical hemolytic uremic syndrome, uncontrolled complement activation may lead to endothelial injury, resulting in microvascular thrombosis. Endothelial injury also may cause stenosis due to endothelial cell swelling, subendothelial expansion, and reactive cellular proliferation. C3a and C5a may cause abnormal vascular permeability by inducing the release of histamine from basophils and mast cells. Organ dysfunction due to thrombosis is accompanied by thrombocytopenia and microangiopathic hemolysis, but only with microangiopathic hemolysis if it is due to nonthrombotic stenosis and with neither when it is due to interstitial edema. MAC membrane attack complex; RBC red blood cell. PATHOPHYSIOLOGY The clinical manifestation of thrombotic thrombocytopenic purpura results from ischemic injury induced by microvascular thrombosis. As a consequence, the platelet count is generally a reliable indicator of disease activity in thrombotic thrombocytopenic purpura. In contrast, the pathophysiology of atypical hemolytic uremic syndrome is more complex (Figure 5). Endothelial injury induced by C5b-9 (membrane attack complex) may expose the thrombogenic components in the intima to the coagulation proteins and platelets, leading to the development of platelet-fibrin thrombosis. The endothelial injury also may lead to stenosis by causing cellular swelling, subendothelial edema, and reactive cellular proliferation. Furthermore, the C3a and C5a generated during complement activation may induce histamine release from basophils or mast cells, increasing the permeability of the microvasculature. Abnormal vascular permeability may cause interstitial edema and fluid accumulation in cavitary spaces that are common in atypical hemolytic uremic syndrome. Organ dysfunction is associated with thrombocytopenia and microangiopathic hemolysis if it is primarily due to thrombosis; with microangiopathic hemolysis but not thrombocytopenia if it is due to nonthrombotic stenosis; and with neither thrombocytopenia nor microangiopathic hemolysis if it is due to abnormal vascular permeability. Therefore, in atypical hemolytic uremic syndrome, mental status change due to cerebral edema or respiratory distress due to pulmonary edema is not invariably associated with worsening thrombocytopenia or microangiopathic hemolysis. The endothelial injury and stenosis at the glomerular arterioles may cause dysregulation of the juxtaglomerular apparatus, leading to severe but often brittle hypertension without concurrent thrombocytopenia in some patients with atypical hemolytic uremic syndrome. CLINICAL FEATURES The incidence of thrombotic thrombocytopenic purpura is estimated to be 2 to 15 cases/10 6 patient-years. Atypical hemolytic uremic syndrome is likely to be one third to one half as frequent as thrombotic thrombocytopenic purpura, although the diagnosis is often unrecognized (Table 2). Both thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome cause mental status changes. Focal deficits are more common in patients with thrombotic thrombocytopenic purpura but also may occur in atypical hemolytic uremic syndrome. Abdominal symptoms such as pain, nausea, vomiting, and diarrhea are common at acute presentation of atypical hemolytic uremic syndrome but also occur albeit less frequently in thrombotic thrombocytopenic purpura.

206 The American Journal of Medicine, Vol 126, No 3, March 2013 Table 2 Comparison of Thrombotic Thrombocytopenic Purpura and Atypical Hemolytic-Uremic Syndrome Disorder Thrombotic Thrombocytopenic Purpura Atypical Hemolytic-Uremic Syndrome Molecular defects ADAMTS13 inhibitors ADAMTS13 mutations CFH antibody Genetic defects in C activators or regulators Mode of transmission Acquired Autosomal recessive Acquired Autosomal dominant** Age Teens - adults Children - adults Children - adults Gender (female vs 2-3:1 1:1 1:1 male) Incidence (relative) 1 5% 1/3-1/2 Pathology Microvascular VWF-platelet thrombosis Thrombotic microangiopathy Common presentation Fatigue, headache, dizziness Focal deficits, confusion Petechiae Abdominal pain Fatigue, headache, dizziness Chest pain, dyspnea Abdominal pain, vomiting, diarrhea Somnolence, confusion Thrombocytopenia Generally precedes symptoms and signs May not reflect disease severity Advanced renal failure Rare Occasional Common Hypertension Rare Occasional Common Vascular permeability Cerebral edema, PRES pulmonary edema, effusions, ascites, anasarca Mortality with plasma therapy Rare Rare Common 10% Very low 20%-30% Treatment of choice Plasma exchange Plasma infusion Eculizumab Long-term prognosis* Relapse Stroke Relapse Stroke Renal failure Relapse End-stage renal disease in 30%-40% CFB complement factor B; CFH complement factor H; CFI complement factor I; MCP membrane cofactor protein; PRES posterior reversible encephalopathy syndrome; THBD thrombomodulin; VWF von Willebrand factor. *Without maintenance therapy. **With variable expression. Renal failure and hypertension tend to be severe in patients with atypical hemolytic uremic syndrome, although the serum creatinine level may be only mildly elevated initially, and some patients may not develop severe renal failure or hypertension. Severe renal failure and hypertension are rare with thrombotic thrombocytopenic purpura without comorbid conditions. Complications of abnormal vascular permeability, such as brain or pulmonary edema, pleural or pericardial effusions, ascites and anasarca, occur in approximately one third of patients with atypical hemolytic uremic syndrome but are rare in thrombotic thrombocytopenic purpura without comorbidity. DIFFERENTIAL DIAGNOSIS OF MICROANGIOPATHIC HEMOLYSIS Thrombotic thrombocytopenic purpura is the diagnosis in patients with severe ADAMTS13 deficiency due to autoimmune inhibitors or, less frequently, genetic mutations. The diagnosis of atypical hemolytic uremic syndrome is less straightforward. The diagnosis may be established by demonstrating genetic mutations or complement factor H antibodies. However, the test results are generally not available for weeks to months. Furthermore, negative test results do not exclude the diagnosis of atypical hemolytic uremic syndrome. Atypical hemolytic uremic syndrome is the presumptive diagnosis in patients with renal function impairment not accompanied with severe ADAMTS13 or comorbid conditions, with 2 potential caveats: the uncertain reliability of clinical ADAMTS13 assays and the unapparent nature of some comorbid conditions. Therefore, the diagnosis of atypical hemolytic uremic syndrome should be constantly reevaluated. The diagnosis of atypical hemolytic uremic syndrome is often uncertain in patients with comorbid conditions. The patients without severe ADAMTS13 deficiency may be classified into 5 groups on the basis of the absence or presence of comorbid conditions (Table 3). Group I is primary atypical hemolytic uremic syndrome without comorbid conditions. Group II includes pregnancy and other conditions that are considered to be triggers of exacerbation in patients with atypical hemolytic uremic syndrome. Group III includes conditions that cause defective regulation of the complement system. Some patients develop atypical hemolytic uremic syndrome due to com-

Tsai Untying the Atypical Hemolytic Uremic Syndrome Knot 207 Table 3 Classification of Comorbid Conditions in Patients With Microangiopathic Hemolytic Anemia but without Severe ADAMTS13 Group Role Examples I None Idiopathic atypical hemolytic uremic syndrome II Trigger of complement activation Pregnancy, surgery, infection, inflammation, pancreatitis III Acquired complement dysregulation* HIV infection, hematopoietic stem cell therapy, autoimmune diseases (eg, lupus, renal scleroderma, catastrophic antiphospholipid antibody syndrome), drugs IVa Thrombotic microangiopathy via other mechanisms Shiga toxin associated hemolytic uremic syndrome Pneumococcal hemolytic uremic syndrome HIV infection, hematopoietic stem cell therapy, autoimmune diseases (eg, lupus, renal scleroderma, catastrophic antiphospholipid antibody syndrome), drugs Severe hypertension IVb Other types of pathology Fibrin thrombosis Disseminated intravascular coagulopathy, catastrophic antiphospholipid antibody syndrome, hemolysis with elevated liver enzymes and low platelet syndrome of pregnancy, heparin-induced thrombocytopenia, paroxysmal nocturnal hemoglobinuria Microvascular cancer cells Vasculitis, autoimmune or infectious V Unrelated Miscellaneous HIV human immunodeficiency virus. *Except for hematopoietic stem cell therapy, there is no direct evidence as yet that acquired complement dysregulation underlies the thrombotic microangiopathy in patients with HIV infection, autoimmune diseases, or catastrophic antiphospholipid syndrome, or taking drugs (eg, gemcitabine, mitomycin, calcineurin inhibitors, quinine, cocaine, bevacizumab). Severe hypertension may be the manifestation of forme fruste atypical hemolytic uremic syndrome rather than a cause of microangiopathic hemolytic anemia. plement factor H autoantibodies after undergoing hematopoietic stem cell therapy. It is assumed that autoreactive B cells targeting complement factor H may emerge because of defective regulation of autoimmunity after myeloablation. Group IV includes thrombotic microangiopathy due to other mechanisms (IVa) and other types of pathology (IVb). Some comorbid conditions such as autoimmune diseases may cause microangiopathic hemolysis via more than one mechanism (group II, III, or IV). Group V includes comorbid conditions that are unrelated to the development of microangiopathic hemolysis but may complicate the clinical presentation. CLINICAL COURSE AND TREATMENT Acquired thrombotic thrombocytopenic purpura is a chronic autoimmune disorder that is characterized by episodes of exacerbation. During clinical remission with normal platelet count, the patients generally do not develop progressive organ injury. The treatment of acquired thrombotic thrombocytopenic purpura during its exacerbation is plasma exchange to replenish the missing ADAMTS13 and remove the inhibitors. With prompt plasma exchange therapy, approximately 90% of the patients are expected to survive the acute episodes. Plasma therapy does not address the underlying autoimmune nature of acquired thrombotic thrombocytopenic purpura. The patients are responsive to the treatment and able to wean from plasma exchange because the autoimmunity to ADAMTS13 is not intense and spontaneously wanes after a few weeks. In fatal cases, the patients have high levels of ADAMTS13 inhibitors that are not amenable to plasma exchange therapy. 6 Some patients may not be able to wean off plasma exchange therapy because their ADAMTS13 does not recover above the threshold level of suppressing platelet thrombosis. Rituximab has been used, with high rates (70%-90%) of remission, initially in patients unable to wean off plasma exchange even after therapy with corticosteroids, antiplatelet drugs, vincristine, cyclophosphamide, azathioprine, or splenectomy. 8 With its efficacy increasingly appreciated, rituximab is often used before the other modalities. 9 A clinical trial has explored the potential role of preemptive rituximab treatment within a few days of presentation. 10 Compared with historical controls, preemptive rituximab treatment seems to shorten the time to sustained remission and postpone the onset of subsequent relapses. Rituximab suppresses but does not eliminate the ADAMTS13 autoimmunity. Most patients, even those with normalized ADAMTS13 levels spontaneously or after rituximab therapy, are expected to develop relapses during their lifetime. For early detection of impending relapses, periodic complete blood count is inadequate and serial monitoring of the ADAMTS13 level is necessary. The patients may be treated

208 The American Journal of Medicine, Vol 126, No 3, March 2013 preemptively with rituximab to prevent relapse if the AD- AMTS13 level is persistently decreased or exhibits a trend of decrease, which often develop over several weeks to months. 6 Patients with hereditary thrombotic thrombocytopenic purpura generally require maintenance plasma infusion therapy every 2 to 3 weeks to prevent exacerbations and progressive renal and other organ injury. Conventionally atypical hemolytic uremic syndrome has been treated as thrombotic thrombocytopenic purpura with plasma exchange, often with unsatisfactory outcomes (Table 2). Maintenance plasma therapy may prevent renal function deterioration in some patients. However, plasma exchange is a technically demanding procedure that is difficult to perform on a long-term basis. Immunosuppressive agents such as rituximab or splenectomy is not expected to be effective for atypical hemolytic uremic syndrome unless it is due to complement factor H autoantibodies. A major advance in the management of atypical hemolytic uremic syndrome is the use of eculizumab, a humanized monoclonal antibody of complement C5 originally approved in 2007 for the treatment of paroxysmal nocturnal hemoglobinuria, another disease due to complement dysregulation. Two prospective clinical trials have demonstrated the efficacy of eculizumab for patients with atypical hemolytic uremic syndrome not responding to plasma exchange or requiring maintenance prophylaxis. 11,12 With eculizumab therapy, renal function improved in the group with active disease and remained stable or slightly improved in the maintenance group. No patients required additional plasma therapy or new dialysis during eculizumab therapy. These trials led to its approval in the United States and European Union for atypical hemolytic uremic syndrome. Eculizumab is generally well tolerated. A major risk is fulminant meningococcal infection. Meningococcal vaccination is mandatory, and the patients should take prophylactic antibiotics at least until 2 weeks after vaccination. Vaccination decreases but does not eliminate the risk of fulminant meningococcal disease. 13 The patients should be well versed on this risk and carry a medication warning card. Long-term prophylactic antibiotics also should be seriously considered. WHO SHOULD BE TREATED WITH ECULIZUMAB AND FOR HOW LONG? Eculizumab is the treatment of choice over plasma exchange for patients with primary atypical hemolytic uremic syndrome, that is, the patients presenting with thrombocytopenia, microangiopathic hemolysis, and renal function impairment but without severe ADAMTS13 deficiency or comorbidity. Eculizumab also should be considered for atypical hemolytic uremic syndrome associated with any of the group II or group III comorbid conditions (Table 3). Patients with end-stage renal disease due to atypical hemolytic uremic syndrome should be treated with eculizumab if they continue to experience extrarenal complications or severe hypertension. For patients with newly diagnosed advanced renal failure, eculizumab treatment may lead to gradual renal function improvement, allowing the discontinuation of dialysis after more than 1 year of treatment. Therefore, patients with advanced renal disease due to atypical hemolytic uremic syndrome should not give up hope of partial recovery or rush prematurely to kidney transplantation. The optimal duration of eculizumab treatment has not been determined. Atypical hemolytic uremic syndrome is a chronic, often progressive disease even in the absence of thrombocytopenia or microangiopathic hemolysis. Therefore, eculizumab probably should be continued indefinitely for patients at the edge of requiring renal replacement therapy and for patients who are likely to have recurrent atypical hemolytic uremic syndrome with serious complications if eculizumab is discontinued. For asymptomatic patients with normal or mildly abnormal renal function, cautious discontinuation of eculizumab may be an option. The patients should be closely monitored for clinical symptoms, blood pressures, blood counts, and renal function. IMPLICATIONS FOR RENAL TRANSPLANTATION Patients with atypical hemolytic uremic syndrome have a high risk of graft failure after kidney transplantation, primarily due to thrombotic microangiopathy affecting the grafts. 14 Unaware of this risk, some patients have received multiple grafts, all ending in failure. Concurrent liver transplantation may prevent thrombotic microangiopathy in the graft by restoring the plasma complement factor H level in patients with CFH mutations. The operation is associated with high rates of perioperative morbidity and mortality unless the patients are pretreated with plasma exchange, most likely due to triggering of the inadequately regulated complement system. Eculizumab therapy, started preoperatively and continued postoperatively, is expected to shift the paradigm. Preliminary experience suggests that with eculizumab therapy excessive morbidity, mortality, and kidney graft failure may be prevented and liver transplantation is not essential. 15-17 Some of the patients with atypical hemolytic uremic syndrome may present with advanced renal disease without history or evidence of microangiopathic hemolysis and thrombocytopenia. Consequently, kidney biopsy or molecular tests for atypical hemolytic uremic syndrome are recommended for patients with renal failure of undetermined causes if kidney transplantation is being considered. Family members of patients with atypical hemolytic uremic syndrome may have the same genetic defects but are asymptomatic. They should refrain from donating their organs.

Tsai Untying the Atypical Hemolytic Uremic Syndrome Knot 209 CONCLUSIONS The last 15 years have witnessed exciting breakthroughs in understanding the causes of microangiopathic hemolysis: the discovery of ADAMTS13 and its deficiency in thrombotic thrombocytopenic purpura and the discovery of defective complement regulation in atypical hemolytic uremic syndrome. Thrombotic thrombocytopenic purpura and atypical hemolytic uremic syndrome are distinct disorders that happen to share the common feature of thrombocytopenia and microangiopathic hemolysis because they both cause arteriolar thrombosis, albeit via completely different mechanisms. The autoimmune nature of acquired thrombotic thrombocytopenic purpura has prompted the use of rituximab to induce remission in patients with persistent disease activity and to prevent its relapse. The development of recombinant ADAMTS13 proteins may make plasma exchange treatment obsolete in the future. Eculizumab prevents death and end-stage renal disease and facilitates renal function recovery in patients with atypical hemolytic uremic syndrome. Many cases of atypical hemolytic uremic syndrome are not recognized or continue to be treated as thrombotic thrombocytopenic purpura. With eculizumab available as an effective treatment, it is imperative to recognize atypical hemolytic uremic syndrome as a distinct disorder that requires its own strategy of management. References 1. Tsai HM, Lian EC. Antibodies to von Willebrand factor-cleaving protease in acute thrombotic thrombocytopenic purpura. N Engl J Med. 1998;339:1585-1594. 2. Levy GG, Nichols WC, Lian EC, et al. Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura. Nature. 2001;413:488-494. 3. Maga TK, Nishimura CJ, Weaver AE, Frees KL, Smith RJ. Mutations in alternative pathway complement proteins in American patients with atypical hemolytic uremic syndrome. Hum Mutat. 2010;31:E1445- E1460. 4. Noris M, Caprioli J, Bresin E, et al. Relative role of genetic complement abnormalities in sporadic and familial ahus and their impact on clinical phenotype. Clin J Am Soc Nephrol. 2010;5:1844-1859. 5. Malina M, Roumenina LT, Seeman T, et al. Genetics of hemolytic uremic syndromes. Presse Med. 2012;41(3 Pt 2):e105-e114. 6. Tsai HM. Autoimmune thrombotic microangiopathy: advances in pathogenesis, diagnosis, and management. Semin Thromb Hemost. 2012;38:469-482. 7. Ermini L, Goodship TH, Strain L, et al. Common genetic variants in complement genes other than CFH, CD46 and the CFHRs are not associated with ahus. Mol Immunol. 2012;49:640-648. 8. Gutterman LA, Kloster B, Tsai HM. Rituximab therapy for refractory thrombotic thrombocytopenic purpura. Blood Cells Mol Dis. 2002;28: 385-391. 9. Elliott MA, Heit JA, Pruthi RK, Gastineau DA, Winters JL, Hook CC. Rituximab for refractory and or relapsing thrombotic thrombocytopenic purpura related to immune-mediated severe ADAMTS13-deficiency: a report of four cases and a systematic review of the literature. Eur J Haematol. 2009;83:365-372. 10. Scully M, McDonald V, Cavenagh J, et al. A phase 2 study of the safety and efficacy of rituximab with plasma exchange in acute acquired thrombotic thrombocytopenic purpura. Blood. 2011;118:1746-1753. 11. Loirat C, Babu S, Furman R, et al. Eculizumab efficacy and safety inpatients with atypical hemolyticuremic syndrome (ahus) resistant to plasma exchange/infusion. Haematologica. 2011;96:S2-0979. 12. Loirat C, Muus P, Legendre C, et al. A phase II study of eculizumab in patients with atypical hemolyticuremic syndrome receiving chronic plasma exchange/infusion. Haematologica. 2011;96:S2-0980. 13. Kelly RJ, Hill A, Arnold LM, et al. Long-term treatment with eculizumab in paroxysmal nocturnal hemoglobinuria: sustained efficacy and improved survival. Blood. 2011;117:6786-6792. 14. Zuber J, Le QM, Sberro-Soussan R, Loirat C, Fremeaux-Bacchi V, Legendre C. New insights into postrenal transplant hemolytic uremic syndrome. Nat Rev Nephrol. 2011;7:23-35. 15. Krid S, Roumenina L, Beury D, et al. Renal transplantation under prophylactic eculizumab in atypical hemolytic uremic syndrome with CFH/CFHR1 hybrid protein. Am J Transplant. 2012;12:1938-1944. 16. Nester C, Stewart Z, Myers D, et al. Pre-emptive eculizumab and plasmapheresis for renal transplant in atypical hemolytic uremic syndrome. Clin J Am Soc Nephrol. 2011;6:1488-1494. 17. Weitz M, Amon O, Bassler D, Koenigsrainer A, Nadalin S. Prophylactic eculizumab prior to kidney transplantation for atypical hemolytic uremic syndrome. Pediatr Nephrol. 2011;26:1325-1329.