Hong Kong Journal of Paediatrics [HK J Paediatr (New Series) 1996;1:151-159]

Royal Hospital for Sick Children, Edinburgh, UK

William Uttley MB, FRCP(Ed), DCH

Correspondence to: Dr William Uttley

Received May 20, 1996

Keyword : Acute renal failure; Haemolytic uraemic syndrome

Mr President, Professor Yeung, Ladies and Gentlemen I am very conscious of the honour done me in being invited to address this meeting and deliver the 7th James H Hutchison Memorial Lecture, particularly so having noted my illustrious predecessors, which include my own one time mentor, Professor JO Forfar as well as some of my long-standing friends.

I am extremely pleased to be reminded of the great esteem in which Professor Hutchison was held in Hong Kong and I know that these feelings were reciprocated. My own contact with Professor Hutchison went back many years, even though I was educated at a Medical School somewhat to the East of Glasgow with which intercourse was sometimes frowned upon. I recall with fondness being entertained in the Hutchison household in 1967 at the Foundation meeting of the European Society for Paediatric Nephrology when I was accompanied by my wife, who just happens also to be a Glasgow medical graduate.

One of the apparently new and certainly topical conditions at that meeting was that of the Gasser syndrome or the haemolytic uraemic syndrome (HUS) and indeed one of my own early papers had discussed the coagulation abnormalities seen in these children.

In 1972, I was appointed Consultant Physician at the Royal Hospital for Sick Children in Edinburgh with a special interest in renal disease. Professor Hutchison produced the third edition of his book in eight years -dedicated as you see to his wife - but reflecting a minimal interest in nephrology, still talking about Bright's disease and uraemia but without specific mention of acute renal failure and certainly nothing relating to the haemolytic uraemic syndrome.

Acute Renal Failure

Acute renal failure can be defined as a failure to maintain biochemical homeostasis within the body and includes high output renal failure. Nevertheless, oliguria is far more common and can itself be defined as urine production less than 300 ml/sq m/day or less than 0.8 ml/kg/hour.

Pre-renal failure which is generally due to reduced renal blood flow but where the kidneys are still reacting normally and producing a urine of high overall concentration or osmolality, but of low sodium content.

Intrinsic renal failure which infers an established failure of kidney function which may be due to arterial, glomerular or tubular causes.

Post-renal failure is essentially a synonym for an obstruction within the urinary tract, the inference of which is to require an ultrasound examination of the kidneys and urinary tract wherever renal failure occurs.

A list of the major causes of intrinsic renal failure will show that the haemolytic uraemic syndrome features as the major cause of those cases where arterial abnormalities predominate.

Where glomerular lesions predominate acute forms of glomerulonephritis and again the haemolytic uraemic syndrome are responsible for most of the cases.

A varied aetiology which includes hypovolaemia and shock, nephrotoxins, drugs and some exanthems is responsible for tubular lesions where the histological findings may be classified as nephrotoxic, acute tubular necrosis (ATN), interstitial and those cases where the microscope is unhelpful.

The Haemolytic Uraemic Syndrome

A better delineation and recognition of the haemolytic uraemic syndrome and a generally accepted increase in incidence has now resulted in it being accepted as the most common cause of acute renal failure in children in the Western world at least, and it is most certainly a major problem for practising paediatric nephrologists and paediatric intensive care units.

Major milestones in its developmental history go back to Moschcowitch (1925) who gave the first description of the somewhat related thrombotic thrombocytopaenic purpura, to Gasser et al (1955) who describe the classical triad of haemolytic anaemia, uraemia and thrombocytopaenia, to Riley et al (1983) and the recognition of the association with E. coli O157 and finally to Karmali et al (1985), who described the production of verotoxin by these organisms and the similarity of the verotoxin to the Shiga or Shiga like toxins (SLT) as seen in bacillary dysentery.

1994 saw a consensus statement by the American Digestive Health Foundation, and 1995 a report on VTEC itself by the Advisory Committee on the microbiological safety of food published by HMSO.

The crucial pathological characteristic of HUS is the thrombotic microangiopathy (TMA) which is to be found in each of the two major sub-groups, diarrhoea positive and diarrhoea negative.

I will not dwell beyond this on the diarrhoea negative cases which are the least common and the more severe, which may be familial and are often recurrent. Some of these diarrhoea negative cases may result from generalised infections such as pneumococcus which can involve a different trigger, such as endotoxaemia or the unmasking of cryptic or T antigens.

I shall concentrate on the diarrhoea positive cases, which account for the vast majority i.e. in excess of 95% of cases seen and most of which are subsequent to a bowel infection with enteropathic E. coli.

A broad outline of the pathogenesis of the diarrhoea associated cases will comprise a cascade from a small innocculum of usually but not exclusively E. coli O157:H7, capable of excreting a verocytotoxin, and hence known also as VTEC. The verotoxin is so called because of the damage it causes to vero monkey cells in culture and is synonymous with Shiga like toxin. The toxin is absorbed onto receptors in the endothelial lining of blood vessels and there will initiate the production of inflammatory mediators which damage the endothelial lining with resultant vasculitis and thrombotic microangiopathy.

In fact VTEC produce two structurely and functionally similar, but immunologically distinct toxins, SLT 1 and SLT2. Both bind to the glycolipid receptor, globotriosylceramide or gb₃.

The receptors are differentially expressed in blood vessels throughout the body but are particularly abundant in renal endothelium. It comes as no surprise therefore that the kidney should be so vulnerable in this condition. Clearly however other major organs as well as blood and platelets will be affected but also one must remember the exacerbation of the inflammation within the bowel wall where receptors are also present.

The observation that the P1 blood antigen also contains the same galactose ceramide linkage as the gb₃ concentrations receptor has aroused much interest in those concerned with the pathogenesis of the syndrome on the basis that toxin may be "mopped up" by the blood cells. However gb₃ concentrations vary indepently of the P1 antigen.

Before describing the syndrome in greater detail we should perhaps first of all look at the epidemiology of the causative organism.

The basic reservoir of the organism appears to be in the intestine of cattle and particularly young calves. The majority of these animals are asymptomatic but some will have diarrhoea. Surveys in Canada for example have shown recovery rates in stool varying from 8% to 19% (Wells et al, 1991) whilst carcase infection was found in 8% to 30% af animals in an abattoir in Sheffield, UK. (Chapman, et al. 1993)

Infection in man is seasonal, the greatest incidence being seen in the summer. Despite an apparent overall increase in isolates the number of cases seen also fluctuate from year to year in various communities, probably due to a tendency to mini epidemics. The first clusterings were described as long ago as 1962 by Lamvik in Canada and Gianantonio et al in the Argentine.

The highest rates of infection in man are currently observed in the UK and North America and the average incidence in Scotland, approximately 4 cases per 100,000 population per annum overall is exceeded only by that of Canada, particularly the province of Alberta! Indeed when we look at the incidence in Scotland there is a further disparity between the east coast and the west coast for reasons which are not entirely clear, but seem to have nothing to do with the reporting or processing of potentially infected samples.

Infection is not currently notifiable in the UK but a voluntary reporting system does exist amongst the laboratories themselves.

Provisional figures from the Public Health Laboratory Service showed that more than 1000 people were infected in Britain during 1995, 247 in Scotland with 10% of the total UK population and 792 in England and Wales. It is of course the very young and the very old who are the most seriously affected, and the most likely to progress through haemorrhagic colitis or bloody diarrhoea to acute renal failure. Caucasians seem more vulnerable than Blacks in this regard. It is known that many cases may remain asymptomatic but, of those developing bloody diarrhoea, some 50% of children will develop renal symptoms and perhaps half of these develop established renal failure which with associated complications has a fatality rate of some 5-10%.

Transmission from the bovine source to man requires but a small inoculum of no more than 10 - 100 bacteria in the most susceptible Ground beef, generally in the form of hamburgers, has been the most notorious vehicle, but coliform transmission has also been described through pigs, poultry, milk, cider and indeed water. Secondary transmission from person to person is well recognised in epidemics.

The best known and largest epidemic occurred in west coast United States between January and May 1993 when more than 700 people caught the infection from undercooked hamburgers and 4 died.

The worst British outbreak occurred in 1994 in West Lothian just to the west of Edinburgh when more than 100 people developed intestinal symptoms from the ingestion of milk sold by a dairy with an ineffective pasteurisation process. This was shown to be due to infection with O157:H7, phage type 2 and organisms were recovered eventually from piping within the dairy and from the faeces of cattle at one of the supplying farms. The eventual outcome of this particular epidemic showed that half the affected subjects were under the age of 15 years, of which one-third were hospitalised, 9 developed haemolytic uraemic syndrome, 6 required dialysis for acute renal failure and 2 remain in end-stage renal failure. 1 child died at home at the onset of the outbreak and 1 elderly adult was treated for thrombotic thrombocytopaenic purpura.

Here I concentrate on the O157, but must remind you that other serotypes may be found. Thus for example O111, O103 have been reported in France and O22, O26, O55 and O111 from Germany as well as the apparently ubiquitous O157.

There is no doubt the epidemiological studies have been significantly aided by developments in bacteriological techniques.

In contrast to most E. coli isolates of human origin E. coli O157 does not ferment D-sorbitol within 24 hours, and sorbitol McConkey agar (SMA) has been developed as a simple rapid screening medium for these organisms.

The sorbitol non-fermenters are then subjected to various further tests including O-agglutination serology, O157 latex agglutination and H-agglutination serology. Verotoxin (VT) production can be demonstrated by testing strains for cytotoxic effects on vero cells. VTEC may also be detected by hybridisation with specific VT1 and VT2 DNA probes.

Phage typing is particularly useful for epidemiological investigation. The West Lothian outbreak previously alluded to as resulting from infection with Phage type 2 (Sharp et al 1994) has now reached the courts, a prosecution has been successful and compensation is being negotiated. Phage typing is therefore of great importance in a legal context, where fault can be shown to have occurred in the preparation and distribution of food.

Clinical Features

The prodrome is characterised by diarrhoea which can be severe, prolonged and which often leads to significant clinical dehydration and shock. However, this relatively conventional diarrhoea is soon superseded by the appearance of bloody diarrhoea as the organism takes hold, releases its toxins and initiates what is essentially a haemorrhagic colitis not dissimilar to ulcerative colitis.

The child suffers severe colicky abdominal pain which generally settles within the first week but may persist, in my experience for several months.

Ultrasound examination of the bowel demonstrates thickening due to oedema, findings which may be confirmed at laparoscopy. This haemorrhagic colitis is therefore not at all similar to a conventional rota virus type of gastro-enteritis with vomiting and diarrhoea, but can proceed to an ischaemic colitis with the potential for stricture, perforation and major haemorrhage.

Prodomal symptoms, which also include lethargy, irritability and low-grade fever, precede the classical triad of the developed phase for anything between 1 and 14 days but generally around 1 week. The exact timing of onset may be difficult to determine by the parents or indeed by the family doctor in a child who has already been having substantial problems.

Haemolysis is thought principally to be of microangiopathic origin producing signs both of anaemia and jaundice which indicate the establishment of the syndrome. However, haemoglobin concentrations vary enormously from the barely abnormal to levels as low as 3 or 4 g/dl. Ongoing haemolysis is present in most cases and repeat blood transfusions may be required. Not all cases with established anaemia may in fact go on to significant oliguria but conversely low grade haemolysis may persist for several weeks after the immediate critical phase has died down.

Laboratory indicators of microangiopathic haemolytic anaemia incorporate the usual concomita of an haemolytic anaemia, but the characteristic and often diagnostic feature is the fragmentation of red blood cells seen on the routine film. This is due to the physical breaking up of the red cell envelope as the blood is driven through the mesh of fibrin strands which have been laid down in the smaller blood vessels as a result of the initiating endothelial damage.

There is evidence also that direct injury to red blood cells is caused by the Shiga like toxins.

The degree or even the existence of oliguria may be difficult to determine and if there is any doubt about the position then short-term catheterisation is indicated. The overall degree of hydration will vary dependent upon the relative amounts of fluid lost through preceding diarrhoea and received through subsequent feeding or medical rehydration. Hypertension is common and may exacerbate the tendency to seizures which are already present due to electrolyte imbalance and cerebral oedema. Biochemical aspects of acute renal failure are again conventional with elevations of creatinine being a particularly useful indicator along with urea, phosphate, uric acid and hydrogen ion along with reductions in bicarbonate, and frequently sodium.

Ultrasound examination of the kidneys and urinary tract at this stage is helpful. The characteristic feature is of a marked increase in echoity of the kidneys, and some idea of renal blood flow can be obtained from Doppler studies. Other diagnostic avenues may be closed by this examination.

I personally feel that renal biopsy has nothing to offer in in respect of management at this or indeed any stage of the illness although it may have a role in research. Once a diagnosis of oliguria has been reached critical decisions in management are required. The duration of the oliguria will not necessarily be predictable at the onset and in view of the potential complications there is general agreement as to the benefits of early dialysis, usually by the peritoneal route.

Thrombocytopaenia defined by a count below 150 x 10⁹/l. is present in 95% of cases and very low levels are often achieved. Bleeding from thrombocytopaenia has not been the major clinical problem in my experience. There seem to be very few clinical correlations with degree or duration of thrombocytopaenia but its resolution can be reassuring during management. Thrombocytopaenia probably results from a combination of platelet trapping and involvement in local intravascular coagulation and the removal of damaged platelets by the reticuloendothelial system.

Potential Complications

Before moving on to management in detail I would like to outline a number of complications which may arise over and above the simple triad I have just considered, and which underline the wide involvement of organs other than the kidney in this syndrome.

The bowel, as we have seen is the initial site of contact with the invading organism and bowel complications may arise after as well as before the establishment of the syndrome. Diarrhoea is superseded by haemorrhagic colitis presumably due to the local effect of toxin. There is oedema of the bowel wall, ischaemic necrosis and ulceration, which may be completely transmural. Perforation is a distinct possibility and at subsequent laparotomy large segments of bowel may have to be removed. In some ways peritoneal dialysis can be reassuring as to the state of the bowel wall. When severe inflammation is present the peritoneal dialysate becomes quite yellow in colour and proteinaceous, but perforation is indicated by the appearance of solid faecal material and even bubbles within the dialysis system.

Longer term complications include large bowel obstruction from stricture due to healing fibrosis of ischaemic areas. Occasionally not only the large bowel is involved but the small bowel also, and two of our own cases experienced long-standing diminution of intestinal activity which resulted in prolonged vomiting, incompetence of the cardia and a need for fundoplication. The implications of severe bowel disease vis a vis nutrition can be imagined.

Hypertension can be a major problem, sometimes not developing until well into the anuric phase. In the main it is due to hyperreninaemia but can be contributed to by an increased blood volume should fluid balance management go awry. Conversely hypotension may result from over-enthusiastic removal of fluid but is a worrying development if sustained and can be an indication of myocardial damage.

Minor neurological symptoms of irritability and lethargy are to be seen in the majority of children whereas more severe neurological complications can occur in up to 25 or 30% of cases. Seizures are in fact the most common manifestation but do not necessarily have a poor prognosis whereas coma, present in about 5% of children has a particularly poor prognosis with regard to overall survival and death occurs in 75% of these children.

The cause of the neurological features is clearly multi-factorial. Many of these relate to the renal failure itself, such as hyponatraemia, hypocalcaemia, acidosis, fluid shifts due to osmolar changes and of course as we have just seen hypertension. In addition however there is the direct effect of toxin on functioning brain cells, i.e. there is a toxic encephalopathy as is seen for example in cases of Shigella dysentery. There are also permanent neurological lesions such as stroke which are due to vascular occlusions subsequent upon vasculitis and thrombotic microangiopathy affecting intra-cerebral arteries and to venous cortical thrombosis.

Clearly these are very worrying developments and it is in this group of cases that the most invasive modalities of treatment such as plasmapheresis are employed. Cerebral aspects of the haemolytic uraemic syndrome have been reviewed by Morris and Coulthard 1994.

Inflammatory lung disease may result from local vasculitis but regular examination of the lung fields is required to detect pathology due to aspiration subsequent upon vomiting or to the effects of abdominal compression from peritoneal dialysis.

These are frequently found in terms of elevation of liver enzymes but major problems with liver failure do not seem to occur.

On the other hand pancreatic lesions are relatively frequent and often troublesome. Needless to say it is the thrombotic microangiopathy again which results in patchy ischaemic necrosis in the pancreas. Transient sugar intolerance is frequently seen during the acute phase and difficulties in managing this are often compounded by the need to use high glucose containing fluids for peritoneal dialysis. However, hyperglycaemia must be controlled and extra care has to be taken with these children who tend to be quite insulin sensitive.

A permanent diabetes mellitus occurs less frequently but may be seen in many as 5% of patients who require life-long glycaemic control.

The heart may become stressed due to hypertension and to an increased fluid over-load. However major anxieties in relation to cardiac function occur when prolonged hypotension develops indicative of ischaemic cardiac damage. This is another particularly bad prognostic group and the heart does not respond well to conventional inotropic therapy. Needless to say this is another subgroup of the children which have to be managed in paediatric intensive care environment.

Management

Being alert to the possible complications of the syndrome we can now progress to look at management.

Previous treatments which have not been shown to be effective include anticoagulants, fibrinolytics, anti platelet therapy, prostacyclin infusions, intravenous infusions of immunoglobulin and of fresh frozen plasma, and the administration of vitamin E.

There is however a suggestion that plasmapheresis has a role to play. This is especially so in adults with thrombotic thrombocytopaenic purpura and in children with severe and principally neurological manifestations.

There is also controversy regarding the role of antibiotic treatment of the E. Coli in the early phase of the bowel infection. The amino quinolones including ciprofloxacin and nalidixic acid have some activity against these organisms but as yet there is little positive evidence of benefit, some of potential harm and the truth of the matter is that probably there is no effect on progression of the condition. (Ostroff et al 1992)

Initial management, however is symptomatic and directed towards the major presenting symptoms.

This is required for the management of anaemia. The use of packed red blood cells limit the strains on an already increased blood volume and there should be no hesitation in the utilisation of repeated small volume infusions necessary to maintain the haemoglobin above 9.5 or 10 g/dl.

There is a theoretical contra-indication to the use of platelet transfusions on the basis that these will provide material for further intravascular platelet aggregation and micro-thrombus formation. Clearly, however invasive procedures such as central line insertion or laparotomy will require platelet cover.

Evidence of a consumption coagulopathy is found in some cases along with a rise in fibrin degradation products but again replacement of coagulation factors is rarely required.

There will inevitably be a short interlude between diagnosis and initiation of active treatment for renal failure.

Scrupulous attention is required to fluid balance, remembering particularly the ongoing fluid losses from residual diarrhoea, vomiting and any urinary output. Insensible water losses within a hospital environment might amount to as much as 25 ml/kg/day. Dangerous hyperkalaemia can be managed with intravenous insulin and glucose treatment - 1 unit/4 g of glucose. Calcium resonium is messy and rarely retained.

There seems little doubt that the early use of dialysis therapy is beneficial and steps to initiate treatment should be taken once oliguria has become established along with uraemia, rising creatinine, fluid retention or hyperkalaemia.

Our own practice is to arrange for surgical placement of a soft silastic Tenchkoff catheter under general anaesthetic, at which time a partial omentectomy is carried out to reduce the risk of catheter blockage (Pumford et al 1994). The anaesthetist should also insert a double, or preferably triple lumen line to facilitate blood withdrawal for biochemical and haematological monitoring, for fluid administration and for measurement of central venous pressure. Commencement of dialysis cycling is determined by clinical criteria but a 24 hour delay to allow embedding of the catheter is desirable if possible. The extra flexibility allowed by manual dialysis has distinct advantages in the early phase. It is often necessary to remove quite substantial amounts of fluid from the patient dependent upon the degree of previous oedema, the need for blood transfusions and for intravenous feeding.

In my experience early dialysis results in a generally better condition of the patient, allows for improved nutrition and I think can be effective in reducing the confusion and toxic encephalopathy present in many cases. There is theoretical evidence that plasminogen activator inhibitor is also removed with a possible improvement on outcome. Certainly there is no one particular biochemical index apart from severe hyperkalaemia, which in itself would indicate the need for dialysis. Other indicators may be present to suggest that a prolonged period of anuria is likely, and it should also be remembered that the early phase of this form of renal failure is quite catabolic and the rate of increase in uraemia and acidosis can be rapid.

The occasional embarrassment caused by the unnecessary placement of the catheter can easily be justified by having been prepared!

More or less continuous peritoneal dialysis will be required for the first few days of treatment, but thereafter progression to overnight cycling dialysis is often possible.

Plasmapheresis involves a plasma exchange and is a means of removing antibodies, presumably some toxins and immune complexes. Objective evidence of benefit for children is hard to come by but it is almost universally employed with beneficial results in adults with thrombotic thrombocytopaenic purpura. A ~multi-centre trial in Germany and Austria appeared to show an improvement in the severity of illness in the acute phase and a reduction in long-term sequelae (Muller-Wiefel 1992). It is best reserved for children in whom predictors of severe disease are present or those with localising neurological abnormality. However, technical problems of access arise, particularly when it proves necessary to apply daily exchanges throughout the period of leucocytosis, thrombocytopaenia or adverse neurological features. For example thrombosis of a 7-French gauge double lumen Medcorp or Cook catheter in the femoral vessel of a 10 kg child can occur with even the most careful attention to detail.

Every centre treating significant numbers of these children require facilities for haemofiltration or haemodiafiltration. This is a technically well proven procedure which in small children requires a venous system with a pump in the extracorporeal circuit ideally with a double lumen vascular cather. Again for reasons of access, need for expert nursing staff, expense and risk of bleeding it need only be applied in cases where peritoneal dialysis is no longer possible. These instances generally follow major abdominal procedures for infarcted or perforated bowel. Removal of fluid is certainly more rapid by this technique but this is not always desirable, and the efficiency of peritoneal dialysis for removal of hydrogen ion and potassium is sufficient for virtually all cases.

An early acceptance by medical staff of the probable need for total parental nutrition (TPN) is required. A combination of lack of appetite from uraemia and a marked tendency to vomit from motility problems of the gut often precludes the possibility of adequate oral or nasogastric feeding. It is important to work up as rapidly as possible to around 110% of normal total energy requirement and to compensate for protein removal by PD by enhancement of the protein intake to a minimum of 1.5 g/kg/24 hours.

The development of hypertension should be anticipated. All patients should have their blood pressure monitored at least two hourly in the acute phase and this is best done by using automatic equipment. Background hypertension should be managed by hydralazine or labetalol. There may be surges of blood pressure increase. We have found that the liquid content of a 10 mg nifedipine capsule given sublingually to be very effective in this regard in a dose of 0.5 mg/kg. Duration of action is short and repeat doses are given as required.

There are good theoretical reasons why long-term established hypertension should be treated with an ACE inhibitor such as captopril.

Needless to say, an increase in the circulating blood volume must be prevented. The clinical give-away for this is a rising pulse rate associated with a triple rhythm but the X-ray features of pulmonary oedema will also be present. Control of hypertension is also particularly important for the prevention of seizures. Should seizures occur, they are best treated conventionally with diazepam and the patient put on a rapid build-up of phenytoin.

The severity of the child's illness and the surrounding evidence of high technology application enhance the fears for their child's survival by the parents. Their anxieties may be compounded by a feeling of guilt once it becomes apparent that an ingested organism was the precipitating cause. The parents should be reassured about this as far as possible and we have found information leaflets for the parents extremely helpful. The parents should be incorporated in the care of the child as far as possible, and arrangements made for them to stay with the patient at least in the short-term. The services of a psychologist are often helpful in assuaging parental guilt, and later the anger which they feel towards people responsible for environmental control, and often eventually to the medical and nursing staff who seem to be ever the carriers of bad news, and the vehicle whereby further invasive, painful and frightening techniques are being applied to their children. Children with significant extra renal complications in addition to those with established chronic renal failure may spend many months in hospital and family management by medical staff can prove extra difficult in these cases. There is also a substantial financial penalty applied to the families in terms of travelling expense, the time off work, care for other members of the family, etc.

Initially it must be remembered that patients, their parents and their siblings are all potential sources of infection, so the patients must be barrier nursed and toileting facilities for their parents separated from that of other visitors to the unit. Regular stool samples must be examined bacteriologically until clearances occurs.

Public Health Departments and Environmental Health Officers require to be notified once a clinical diagnosis has been made even before confirmation of VTEC status. This allows family and other close contacts to be screened and geographical pin-pointing of cases in a potential epidemic to take place. Good contemporaneous recording of potential source of infection is required as there is frequently a requirement to produce legal reports when claims for compensation arise. Public Health and Environmental Health Departments require to maintain close contact with laboratories and conversely paediatric units given advance warning of potential epidemics.

Outcome

We now come on to the sort of outcomes which might be anticipated in this condition.

I would like to illustrate these aspects by reference to some of our own data derived over a six year period. Twenty seven children have been analysed, 16 boys and 11 girls. The majority of these were under the age of 4 at presentation and their year of presentation confirms the variable incidence of this condition, most children having been seen in 1994 and 1995.

Firstly there is an association between the degree of leucocytosis and the severity of the illness (Robson 1989). Our studies have confirmed a highly significant negative correlation, r = -0.74 p<0.05, between the highest white cell count and a radio isotope chromium slope determination of GFR obtained between 6 and 12 months after onset.

Although less useful at the time of diagnosis there is also a strong correlation (r =-0.73) between the number of days spent on dialysis in the acute phase and the follow-up glomerular filtration rate.

Finally there is also an inverse relationship between the haemoglobin level at presentation and the long-term outcome.

Controversy exists in the literature about this and the negative correlation has been put forward as evidence for the mopping up of verotoxin by the P1 blood group antigens present in the vast majority of Caucasian children. Thus, a low haemoglobin is said to be associated with high toxin uptake and subsequent red cell destruction, and thereby protection of renal tissue by this competitive binding. We found a less powerful but also significant correlation, r = -0.51 between haemoglobin concentration at the time of presentation and the follow-up GFR, which gives some support to the figures of Taylor (1990) regarding the protective effect of the P 1 blood group.

However, we found that the follow-up glomerular filtration rate in our 21 P1 positive patients (92.4 + 40 ml/min/1.73 m²) was not significantly different from the 5 P1 negative children (80.2 + 47.5 ml/min/1.73 m²).

We were not able to show any difference with age although the extremes of age are said to be the most vulnerable.

Twenty five per cent of the diarrhoea associated children are left with long-term complications and there is a 5% mortality. This compares with a 75% incidence of sequelae in the non-enteropathic cases with a 50% mortality.

Our own figures are quite representative of these described complications classified here into short-term and long-term.

It must be admitted that management thus far has been reactive and supportive. Little of fundamental value seems to be able to be done in the early phases which will prevent progression of the disease in those fated so to do.

Perhaps I could remind you at this stage of the pathogenic cascade which actually results in the thrombotic microangiopathy.

The enteropathic E.Coli produces lipopolysaccharide (LPS) endotoxins on the one hand which can directly damage renal endothelium. On the other hand Shiga like toxins, particularly the B component is taken up by the gb₃ receptors of endothelial cells. The toxin is then internalised by endocytosis, damages ribosomes and induces cell death.

Thus renal endothelial damage from these two mechanisms causes the release of further mediators which result in activation of white cells, of the coagulation cascade and of platelets. Renal vessels are constricted. A thrombotic microangiopathy ensues as we have seen at length.

Current research involves absorption of toxin by a compound, Synsorb-Pk which contains gb₃ whereby the oral administration of which to children in the prodromal phase may reduce disease severity.

Injectible anti toxins raised against the SLTs are also being discussed as is antibody against LPS endotoxin.

Agents which may prevent the eventual development of thrombotic microangiopathy include Oxpentyfylline which inhibits endotoxin induced tumour necrosis factor (TNF) production, ginkgolide mixture which inhibits the effect of platelet activation factor (PAF) and more recently described endothelin and platelet inhibitors.

The problem with current and future treatment strategies is that none of them are preventative, most will be expensive and some will be toxic. Thus primary prevention is paramount and I think it is up to the medical profession to ensure that sufficient pressure is put on to government and agricultural authorities to reduce the bovine source whilst at the same time increasing public awareness regarding these illnesses.

The majority of cattle harbouring the infection are asymptomatic. Improvement in awareness by the agricultural industry and veterinary surveillance must be demanded.

High standards of practice in abattoirs must be in place if unaffected carcasses are to remain uncontaminated from infected material.

The food industry must ensure thorough cooking of meat products of all types, the pasteurisation of milk and the cleanliness of water. Finally increased education of the public must be achieved so that their own standards of food hygiene, particularly in respect of picnics and barbecues where chopped meat products are likely to be used. For example cooking chopped meat products at 70 degrees Celsius for two minutes will eliminate the threat from hamburgers. At a barbecue this would be achieved by heating out the juices and unsuring all pink meat has been cooked out. The possibility of person to person transmission cannot be ignored and vigilance regarding personal hygiene must be maintained by those who prepare food within the families.

President, Ladies and Gentlemen, the haemolytic uraemic syndrome presents a world-wide problem which is on the increase. In children it is the most important cause of acute renal failure and an important cause of chronic renal failure. It is already an important consumer of healthcare resources, calculated in the United States as causing the expenditure of between 10 and 25 million dollars on care during the acute phase with an additional comparative sum required for children requiring treatment for end-stage renal failure.

In the individual child it is responsible for severe extra renal disease in addition to acute renal failure. Many of these children have to be managed in an intensive care situation and often prolonged hospitalisation is required. Enormous family stress and frequently guilt are engendered in the short term whilst in the long term there may be a burden of care for a damaged child. Longterm residua are frequent and may not be apparent without prolonged follow up.

Finally I should warn you that doctors are often drawn into the medical legal context where claims for compensation arise where fault in the preparation or distribution of food can be shown.

Recommended Reviews

Kaplan BS, McGowan KL. The hemolytic uremic syndrome. Curr Opin Infect Dis 1994;7:351-7.

Seigler RL. The hemolytic uremic syndrome. Ped Clin N Americ 1995;42(6):1505-29.

References

Chapman PA, Siddons CA, Wright DJ, Norman P, Fox J, Crick E. Cattle as a possible source of verocytotoxin producing Escherichia Coli 0157 infections in man. Epidemiol. Infect 1993;111(3):439-47.

Gianantonio CA, Vitacco M, Mendilaharzu Jmendilaharzu F, Rutty A. Acute renal failure in infancy and childhood. J Pediatr 1962;61:660-78.

Karmali MA, Steele BT, Petric M, Lim C. Sporadic cases of haemolytic uraemic syndrome associated with faecal cytotoxin end cytotoxin producing Escherichia Coli in stools. Lancet 1983;1:619-20.

Karmali MA, Petric M, Lim C. The association between idiopathic haemolytic uraemic syndrome and infection by verotoxin producing Escherichia Coli J Infect Dis 1985;151:775-82.

Leonard M, Ruchelli E, Kaplan BS. The pathogenesis of typical diarrhea associated hemolytic uraemic syndrome. In International Year Book of Nephrology 1993 ed Andeucci VE and Fine LG London: Springer Verlag 1992:199-216.

Morris KP, Coulthard MG. Renal encephalopathy in Clinical Paediatrics, Coma. ed Eyre JA. Balliere London 1994;2:1109-48.

Moschcowitch E. An acute febrile pleochromic anemia with hyaline thrombosis of the terminal arterioles and capillaries. Arch Int Med 1925;36:89-93.

Muller-Weifel DE, Tonshoff B, Bulla M, et al. The beneficial effect of plasma exchange (TPE) on the clinical course of children with the haemolytic uraemic syndrome (HUS) (abstract) Pediatr Nephr 1992;6:C165.

Ostroff SM, Kobayashi JM, Lewis JH. Infections with Escherichia Coli 0157:H7 in Washington State, the first year of statewide surveillance. J Am Med Assoc 1989;262:355-9.

Pumford N, Cassey J, Uttley WS. Omentectomy with peritoneal catheter placement in children with acute renal failure. Nephron 1994;68(3):327-8.

Riley LW, Remis RS, Helgerson SD, et al. Haemorrhagic colitis associated with a rare Escherichia Coli serotype. N Engl J Med 1983;308:681-5.

Robson WLM, Fick G, Wilson R. Prognostic factors in typical post diarrhoea hemolytic uremic syndrome Child Nephrol Urol 1980;9:203-7.

Sharp JCM, Coia JE, Curnow J. Reilly Escherichia Coli infections in Scotland J Med Microbiol 1994;40:3-9.

Taylor CM, Milford DV, Rose FE, Roy TCP, Rowe B. The expression of the blood group P1 in post enteropathic haemolytic uraemic syndrome. Pediatr Nephrol l990;4:59-61.

Wells JG, Shipman LD, Greene KID, Sowers EG, Green JH, Cameron DN, et al. Isolation of Escherichia Coli serotype 0157:H7 and other Shiga like toxin producing Escherichia from dairy cattle. J Clin Microbiol 1991;29(5):985-9.