ӭȨڿ վ


ҽѧӢҽѧڿij塣ӢTheLancet־Hong Kong Medical JournalصԼڿǵ¼:⡢ժҪԡۡлοסӢժҪдʽڱ20034Ѿ־ĵԡۡлͲοһϸ϶ߡ




The most frequent complication associated with plasma exchangewas pulmonary oedemawhich was diagnosed on clinical andradiological grounds in 11 cases.Pulmonary oedema was not confinedto patients undergoing TPE;three of six HUS/TTP cases not treatedwith TPE had pulmonary oedema.Hypocalcaemia(calcium2.12mmol/L)occurred in 15 of the 16 patients treated with TPE.

Although severe(minimum serum calcium 1.32mmol/L)in manycasesintravenous magnesium was given when appropriate and noclinical effects were observed.Other complications associated withTPEwere line infectionwithmeticillin-resistantStaphylococcus aureusand extravasation infusion.




HUS/TTP used to be a rare disease in adultswith an estimatedfrequency of one case per million per year.In 50%of cases it wasassociated with pregnancymalignant hypertensionHIV infectioncanceror chemotherapyand the remainderof caseswere familial orof unknown cause.In 1986 the first association of HUS/TTP withEcoliO157 infection was made and the incidence of the disorder hassince continued to rise in parallel with the global rise inE coliO157infections.After exposure toE coliO157between 3%and 7%ofall patients progress to overt HUS/TTP.The incidence of HUS/TTPis highest in children and elderly people.

The course and prognosis of HUS/TTP differ substantiallybetween adults and children.Children with HUS develop acute renalfailure precipitately and the treatment of choice is dialysiswhich isinitiated when the child becomes oliguric.Most children respond todialysisand mortality rates of less than 5%are nowreported.In thecentral Scotland outbreak there were no deaths in children.Adultsseem to develop neurological or cardiovascular complications beforethe onset of oliguria.Neurological features are associated withincreased mortalityand neurological and cardiovascularcomplications of HUS/TTP were the most frequent causes of death inthe central Scotland outbreak.

Plasma exchange is an expensive(2500 per person treated inour hospital)and intensive procedure.Its effectiveness in thetreatment of HUS/TTP induced byE coliO157 needs to be showndefinitively in a multicentrerandomised controlled trial.Howeverfor a disease with very high mortality and just one potentiallybeneficial treatment optiona trial thatwithholds this optionwould behard to justify.It would also be extremely difficult to organise sincecases ofE coliO157 occur sporadically.There will always be anunavoidable selection bias within such a trialwith patients who areexcluded from treatment because they have contraindications to TPEorwho die before treatment can be initiated.

If 5%of all cases ofEcoliO157 develop HUS/TTPwe wouldexpect about 40 adult cases of HUS/TTP per year in the UK(datafrom the Communicable Disease Surveillance Centre and ScottishCentre for Infection and Environmental Health).We suggest that anational register be established for adult cases of HUS/TTPascurrently operates for cases in children.This database would enablemonitoring of treatment and outcomes in adultsproviding definitiveevidence of the effectiveness of TPEwithin about 5 years.

There is no evidence from our experience that TPE is harmful.

A national register of HUS/TTP secondary toE coliO157 coulddefine the role of TPE in the treatment of this serious disorder.



AcknowledgementsWe thank AK R Chaudhuri and W HWatson for their clinicalcontribution;the renal physicians and haematoligists at GlasgowRoyal Infirmary and Stobhill Hospital for clinical assistance in themanagement of cases;M Drummond for data collection;and theCentral ScotlandE coliO157 Research Group for the laboratorydatabase.




The feasibility of ultrasonography for diagnosis of fetal cardiacabnormality was recognised in the early 1980sand cardiac scanningis gradually being incorporated into fetal screening protocols.Theeffect of the screening process on the incidence and types ofcongenital heartdisease atterm has been difficultto ascertain becausemany pregnant women and infants travel great distances to specialistcentres which are farfrom their health authority.For a single centrethe geographical area from which its fetal referrals arrive is generallynot the same as the area attracting postnatal referralsand the numberof births that each serves is impossible to define.The BritishPaediatric Cardiac Association(BPCA)undertook a nationalcollaborative study of fetal cardiac screening.The aim was to assessthe effect of fetal diagnosis of congenital heart disease on the patternof serious congenital heart disease at term.


òֿоĶʹõIJϺͲõķҲɷֱ֮Ϊ:뷽(Subjects and methods or Patients and methods)뷽(Materials and methods)ʵĵǶĵݺͲõķϸ˳Ϊ:ʹõIJϻоĶdzǽͳƷһΪعʱ̬϶һȥʱżҲùȥʱǶͼݼֵڿ͹ʿɲһʱ:

Patients and methods

The Information and Statistics Department of the Scottish Homeand Health Department collected data on the demographics andlaboratory results of all possible outbreak cases.We collected clinicaldata by reviewing the case notes of all cases admitted to hospital inthe Lanarkshire area.

All confirmed or probable cases ofEscherilchia coli(E coli)0157 infectionidentified in the Lanarkshire area during the outbreakperiodwere included in the assessment and analysis.Confirmedcaseswere those in whom the outbreak strain ofE coliO157 wasisolated from stool samples.If stool cultureswere negative atthe locallaboratoriesspecimens were sent to Scotland'sE colireferencelaboratory in Aberdeenfor the more sensitive isolation method ofimmunomagnetic separation.Probable cases were those with bloodydiarrhoea or haemolytic uraemic syndrome(HUS)/thromboticthrombocytopenic purpura(TTP)an association with food sourcesimplicated in the outbreaknoE coliO157 isolatedand no otherorganism isolated.Adults were defined as patients 15 years of age orolder.

To allow standardisation of diagnosis in the face of a hugeclinical workloada case definition for HUS and TTP was developedat the beginning of the outbreak.HUS was defined as evidence ofred-cell haemolysis(red-cell fragmentation on blood film and lactatedehydrogenase1.5 times the upper limitof normal[our laboratory 0480 IU/L])plus thrombocytopenia(platelets150109/L)with rising urea and creatinine concentrations.All three criteria hadto be met before the diagnosis could be madebut not necessarily onthe same blood sample.A diagnosis of TTPwas given to patientswhomet these laboratory criteria and developed new neurologicalsymptoms and signs.One patient was included as having developedHUS despite a minimum platelet count of 228109/L(on death).

He had bloody diarrhoeaan association with an implicated foodsourceacute renal failurethe criteria for red-cell haemolysisand afalling platelet count.

In the assessment of premorbid illnessmedical historiesincluded as relevant were ischaemic heart diseasecardiac failurehypertentioncerebrovascular diseaserenal diseasediabetesandimmunosuppression.Pulmonary oedemawas diagnosed on clinical andradiological evidence.

TPE was performed at three centres with three Cobe SpectraApheresis Systems(Cobe Laboratories LtdGloucesterUK)and aBaxter Fenwal CS-3000 Plus Cell Separator(Baxter HealthcareNewberryUK).Plasma was exchanged with 2.02.4 Lfresh frozenplasma or cryosupernatant in refractory patients.The anticoagulantused was ACD-A.A combination of central and peripheral venousaccess was used.Intravenous hydrocortisone was given with eachexchange.Intravenous prostacyclin was also given to cases receivingTPEat doses between 40 mg/h and 200 mg/hwhere tolerated.Datawere analysed by means of SPSS(version 7.5).




There were 262 cases ofE coliO157 infection in theLanarkshire area:200 confirmed cases and 62 probable cases.Themedian age of all affected was 53 yearsbut there were highernumbers at the extremes of age.47%of infectedindividualswere over 55 years of age.13(5%)people died.In 10cases death was associated with the systemic complications ofE coliO157 infection.

28(11%)of the Lanarkshire cases ofE coliO157 met thediagnostic criteria forHUS/TTP.Casesmet the criteria forHUS/TTPa median of 7 days(range 415)after the onset of gastrointestinalsymptoms.A further eight cases had evidence of thromboticmicroangiopathy but did not meet the criteria for HUS/TTP and werenot eligible for TPE.22(79%)cases with HUS/TTP were adultsand six(21%)were children.The median age of adults whodeveloped HUS/TTP was 71 years and the median age of children 6years.The demographicsclinical featurestreatmentlaboratoryresultsand outcome of the adult cases with HUS/TTP are shown intable 1.Blood results are taken from the day that the diagnosticcriteria for HUS/TTP were metbefore TPE in cases so treated.

The mortality rate in adults with HUS/TTP was 45%(ten of22).Seven of 12 cases aged over 70 years and three of ten aged 70years or less died.There were no deaths in children.Necropsiesweredone for all cases who died.Causes of death in patients with HUS/TTPwere acute renal failure secondary to HUS(two cases)cardiacarrest(two cases)intracerebral haemorrhagecerebral infarctionacute myocardial infarctionmultiple organ failurehepatorenalsyndrome secondary to macronodular cirrhosis and septic shock.

TPE was used in 16 of the 22 adultpatientswithHUS/TTP.Forpatients treated with TPE later received haemodialysisbecause ofdeteriorating renal function.Patients who did not receive TPE wereeither too unwell to tolerate the procedure or died before TPE couldbe carried out.

In all 16 cases treated with TPEthe first exchange was firstdone within 24h of the criteria for HUS/TTP being met.Theminimum number of changes was onethe maximum 16and themedian six.Patients underwent a total of 107 proceduresand 1100units of fresh frozen plasmawere used.Two patients proved refractoryto treatment with fresh frozen plasmaafter five and six exchangesbut were successfully treated by additional TPE with cryosupernatantas the exchange fluid.Five of the 16(31%)TPE-treated patientsdiedfour of eight aged over 70 years and one of eight aged 70 yearsor less.Premorbid illnessneurological featurestreatment withciprofloxacin or prostacyclinand the laboratory severity of HUS/TTPwere not associated with deathalthough the number of caseswas toosmall to allow statistical conclusion.