Hemolytic Uremic Syndrome

Updated: Dec 17, 2013
  • Author: Robert Stanley Rust, Jr, MD, MA; Chief Editor: Amy Kao, MD  more...
  • Print
Overview

Background

When Gasser et al first described hemolytic uremic syndrome (HUS) in 1955, it was usually a fatal illness. [1] HUS typically appeared in early childhood and included the combination of Coombs-negative (nonimmune) thrombocytopenic microangiopathic hemolytic anemia and irreversible acute renal failure. Survival greatly improved with the advent and improvement of dialysis and kidney transplantation. However, HUS remains a leading cause of acute renal failure in North American children and is increasingly recognized as a cause of renal failure in adults. Unfortunately, little advance has been made in preventing or acutely reversing this most serious aspect of HUS. HUS accounts for 7% of cases of hypertension in infants younger than 12 months. [2, 3, 4, 5]

Peripheral smear in hemolytic uremic syndrome, wit Peripheral smear in hemolytic uremic syndrome, with findings of microangiopathic hemolytic anemia. Note schistocytes/helmet cells, as well as decrease in platelets. Image courtesy of Emma Z Du, MD.

Clinical and pathologic similarities between HUS and thrombotic thrombocytopenic purpura (TTP)–the other major thrombotic microangiopathy (TMA)–have long been appreciated. However, certain features have been relied on to distinguish most cases labeled HUS, which is predominantly a disease of children younger than 5 years, from most cases labeled TTP, which is predominantly a disease of adults. Renal manifestations are more prominent than neurologic ones in HUS, whereas neurologic findings are more prominent than renal findings in TTP. In industrialized nations, fever precedes the onset of TTP more commonly than it precedes HUS. [6] In most nonfamilial cases of HUS in industrialized or nonindustrialized nations, dysentery is an important hallmark of HUS.

The HUS label has long been applied to individuals older than ten years of age who have developed thrombotic microangiopathy with manifestations that were predominantly renal. Conversely, thrombotic microangiopathy in small children who have predominantly neurologic manifestations has been labeled TTP. During the early phases of disease recognition, recognition of atypical cases and difficult-to-classify cases eroded confidence that objective criteria other than age could be used to distinguish atypical HUS from atypical TTP.

Two conflicting tendencies followed. The first was awkward expansion of labeled entities within a presumed continuum between classic TTP and classic HUS. The second was the broad application of a single, nonspecific, and unsatisfactory term TTP/HUS to cases of thrombotic microangiopathy associated with renal failure and various degrees of involvement of additional organ systems, particularly the nervous system. The recognition of phenotypic instability in recurrent cases encouraged the tendency to consider HUS and TTP to be due to the same underlying mechanisms though variously manifested in part because of age-related vulnerabilities. [7] An example of this apparent phenotypic instability was a patient who had 5 episodes of the HUS phenotype before the age of 15 years and who had 9 episodes of the TTP phenotype after 20 years of age. [8] However, in 1988 Wardle argued that in most cases, HUS and TTP were separate entities of distinct pathogenesis. [9]

Recent advances in the understanding of the pathogenesis of HUS or TTP have supported Wardle's point of view and have clarified the boundary between these illnesses and produced useful diagnostic tests to identify discrete processes that clearly define a particular pathogenic process. However, such tests do not help in distinguishing some clinical syndromes in the TTP-HUS spectrum. Moreover, the speed of this diagnostic progress has outpaced the establishment of a consensus regarding diagnostic categories and boundaries. Old systems of classification have been variously amended.

Pending the establishment of a widely accepted system of classification, this article considers HUS on the basis of a tentative practical system of classification with reference to areas of overlap with TTP, whereas the subject of TTP and its classification is considered in a separate articles (see Medscape Reference article Thrombotic Thrombocytopenic Purpura).

Next:

Pathophysiology

Recent retrospective investigations have demonstrated that some and perhaps most fatal cases of HUS and TTP are pathologically distinct entities. Both conditions manifest microangiopathy with thromboses. Immune mechanisms play a role in some instances; however, the microangiopathies of these conditions are primarily the result of different combinations of developmental, toxic, or mechanical and/or rheologic processes rather than primary immune-mediated processes.

Therefore, the arteriolar and capillary microthrombotic process found in most cases of HUS is the result of the activity of specific toxins with ensuing injury to endothelial cells. On the other hand, most cases of TTP are the result of one of several possible abnormalities of platelet function. Microangiopathic anemia is not associated with Coombs positivity in either condition. In both conditions, it is chiefly the result of rheological disturbances produced by clots. In addition, in HUS, an additional effect of vascular endothelial swelling occurs.

Both HUS and TTP are families of illness comprising a large core of typical cases and additional atypical examples mediated by a broad variety of heritable or acquired conditions. Conditions that produce the various atypical forms overlap, and examples of particular stimuli (eg, verocytogenic Escherichia coli gastrointestinal infection) that classically produce HUS in children younger than 5 years are described. However, in adults, some of these conditions may provoke TTP. Therefore, these 2 families of conditions cannot be separated entirely.

In most cases of HUS, the cause is activity of toxigenic proteins that have deleterious effects on endothelial cells, particularly those of colon and kidney. The 2 most important toxins were initially identified in studies of Shigella dysenteriae and therefore named Shiga toxin (Stx), specifically Stx1 and Stx2. Because the assay for these toxins used verocytes, they are also called verocytotoxins (ie, VT-1 and VT-2). In this discussion, these toxins are called Stx1 and Stx2.

Ensuing studies have identified STX (Stx-E coli) as the most important toxic protein in E coli -associated postinfectious HUS (IStx-HUS). The most commonly identified environmental source of Stx-E coli – producing HUS in humans is the stool of various animals, particularly cattle, sheep, goats, horses, dogs, domestic fowl, and wild birds, as well as humans. These bacteria are also found in flies that feed on the feces of these various animals.

Classification

Clinical and laboratory information concerning the presence of infectious agents that may or may not elaborate verocytotoxins, and tests of ADAMTS-13 enzymatic function or other disturbances associated with TTP (when indicated) permit confident diagnosis of most cases of HUS. On this basis, TMAs have been classified in several ways, including the following:

  • Hereditary, recurrent TTP - Idiopathic or ADAMTS-13 deficient
  • Postinfectious TTP - Acquired, anti-ADAMTS-13 immunoglobulin G (IgG)–mediated
  • TTP-like illness without identifiable ADAMTS-13 deficiency
  • Postinfectious HUS - Stx-related (IStx-HUS)
  • Postinfectious HUS - Non–Stx-related (INon-Stx HUS)
  • Sporadic or immunologic HUS - Diarrheal and nondiarrheal
  • Familial HUS - H-factor normal or H-factor deficient
  • TMA not otherwise specified

As far as HUS is concerned, this classification system is a considerable simplification of old systems of classification. The overwhelming majority of cases of what might be regarded as typical or postinfectious HUS, ie, microangiopathy and renal failure after Stx- or non-Stx–elaborating infectious illnesses, with diarrhea more commonly than without. This simplified scheme readily accommodates these cases, which may be referred to as IStx and non – IStx-HUS. In the venerable Drummond scheme for the classification of HUS, such cases were called classic infantile HUS or postinfectious HUS and subclassified according to whether diarrhea was present. [10]

Sporadic or immunologic cases may or may not be associated with a diarrheal prodrome, and they include acquired transient abnormalities of complement regulation. Clinically defined familial cases are subclassified according to whether the individual is constitutionally deficient in H factor or activity of the third component of complement (C'3). Non-postinfectious HUS– illnesses that occur after inflammatory, immunologic, oncologic, endocrine or obstetric, toxic, and other settings are included in the familial category based on clinical grounds. Otherwise, they are placed in the sporadic or immunologic group. Individuals with TMA in association with such presumed provocations are identified as having TTP when the clinical and laboratory syndrome is consistent with that entity. Remaining cases are TMA not otherwise specified.

Stx-related postinfectious HUS, or IStx-HUS

The mechanism of IStx-HUS is increasingly well defined, whereas the mechanisms of INon-Stx HUS is less well understood. Cases in most young children with a respiratory or other presumed viral prodromes to microangiopathic acute renal failure are also readily accommodated, especially because specific testing may help in distinguishing such cases from infantile TTP (see the Medscape Reference articles Thrombotic Thrombocytopenic Purpura and Hemolytic Uremic Syndrome in Emergency Medicine).

Most other cases that are likely to represent forms of HUS may be classified as sporadic or familial HUS. Some mechanisms have been assigned to patients in these groups. Certain provocations associated with HUS are labeled as sporadic or familial. Identification of a family history of similar events has traditionally permitted the diagnosis of familial HUS. Identification of 1 of several known heritable mechanisms for the occurrence of HUS should also permit the diagnosis of familial HUS.

Future identification of additional heritable mechanisms will likely increase the percentage of sporadic cases that are transferred to the heritable category. Individuals with microangiopathic diseases and various combinations of renal, gastroenterologic, neurologic, and other manifestations that cannot be confidently classified as sporadic or clinically familial acute microangiopathic renal disease might be identified as having HUS, TTP, or another microangiopathic entity.

Most cases of HUS arise in previously healthy children younger than 5 years and cause the typical combination of hematologic, gastroenterologic, and renal disease. Most cases occur after an apparently infectious process. Pulmonary findings are not uncommon, and neurologic manifestations arise in approximately one third of all patients with HUS. These problems tend to be mild and transient. Most instances of IStx-HUS are induced by E coli (the most common agent in industrialized nations). Persistent renal disease is unfortunately common and often severe. Outcomes with regard to nonrenal systems are considerably worsened in Streptococcus pneumoniae – or S dysenteriae type 1–induced Stx-HUS. Most cases of INon-Stx HUS have a relatively favorable outcome.

In general, the clinical involvement of various organ systems is less widespread in HUS than in TTP. The hemolytic anemia and associated thrombocytopenia of HUS is typically due to a mechanical microangiopathic processes rather than to some directly immune-mediated hemolytic process such as that causing a Coombs-positive hemolytic anemia. Coombs testing is negative in HUS, and the abnormalities of platelet production that characterize most cases of TTP are not found in HUS.

Laboratory testing for heritable or acquired deficiency of ADAMTS-13 activity permits the distinction of most infantile or childhood cases of TTP from HUS, although the boundary between infantile TTP and HUS is occasionally uncertain. HUS in the elderly remains a condition for which additional pathophysiologic characterization is necessary. It may well be a condition that can be differentiated from HUS, the result of unique pathogenicity. HUS-like illness in the elderly tends to be unresponsive to therapies that are usually effective in childhood HUS. [11, 12]

Postinfectious HUS

IStx-HUS is the largest category of HUS, accounting for as many as 60-75% of all cases of HUS. [13, 14] The 2 important varieties of toxins are Stx1 and Stx2. These are also called verocytotoxins (hence, the alternative designations VT-1 and VT-2) because they may be identified by their toxic effects on vero cells. These toxins were initially identified as products of Shigella organisms, hence the term Stx, although in much of the world, verocytotoxin-producing E Coli (VETC) are the most common cause of postinfectious HUS.

The toxins are usually elaborated by toxigenic bacteria that have been ingested and that become transiently established in the colon. In a few cases, other routes of infection (eg, through the respiratory system) establish transient infections with Stx-elaborating pathogens such as S pneumoniae, the cause of some particularly virulent cases of HUS.

Most cases of IStx-HUS occur in children younger than 5 years and are of gastroenteric origin with associated diarrhea. In developed regions of the Western hemisphere and Europe, 60-70% of all cases of HUS are caused by Stx-producing strains of E coli. In nearly 50% of the cases of IStx-E coli HUS, the O157:H7 E coli serotype is found. This particular serotype was termed enterohemorrhagic E coli (EHEC). However, other enterohemorrhagic serotypes have been subsequently identified, including 026 (25%), 0111 (11%), 0145 (11%), and 0103 (6%). Serotypes 055, 086, 0118, and 0120 together account for less than 1%. [14]

In Argentina and Uruguay, where endemicity of Stx-E coli HUS is highest in the world, the 08, 025, 0112, 0103, 0113, 0145, 0171, and 0174 serotypes are most likely to provoke HUS. Approximately 39% of Argentinian beef cattle are chronically colonized by E coli manifesting these various serotypes. [15, 16]

In industrialized nations, toxigenic E coli bacteria are ingested from a variety of sources (chiefly water, milk, or foodstuffs contaminated with fecal material), from contact with animals or their excreta, or from fecal-oral transmission from human to human. An epizootic reservoir for O157:H7 E coli accounts for the high prevalence of that serotype, particularly in cattle herds and hence in undercooked hamburger. These examples are frequently responsible for HUS in North America. [17] Recently, a worrisome trend of increasing prevalence of particularly virulent EHEC has been identified in evolutionary biological studies carried out in Michigan. [18] These findings offer a possible explanation for recent severe spinach-related outbreaks in the United States as well as outbreaks of severe HUS in Japan.

Nonevolutionary changes in bacterial genome must also be considered as an explanation. A dramatic increase in EHEC-related HUS in Sweden was linked to high rates of infection of beef cattle with EHEC. Investigation supported the conclusion that the increase was due to importation of beef cattle harboring EHEC in their guts. The prevalence was 15% in imported cattle as compared with 1% in domestic beef. [19] This suggests the importance of public health measures to identify infections in imported beef.

The finding that as many as 80% of household contacts of children with HUS have circulating Stx supports the feasibility of human-to-human transmission. Factors beyond mere intestinal acquisition of a toxigenic strain of E coli appears to regulate susceptibility to HUS. Shedding of toxigenic E coli may persist for weeks in humans or in animals who have acquired the organism, whether or not they develop diarrhea or other features of HUS.

Ingested toxigenic E coli multiply in the colon. In 38-75% of exposures, cramping and diarrhea ensue after a mean latency of 3 days after ingestion. The diarrhea is initially nonbloody. However, in 70% of patients, it becomes bloody in 1-2 days, and it may be associated with vomiting. Of note, in the remaining cases of HUS associated with Stx-E coli, no premonitory diarrhea is observed. In diarrheal cases, large-bowel inflammation occurs, and submucosal hemorrhages often develop, especially in the ascending and transverse colonic segments. In some instances, toxigenic E coli, particularly the O157:H7 serotype, may induce hemorrhagic colitis without ensuing HUS. [13, 20, 21]

Particular proteins play roles in establishing intestinal infection and in ensuing inflammation, entry into the bloodstream, adhesion to circulating cells, and transition to binding at the sites of intimal injury in the kidney and elsewhere. E coli O157:H7 produce an adhesion intimin in addition to Stx1 and Stx2. Intimin mediates attachment of the ingested organism to colonocytes. Stx2, in turn, mediates attachment to cell surface globotriaosylceramide (Gb3) receptors of other cells, specifically those of polymorphonuclear cells (PMNs), monocytes, erythrocytes, platelets, and endothelial cells.

Stx2 is the more toxic of the 2 Stxs, and it is the toxin most likely to account for renal disease. Exposure to the Stx1 toxin alone may provoke diarrhea without associated renal disease. Clinical series show a 55-70% likelihood that acute renal failure will follow E coli verotoxigenic colitis in children.

Attachment to circulating PMNs appears to be especially responsible for distributing the Stx toxins throughout the body, with particular apparent tropism for attachment to endothelial receptors in kidney. [22] Stx2 binding to leukocytes is of relatively low affinity and permits reattachment to other cell surfaces, particularly those of the kidney [23] ; this produces renal dysfunction in susceptible individuals.

Stx2 binding is particularly likely to occur in blood vessels of the distal convoluted tubules, especially those adjacent to glomeruli and collecting ducts. Selection of this site in children but not adults may have something to do with age-related expression of endothelial Gb3 receptors in this particular anatomic location. [24] This particular regional vulnerability may explain the tendency of HUS TMA to manifest limited organ-system confinement compared with TTP.

Stx1 and Stx2 are made of one A and two small B subunits. One of the B subunits mediates binding to bowel, and a Stx B subunit mediates binding to kidney Gb3 endothelial receptors. On binding, the A unit is internalized and disrupts endothelial function by inhibiting protein synthesis. The development and use of specific antibodies to the A subunit to prevent HUS are among areas of considerable research interest. The disruption of cellular function mediated by the A subunit is the likely proximate cause of injury to the colonic wall and the renal glomerular endothelium. [25, 26]

The microvascular injury in the kidney appears to provoke a procoagulant state that may not have resulted merely based on injury to vascular endothelium of the colon. Individuals without substantial renal involvement do not appear to manifest this same procoagulant picture. Onset of the procoagulant state occurs early in the course of HUS and is indicated by marked elevation of serum thrombomodulin. [27] Therefore, this elevation is likely a valuable laboratory indicator of the onset of renal disease. In the converse, decreasing levels of serum thrombomodulin indicate the onset of recovery or renal function.

The marker for the renal phase of HUS pathogenesis is endothelial swelling in the blood vessels of the renal distal convoluted tubules, which is worsened by the accumulation of proteinaceous material and cellular debris in the vascular subendothelial space. Thrombus formation ensues. In Stx-E coli –related HUS, the fibrin deposition in clots is explained by activation of both prothrombin peptide F1+2 and increases the prevalence of the D-dimer before the microangiopathic stage develops. [28]

Flow obstruction likely participates in the formation of the erythrocyte-rich fibrin clots. These clots may entrap some platelets, but the rate of platelet loss due to this mechanism is far lower than in TTP, in which typically defectively cleaved platelets constitute a principle element of clots. The pathologic changes of Stx-E coli HUS are typically confined to capillary subendothelial spaces in the region of the distal convoluted tubules, although in particularly severe cases, thrombi show anterograde propagation into renal arterioles.

These pathogenic events render portions of the nephron ischemic and compromise function. They also produce a rheologically unfavorable situation that promotes red blood cell shearing with schistocyte formation. [29]

HUS-associated thrombi are largely confined to the kidney, although they may be found in liver, lung, heart, or brain. Thrombi in such extrarenal organs tend to produce only mild symptoms. This is unlike TTP, where thrombi are found in heart, pancreas, kidney, adrenal, and brain (in decreasing order of severity) and often produce signs or symptoms of their presence in these various organs. [30]

In the most severe forms of IStx-HUS (eg, S pneumoniae or S dysenteriae –associated Stx-HUS), leukocyte entrapment may be seen. HUS thrombi do not contain the von Willebrand factor (vWF) multimers that are characteristic of ADAMTS-13 deficiency TTP. [31, 30, 12]

Neither inherited or acquired deficiency of ADAMTS-13 activity, the defining pathogenetic basis of a considerable number of TTP cases, is a feature of postinfectious or other forms of HUS. [32, 33, 34] Young children with Stx-E coli –associated HUS may have elevated rates of ADAMTS-13 cleavage of vWF multimers. This, in turn, may result in smaller-than-normal rather than larger-than-normal vWF multimers.

The enhanced cleavage in HUS may be due to increased availability of ADAMTS-13 cleavage-mediating receptor sites on the vWF multimers. This may be the result of abnormal unfolding of the multimer receptor site areas due to the increased sheer stress vWF multimers may experience during circulation through regions containing HUS thrombi and capillary or arterial microangiopathy. [29]

Other incompletely understood elements are likely to play roles in vulnerability to Stx-HUS. As many as 82% of the household contacts of a child with IStx-E coli HUS (many of whom have hemorrhagic diarrhea) have Stx bound to their PMNs. Despite this binding, such individuals often do not have evidence of renal dysfunction. Therefore, an additional mediator is hypothesized to be necessary for the development of Stx-HUS; this is presumed to be a particular lipopolysaccharide. [35] Why the risk for HUS in individuals with sporadic Stx-E coli colitis is only 3-9%, while the risk is as high as 20% is some epidemics is unclear.

The observation that the use of an antimotility agent increases the risk for HUS suggests that prolonged contact of organism with colonocytes or with inflammation-associated PMNs may play an important role in pathogenesis. [21]

The administration of antibiotics, such as trimethoprim sulfa, to children with E coli O157-associated diarrhea may increase their risk for HUS. [36, 21] The importance of this observation with regard to pathogenesis of HUS is not entirely unclear, although the possibility that such treatment may increase risk for HUS and the often questionable role of antibiotic treatment of diarrhea have caused many clinicians to avoid such treatment.

Special circumstances must, however, be considered. In rare instances, individuals may harbor not only Stx-E Coli, but also Clostridium septicum. This situation may arise as the result of dirty lacerations (in which case localized gas gangrene may provide an important sign) but may also arise as gut infection in individuals exposed to sheep or individuals with carcinoma of the cecum. Failure to adequately treat individuals harboring such dual infection has been associated with serious complications including intracranial C septicum infection. [37]

Levels of C'3 complement are low in approximately one half of all cases of diarrheal HUS; this finding suggests that activation of the alternative complement pathway occurs in postinfectious HUS.

Acute renal failure, the second and most serious cardinal manifestation of HUS, develops in 50-70% of patients with Stx-induced hemorrhagic gastroenteritis. Acute renal failure is a potentially life-threatening complication that often leads to permanently impaired renal function, the principal serious consequence of Stx-E coli HUS.

Insofar as the cerebral manifestations of HUS are concerned, E coli Stx2 induced cerebral microcirculatory endothelial injury in piglets, suggesting a particular age-related vulnerability. [38, 39] Both arteriolar and capillary thrombi are seen in the brains of 50-75% of individuals who had fatal HUS, and thrombi are found in their livers and lungs.

In industrialized nations, IStx-HUS tends to occur in mid summer. [40] Undercooked hamburger appears to be a major vehicle for food-borne E coli O157:H7 outbreaks in children. One study accounted for as many as 46% of cases in children and suggested an epizootic reservoir. [17]

The other major agent for inducing IStx-HUS is S dysenteriae type 1. It is the major cause of HUS in most nonindustrialized tropical, subtropical, and some temperate zones of the world. Therefore, it may the most important cause of HUS worldwide, and it is further distinguished as the cause of the most severe form of HUS. It is acquired by ingesting bacteria from the various sources likely to contain E coli.

The mechanisms are likely similar to those of IStx-E coli HUS. To these is added the increased probability for prerenal dehydration or septic shock with Stx-Shig1, which likely worsens glomerular ischemia and which leads to acute cortical necrosis of kidney. In addition, the comparatively poor availability of intensive medical care in regions with a high prevalence of Stx-Shig1 HUS accounts for the high mortality rate of 30% and the morbidity seen in this form of HUS. Septic shock also plays a role in S pneumoniae HUS, which may also produce acute cortical necrosis of the kidney.

Other pathogens are associated with diarrhea-associated postinfectious HUS. Among them are toxin-elaborating bacteria such as Salmonella and Yersinia species. Yersinia organisms may provoke Stx-associated HUS of severity rivaling that due to S dysenteriae or S pneumoniae. Viruses may provoke early-childhood HUS with diarrhea or a respiratory prodrome (ie, an INon-Stx HUS). Examples include echoviruses, adenovirus, HIV, or Coxsackie virus.

INon-Stx HUS corresponds to many cases of HUS that were included under the old category of classic infantile HUS. These cases account for 10% of all cases of HUS in industrialized nations and tend to be infants. A febrile prodrome in the classic cases is associated with the development of diarrhea. However, bacterial blood cultures are negative and a Stx-elaborating pathogen is not identified. A viral pathogen is sometimes isolated from appropriate cultures or identified serologically. In other instances, no associated diarrhea occurs.

Identified viral pathogens include echoviruses or Coxsackie viruses, adenovirus, and HIV. Viruses may directly mediate vascular endothelial injury in the kidney, but the process is little understood. Likewise, whether the peculiar age-related and regional vascular susceptibility has the same basis as that which occurs in IStx-HUS is not fully understood. In most cases, HUS tends to be mild and is likely to have a relatively good prognosis. However, severe cases are described.

In children older than 5 years of age and in adults with HUS, an increased degree of glomerular endothelial abnormality is found with HUS, and necrotizing arterial thrombosis rather than capillary microangiopathy may predominate. [25] The disease also has a peculiar tendency to manifest pulmonary thrombosis, which may be seen in early childhood HUS. This is especially prominent in HUS associated with the use of cyclosporin A or various cancer chemotherapeutic regimens. These thrombi tend to entrap leukocytes and erythrocytes, and they may be associated with regional tissue necrosis. [41]

Of interest, in some individuals (especially adults), Stx-E coli O157:H7 infection provokes a TTP rather than an HUS phenotype. [42] Disease in older individuals with HUS is most likely to fit in the sporadic or familial categories of HUS.

Familial HUS

Familial HUS is clinically defined by the occurrence of HUS in 2 or more family members. The occurrence is not usually associated with diarrhea, although diarrhea-associated cases do occur. Cases of sporadic HUS may be reclassified as familial when the process is identified in other family members, although co-exposure to Stx-expressing bacteria such as EHEC must of course be excluded, in which case diarrhea is usually prominent. The first description of familial HUS was provided in 1956, although the careful studies of Kaplan and associates in 1975 made a considerable additional contribution. Familial HUS is thought to account for 5-10% of HUS cases.

Recognition of this entity is important because mortality is as high as 50%, despite modern management. This rate is much higher than the mortality of postinfectious HUS. Thus, severity provides another distinguishing characteristic that may prompt testing to identify the identification of mutations associated with this form of HUS.

Both autosomal dominant and autosomal recessive patterns of inheritance are described. The pathophysiology of familial HUS is less well understood than that of postinfectious HUS. Familial cases are subclassified by whether they manifest (D+) or lack (D-) a diarrheal prodrome. Kindreds may contain individuals who manifest TTP rather than HUS.

Approximately 10-20% of familial or sporadic (D-) cases are associated with mutations in a region of chromosome 1 that encodes for various complement regulatory proteins. Some familial cases have heritable deficiencies of C'3. Others involve deficiencies of liver-synthesized complement factor H (HF1). [43, 44] Cases are subclassified according to whether an identifiable defect in HF1 expression is present. Patients with HF1 mutations tend to have low C'3 levels as well with a mutation in MCP, a surface-bound complement regulator. Factor H deficiency is associated with type II mesangiocapillary glomerulonephritis, a particular type of kidney disease, that may develop with or without an HUS presentation. [45, 46] Defective HF1 expression may be found in affected individuals whose families pass on the disease trait in either autosomal dominant or autosomal recessive patterns of inheritance.

Individuals with either of these deficiencies tend to have HUS of greater-than-average severity. A missense mutation in the gene that encodes factor H was identified in some cases of familial HUS [43] , with the subsequent demonstration of genetic heterogeneity in affected individuals. Furthermore, mutations of this same gene may be associated with familial or sporadic forms of HUS (defined by the absence of family history) without diarrheal prodrome. [47]

Factor H is a fluid-phase regulator of the activation of the alternative complement pathway, which plays a critical role in regulating the discernment of host from foreign tissues. The various missense mutations associated with HUS result in abnormalities in the carboxy terminal of factor H, a region important for binding to C'3 complement receptors and cell-surface polyanionic structures. Early procoagulant activation is hypothesized to occur, as in diarrheal cases, because of injury to the endothelial cells. Dysregulation of the alternative complement pathway, due to abnormal binding function of factor H, then prolongs the abnormal procoagulant state. How defective HF1 expression participates in HUS (or TTP) pathogenesis is not really understood.

Patients with HUS usually have normal levels of factor H with normal or low levels of complement or C'3. [48, 49] A normal factor H level does not exclude mutation of the factor H gene. How many cases of (D-) HUS have demonstrable abnormalities in the factor H gene is unclear. One recent extensive literature review of found factor H gene abnormalities in less than 15% of all nondiarrheal cases of severe HUS in which a renal transplant was required.

C'3 levels are inversely associated with disease severity and outcome in both (D+) and (D-) sporadic or familial HUS. [50, 51, 52] Although HUS occurs in individuals with specific abnormalities of only the carboxy terminus of factor H, complete absence of factor H in pigs, mice, and even humans is not associated with increased susceptibility to HUS. Rather, it portends the possibility of developing mesangiocapillary glomerulonephritis. [53]

Sporadic HUS

A wide variety of stimuli can provoke endothelial injury with an HUS phenotype. Adults, particularly the elderly are at higher risk for sporadic HUS than children. Sporadic HUS includes examples of the former immunologic HUS category that are nonfamilial. It includes HUS with an acute acquired decrease in the concentration of C'3 or a deficiency of H factor activity. [43, 54] A confusion is that some authorities include some individuals who harbor GI Stx-E coli infections [14] or some who have diarrheal prodromes in the sporadic category.

However, most cases in the sporadic category do not have an infectious diarrheal prodrome. Among the most common provocative illnesses are noninfectious vasculitic and inflammatory illnesses, such as Henoch-Schönlein syndrome, systemic lupus erythematosus (SLE), scleroderma, polyarteritis nodosa, and Wegener granulomatosis. In some individuals, these illnesses may provoke TTP rather than HUS. In others, they may result in rapidly progressive vasculitic glomerulonephritis rather than the peculiar glomerulopathies of HUS or of TTP. The elderly are at higher risk for HUS due to these provocations than children perhaps because of an prevalence of these illnesses and their treatments. However, other age-related factors may also be at work.

Other provocations of sporadic HUS are malignant hypertension, kidney irradiation, bone marrow transplantation, immunosuppressants (cyclosporine, tacrolimus, methylprednisolone), snake-venom or diethylene glycol intoxication, and chemotherapy drugs (eg, mitomycin). [55]

Sporadic HUS due to these stimuli usually occurs without a diarrheal prodrome. The identification of additional individuals in a kindred who develop HUS due to these or some other provocations listed causes their HUS to be reclassified as familial rather than sporadic. The reclassification is also true of kindreds in whom familial factor H deficiency is identified.

Tacrolimus-associated HUS, for which renal transplant patients are at risk, tends to arise in adults rather than children. This is consistent with the general rule that HUS tends to be more severe and difficult to treat in adults than children. Tacrolimus-associated HUS occurs idiosyncratically slightly more often in men than women and has a mean onset at about 40 years of age or at about 7 months after receipt of renal allograft. Only 45% of patients improve with various combinations of anticoagulation, use of antiplatelet agents, dialysis, and plasma exchange.

After HUS develops, tacrolimus is usually replaced with cyclosporine. However, in some instances, an initial dose reduction of tacrolimus is tried. Graft loss occurs in 25% of patients. Without successful retransplantation, 100% of patients die. Even with transplantation, approximately one third of patients die. If associated liver failure occurs, 60% die. [56]

The immunologic form of HUS is associated with a decrease in serum concentration C'3, an event that can be detected only after presentation. Other secondary forms of HUS include those associated with SLE, scleroderma, malignant hypertension, kidney radiation, immunosuppression, snake-venom intoxication, diethylene glycol intoxication, or chemotherapy with mitomycin or cyclosporin. Endocrine provocations for HUS include pregnancy and use of oral contraceptives.

Sporadic HUS tends to be associated with greater rates of recurrence and greater prevalence of kidney failure. It is also associated with an increased risk for seizures and other neurologic complications. It tends to be a severe disease and the response to supportive therapy may be poor. Hypertension may be severe in affected individuals. Transplantation after sporadic HUS may be followed by recurrence.

Previous
Next:

Epidemiology

Frequency

United States

The annual incidence of hemolytic uremic syndrome in the United States is approximately 2.2 cases per 100,000 population. The highest incidence in the United States is in children younger than 5 years. Incidence as high as 6.1/100,000 population per year has been estimated for children 5 years of age or younger, although an estimate as low as 1.08/100,000 children less than 5 years of age has been provided. [57] The incidence tends to decline with age, with lowest incidence in adults aged 55-59 years (0.5/100,000 population per year). HUS may be an under-reported disease. In one study, only 43% of identified cases had been reported to public health agencies. This California study additionally found that despite strenuous public health measures, the prevalence of STEC-associated HUS had not changed from the 0.67/100,000 rate. [58]

HUS is associated with verotoxigenic E coli O157:H7, which accounts for nearly half of childhood cases of HUS. This form tends to occur in midsummer, with most cases occurring between June and September. Summertime predominance is likely to be found in most other developed nations located in temperate climates.

Undercooked hamburger is a particularly important source of verotoxigenic E coli O157:H7. Undercooked ground meats processed by using insufficiently cleaned grinders in which beef was previously ground are another course of infection. Milk, water (from drinking, swimming, or tooth brushing), cider, juices, vegetables washed in water, and human excreta are additional important sources of infection. Other sources include deer, sheep, goats, horses, dogs, and birds.

Of note, several population-based studies showed that the prevalence of HUS substantially increased in the 1980s in the United States on the West Coast; this observation may also be true in other developed nations. [59, 13] However, this suggestion was not supported in one careful study. [57]

When these data are considered with the 40-year increasing prevalence of other autoimmune diseases (eg, juvenile rheumatoid arthritis, asthma, SLE, multiple sclerosis in women) in industrialized nations, one might conclude that a common set of influences is disturbing the development of immunoregulation and tolerance. Current research into the genetic and immunoexperiential factors that determine the competence of immunoregulatory T cells is likely to prove relevant to these worrisome observations.

International

E coli-related HUS

Data concerning the prevalence of HUS in many parts of the world are incomplete. Consumption of improperly stored or prepared meats and other foodstuffs in warm seasons or warm climates increases the risk of exposing individuals, especially children, to Stx-producing E coli. This risk is greatest where sanitation is poor.

  • Verotoxigenic E coli, particularly the O157:H7 strain, accounts for at least 75% of all cases of postinfectious Stx-HUS in Western Europe, where the incidence may be on the order of 0.5 case per 100,000 population per year. Incidences in Scandinavia, Switzerland, and perhaps other individual countries may be lower than this.
  • In Japan, the O157:H7 Stx strain of E coli is the most important cause of postinfectious HUS.
  • The incidence of postinfectious HUS prevalence is approximately 5-fold higher in Argentina and Uruguay, at 10.5 cases per 100,000 population per year, than in the United States. This high prevalence is ascribed to an epizootic reservoir in Argentine beef, though the manner in which beef is handled and cooked must also play a role in this high incidence.
  • Stx- E coli are found in approximately 0.6-1.4% of diarrheal stool samples in individuals from Calcutta. However, these organisms are found in as many as 50% of raw-beef samples from the region. They are overwhelmingly non-O157:H7 strains. In such tropical regions, S dysentaeriae is a more important cause of infectious HUS. However, sometimes individuals with diarrheal HUS who are found to have Stx-elaborating Shigella in stool have that organism as well as E coli in blood.

S dysenteriae-related HUS

Data concerning the incidence of S dysenteriae -related postinfectious Stx-HUS is limited. However, in developing nations a very large number of cases of Shigella Stx-HUS likely occur, with an appalling fatality rate. Shigella Stx-HUS may or may not be associated with diarrhea, however. One careful study of tropical Shigella -Stx HUS noted diarrhea in 68% of cases and similar rates of mortality (55%) whether or not diarrhea was present. Among individuals with diarrhea, 16% had neurologic abnormalities. Shigella Stx-HUS without diarrhea tends to have lower hemoglobin and platelet counts than Shigella Stx-HUS with diarrhea. [60]

  • In developing nations, approximately 10 million cases of diarrhea occur in children younger than 5 years. About 1 million of these children develop dysentery (bloody diarrhea), and approximately 100,000 of these children have Shigella infection. How many of these children develop postinfectious HUS is unknown.
  • Travelers' diarrhea particularly occurs in individuals who have visited tropical countries. Travelers' diarrhea represents an important potential source of sporadic outbreaks of postinfectious Shigella Stx-HUS when these individuals travel to developed nations.
  • Between 1993 and 1998, about 5% of individuals returning to Barcelona with traveler's diarrhea harbored enterotoxigenic Shigella species, chiefly Shigella flexneri or Shigella sonnei. In approximately 20%, the Shigella organism could elaborate Stx1, the toxin most likely to produce severe dysentery, bacteremia, shock, disseminated intravascular coagulation (DIC), or HUS. Clustered cases of Shigella Stx-HUS traceable to an index case of traveler's Shigella dysentery appear to be largely due to person-to-person (fecal-oral) transmission.
  • North Africa is a region where the risk for severe Shigella Stx-HUS, Stx-sepsis, and shock is increased in children younger than 5 years.

Mortality/Morbidity

Mortality

When originally described, the mortality rate of menolytic uremic syndrome (HUS) was 50% or greater. Improved supportive therapy, including transfusion; dialysis; and careful management of fluids, electrolytes, and hypertension, where such approaches are readily available, have significantly reduced the high mortality rate for children with HUS.

Since the 1970s, the acute case-mortality rate of HUS (including all subtypes) in developed nations has been approximately 5-10%. A California study of patients hospitalized with HUS showed an acute phase mortality of 2.7%. [58] Comparable data on children with familial HUS shows an acute phase mortality rate of 5% or higher. The mortality rate for early childhood Asian or African S dysenteriae HUS may be as high as 30-55%. Severe hyponatremia was identified as a factor predictive of higher mortality in the Kenyan series. [60] Non–Stx-HUS may have an acute mortality risk as high as 25%, even in developed nations.

Mortality and morbidity rates are distinctly greater in children who develop HUS after a prodromal respiratory illness without GI disturbance than in those who develop HUS after a diarrheal prodrome. Children who have neurologic signs in association with HUS are most likely to die or to have residual hypertension or chronic renal dysfunction.

Adults account for a large percentage of non–Stx-HUS cases. They are most likely to have HUS as a secondary complication of a serious underlying systemic disease; for this reason, the adult case-mortality rate remains higher than that for children.

In tacrolimus-associated HUS, which is chiefly a disease of adults with kidney allografts, graft loss occurs in 25% of all patients. Without successful retransplantation, 100% of these individuals die. Even with successful retransplantation, approximately one third die. If associated liver failure is present, 60% die. [56]

Morbidity

The chief morbidity of HUS is chronic renal failure. In the United States, HUS is the leading cause of acquired renal failure in children. Various degrees of permanent renal injury occur in approximately one third of all cases of HUS. Individuals, usually children, who develop HUS after an S pneumoniae or S dysenteriae infection are most likely to develop severe kidney dysfunction and end-stage renal disease due to the renal necrosis and severe glomerulosclerosis.

Only 45% of adults with tacrolimus-associated HUS improve, despite treatment with various combinations of anticoagulation, antiplatelet agents, dialysis, and plasma exchange. Graft loss occurs in 25% of patients.

HUS accounts for approximately 7% of all cases of hypertension in infants younger than 12 months.

Race

No definite racial predilection for HUS has been identified beyond the elevated risk sustained by individuals of particular ancestry whose standard of living or place of residence may account for that elevated risk. One study found that white individuals were more likely than black individuals to be hospitalized for their HUS. [57]

Sex

See the list below:

  • Some data suggest that girls are at greater risk for sporadic postinfectious HUS than boys are. One study found that among children younger than 5 years, girls are more likely to be hospitalized for HUS than boys are. [57]
  • Predominantly an adult disease, tacrolimus-associated HUS occurs slightly more often in men than in women.

Age

Two thirds of all cases occur in children younger than 3 years, and few cases occur after 5 years of age. [14] HUS occurs less commonly in neonates than in children.

  • HUS may occur in adults (especially in elderly adults), usually as the result of an identifiable provocation.
  • Tacrolimus-associated HUS has a mean onset at about 40 years of age or about 7 months after the receipt of a renal allograft. Other factors governing this apparently idiosyncratic medication reaction are not well understood.
  • In the elderly, the pathogenesis of HUS usually differs from that of HUS in childhood. Elderly individuals respond relatively poorly to support and management that are effective in childhood cases. [11, 12]
Previous