Thrombotic microangiopathy (TMA) is characterized by injured endothelial cells that are thickened, swollen, or detached mainly from arterioles and capillaries. Platelet and hyaline thrombi causing partial or complete occlusion are integral to the histopathology of TMA. TMA is usually the result of microangiopathic hemolytic anemia (MAHA), with its typical features of thrombocytopenia and schistocytes. In the kidney, TMA is characterized by swollen endocapillary cells (endotheliosis), fibrin thrombi, platelet plugs, arterial intimal fibrosis, and a membranoproliferative pattern. Fibrin thrombi may extend into the arteriolar vascular pole, producing glomerular collapse and at times cortical necrosis. In kidneys that recover from acute TMA, secondary focal segmental glomerulosclerosis may be seen. Diseases associated with this lesion include thrombotic thrombocytopenic purpura (TTP), hemolytic-uremic syndrome (HUS), malignant hypertension, scleroderma renal crisis, antiphospholipid syndrome, preeclampsia/HELLP (hemolysis, elevated liver enzymes, low platelet count) syndrome, HIV infection, and radiation nephropathy.
HEMOLYTIC-UREMIC SYNDROME/THROMBOTIC THROMBOCYTOPENIC PURPURA
HUS and TTP are the prototypes for MAHA. Historically, HUS and TTP were distinguished mainly by their clinical and epidemiologic differences. TTP develops more commonly in adults and was thought to include neurologic involvement more often. HUS occurs more commonly in children, particularly when associated with hemorrhagic diarrhea. However, atypical HUS (aHUS) can first appear in adulthood, and better testing has revealed that neurologic involvement is as common in HUS as in TTP. Accordingly, HUS and TTP now should be differentiated and treated according to their specific pathophysiologic features.
HUS is loosely defined by the presence of MAHA and renal impairment. At least four variants are recognized. The most common is Shiga toxin–producing Escherichia coli (STEC) HUS, which is also known as D+HUS or enterohemorrhagic E. coli (EHEC) HUS. Most cases involve children <5 years of age, but adults also are susceptible, as evidenced by a 2011 outbreak in northern Europe. Diarrhea, often bloody, precedes MAHA within 1 week in >80% of cases. Abdominal pain, cramping, and vomiting are frequent, whereas fever is typically absent. Neurologic symptoms, including dysphasia, hyperreflexia, blurred vision, memory deficits, encephalopathy, perseveration, and agraphia, often develop, especially in adults. Seizures and cerebral infarction can occur in severe cases. STEC HUS is caused by the Shiga toxins (Stx1 and Stx2), which are also referred to as verotoxins. These toxins are produced by certain strains of E. coli and Shigella dysenteriae. In the United States and Europe, the most common STEC strain is O157:H7, but HUS due to other strains (O157/H−, O111:H−, O26:H11/H−, O145:H28, and O104:H4) has occurred. After entry into the circulation, Shiga toxin binds to the glycolipid surface receptor globotriaosylceramide (Gb3), which is richly expressed on cells of the renal microvasculature. Upon binding, the toxin enters the cells, inducing inflammatory cytokines (interleukin 8 [IL-8], monocyte chemotactic protein 1 [MCP-1], and stromal cell–derived factor 1 [SDF-1]) and chemokine receptors (CXCR4 and CXCR7); this action results in platelet aggregation and the microangiopathic process. Streptococcus pneumoniae can also cause HUS. Certain strains produce a neuraminidase that cleaves the N-acetylneuraminic acid moieties covering the Thomsen-Friedenreich antigen on platelets and endothelial cells. Exposure of this normally cryptic antigen to preformed IgM results in severe MAHA.
Atypical HUS is the result of congenital complement dysregulation. The affected patients have the low C3 and normal C4 levels characteristic of alternative pathway activation. Factor H deficiency, the most common defect, has been linked to families with aHUS. Factor H competes with factor B to prevent the formation of C3bBb and acts as a cofactor for factor I, which proteolytically degrades C3b. More than 70 mutations of the factor H gene have been identified. Most are missense mutations that produce abnormalities in the C-terminus region, affecting its binding to C3b but not its concentration. Other mutations result in low levels or the complete absence of the protein. Deficiencies in other complement-regulatory proteins, such as factor I, factor B, membrane cofactor protein (CD46), C3, complement factor H–related protein 1 (CFHR1), CFHR3, CFHR5, and thrombomodulin, have also been reported. Finally, an autoimmune variant of aHUS has been discovered. DEAP (deficient for CFHR protein and positive for factor H autoantibody) HUS occurs when an autoantibody to factor H is formed. DEAP HUS is often associated with a deletion of an 84-kb fragment of the chromosome that encodes for CFHR1 and CFHR3. The autoantibody blocks the binding of factor H to C3b and surface-bound C3 convertase.
Thrombotic Thrombocytopenic Purpura
Traditionally, TTP is characterized by the pentad: MAHA, thrombocytopenia, neurologic symptoms, fever, and renal failure. The pathophysiology of TTP involves the accumulation of ultra-large multimers of von Willebrand factor as a result of the absence or markedly decreased activity (<5–10%) of the plasma protease ADAMTS13, a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13. These ultra-large multimers form clots and shear erythrocytes, resulting in MAHA; however, the absence of ADAMTS13 alone may not itself produce TTP. Often, an additional trigger (such as infection, surgery, pancreatitis, or pregnancy) is required to initiate clinical TTP.
Data from the Oklahoma TTP/HUS Registry suggest an incidence rate of 11.3 cases/106 patients in the United States. The median age of onset is 40 years. The incidence is more than nine times higher among blacks than among non-blacks. Like that of systemic lupus erythematosus, the incidence of TTP is nearly three times higher among women than among men. If untreated, TTP has a mortality rate exceeding 90%. Even with modern therapy, 20% of patients die within the first month from complications of microvascular thrombosis.
The classic form of TTP is idiopathic TTP, which is usually the result of a deficiency in ADAMTS13. While TTP had traditionally been associated with infection, malignancy, and intense inflammation (e.g., pancreatitis), ADAMTS13 activity usually is not decreased in these conditions. In idiopathic TTP, the formation of an autoantibody to ADAMTS13 (IgG or IgM) either increases its clearance or inhibits its activity. Upshaw-Schülman syndrome is a hereditary condition characterized by congenital deficiency of ADAMTS13. TTP in these patients can start within the first weeks of life but in some instances may not present until the patient is several years of age. Both environmental and genetic factors are thought to influence the development of TTP. Plasma transfusion is an effective strategy for prevention and treatment.
Drug-induced TMA is a recognized complication of treatment with some chemotherapeutic agents, immunosuppressive agents, antiplatelet agents, and quinine. Two different mechanisms have been described. Endothelial damage (pathologically similar to that in HUS) is the main cause of the TMA that develops in association with chemotherapeutic agents (e.g., mitomycin C, gemcitabine) and immunosuppressive agents (cyclosporine, tacrolimus, and sirolimus). This process is usually dose-dependent. Alternatively, TMA may develop as a result of drug-induced autoantibodies. This form is less likely to be dose-dependent and can, in fact, occur after a single dose in patients with previous exposure. Ticlopidine produces TTP by inducing an autoantibody to ADAMTS13, but ADAMTS13 deficiency is found in fewer than half of patients with clopidogrel-associated TTP. Quinine appears to induce autoantibodies to granulocytes, lymphocytes, endothelial cells, and platelet glycoprotein IbB/IX or IIb/IIIa complexes, but not to ADAMTS13. Quinine-associated TTP is more common among women. TMA has been reported with drugs that inhibit vascular endothelial growth factor, such as bevacizumab; the mechanism is not completely understood.
Treatment should be based on pathophysiology. Autoantibody-mediated TTP and DEAP HUS respond to plasma exchange or plasmapheresis. In addition to removing the autoantibodies, plasma exchange with fresh-frozen plasma replaces ADAMTS13. Twice-daily plasma exchanges with administration of vincristine and rituximab may be effective in refractory cases. Plasma infusion is usually sufficient to replace the ADAMTS13 in Upshaw-Schülman syndrome. Plasma exchange should be considered if larger volumes are necessary. Drug-induced TMA secondary to endothelial damage typically does not respond to plasma exchange and is treated primarily by discontinuing use of the agent and providing supportive care. Similarly, STEC HUS should be treated with supportive measures. Plasma exchange has not been found to be effective. Antimotility agents and antibiotics increase the incidence of HUS among children, but azithromycin was recently found to decrease the duration of bacterial shedding by adults. Eculizumab is a monoclonal antibody to C5 that is approved for use in aHUS, for which ongoing therapy may be necessary. Plasma infusion/exchange may play a role in aHUS by replacing complement-regulatory proteins. Antibiotics and washed red cells should be given in neuraminidase-associated HUS, and plasmapheresis may be helpful. However, plasma and whole-blood transfusion should be avoided since these products contain IgM, which may exacerbate MAHA. Finally, combined factor H and ADAMTS13 deficiency have been reported. The affected patients are generally less responsive to plasma infusion, a result illustrating the complexity of the management of these cases.
HEMATOPOIETIC STEM CELL TRANSPLANTATION–ASSOCIATED THROMBOTIC MICROANGIOPATHY (HSCT-TMA)
HSCT-TMA develops after HSCT, with an incidence of 8.2%. Etiologic factors include conditioning regimens, immunosuppression, infections, and graft-versus-host disease. Other risk factors include female sex and human leukocyte antigen (HLA)–mismatched donor grafts. HSCT-TMA usually occurs within the first 100 days of HSCT. Table 341-1 lists definitions of HSCT-TMA currently used for clinical trials. Diagnosis may be difficult since thrombocytopenia, anemia, and renal insufficiency are common after HSCT. HSCT-TMA carries a high mortality rate (75% within 3 months). The majority of patients have >5% ADAMTS13 activity, and plasma exchange is beneficial in <50% of patients. Discontinuation of calcineurin inhibitors and substitution with daclizumab (antibody to the IL-2 receptor) are recommended. Treatment with rituximab and defibrotide may also be helpful.
TABLE 341-1Criteria for Establishing Microangiopathic Kidney Injury Associated with Hematopoietic Stem Cell Transplantation |Favorite Table|Download (.pdf) TABLE 341-1 Criteria for Establishing Microangiopathic Kidney Injury Associated with Hematopoietic Stem Cell Transplantation
|International Working Group ||Blood and Marrow Transplant Clinical Trials Network Toxicity Committee |
|>4% schistocytes in the blood ||RBC fragmentation and at least 2 schistocytes per high-power field |
|De novo, prolonged, or progressive thrombocytopenia ||Concurrent increase in LDH concentration above baseline |
|A sudden and persistent increase in LDH concentration ||Negative direct and indirect Coombs test |
|Decrease in hemoglobin level or increased RBC transfusion requirement ||Concurrent renal and/or neurologic dysfunction without other explanations |
|Decrease in haptoglobin concentration || |
HIV-related TMA is a complication encountered mainly before widespread use of highly active antiretroviral therapy. It is seen in patients with advanced AIDS and low CD4+ T cell counts although it can be the first manifestation of HIV infection. The presence of MAHA, thrombocytopenia, and renal failure are suggestive, but renal biopsy is required for diagnosis since other renal diseases are also associated with HIV infection. Thrombocytopenia may prohibit renal biopsy in some patients. The mechanism of injury is unclear, although HIV can induce apoptosis in endothelial cells. ADAMTS13 activity is not reduced in these patients. Cytomegalovirus co-infection may also be a risk factor. Effective antiviral therapy is key, while plasma exchange should be limited to patients who have evidence of TTP.
Either local or total body irradiation can produce microangiopathic injury. The kidney is one of the most radiosensitive organs, and injury can result with as little as 4–5 Gy. Such injury is characterized by renal insufficiency, proteinuria, and hypertension usually developing ≥6 months after radiation exposure. Renal biopsy reveals classic TMA with damage to glomerular, tubular, and vascular cells, but systemic evidence of MAHA is uncommon. Because of its high incidence after allogeneic HSCT, radiation nephropathy is often referred to as bone marrow transplant nephropathy. No specific therapy is available, although observational evidence supports renin-angiotensin system blockade.
SCLERODERMA (PROGRESSIVE SYSTEMIC SCLEROSIS)
Kidney involvement is common (up to 52%) in patients with widespread scleroderma, with 20% of cases resulting directly from scleroderma renal crisis. Other renal manifestations in scleroderma include transient (prerenal) or medication-related forms of acute kidney injury (e.g., associated with D-penicillamine, nonsteroidal anti-inflammatory drugs, or cyclosporine). Scleroderma renal crisis occurs in 12% of patients with diffuse systemic sclerosis but in only 2% of those with limited systemic sclerosis. Scleroderma renal crisis is the most severe manifestation of renal involvement, and is characterized by accelerated hypertension, a rapid decline in renal function, nephrotic proteinuria, and hematuria. Retinopathy and encephalopathy may accompany the hypertension. Salt and water retention with microvascular injury can lead to pulmonary edema. Cardiac manifestations, including myocarditis, pericarditis, and arrhythmias, denote an especially poor prognosis. Although MAHA is present in more half of patients, coagulopathy is rare.
The renal lesion in scleroderma renal crisis is characterized by arcuate artery intimal and medial proliferation with luminal narrowing. This lesion is described as “onion-skinning” and can be accompanied by glomerular collapse due to reduced blood flow. Histologically, scleroderma renal crisis is indistinguishable from malignant hypertension, with which it can coexist. Fibrinoid necrosis and thrombosis are common. Before the availability of angiotensin-converting enzyme (ACE) inhibitors, the mortality rate for scleroderma renal crisis was >90% at 1 month. Introduction of renin-angiotensin system blockade has lowered the mortality rate to 30% at 3 years. Nearly two-thirds of patients with scleroderma renal crisis may require dialysis support, with recovery of renal function in 50% (median time, 1 year). Glomerulonephritis and vasculitis associated with antineutrophil cytoplasmic antibodies and systemic lupus erythematosus have been described in patients with scleroderma. An association has been found with a speckled pattern of antinuclear antibodies and with antibodies to RNA polymerases I and III. Anti-U3-RNP may identify young patients at risk for scleroderma renal crisis. Anticentromere antibody, in contrast, is a negative predictor of this disorder. Because of the overlap between scleroderma renal crisis and other autoimmune disorders, a renal biopsy is recommended for patients with atypical renal involvement, especially if hypertension is absent.
Treatment with ACE inhibition is the first-line therapy unless contraindicated. The goal of therapy is to reduce systolic and diastolic blood pressure by 20 mmHg and 10 mmHg, respectively, every 24 h until blood pressure is normal. Additional antihypertensive therapy may be given once the dose of drug for ACE inhibition is maximized. Both ACE inhibitors and angiotensin II receptor antagonists are effective, although data suggest that treatment with ACE inhibitors is superior. ACE inhibition alone does not prevent scleroderma renal crisis, but it does reduce the impact of hypertension. Intravenous iloprost has been used in Europe for blood pressure management and improvement of renal perfusion. Kidney transplantation is not recommended for 2 years after the start of dialysis since delayed recovery may occur.
Antiphospholipid syndrome (Chap. 379) can be either primary or secondary to systemic lupus erythematosus. It is characterized by a predisposition to systemic thrombosis (arterial and venous) and fetal morbidity mediated by antiphospholipid antibodies—mainly anticardiolipin antibodies (IgG, IgM, or IgA), lupus anticoagulant, or anti-β-2 glycoprotein I antibodies (antiβ2GPI). Patients with both anticardiolipin antibodies and antiβ2GPI appear to have the highest risk of thrombosis. The vascular compartment within the kidney is the main site of renal involvement. Arteriosclerosis is commonly present in the arcuate and intralobular arteries. In the intralobular arteries, fibrous intimal hyperplasia characterized by intimal thickening secondary to intense myofibroblastic intimal cellular proliferation with extracellular matrix deposition is frequently seen along with onion-skinning. Arterial and arteriolar fibrous and fibrocellular occlusions are present in more than two-thirds of biopsy samples. Cortical necrosis and focal cortical atrophy may result from vascular occlusion. TMA is commonly present in renal biopsies, although signs of MAHA and platelet consumption are usually absent. TMA is especially common in the catastrophic variant of antiphospholipid syndrome. In patients with secondary antiphospholipid syndrome, other glomerulopathies may be present, including membranous nephropathy, minimal change disease, focal segmental glomerulosclerosis, and pauci-immune crescentic glomerulonephritis.
Large vessels can be involved in antiphospholipid syndrome and may form the proximal nidus near the ostium for thrombosis of the renal artery. Renal vein thrombosis can occur and should be suspected in patients with lupus anticoagulant who develop nephrotic-range proteinuria. Progression to end-stage renal disease can occur, and a thrombosis may form in the vascular access and the renal allografts. Hypertension is common. Treatment entails lifelong anticoagulation. Glucocorticoids may be beneficial in accelerated hypertension. Immunosuppression and plasma exchange may be helpful for catastrophic episodes of antiphospholipid syndrome but by themselves do not reduce recurrent thrombosis.
HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome is a dangerous complication of pregnancy associated with microvascular injury. Occurring in 0.2–0.9% of all pregnancies and in 10–20% of women with severe preeclampsia, this syndrome carries a mortality rate of 7.4–34%. Most commonly developing in the third trimester, 10% of cases occur before week 27 and 30% post-partum. Although a strong association exists between HELLP syndrome and preeclampsia, nearly 20% of cases are not preceded by recognized preeclampsia. Risk factors include abnormal placentation, family history, and elevated levels of fetal mRNA for FLT1 (vascular endothelial growth factor receptor 1) and endoglin. Patients with HELLP syndrome have higher levels of inflammatory markers (C-reactive protein, IL-1Ra, and IL-6) and soluble HLA-DR than do those with preeclampsia alone.
Renal failure occurs in half of patients with HELLP syndrome, although the etiology is not well understood. Limited data suggest that renal failure is the result of both preeclampsia and acute tubular necrosis. Renal histologic findings are those of TMA with endothelial cell swelling and occlusion of the capillary lumens, but luminal thrombi are typically absent. However, thrombi become more common in severe eclampsia and HELLP syndrome. Although renal failure is common, the organ that defines this syndrome is the liver. Subcapsular hepatic hematomas sometimes produce spontaneous rupture of the liver and can be life-threatening. Neurologic complications such as cerebral infarction, cerebral and brainstem hemorrhage, and cerebral edema are other potentially life-threatening complications. Nonfatal complications include placental abruption, permanent vision loss due to Purtscher-like (hemorrhagic and vaso-occlusive vasculopathy) retinopathy, pulmonary edema, bleeding, and fetal demise.
Many features are shared by HELLP syndrome and MAHA. Diagnosis of HELLP syndrome is complicated by the fact that aHUS and TTP also can be triggered by pregnancy. Patients with antiphospholipid syndrome also have an elevated risk of HELLP syndrome. A history of MAHA before pregnancy is of diagnostic value. Serum levels of ADAMTS13 activity are reduced (by 30–60%) in HELLP syndrome but not to the levels seen in TTP (<5%). Determination of the ratio of lactate dehydrogenase to aspartate aminotransferase may be helpful; this ratio is 13:1 in patients with HELLP syndrome and preeclampsia as opposed to 29:1 in patients without preeclampsia. Other markers, such as antithrombin III (decreased in HELLP syndrome but not in TTP) and d-dimer (elevated in HELLP syndrome but not in TTP), may also be useful. HELLP syndrome usually resolves spontaneously after delivery, although a small percentage of HELLP cases occur post-partum. Glucocorticoids may decrease inflammatory markers, although two randomized controlled trials failed to show much benefit. Plasma exchange should be considered if hemolysis is refractory to glucocorticoids and/or delivery, especially if TTP has not been ruled out.
Renal complications in sickle cell disease result from occlusion of the vasa recta in the renal medulla. The low partial pressure of oxygen and high osmolarity predispose to hemoglobin S polymerization and erythrocyte sickling. Sequelae include hyposthenuria, hematuria, and papillary necrosis (which can also occur in sickle trait). The kidney responds by increases in blood flow and glomerular filtration rate mediated by prostaglandins. This dependence on prostaglandins may explain the greater reduction of glomerular filtration rate by nonsteroidal anti-inflammatory drugs in these patients than in others. The glomeruli are typically enlarged. Intracapillary fragmentation and phagocytosis of sickled erythrocytes are thought to be responsible for the membranoproliferative glomerulonephritis–like lesion, and focal segmental glomerulosclerosis is seen in more advanced cases. Proteinuria is present in 20–30%, and nephrotic-range proteinuria is associated with progression to renal failure. ACE inhibitors reduce proteinuria, although data are lacking on prevention of renal failure. Patients with sickle cell disease are also more prone to acute renal failure. The cause is thought to reflect microvascular occlusion associated with nontraumatic rhabdomyolysis, high fever, infection, and generalized sickling. Chronic kidney disease is present in 12–20% of patients. Despite the frequency of renal disease, hypertension is uncommon in patients with sickle cell disease.