Haemolytic uraemic syndrome (HUS), comprising microangiopathic haemolytic anaemia, thrombocytopaenia and acute kidney injury, remains the leading cause of paediatric intrinsic acute kidney injury, with peak incidence in children aged under 5 years. HUS most commonly occurs following infection with Shiga toxin-producing Escherichia coli (STEC-HUS). Additionally, HUS can occur as a result of inherited or acquired dysregulation of the alternative complement cascade (atypical HUS or aHUS) and in the setting of invasive pneumococcal infection. The field of HUS has been transformed by the discovery of the central role of complement in aHUS and the dawn of therapeutic complement inhibition. Herein, we address these three major forms of HUS in children, review the latest evidence for their treatment and discuss the management of STEC infection from presentation with bloody diarrhoea, through to development of fulminant HUS.
- general paediatrics
Statistics from Altmetric.com
Faecal culture, to investigate for E. coli O157:H7, should be performed on all children presenting with acute bloody diarrhoea.
Blood tests (full blood count, blood film, and urea and creatinine) should be performed in children presenting with bloody diarrhoea.
Efforts should be made to ensure adequate hydration in children with confirmed haemolytic uraemic syndrome (HUS) following Shiga toxin-producing Escherichia coli infection (STEC-HUS).
Antibiotics and antidiarrhoeal agents should be avoided in children with suspected STEC.
Intravenous fluid at presentation with STEC-HUS should be considered if there is no evidence of fluid overload.
There is currently no specific treatment that has been shown to improve the outcome of STEC-HUS.
Atypical HUS should be considered in children presenting with HUS without preceding bloody diarrhoea.
HUS may occur in the setting of severe pneumococcal disease.
Haemolytic uraemic syndrome (HUS) describes the triad of microangiopathic haemolytic anaemia, thrombocytopaenia and acute kidney injury (AKI). HUS remains the leading cause of paediatric intrinsic AKI, affecting approximately 1 in 100 000 children a year in the UK, with a peak incidence in children aged under 5 years. HUS following infection with Shiga toxin-producing Escherichia coli (STEC-HUS or typical HUS, formally D+HUS) is the most common cause of HUS, responsible for up to 90% of cases in children.1 Atypical HUS (aHUS, formally D−HUS) describes HUS in the absence of evidence of STEC infection. HUS can also occur as a rare complication of invasive infection with Streptococcus pneumoniae. While these forms of HUS share common clinical features, their management differs significantly. This article seeks to highlight the clinical features of each form of HUS and guide clinicians in their management of children presenting with this rare disease.
STEC-HUS (formally D+HUS)
STEC-HUS has become the accepted term for HUS secondary to infection with Shiga toxin (stx)-producing bacteria. This change in nomenclature from D+HUS/D−HUS to STEC-HUS and aHUS, respectively, has occurred due to an increased understanding of the various aetiologies of HUS2; it also highlights the fact that patients with aHUS may present with preceding diarrhoea. In the UK, STEC-HUS is most commonly secondary to E. coli O157:H7, with an annual incidence of 0.71 per 100 000.3 STEC-HUS is much more common in children compared with adults, and the exact reason for this is unclear; one explanation is the increased incidence of anti-stx antibodies in adults, which may be protective against the development of HUS4; an alternative explanation is a difference in the glomerular expression of the stx receptor (Gb3) in children.5 The mortality rate from acute STEC-HUS is 3%–5%, with long-term morbidity occurring in approximately 30%.3 In addition to E. coli O157:H7, which is responsible for 80% of cases in the UK,3 other Shiga toxin-producing E. coli serotypes (O111, O26, O145, O103, O104) and Shigella dysenteriae have been identified as a cause of HUS, which occur endemically in parts of the world. Cattle are the natural reservoir of STEC; isolation of STEC in bowel flora from cattle varies worldwide from almost 0% in Scandinavia to nearly 66% in Japan.6 There has been a lot of interest in developing a vaccine for E. coli that could be used in cattle to reduce the risk of human infection; to that end, there are now several commercially available vaccines. Vaccination of livestock has the potential to reduce the risk of human infection by up to 85%7. Adoption of these vaccines has been slow due to the inherent difficulty of conducting trials showing a reduction in human infection and economic issues.7
STEC infection is most commonly acquired through ingestion of undercooked contaminated minced beef.8 Other sources of infection include water contamination, direct contact with infected animals (eg, petting zoos and country shows) and person–person spread.8 Contamination of vegetables is an emerging source of E. coli; in 2011 a large German outbreak of HUS was eventually traced to ingestion of contaminated bean sprouts.9
STEC continues to shed from the bowel of an infected individual up to 1 month after the resolution of symptoms, underpinning the need for meticulous infection control and source isolation. Currently in the UK, children aged under 5 years must not return to nursery/school until they have two negative stool cultures 24 hours apart, while children older than 5 years must have 48 hours of normal stool prior to returning to school.10
Following infection with STEC, the majority of children develop diarrhoea, which is often bloody in nature. In 85%–90% this resolves with no further sequelae. In 10%–15% HUS develops typically occurring 1 week after the diarrhoea.11 Children presenting with acute bloody diarrhoea should be investigated for early signs of HUS (table 1). It is vital that robust public health systems are in place to ensure follow-up for patients with positive STEC cultures (blood and/or stool). In the UK, as in many other countries, STEC is a notifiable disease and must be reported to public health services to ensure patients are followed up and reduce the risk of an outbreak. As HUS evolves, with ongoing endothelial and platelet activation, microthrombi formation and progressive renal failure, renal replacement becomes necessary in 50%–60% of cases.
Extrarenal manifestations occur in approximately 20% of children with STEC-HUS, including cardiac (ischaemia, raised cardiac enzymes, overt cardiac failure), endocrine (diabetes), neurological (encephalopathy, seizures, coma, permanent haemiparesis, cortical blindness) and gastroenterological (necrosis, perforation, stricture formation, pancreatitis) involvement; which are associated with increased risk of mortality.
Long-term morbidity is common following the resolution of the acute phase of HUS, and approximately 30% of children show evidence of long-term renal complications including hypertension, proteinuria and decreased glomerular filtration rate.12
Management on presentation with bloody diarrhoea and isolation of STEC
The diarrhoeal prodrome represents a vital period where timely intervention can influence the disease course and outcome. Modifiable risk factors have been extensively investigated during this phase of the disease.
Antibiotics should be avoided if possible; in vitro work has demonstrated that certain antibiotics induce Shiga toxin production and release,13 and therefore may increase the risk of HUS. This hypothesis is supported by retrospective studies demonstrating an increased risk of HUS in children who have received antibiotics during the diarrhoeal phase.14 This effect was not seen in a 2002 meta-analysis,15 although the findings of this analysis are debated.11
Drugs that slow gastric transit, such as loperamide and opioid analgaesia, should be avoided as they theoretically delay colonic excretion of E. coli, and their use has been linked with an increased risk of developing HUS.16
Efforts to maintain glomerular perfusion are vital and therefore drugs, such as non-steroidal anti-inflammatory drugs (NSAIDs), which affect glomerular blood flow must be avoided.
Intravenous fluid expansion during the diarrhoeal prodrome is one strategy that has been shown to attenuate the renal injury and reduce the risk of oligoanuria in children who subsequently develop HUS.17 However, as only 10%–15% of children infected with STEC convert to HUS, this strategy must be balanced against the risk of admitting many children who will not develop HUS.
Management on presentation with HUS
As previously stated, 10%–15% of STEC infected children will progress to HUS. Children with HUS may present with coexisting dehydration due to the preceding diarrhoeal illness. This results in renal hypoperfusion, which will exacerbate the renal injury, elevating serum creatinine and aggravating oligoanuria. To investigate whether hyperhydration could attenuate AKI in children with HUS, Ardissino et al used intravenous fluids to reach a ‘target weight’, calculated as 7%–10% over their working weight (determined by previously documented weight or parental reporting). In this prospective cohort study, there was a significant reduction in central nervous system complications, need for renal replacement and paediatric intensive care unit support, and improved long-term outcomes.18 This strategy must be balanced against the risk of volume overload in oligoanuric children. It is advisable that children presenting with HUS are discussed with paediatric nephrology services. Assessment of volume status is a key part of early management. Children may present with either dehydration (due to diarrhoea) or with volume overload (secondary to oligoanuria). In children with signs of dehydration, cautious rehydration with 0.9% saline should be considered, with regular reassessment to avoid volume overload in the oligoanuric child. In children presenting with signs of volume overload and oligoanuria, intravenous furosemide (often requiring doses up to 5 mg/kg) may be useful in inducing diuresis. Once euvolaemic, management should aim to keep a neutral fluid balance by replacing insensible losses (300–400 mL/m2/day) and measured losses. Ongoing discussion with the nephrology team is advisable, and in the oligoanuric child, transfer to a tertiary unit for consideration of renal replacement therapy is usually appropriate.
Blood products are often necessary during the acute phase of HUS; up to 80% of children require at least one red cell transfusion to counteract the severe anaemia that occurs due to ongoing haemolysis.19 Transfusion thresholds vary between clinicians; however, the ECUSTEC trial (discussed below) uses the following limits, either a haemoglobin <70 g/L or <75 g/L with fall of greater than 20 g/L in the previous 24 hours. The decision to transfuse should be made in liaison with nephrology services due to the risk of hyperkalaemia and volume overload in an oligoanuric child. Early use of recombinant human erythropoietin has been proposed as a treatment to reduce the requirement for transfusion. The results of this are mixed; one small randomised study demonstrated a reduced need for blood transfusion, although there was no effect on the disease course.20 The same benefit was not seen in a larger retrospective cohort study.21
Ongoing platelet consumption commonly leads to very severe thrombocytopaenia. Platelet transfusions are sometimes given perioperatively to increase platelet count for haemodialysis (HD) or peritoneal dialysis (PD) catheter placement. A retrospective study by Weil et al demonstrated no bleeding in children, regardless of whether or not they received platelets, during HD or PD catheter placement despite platelet counts as low as 13×109 cells/L.22 There is concern regarding platelet transfusion in HUS due to the theoretical risk of propagating thrombus formation. The only retrospective trial investigating the use of platelets in children with HUS found no statistically significant evidence of worse disease in children who received platelets. There was, however, a trend towards prolonged need for dialysis in the group receiving platelets.23
Approximately half of children require a period of renal replacement therapy; however, there is currently no evidence to suggest that one modality is superior to another (HD or PD). The choice is, therefore, dependent on the expertise available.
The nutritional requirements of critically ill children are commonly unmet, and this seems to be exaggerated in children with AKI,24 possibly due to concerns over fluid and electrolyte management. Early initiation of adequate enteral feeds in critical illness is linked to a decrease in short-term and long-term mortality.25 This is especially important in AKI as children are in a highly catabolic state, potentially contributing to further metabolic derangement, including uraemia. It is important that specialist dietetic input and initiation of balanced electrolyte feed are sought early.
Numerous therapies have been trialled to affect the disease course including Shiga toxin-binding agent and Shiga toxin-neutralising agent, which have shown promise in mouse models of HUS.26 The only published data of Shiga toxin-binding agents (SYNSORB Pk) failed to show either short-term or long-term benefit.27 Anticoagulants, antiplatelet and thrombolytic agents have been used to reduce the burden of thrombi.28–30 Steroids have been used twice daily to reduce inflammation.31 Intravenous immunoglobulin has been used to try and neutralise Shiga toxin.32 Unfortunately, none of these therapies have shown any effect on the disease course, although it is difficult to draw any conclusions from these studies due to the low numbers investigated and the low event rates (eg, death or end-stage renal failure) in treatment arms.
Plasma therapy with or without exchange has been used in an effort to replace and remove circulating factors leading to HUS, most commonly as a salvage therapy for patients who develop neurological manifestations. Several small observational studies and single case reports have reported rapid recovery following plasma exchange.33 These results have not been seen in other larger series nor in controlled trials.34 35 The practice of plasma therapy in severe STEC-HUS varies widely, and there is no consensus regarding its use; it should be noted that plasma exchange is not without appreciable morbidity and mortality, particularly in children.36
During a large outbreak of STEC, immunoadsorption was used in a small number of adult patients with severe neurological complications with substantial benefit.37 It is worth noting that the patients included in these series had all received prior treatment with plasma exchange and eculizumab (see "Eculizumab in STEC-HUS".
aHUS (formally D−HUS)
aHUS is a much rarer disorder, with an incidence approximately one-tenth of STEC-HUS.38 The initial presentation is similar to STEC-HUS, without preceding bloody diarrhoea. However, compared with STEC-HUS, the prognosis is poor, with death or dialysis-dependent renal failure occurring in 50% following the initial presentation.39 There are many causes of aHUS, including inborn errors of cobalamin metabolism, drugs and non-Shiga toxin infection (secondary aHUS) (summarised in Kavanagh et al 39); discussion of these is outside the scope of this article.
One important differential diagnosis in patients presenting with aHUS is thrombotic thrombocytopaenic purpura (TTP). TTP is a rare diagnosis in children; it occurs secondary to congenital absence of the metalloprotease, ADAMSTS13 (Upshaw-Schulman syndrome), or acquired anti-ADAMSTS13 antibodies.40 The pathogenesis of TTP is significantly different to HUS despite the indistinguishable clinical phenotype. von Willebrand factor (VWF) is secreted as a large multimeric glycoprotein; this is subsequently cleaved by ADAMSTS13, and in the absence of ADAMSTS13, there is accumulation of ultra-large VWF multimers, which bind and activate platelets in the microcirculation. Patients presenting with aHUS/TTP should have measurement of serum ADAMSTS13, and an ADAMSTS13 activity >10% effectively excludes the diagnosis of TTP. aHUS most commonly results from inherited or acquired dysregulation of the alternative complement pathway (primary complement-mediated aHUS). The complement pathway forms part of the innate immune system, providing protection against invading pathogens. The alternative complement pathway (figure 1) is constituently active due to spontaneous hydrolysis of C3 generating C3(H2O), the first step in alternative complement cascade (figure 1). In brief, activated C3 is deposited on cell surfaces, generating an amplification loop that culminates in cleavage of C5 to C5a (a potent anaphylatoxin) and C5b. C5b complexes with C6–C9 to form the membrane attack complex, which punches a hole in the cell surface causing osmotic lysis. To protect host cells from complement-mediated damage, a number of regulatory proteins exist (figure 2) including complement factor H (CFH), complement factor I (CFI) and membrane cofactor protein (CD46/MCP). It is failure to regulate the alternative complement pathway that underlies the pathogenesis of aHUS. Mutations in complement components CFH, CFI, CD46, CFB and C3 have been identified as well as anti-CFH autoantibodies.39
Until recently, the only treatment for aHUS was regular plasma exchange or plasma infusion to remove antibodies and replace deficient complement regulators.41 Response to plasma therapy is variable, depending on the underlying complement abnormality.42 Overall, plasma exchange leads to remission in the majority of children, but over half suffer relapses.42
Eculizumab is a humanised monoclonal antibody against complement C5. It blocks the conversion of C5 to C5a and C5b, therefore preventing the generation of the membrane attack complex. In 2009, the first reports of eculizumab in aHUS were published.43 44 Due to the ultra-rare nature of aHUS, placebo-controlled randomised trials are not possible; however, two phase II trials (trial 1 n=17, trial 2 n=20) demonstrated normalisation of haematological parameters in 88% and 90%, and improved estimated glomerular filtration rate ≥15 mL/min/1.73 m2 in 59% and 40%, respectively, at 2 years. Only one patient died during follow-up, and this death was not related to aHUS45; there were no new episodes of aHUS in patients treated with eculizumab during follow-up. Data on the long-term use of eculizumab, in children with aHUS, show good efficacy and safety profile.46 Eculizumab effectively blocks the alternative complement pathway; patients are therefore at increased risk of infection from Neisseria meningitidis. To reduce this risk, it is recommended that children receive the meningococcal ACWY and B vaccines. Many clinicians also advocate continuous antibiotic prophylaxis (penicillin V or erythromycin if penicillin allergic) for patients treated with eculizumab, although there is no direct evidence for a benefit of prophylaxis. As with many novel biologic therapies, the cost of eculizumab means it is not a viable option in many parts of the world (in the UK, the cost per adult is approximately £340 200 a year). In 2014, eculizumab was approved by the National Institute for Health and Care Excellence, in the UK, for the treatment of organ or life-threatening aHUS, after concluding that ‘eculizumab represents a step change in the treatment of patients with aHUS and could be considered a significant innovation for a disease with a high unmet clinical need’. In the UK, this is coordinated through a national specialist service (National Renal Complement Therapeutic Centre) based in Newcastle upon Tyne, which facilitates rapid access to treatment.
Transplantation has been a challenging area in aHUS. Prior to eculizumab, the outcome following isolated kidney transplantation was poor with recurrence of aHUS post-transplant in 50%–60%.47 Combined liver–kidney transplant is one option to circumvent disease recurrence. This is based on the fact that fluid phase complement regulators are produced in the liver; by replacing the liver, normal complement factors can be produced. Liver–kidney transplantation in aHUS has had mixed results; early attempts were largely unsuccessful, with children developing graft failure in the liver, thought to be due to uncontrolled complement activation secondary to the transplantation process. More recently, the success rate has improved to around 80%.48
Isolated kidney transplantation has become possible due to the use of eculizumab preoperatively and continued post-transplant. Short-term and long-term data show favourable transplant outcomes in patients treated with eculizumab.48 In the UK, the current recommendation is that any child with a medium/high risk of recurrence is treated prophylactically with eculizumab.
It is clear that the outlook for aHUS has been transformed by therapeutic complement inhibition, following successful translational research in this rare disease.
Eculizumab in STEC-HUS
Unlike aHUS, the role of eculizumab in STEC-HUS remains unclear. In vitro studies have demonstrated a role for complement overactivation in the pathogenesis of STEC-HUS.49 This has led several investigators to use eculizumab for severe STEC-HUS. The first reported use was in three children requiring dialysis with severe neurological involvement.50 Eculizumab was administered to all three children, who exhibited reversal of their neurological disease within 24 hours and complete resolution of their renal disease within 3 weeks. This case report was used as the rationale for eculizumab administration to severely affected adult patients during the 2011 German O104:H4 outbreak.34 Closer analysis of this data did not show improved outcome in patients who received eculizumab. There continue to be small case series and single case reports of patients with STEC-HUS successfully treated with eculizumab, raising the possibility that there are subgroups of patients with STEC-HUS who may benefit from complement blockade with eculizumab. There are two current placebo-controlled randomised trials (ECUSTEC (UK) and ECULISHU (NCT02205541 France)), which will examine the role of eculizumab in STEC-HUS. If there is a benefit, eculizumab has the potential to transform the management of children with STEC-HUS.
Pneumococcal HUS (p-HUS)
HUS is a rare complication of invasive infection with S. pneumoniae, with an incidence of approximately 0.5% following invasive infection.51 The acute mortality from p-HUS is higher than STEC-HUS at approximately 10%, although the vast majority of deaths are in children with associated pneumococcal meningitis.52 p-HUS occurs as a result of neuraminidase secreted by S. pneumoniae, resulting in desialylation of glycoproteins on the surface of red blood cells, platelets and endothelial cells. This reveals an element of the glycoprotein that is usually hidden, the Thomsen-Friedenreich antigen or T-antigen. It is proposed that this T-antigen reacts with circulating IgM resulting in p-HUS.53 Interestingly, the T-antigen is also found in infants with necrotising enterocolitis (NEC), where it is associated with increased severity of disease, although the clinical role for T-antigen in NEC is unclear.54
The principle of p-HUS management is supportive with antibiotics for the underlying S. pneumoniae and renal replacement if required (approximately 80% of children require a period of dialysis52). Long-term renal morbidity in survivors is comparable to that of STEC-HUS in that approximately 25% have renal dysfunction at follow-up.52 There is evidence of benefit with plasma exchange; however, this is in the form of single case reports or at best small case series.52 There is theoretical concern about administering blood products containing anti-T IgM; some authors propose using anti-T negative plasma products. Interestingly, since the publication of the last UK data,52 Prevenar 13 has been introduced to the routine immunisation schedule in the UK; of the six serotypes detected in this study, five are contained in Prevenar 13. It is yet to be determined the effect this has on the incidence of p-HUS in children.
The management of HUS has evolved greatly since it was first identified in 1955. The acute mortality rate from STEC-HUS has reduced to less than 5%. However, there is still significant long-term morbidity. There is work to be done if we are to achieve a further reduction in the number of children suffering from these long-term complications. First, efforts to reduce the number of children acquiring STEC through public education and livestock interventions. Second, management strategies to identify children with STEC by ensuring children presenting with bloody diarrhoea are assessed for STEC-HUS, including blood tests and importantly faecal culture. In children with confirmed STEC, there is a vital window where simple interventions to maintain renal perfusion by ensuring adequate hydration and avoiding drugs that increase the risk of developing HUS such as antibiotics, NSAIDs and antidiarrhoeal agents can affect the disease course.
Effective management of established HUS remains challenging. Improved supportive care has led to a reduction in mortality, but there is still a real need for targeted therapies that affect the disease course. aHUS, a disease that only a decade ago caused end-stage renal failure or death in half of patients within a year, now has an effective treatment, but there is significant variation in access to treatment globally due to economic considerations.
Finally, ongoing trials should clarify whether there is a role for complement inhibition in STEC-HUS.
Contributors SJ devised the outline and overview for the article, provided regular supervision during the writing of the article and edited the final version. PRW performed literature review and appraisal and wrote the text of the article under supervision.
Competing interests SJ is a member of the scientific advisory board for the Alexion global aHUS Registry.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.