- Email: [email protected]
Veno-venous haemodiafiltration in meningococcal septicaemia
Letters to the Editor
An unusual
gallstone
SiR-In November, 1990, a 25-year old Filipino woman was found at routine medical evaluation by the Andorran Immigration Office to have an enlarged spleen. Blood count showed white-blood-cell count of 3-5X10"/L; platelets
63X10"/L; prothrombin time, albumin, and liver function normal. Abdominal echography revealed a linear branching pattern of increased echogenicity, thickness of portal venous branches, and widening of the portal and splenic veins and collateral vessels. Stool examination was negative for ova and parasites. Rectal biopsy found ovalshaped eggs of Schistosoma japoraicum. She mentioned that she had received treatment for schistosomiasis when she was a child. In July, 1994, she attended the casualty department of our hospital with typical biliary colic, without fever. On clinical examination the abdomen was soft, with tenderness in the right hyppocondrium, but no liver enlargement. The spleen was firm, and increased in size. Plain abdominal radiography showed a radio-opaque gallstone (figure, top). Abdominal ultrasonography showed the same abnormalities as previously, and a slightly distended gallbladder with increased wall thickness, containing a single stone. At laparotomy, the gallbladder
were
measured 9X4 cm; the wall was thickened and the mucosa was velvety and bile stained. Microscopic examination showed many oval-shaped calcified eggs of Sjaponicum in the serosa. A wedge liver biopsy showed fibrosis in the portal tracts with the presence of calcified eggs and minimal granulomatous reaction. The stone (figure, bottom) consisted of calcium carbonate (calcite). Pathological lesions due to Schistosoma species are not commonly reported in the gallbladder, and in a comprehensive review,’ nothing is said about gallbladder involvement with S japonicum. This may be because disease manifestations are less well studied in Sjaponicum, or due to individual variations in disease pattern.2 *G de Celis, D Gómez, C Llebaria, X Latorre Hospital Nostra Senyora de Mentxell, Avda Fiter 1 Rossell, 1-13 EscaldesEngordany, Principat d’Andorra
1
2
Hatz C, Murakami H, Jenkins JM. A review of the literature on the use of ultrasonography in schistosomiasis with special reference to its use in field studies. 3. Schistosoma japonicum. Acta Trop 1992; 51: 29-36. Chen MG, Mott KE. Progress in assessment of morbidity to Schistosoma japonicum infection. Trop Dis Bull 1988; 85: R1-44.
Veno-venous haemodiafiltration in
meningococcal septicaemia SIR-Huge media attention was given to the publication of Best and colleagues’ report (Jan 20, p 202)’ about the use of haemodiafiltration for meningococcal septicaemia. On the next day we admitted a 22-year-old child to our intensive care unit who had presented in shock and later developed a purpuric rash despite fluid resuscitation and appropriate intravenous antibiotics at a local district hospital. On arrival metabolic acidosis and evidence diffuse intravascular haematological coagulation. The parents were aware of and keen for us to embark on life-saving haemodiafiltration, as reported in Glasgow. In view of the child’s poor outlook and the high level of media attention that had been focused on meningococcal disease, we felt compelled to start this treatment. Therapy was started within 3 h of admission to our unit and 7 h after presentation at the district hospital. Regrettably the child’s condition did not improve and he was judged brain-stem dead 36 h later. Blood cultures were positive for meningococcus. When dealing with critical illnesses, physicians are frequently under pressure to depart from evidence-based medicine and embark on novel therapies of unproven benefit. Increasing media attention encourages this practice. We support the recommendations from Glasgow that a randomised controlled trial to evaluate the benefits of haemodiafiltration should be conducted before this potential therapy slips into widespread use. at
our
unit
he
had
severe
for
*G Connett, M Waldron, T Woodcock Southampton
1
Figure: Top--plain abdominal radiograph; bottom-calcite
stone
General
Hospital, Southampton SO16 6YD, UK
Best C, Walsh J, Sinclair J, Beattie J.
meningococcal septicaemia.
Lancet
Early haemo-diafiltration
in
1995; 347: 202.
611
SiR-Best and colleagues’ report the use of veno-venous haemodiafiltration as a successful treatment strategy for children with meningococcal septicaemia. Several important questions remain unanswered. First, how long did each child stay on continuous haemodiafiltration, and, second, what selection criteria did they use to decide which patients should receive this treatment? Did they use a Glasgow meningococcal septicaemia score (GMSPS) of 9 or greater, as implied from their table? We agree with Best and coworkers that this treatment strategy remains speculative. Indeed Pollack2 reached the same conclusion after the use of plasma exchange or blood exchange in the treatment of children and adults with septicaemia.3 There have been many case reports in children, and a series in which haemofiltration was used with or without plasmafiltration in the treatment of 27 children with sepsis.4 The table shows the pertinent points of these studies. We urge caution before embarking on a randomised controlled trial assessing the efficacy of various blood filtering techniques, be they plasma exchange, blood exchange, or veno-venous haemodiafiltration, especially if GMSPS of 8 or 9 or greater was used as the sole selection criterion.
Number Median (range) age Mode of treatment (no) Volume exchanged
15
Specific
20% mortality vs 60% in historical controls (n=10)
comments
9 (2-23) yr PE or BE (15) 30-60 mL/kg*
27 26.6
(0.33-185) mo HF (18); HF+PF (9) Median duration 36 h
(range 2-145) 70% mortality as predicted by PRISM. No difference between HF & HFPF
PE=plasma exchange, BE=blood exchange, HF=haemofiltration, PF=plasmafiltration, PRISM=paediatric risk of mortality score. *Multiple exchanges, 12 hourly (median 3, range 1-4). Table: Details of treatment of two studies
At
have treated 47 children with over the past 3 years, with an overall mortality of 16%. Of these, 22 children with a GMSPS of 9 or greater were treated conventionally, with a mortality of 32% (five of the seven deaths occurring within 1 h of admission to our unit). The mortality of 32% compares favourably with the predicted minimum mortality of 74% (range 74-100%) for this group of patients.s It is noteworthy that of the 15 survivors treated conventionally, their median length of stay in our paediatric intensive care unit (ICU) was 5 days (first and third quartiles 3-8 days), which is considerably shorter than the median length of stay of the four children reported by Best (16-5 days, p<003 Mann-Whitney). Although, we agree that decreased mortality is the ultimate aim when assessing new applications of a recognised technique (in this case, veno-venous haemodiafiltration), other factors such as frequency of complications and length of stay must be considered before a treatment strategy becomes commonplace, especially one which could conceivably be used in every adult and paediatric ICU. our
institution
patients with septicaemia, from
we
meningococcal septicaemia
*S Tibby, M Champion, M Hatherill, M J Marsh, I A Murdoch Guy’s Paediatric Intensive Care Unit, 9th Floor Guy’s Tower, Guy’s Hospital, St Thomas Street, London SE1 9RT, UK 1 2 3
4 5
C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Pollack M. Editorial response: blood exchange and plasmapheresis in sepsis and septic shock. Clin Infect Dis 1992; 15: 431-33. van Deuran M, Santman FW, van Dalen R, Sauerwein RW, Span LFR, van der Meer JWM. Plasma and whole blood exchange in meningococcal sepsis. Clin Infect Dis 1992; 15: 424-30. Reeves JH, Butt WW. Blood filtration in children with severe sepsis: safe adjunctive therapy. Intens Care Med 1995; 21: 500-04. Thomson APJ, Sills JA, Hart CA. Validation of the Glasgow meningococcal septicaemia prognostic score: a 10 year retrospective survey. Crit Care Med 1991; 19: 26-30. Best
612
SiR-Best and colleagues’ report remarkably similar findings those of a trial over 10 years ago. Bjorvatn and cocombined workers2 plasmapheresis and reported in four children with or blood exchange, leucapheresis, all survived with no who meningococcal septicaemia, and colleagues3 recorded sequelae. Additionally, Westendorp 24% mortality in 13 patients with meningococcal septic with a shock leucaplasmapheresis, undergoing in a control of The 78% group. contemporaneous mortality theory that the mortality associated with this and similar conditions of overwhelming sepsis results from excessive, as opposed to adequate, inflammatory host response, is gaining acceptance.4 Removal of factors contributing to such excessive and inappropriate inflammation in a controllable fashion in combination with antibiotic therapy seems logical and the clinical results reported by these three groups lend support to such therapeutic manoeuvres. The factors responsible for the apparent benefit from such extracorporeal procedures continue to remain undefined. What seems to be clear are the disappointing results of large pharmaceutical-industry-sponsored trials, which are often based on animal models of sepsis and therapy and bear little relation to real life, suggesting that no one proinflammatory mediator is responsible for mortality. The additional observation by many intensive care specialists of rapid (if often only temporary) improvement in clinical condition after haemofiltration for renal failure in patients with septic shock lends further support to the beneficial effects of extracorporeal pheresis or filtration procedures. We have noted that haemofiltration in the context of sepsis often results in the rapid accumulation of insoluble deposits in the air trap and filter that are large enough to necessitate the premature replacement of the extracorporeal circuit; these insoluble deposits consist of tightly packed polymorphs. Polymorphs are killer cells that contribute to organ damage and destruction in both animals and man,s presumably subsequent to excessive and inappropriate exposure to proinflammatory mediators. In view of clinical observations, encouraging experience in meningococcal septicaemia, and disappointing trial data, we are studying the safety and efficacy of a purpose-designed extracorporeal leucofiltration device adapted for use with existing haemofiltration systems in patients fulfilling modified criteria for the systemic inflammatory response syndrome. Our data so far suggest that such a device removes leucocytes safely and effectively in an easily controlled and monitored fashion; such leucocyte removal may have contributed to the encouraging preliminary results. to
*Vanya Gant, Cathy Maciver, David F Treacher *Department of Microbiology and Infection, UMDS St Thomas’ Hospital, London SE1 7EH, UK; and Department of Intensive Care, Guy’s and St Thomas’ Trust, London
1 2
3 4
5
Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Bjorvatn B, Bjertnaes L, Fadnes HO, et al. Meningococcal septicaemia treated with combined plasmapheresis and leucapheresis or with blood exchange. BMJ Res Ed 1984; 288: 439-41. Westendorp RG, Brand A, Haanen J, et al. Leukaplasmapheresis in meningococcal septic shock. Am J Med 1992; 92: 577—78. Gant VA. Cytokines in sepsis—whither therapy? In: Holgate ST, ed. Horizons in medicine, vol 6. Oxford: Blackwell Science, 1995: 291-310. Welbourn CRB, Young Y. Endotoxin, septic shock, and acute lung injury: neutrophils, macrophages and inflammatory mediators. Br J Surg 1992; 79: 998-1003.
SIR-I have been treating infants and children with septic shock with plasma exchange because there is evidence that this method is effective.1,c Hard evidence of benefit is difficult to obtain and patients are too few and too
easily achieved in small children as Best and colleagues suggested in a national newspaper, yet after this publicity and with the high mortality of this condition it could well prove impossible to obtain consent for the controlled trial that would give us the answer to its is
BP=blood pressure; CI=cardlac mdex (L min 1 m 2); SVR=systemic vascular resistance. All inotrope dosages are in wg kg 1 mm 1.
Table: Cardiovascular indices before and after
plasma
exchange (PE) individual to randomise treatment, especially within a single centre. The table shows my results in five patients-all of whom I was asked to see at a time when death was felt to be inevitable. Patients 1 (female, aged 1-3 years, weight 9 kg) and 2 (male, 9-5 years, 37 kg) had Staphylococcus aureus
septicaemia, patient 3 (male, 10 years, 32 kg) gram-negative septicaemia, and patients 4 (female, 1-7 years, 12 kg) and 5 (female, 6 months, 8 kg) meningococcal septicaemia. Three (patients 1, 3, and 4) survived. The aim in each patient was to give three daily plasma exchanges of 100 mLJkg up to a maximum of 5 L. Replacement was with 4-5% human albumin solution and fresh frozen plasma (FFP), with a minimum of 600 mL FFP being used in each exchange. Patient 1 needed inotropes and had developing ischaemia and gangrene extending down from both forearms and from both knees. After the first plasma exchange striking reperfusion of all four limbs took place, in addition to a reduction in the dosages of inotropes required and increased urine output. She survived with the loss of only the fingertips from her left hand and gangrene from her right mid-foot down, requiring a below knee amputation for the fitting of a prosthesis. Patient 2 also had septic shock requiring high doses of inotropes to maintain his blood pressure. His cardiovascular indices were measured with a pulmonary artery catheter, and calculation of cardiac output with standard thermodilution techniques. The table shows the effect of the first plasma exchange. There was a substantial reduction in inotrope dosage during the course of the exchange, during which the patient’s blood pressure rose, with a fall in cardiac index and a rise in systemic vascular resistance. Unfortunately the patient went on to develop adult respiratory distress syndrome (ARDS) and died. Patient 3 also had a striking reduction in inotropic requirement, with a fall in cardiac output and rise in systemic vascular resistance during the procedure. As with patient 2, patient 4 showed a pronounced haemodynamic improvement as a response to plasma exchange, in addition to improved tissue perfusion. She went on to survive without handicap. Patient 5 was both haemodynamically unstable and had incipient gangrene of both hands and below the knee of both legs. Again there was an improvement of haemodynamic status after the first plasma exchange and after the three exchanges there was reperfusion of both hands and her left leg. She remained in renal failure, her biochemical indices being controlled with haemofiltration. Unfortunately, after 1 week she developed secondary pneumonia, ARDS, and died from respiratory failure. Best and colleagues3 do not mention the clearance obtained with the filtration through haemofiltrate removal or the countercurrent dialysis flow rates. Nor do they give any indication of how haemofiltration altered the clinical course other than to mention that all patients survived, as opposed to only one of four expected (p=0-15, Fischer’s exact test). Despite this deficiency the press and public reaction to their report has been immense. Safe and effective haemofiltration
not
as
effectiveness. I suggest that the answer to this dilemma is to start a multicentre trial in selected paediatric ICUs, randomising patients with meningococcaemia or septic shock and cardiovascular instability requiring inotropic support to either haemofiltration or plasma exchange on admission. This, combined with appropriate recording of changes in patient status during treatment and measurement of effluent cytokines might begin to give some answers to how this awful and unpredictable condition ought to be treated. Malcolm A Lewis Paediatric
Nephrology Department, Royal Manchester Children’s Hospital, Pendlebury, Manchester M27 4HA, UK 1
2
3
Gardlund B, Sjolin J, Nilsson A, Roll M, Wickerts CJ, Wretlind B. Plasma levels of cytokines in primary septic shock in humans: correlation with disease severity. J Infect Dis 1995; 172: 296-301. Gardlund B, Sjolin J, Nilsson A, et al. Plasmapheresis in the treatment of primary septic shock in humans. Scand J Infect Dis 1993; 25: 757-61. Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202.
SiR-Best and colleagues’ regard haemodiafiltration as a promising method of treating acute meningococcaemia, but survivors seem predisposed to thrombotic complications. This tendency is not surprising in view of the behaviour of antithrombin III (ATIII) in meningococcaemia, a condition in which those destined to die from the disease develop very low serum concentrations of ATIII compared with survivors:2
*Percentages of mean control value of 6.93 pmol/L (SD). In view of the inefficiency of heparin when there is a lack of ATIII/ this index is relevant to the handling of any potential coagulopathy in such patients. Since rescue by haemodiafiltration of more severely affected patients might result in more survivors with thrombotic complications, it would at least be appropriate for ATIII to be measured in those under critical observation. The concentrations might then be better corrected with ATIII concentrate4°5 rather than fresh plasma: the latter would unfortunately also supply fresh complement, which on endotoxic activation might contribute to further secondary endothelial and platelet damage. Much is still to be learned about the optimum treatment.’’5 Progress will only be made if as much as possible is gleaned in an informed environment from each meningococcaemic
patient. I thank the editor of the Journal of Clinical Pathology for pertnission to reproduce data from reference 2.
J M Stark Department of Medical Microbiology, University of Wales College of Medicine, Cardiff CF4 4XN, UK
1
2
3
Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1995; 347: 202. Stark JM, Matthews N, Ryley HC, Greenwood BM, Lewis S, Whittle HC. Meningococcal infection and proteolytic control. J Clin
Pathol 1978; 31: 1177-81. Marder VJ, Feinstein DI, Furness C, Colman RW. Consumptive haemorrhagic disorders. In: Coleman RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis and thrombosis, 3rd ed. Philadelphia: J B Lippincott, 1993.
613
4
5
Fourrier F, Lestavel P, Chopin C, et al. Meningococcaemia and purpura fulminans in adults: acute deficiencies of proteins C and S and early treatment with antithrombin III. Intens Care Med 1990; 16: 121-24. Cobcroft R, Henderson A, Solano C, Scott D. Meningococcal purpura fulminans treated with antithrombin III concentrate: what is the optimal treatment? Aust N Z J Med 1994; 24: 575-76.
SIR-During the past 3 years, 26 children with meningococcal septicaemia have required admission to our regional paediatric ICU. During this time we have also favoured the early use of haemofiltration, even in the presence of adequate urine output. Eight patients have been so treated, although we have tended to use arteriovenous filtration without diafiltration. These patients were the severest cases of meningococcal disease seen in this hospital, but only one died. The 4% mortality in the whole group is also substantially less than expected. The use of filtration techniques in critically ill patients is not without complications. The cases detailed by Best and colleagues’ were clearly severely compromised by their disease; even so the number of amputations required (two of the four patients needed below-knee amputations and one lost several digits) seems unusually high: the local and systemic effects of veno-venous diafiltration might have contributed. Large intravenous cannulae are required (usually two 5 F or one 8-5 F gauge catheters), which might compromise poorly perfused lower limbs. Furthermore, the filtration circuit creates a large extracorporeal blood volume, which may increase cardiovascular instability, especially in small
children. Arteriovenous filtration can function adequately with two 4 F lines, and does not require a blood therefore mechanical pump, creating little volume. extracorporeal Best and colleagues suggest that cytokine removal might contribute to the benefits seen, citing work by Bellomo et aF in support. However, although this work shows the removal of cytokines by haemofiltration, no reduction in plasma cytokine concentrations was demonstrated.2 There are several possible explanations: first, cytokines have a rapid turnover and are largely found in tissues, with small amounts in the circulation,3 limiting the benefit of plasma clearance; second, diffusion of large molecular weight cytokines through dialysis membranes is poor (simple haemofiltration may provide more effective clearance3). Dialysis membranes may themselves lead to enhanced cytokine production, an effect that may be exacerbated by the presence of endotoxin.4The effect of haemofiltration on survival in multiple organ system failure is also unclear, with recent work failing to demonstrate an improved survival in septic burned patients,’ despite evidence of interleukin-6 clearance from plasma. Haemofiltration has, however, been shown to confer other benefits, and we echo Best’s comments concerning the increased control of fluid balance and reduction of temperature. We feel that haemofiltration is useful in the management of meningococcal sepsis. Exact mechanisms for this benefit remain obscure. Further trials are necessary to determine the true benefits and complications of the various techniques available. *S
Morley,
A D’Amore, R I Ross Russell
Paediatric Intensive Care Unit, Addenbrooke’s Hospital,
1 2
3
Cambridge CB2 2QQ,
UK
C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Bellomo R, Tipping P, Boyce N. Continuous veno-venous haemofiltration with dialysis removes cytokines from the circulation of septic patients. Crit Care Med 1993; 21: 522-26. DiCarlo JV, Dudley TE, Sherbotie JR, Kaplan BS, Costarino AT. Continuous arterio-venous haemofiltration/dialysis improves pulmonary gas exchange in children with multiple organ system failure.
Best
614
Crit Care 4
Med 1990; 18: 822—26. M, Cozzolino F, Bergesio F, et al. In vitro interleukin-1 production by different dialysis membranes. Nephrol Dial Transpl 1988;
Amato
3: 432-34. 5
Gueugniaud P, Bertin-Maghit M, Hirschauer C, Petit P. Removal of cytokines in septic patients using continuous veno-venous haemodiafiltration. Crit Care Med 1994; 22: 717.
A uthors’
reply and colleagues raise several points. The patients continuous veno-venous haemodiafiltration as
SiR-Tibby were
on
follows:
Veno-venous haemodiafiltration was stopped when the patients stabilised. Patient C was initially on for 18 h, then the procedure was stopped for 18 h. However, she became haemodynamically unstable with hyperpyrexia and so haemodiafiltration was re-started with good effect. Patient D stabilised after 5 days but subsequently developed renal failure. Her renal function is now normal. Haemodiafiltration was initiated if the GMSPS was 9 or greater on admission and did not improve with fluid loading and increasing inotropic support. Other observations that correlate with a high mortality played a part-for example, platelet count of less than 100X10"/L, white blood count less than lOX lO9/L, and cyanosis on admission,’ combined with our clinical opinion that the patients condition was deteriorating. In particular, we did not wait for established renal failure before starting the procedure. We are not aware of any case reports in which early haemodiafiltration has been given for meningococcal sepsis per se, although the technique has been used many times for the treatment of established renal failure with data on tumour necrosis factor alpha and interleukin clearances.z We emphasise that, in our opinion, control of fluid volume and temperature is probably just as important as cytokine removal. The report to which Tibby and colleagues refer3 is an excellent retrospective analysis of the safety of haemofiltration and plasmafiltration in children with severe sepsis who developed multiorgan failure. Reeves and Butt3 in their conclusion state that they cannot comment on the effect the techniques may have had on outcome, only that the therapy seemed safe. Of the three patients who required lower limb surgery, two were affected on the opposite side to the double lumen femoral venous cannula. Pumped venovenous haemodiafiltration was used because the patients were haemodynamically unstable. Tibby and colleagues’ results with conventional therapy in children with an initial GMSPS of 9 or greater are excellent, but highlight the difficulties of comparing outcomes in different groups of patients.4 Obviously the sickest group of patients stay in paediatric ICU for the longest time. In conventional in addition, meningococcal therapy septicaemia is difficult to define. We note that Champion and colleagues,’ who describe the use of extra corporeal membrane oxygenation in meningococcal septicaemia also used protein C, ATIII, epoprostenol, and continuous arteriovenous haemofiltration, as well as inotropes and antibiotics. None of us expected the degree of press coverage that Connett and colleagues remark upon, but this was perhaps inevitable in view of the very high public profile of meningococcal sepsis. In our report we were at pains to emphasise the speculative nature and need for assessment of veno-venous haemodiafiltration, and believed The Lancet to be the most appropriate forum to raise this issue. It is
deal with the hopes of patients when the media reports a treatment that can "save their child". Sadly, there will always be some patients who present too late for any form of therapy to help. Novel therapies tend to be applied to meningococcal sepsis more after case reports than as the result of formal scientific assessment. Until there is agreement on how to assess illness severity on presentation and what constitutes conventional treatment, studies will always suffer from having too few patients in non-comparable groups. A national data collection project is needed to make sure we are all talking about the same thing before we can look at any advances.
obviously difficult
to
*Paul Cockwell, Caroline O S
1
2 3
5
1 Gedde-Dahl TW, Bjark P, Arne Høiby E, Host JH, Brunn JN. Severity of meningococcal disease: assessment by factors and scores and implications for patient management. Rev Infect Dis 1990; 12:
SiR-Reports of six children with severe meningococcal sepsis suggest that mortality may be lowered by the use of veno-arterial extracorporeal membrane oxygenation or continuous haemofiltration.1,2 If these observations are confirmed by clinical trials they will extend available therapies for this devastating disease. Best and colleagues2 state that one possible mechanism for the success of their treatment is the removal of proinflammatory mediators by haemofiltration. The assumption that the immunomodulating effect of haemofiltration is downregulatory is speculative, however, because removal of soluble mediators may be counterbalanced by activation of circulating leucocytes on the filter membrane. An additional explanation is the use of heparin in these patients. Heparin competes for glycosaminoglycan (GAG) binding sites with inflammatory mediators that have a necessity for GAG engagement to exert their biological effects. These molecules include several chemotactic cytokines (chemokines) which, when immobilised on endothelial cell surface GAGs, direct transendothelial leucocyte migration and are thus pivotal in the development of the inflammatory cascaded Displacement of these molecules by heparin or other GAG-binding proteins may abrogate their effect. Indeed, in-vitro studies have shown that heparin downregulates the action of mediators that have a
GAG-binding requirement.4
Protamine infused into the post-filter line reversed the anticoagulant effect of heparin in four of these patients. This technique is not exact, and even with a normal partial thromboplastin time there may remain sufficient free circulating heparin to exert a substantial immunomodulating In effect. addition, little is known about the pharmacokinetics of the protamine-heparin salt. Breakdown products from this complex may themselves exert a biological effect. Although currently out of vogue, heparin has previously been used as part of treatment regimens for aggressive immune
or
inflammatory
disorders." We do
not
advocate its
reintroduction, but, as a clearer understanding of the role of GAGs in regulating inflammatory and immune processes
Savage
Champion MP, Murdoch IA, Sajjanhar T, Marsh MJ. Extracorporeal membrane oxygenation for refractory shock in fulminant meningococcal sepsis. Lancet 1996; 347: 201-02. Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1995; 347: 202. Tanaka Y, Adams DH, Shaw S. Proteoglycans on endothelial cells present adhesion inducing cytokines to leukocytes. Immunol Today 1993; 14: 111-15.
Royal Hospital for Sick Children, Yorkhill, Glasgow G3 8SJ, UK
973-92. 2 Bellomo R, Tipping P, Boyce N. Continuous veno-venous haemofiltration with dialysis removes cytokines from the circulation of septic patients. Crit Care Med 1993; 21: 522-26. 3 Reeves JH, Butt WW. Blood filtration in children with severe sepsis: safe adjunctive therapy. Intens Care Med 1995; 21: 500-04. 4 Pollack M. Editorial response: blood exchange and plasmapheresis in sepsis and septic shock. Clin Infect Dis 1992; 15: 431—33. 5 Champion MP, Murdoch IA, Sajjanhar T, Marsh MJ. Extracorporeal membrane oxygenation for refractory shock in fulminant meningococcal sepsis. Lancet 1996; 347: 201.
and other anti-GAG
Centre for Clinical Research in Immunology and Signalling, Medical School, University of Birmingham, Birmingham B15 2TT, UK
4
*Crispin Best, Jo Walsh, John Sinclair, Jim Beattie Intensive Care Unit,
develops, further evaluation of heparin therapies might be appropriate
Daubener W, Nockemann S, Gutsche M, Hadding U. Heparin inhibits the antiparasitic and immune modulatory effects of human recombinant interferon-&ggr;. Eur J Immunol 1995; 25: 688—92. Fye KH, Hancock D, Moutsopoulos H, Dumes DH, Arieff AI. Low dosage heparin in rapidly progressive glomerulonephritis. Arch Intern Med 1976; 136: 995—99.
Vertical transmission of SIR-A chronic G virus
hepatitis G
agent that may be responsible for acute and hepatitis has been identified,I,2 and named hepatitis (HGV). At present, HGV infection can be detected only by reverse-transcription polymerase chain reaction (RT-PCR), with primers of the helicase-like region. The risk of HGV infection seems to be increased in people who are also infected with hepatitis B (HBV) or hepatitis C (HCV) virus and in individuals with a history of intravenous drug addiction (IVDA).’3 The risk of mother-to-infant transmission of HGV is unclear. We studied 61 pregnant women and their newborn children. From these, 47 women were with and 14 without risk factors for HGV infection. The 61 infants were followed up for over a year for vertical transmission of HGV, HCV, and HIV-1 by RT-PCR every 3 months. 30 (49-2%) women were infected with HCV and 17 (27-9%) with HIV-1. 14 (22-9%) women had no detectable infection with HBV, HCV, or HIV-1 and had no history of IVDA. The rate of vertical transmission for HIV-1 was 11-8% (2/17) and for HCV 6-7% (2/30), which correspond to reported HCV and HIV-1 vertical transmission rates. 4,5 Nine (14-8%) women had detectable HGV viremia by PCR. Six of these belonged to the 30 HCV-infected women (20%) and three to the 17 HIV-1-infected women (17-6%). Six (66-7%) of the nine HGV-positive women had a history. of IVDA. Vertical HGV transmission was shown in three of nine infants (33-3%). Two of the three mothers of the three HGV-positive infants were infected with HIV-1 and the third with HCV. Cotransmission of two viruses could not be demonstrated. During follow-up for a mean of 13 months none of the three infants with HGV infection and none of the two infants with HCV infection became icteric or showed other biochemical or clinical signs of hepatitis. In the high-risk population we studied, there was a higher rate of HGV mother-to-infant transmission compared with HIV-1or HCV. new
*Heinz-Hubert Feucht, Bernhard Zöllner, Susanne Polywka, Rainer Laufs *Institute of Medical Microbiology and Immunology, Universitätskrankenhaus Eppendorf, 20246 Hamburg, Germany
1
2
Linnen J, Wages J, Zhang-Keck Z-Y, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science 1996; 271: 505-08. Simons JN, Leary TP, Dawson GJ, et al. Isolation of novel virus-like sequences associated with human hepatitis. Nature Med 1995; 1: 564-69.
615
An unusual
gallstone
SiR-In November, 1990, a 25-year old Filipino woman was found at routine medical evaluation by the Andorran Immigration Office to have an enlarged spleen. Blood count showed white-blood-cell count of 3-5X10"/L; platelets
63X10"/L; prothrombin time, albumin, and liver function normal. Abdominal echography revealed a linear branching pattern of increased echogenicity, thickness of portal venous branches, and widening of the portal and splenic veins and collateral vessels. Stool examination was negative for ova and parasites. Rectal biopsy found ovalshaped eggs of Schistosoma japoraicum. She mentioned that she had received treatment for schistosomiasis when she was a child. In July, 1994, she attended the casualty department of our hospital with typical biliary colic, without fever. On clinical examination the abdomen was soft, with tenderness in the right hyppocondrium, but no liver enlargement. The spleen was firm, and increased in size. Plain abdominal radiography showed a radio-opaque gallstone (figure, top). Abdominal ultrasonography showed the same abnormalities as previously, and a slightly distended gallbladder with increased wall thickness, containing a single stone. At laparotomy, the gallbladder
were
measured 9X4 cm; the wall was thickened and the mucosa was velvety and bile stained. Microscopic examination showed many oval-shaped calcified eggs of Sjaponicum in the serosa. A wedge liver biopsy showed fibrosis in the portal tracts with the presence of calcified eggs and minimal granulomatous reaction. The stone (figure, bottom) consisted of calcium carbonate (calcite). Pathological lesions due to Schistosoma species are not commonly reported in the gallbladder, and in a comprehensive review,’ nothing is said about gallbladder involvement with S japonicum. This may be because disease manifestations are less well studied in Sjaponicum, or due to individual variations in disease pattern.2 *G de Celis, D Gómez, C Llebaria, X Latorre Hospital Nostra Senyora de Mentxell, Avda Fiter 1 Rossell, 1-13 EscaldesEngordany, Principat d’Andorra
1
2
Hatz C, Murakami H, Jenkins JM. A review of the literature on the use of ultrasonography in schistosomiasis with special reference to its use in field studies. 3. Schistosoma japonicum. Acta Trop 1992; 51: 29-36. Chen MG, Mott KE. Progress in assessment of morbidity to Schistosoma japonicum infection. Trop Dis Bull 1988; 85: R1-44.
Veno-venous haemodiafiltration in
meningococcal septicaemia SIR-Huge media attention was given to the publication of Best and colleagues’ report (Jan 20, p 202)’ about the use of haemodiafiltration for meningococcal septicaemia. On the next day we admitted a 22-year-old child to our intensive care unit who had presented in shock and later developed a purpuric rash despite fluid resuscitation and appropriate intravenous antibiotics at a local district hospital. On arrival metabolic acidosis and evidence diffuse intravascular haematological coagulation. The parents were aware of and keen for us to embark on life-saving haemodiafiltration, as reported in Glasgow. In view of the child’s poor outlook and the high level of media attention that had been focused on meningococcal disease, we felt compelled to start this treatment. Therapy was started within 3 h of admission to our unit and 7 h after presentation at the district hospital. Regrettably the child’s condition did not improve and he was judged brain-stem dead 36 h later. Blood cultures were positive for meningococcus. When dealing with critical illnesses, physicians are frequently under pressure to depart from evidence-based medicine and embark on novel therapies of unproven benefit. Increasing media attention encourages this practice. We support the recommendations from Glasgow that a randomised controlled trial to evaluate the benefits of haemodiafiltration should be conducted before this potential therapy slips into widespread use. at
our
unit
he
had
severe
for
*G Connett, M Waldron, T Woodcock Southampton
1
Figure: Top--plain abdominal radiograph; bottom-calcite
stone
General
Hospital, Southampton SO16 6YD, UK
Best C, Walsh J, Sinclair J, Beattie J.
meningococcal septicaemia.
Lancet
Early haemo-diafiltration
in
1995; 347: 202.
611
SiR-Best and colleagues’ report the use of veno-venous haemodiafiltration as a successful treatment strategy for children with meningococcal septicaemia. Several important questions remain unanswered. First, how long did each child stay on continuous haemodiafiltration, and, second, what selection criteria did they use to decide which patients should receive this treatment? Did they use a Glasgow meningococcal septicaemia score (GMSPS) of 9 or greater, as implied from their table? We agree with Best and coworkers that this treatment strategy remains speculative. Indeed Pollack2 reached the same conclusion after the use of plasma exchange or blood exchange in the treatment of children and adults with septicaemia.3 There have been many case reports in children, and a series in which haemofiltration was used with or without plasmafiltration in the treatment of 27 children with sepsis.4 The table shows the pertinent points of these studies. We urge caution before embarking on a randomised controlled trial assessing the efficacy of various blood filtering techniques, be they plasma exchange, blood exchange, or veno-venous haemodiafiltration, especially if GMSPS of 8 or 9 or greater was used as the sole selection criterion.
Number Median (range) age Mode of treatment (no) Volume exchanged
15
Specific
20% mortality vs 60% in historical controls (n=10)
comments
9 (2-23) yr PE or BE (15) 30-60 mL/kg*
27 26.6
(0.33-185) mo HF (18); HF+PF (9) Median duration 36 h
(range 2-145) 70% mortality as predicted by PRISM. No difference between HF & HFPF
PE=plasma exchange, BE=blood exchange, HF=haemofiltration, PF=plasmafiltration, PRISM=paediatric risk of mortality score. *Multiple exchanges, 12 hourly (median 3, range 1-4). Table: Details of treatment of two studies
At
have treated 47 children with over the past 3 years, with an overall mortality of 16%. Of these, 22 children with a GMSPS of 9 or greater were treated conventionally, with a mortality of 32% (five of the seven deaths occurring within 1 h of admission to our unit). The mortality of 32% compares favourably with the predicted minimum mortality of 74% (range 74-100%) for this group of patients.s It is noteworthy that of the 15 survivors treated conventionally, their median length of stay in our paediatric intensive care unit (ICU) was 5 days (first and third quartiles 3-8 days), which is considerably shorter than the median length of stay of the four children reported by Best (16-5 days, p<003 Mann-Whitney). Although, we agree that decreased mortality is the ultimate aim when assessing new applications of a recognised technique (in this case, veno-venous haemodiafiltration), other factors such as frequency of complications and length of stay must be considered before a treatment strategy becomes commonplace, especially one which could conceivably be used in every adult and paediatric ICU. our
institution
patients with septicaemia, from
we
meningococcal septicaemia
*S Tibby, M Champion, M Hatherill, M J Marsh, I A Murdoch Guy’s Paediatric Intensive Care Unit, 9th Floor Guy’s Tower, Guy’s Hospital, St Thomas Street, London SE1 9RT, UK 1 2 3
4 5
C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Pollack M. Editorial response: blood exchange and plasmapheresis in sepsis and septic shock. Clin Infect Dis 1992; 15: 431-33. van Deuran M, Santman FW, van Dalen R, Sauerwein RW, Span LFR, van der Meer JWM. Plasma and whole blood exchange in meningococcal sepsis. Clin Infect Dis 1992; 15: 424-30. Reeves JH, Butt WW. Blood filtration in children with severe sepsis: safe adjunctive therapy. Intens Care Med 1995; 21: 500-04. Thomson APJ, Sills JA, Hart CA. Validation of the Glasgow meningococcal septicaemia prognostic score: a 10 year retrospective survey. Crit Care Med 1991; 19: 26-30. Best
612
SiR-Best and colleagues’ report remarkably similar findings those of a trial over 10 years ago. Bjorvatn and cocombined workers2 plasmapheresis and reported in four children with or blood exchange, leucapheresis, all survived with no who meningococcal septicaemia, and colleagues3 recorded sequelae. Additionally, Westendorp 24% mortality in 13 patients with meningococcal septic with a shock leucaplasmapheresis, undergoing in a control of The 78% group. contemporaneous mortality theory that the mortality associated with this and similar conditions of overwhelming sepsis results from excessive, as opposed to adequate, inflammatory host response, is gaining acceptance.4 Removal of factors contributing to such excessive and inappropriate inflammation in a controllable fashion in combination with antibiotic therapy seems logical and the clinical results reported by these three groups lend support to such therapeutic manoeuvres. The factors responsible for the apparent benefit from such extracorporeal procedures continue to remain undefined. What seems to be clear are the disappointing results of large pharmaceutical-industry-sponsored trials, which are often based on animal models of sepsis and therapy and bear little relation to real life, suggesting that no one proinflammatory mediator is responsible for mortality. The additional observation by many intensive care specialists of rapid (if often only temporary) improvement in clinical condition after haemofiltration for renal failure in patients with septic shock lends further support to the beneficial effects of extracorporeal pheresis or filtration procedures. We have noted that haemofiltration in the context of sepsis often results in the rapid accumulation of insoluble deposits in the air trap and filter that are large enough to necessitate the premature replacement of the extracorporeal circuit; these insoluble deposits consist of tightly packed polymorphs. Polymorphs are killer cells that contribute to organ damage and destruction in both animals and man,s presumably subsequent to excessive and inappropriate exposure to proinflammatory mediators. In view of clinical observations, encouraging experience in meningococcal septicaemia, and disappointing trial data, we are studying the safety and efficacy of a purpose-designed extracorporeal leucofiltration device adapted for use with existing haemofiltration systems in patients fulfilling modified criteria for the systemic inflammatory response syndrome. Our data so far suggest that such a device removes leucocytes safely and effectively in an easily controlled and monitored fashion; such leucocyte removal may have contributed to the encouraging preliminary results. to
*Vanya Gant, Cathy Maciver, David F Treacher *Department of Microbiology and Infection, UMDS St Thomas’ Hospital, London SE1 7EH, UK; and Department of Intensive Care, Guy’s and St Thomas’ Trust, London
1 2
3 4
5
Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Bjorvatn B, Bjertnaes L, Fadnes HO, et al. Meningococcal septicaemia treated with combined plasmapheresis and leucapheresis or with blood exchange. BMJ Res Ed 1984; 288: 439-41. Westendorp RG, Brand A, Haanen J, et al. Leukaplasmapheresis in meningococcal septic shock. Am J Med 1992; 92: 577—78. Gant VA. Cytokines in sepsis—whither therapy? In: Holgate ST, ed. Horizons in medicine, vol 6. Oxford: Blackwell Science, 1995: 291-310. Welbourn CRB, Young Y. Endotoxin, septic shock, and acute lung injury: neutrophils, macrophages and inflammatory mediators. Br J Surg 1992; 79: 998-1003.
SIR-I have been treating infants and children with septic shock with plasma exchange because there is evidence that this method is effective.1,c Hard evidence of benefit is difficult to obtain and patients are too few and too
easily achieved in small children as Best and colleagues suggested in a national newspaper, yet after this publicity and with the high mortality of this condition it could well prove impossible to obtain consent for the controlled trial that would give us the answer to its is
BP=blood pressure; CI=cardlac mdex (L min 1 m 2); SVR=systemic vascular resistance. All inotrope dosages are in wg kg 1 mm 1.
Table: Cardiovascular indices before and after
plasma
exchange (PE) individual to randomise treatment, especially within a single centre. The table shows my results in five patients-all of whom I was asked to see at a time when death was felt to be inevitable. Patients 1 (female, aged 1-3 years, weight 9 kg) and 2 (male, 9-5 years, 37 kg) had Staphylococcus aureus
septicaemia, patient 3 (male, 10 years, 32 kg) gram-negative septicaemia, and patients 4 (female, 1-7 years, 12 kg) and 5 (female, 6 months, 8 kg) meningococcal septicaemia. Three (patients 1, 3, and 4) survived. The aim in each patient was to give three daily plasma exchanges of 100 mLJkg up to a maximum of 5 L. Replacement was with 4-5% human albumin solution and fresh frozen plasma (FFP), with a minimum of 600 mL FFP being used in each exchange. Patient 1 needed inotropes and had developing ischaemia and gangrene extending down from both forearms and from both knees. After the first plasma exchange striking reperfusion of all four limbs took place, in addition to a reduction in the dosages of inotropes required and increased urine output. She survived with the loss of only the fingertips from her left hand and gangrene from her right mid-foot down, requiring a below knee amputation for the fitting of a prosthesis. Patient 2 also had septic shock requiring high doses of inotropes to maintain his blood pressure. His cardiovascular indices were measured with a pulmonary artery catheter, and calculation of cardiac output with standard thermodilution techniques. The table shows the effect of the first plasma exchange. There was a substantial reduction in inotrope dosage during the course of the exchange, during which the patient’s blood pressure rose, with a fall in cardiac index and a rise in systemic vascular resistance. Unfortunately the patient went on to develop adult respiratory distress syndrome (ARDS) and died. Patient 3 also had a striking reduction in inotropic requirement, with a fall in cardiac output and rise in systemic vascular resistance during the procedure. As with patient 2, patient 4 showed a pronounced haemodynamic improvement as a response to plasma exchange, in addition to improved tissue perfusion. She went on to survive without handicap. Patient 5 was both haemodynamically unstable and had incipient gangrene of both hands and below the knee of both legs. Again there was an improvement of haemodynamic status after the first plasma exchange and after the three exchanges there was reperfusion of both hands and her left leg. She remained in renal failure, her biochemical indices being controlled with haemofiltration. Unfortunately, after 1 week she developed secondary pneumonia, ARDS, and died from respiratory failure. Best and colleagues3 do not mention the clearance obtained with the filtration through haemofiltrate removal or the countercurrent dialysis flow rates. Nor do they give any indication of how haemofiltration altered the clinical course other than to mention that all patients survived, as opposed to only one of four expected (p=0-15, Fischer’s exact test). Despite this deficiency the press and public reaction to their report has been immense. Safe and effective haemofiltration
not
as
effectiveness. I suggest that the answer to this dilemma is to start a multicentre trial in selected paediatric ICUs, randomising patients with meningococcaemia or septic shock and cardiovascular instability requiring inotropic support to either haemofiltration or plasma exchange on admission. This, combined with appropriate recording of changes in patient status during treatment and measurement of effluent cytokines might begin to give some answers to how this awful and unpredictable condition ought to be treated. Malcolm A Lewis Paediatric
Nephrology Department, Royal Manchester Children’s Hospital, Pendlebury, Manchester M27 4HA, UK 1
2
3
Gardlund B, Sjolin J, Nilsson A, Roll M, Wickerts CJ, Wretlind B. Plasma levels of cytokines in primary septic shock in humans: correlation with disease severity. J Infect Dis 1995; 172: 296-301. Gardlund B, Sjolin J, Nilsson A, et al. Plasmapheresis in the treatment of primary septic shock in humans. Scand J Infect Dis 1993; 25: 757-61. Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202.
SiR-Best and colleagues’ regard haemodiafiltration as a promising method of treating acute meningococcaemia, but survivors seem predisposed to thrombotic complications. This tendency is not surprising in view of the behaviour of antithrombin III (ATIII) in meningococcaemia, a condition in which those destined to die from the disease develop very low serum concentrations of ATIII compared with survivors:2
*Percentages of mean control value of 6.93 pmol/L (SD). In view of the inefficiency of heparin when there is a lack of ATIII/ this index is relevant to the handling of any potential coagulopathy in such patients. Since rescue by haemodiafiltration of more severely affected patients might result in more survivors with thrombotic complications, it would at least be appropriate for ATIII to be measured in those under critical observation. The concentrations might then be better corrected with ATIII concentrate4°5 rather than fresh plasma: the latter would unfortunately also supply fresh complement, which on endotoxic activation might contribute to further secondary endothelial and platelet damage. Much is still to be learned about the optimum treatment.’’5 Progress will only be made if as much as possible is gleaned in an informed environment from each meningococcaemic
patient. I thank the editor of the Journal of Clinical Pathology for pertnission to reproduce data from reference 2.
J M Stark Department of Medical Microbiology, University of Wales College of Medicine, Cardiff CF4 4XN, UK
1
2
3
Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1995; 347: 202. Stark JM, Matthews N, Ryley HC, Greenwood BM, Lewis S, Whittle HC. Meningococcal infection and proteolytic control. J Clin
Pathol 1978; 31: 1177-81. Marder VJ, Feinstein DI, Furness C, Colman RW. Consumptive haemorrhagic disorders. In: Coleman RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis and thrombosis, 3rd ed. Philadelphia: J B Lippincott, 1993.
613
4
5
Fourrier F, Lestavel P, Chopin C, et al. Meningococcaemia and purpura fulminans in adults: acute deficiencies of proteins C and S and early treatment with antithrombin III. Intens Care Med 1990; 16: 121-24. Cobcroft R, Henderson A, Solano C, Scott D. Meningococcal purpura fulminans treated with antithrombin III concentrate: what is the optimal treatment? Aust N Z J Med 1994; 24: 575-76.
SIR-During the past 3 years, 26 children with meningococcal septicaemia have required admission to our regional paediatric ICU. During this time we have also favoured the early use of haemofiltration, even in the presence of adequate urine output. Eight patients have been so treated, although we have tended to use arteriovenous filtration without diafiltration. These patients were the severest cases of meningococcal disease seen in this hospital, but only one died. The 4% mortality in the whole group is also substantially less than expected. The use of filtration techniques in critically ill patients is not without complications. The cases detailed by Best and colleagues’ were clearly severely compromised by their disease; even so the number of amputations required (two of the four patients needed below-knee amputations and one lost several digits) seems unusually high: the local and systemic effects of veno-venous diafiltration might have contributed. Large intravenous cannulae are required (usually two 5 F or one 8-5 F gauge catheters), which might compromise poorly perfused lower limbs. Furthermore, the filtration circuit creates a large extracorporeal blood volume, which may increase cardiovascular instability, especially in small
children. Arteriovenous filtration can function adequately with two 4 F lines, and does not require a blood therefore mechanical pump, creating little volume. extracorporeal Best and colleagues suggest that cytokine removal might contribute to the benefits seen, citing work by Bellomo et aF in support. However, although this work shows the removal of cytokines by haemofiltration, no reduction in plasma cytokine concentrations was demonstrated.2 There are several possible explanations: first, cytokines have a rapid turnover and are largely found in tissues, with small amounts in the circulation,3 limiting the benefit of plasma clearance; second, diffusion of large molecular weight cytokines through dialysis membranes is poor (simple haemofiltration may provide more effective clearance3). Dialysis membranes may themselves lead to enhanced cytokine production, an effect that may be exacerbated by the presence of endotoxin.4The effect of haemofiltration on survival in multiple organ system failure is also unclear, with recent work failing to demonstrate an improved survival in septic burned patients,’ despite evidence of interleukin-6 clearance from plasma. Haemofiltration has, however, been shown to confer other benefits, and we echo Best’s comments concerning the increased control of fluid balance and reduction of temperature. We feel that haemofiltration is useful in the management of meningococcal sepsis. Exact mechanisms for this benefit remain obscure. Further trials are necessary to determine the true benefits and complications of the various techniques available. *S
Morley,
A D’Amore, R I Ross Russell
Paediatric Intensive Care Unit, Addenbrooke’s Hospital,
1 2
3
Cambridge CB2 2QQ,
UK
C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1996; 347: 202. Bellomo R, Tipping P, Boyce N. Continuous veno-venous haemofiltration with dialysis removes cytokines from the circulation of septic patients. Crit Care Med 1993; 21: 522-26. DiCarlo JV, Dudley TE, Sherbotie JR, Kaplan BS, Costarino AT. Continuous arterio-venous haemofiltration/dialysis improves pulmonary gas exchange in children with multiple organ system failure.
Best
614
Crit Care 4
Med 1990; 18: 822—26. M, Cozzolino F, Bergesio F, et al. In vitro interleukin-1 production by different dialysis membranes. Nephrol Dial Transpl 1988;
Amato
3: 432-34. 5
Gueugniaud P, Bertin-Maghit M, Hirschauer C, Petit P. Removal of cytokines in septic patients using continuous veno-venous haemodiafiltration. Crit Care Med 1994; 22: 717.
A uthors’
reply and colleagues raise several points. The patients continuous veno-venous haemodiafiltration as
SiR-Tibby were
on
follows:
Veno-venous haemodiafiltration was stopped when the patients stabilised. Patient C was initially on for 18 h, then the procedure was stopped for 18 h. However, she became haemodynamically unstable with hyperpyrexia and so haemodiafiltration was re-started with good effect. Patient D stabilised after 5 days but subsequently developed renal failure. Her renal function is now normal. Haemodiafiltration was initiated if the GMSPS was 9 or greater on admission and did not improve with fluid loading and increasing inotropic support. Other observations that correlate with a high mortality played a part-for example, platelet count of less than 100X10"/L, white blood count less than lOX lO9/L, and cyanosis on admission,’ combined with our clinical opinion that the patients condition was deteriorating. In particular, we did not wait for established renal failure before starting the procedure. We are not aware of any case reports in which early haemodiafiltration has been given for meningococcal sepsis per se, although the technique has been used many times for the treatment of established renal failure with data on tumour necrosis factor alpha and interleukin clearances.z We emphasise that, in our opinion, control of fluid volume and temperature is probably just as important as cytokine removal. The report to which Tibby and colleagues refer3 is an excellent retrospective analysis of the safety of haemofiltration and plasmafiltration in children with severe sepsis who developed multiorgan failure. Reeves and Butt3 in their conclusion state that they cannot comment on the effect the techniques may have had on outcome, only that the therapy seemed safe. Of the three patients who required lower limb surgery, two were affected on the opposite side to the double lumen femoral venous cannula. Pumped venovenous haemodiafiltration was used because the patients were haemodynamically unstable. Tibby and colleagues’ results with conventional therapy in children with an initial GMSPS of 9 or greater are excellent, but highlight the difficulties of comparing outcomes in different groups of patients.4 Obviously the sickest group of patients stay in paediatric ICU for the longest time. In conventional in addition, meningococcal therapy septicaemia is difficult to define. We note that Champion and colleagues,’ who describe the use of extra corporeal membrane oxygenation in meningococcal septicaemia also used protein C, ATIII, epoprostenol, and continuous arteriovenous haemofiltration, as well as inotropes and antibiotics. None of us expected the degree of press coverage that Connett and colleagues remark upon, but this was perhaps inevitable in view of the very high public profile of meningococcal sepsis. In our report we were at pains to emphasise the speculative nature and need for assessment of veno-venous haemodiafiltration, and believed The Lancet to be the most appropriate forum to raise this issue. It is
deal with the hopes of patients when the media reports a treatment that can "save their child". Sadly, there will always be some patients who present too late for any form of therapy to help. Novel therapies tend to be applied to meningococcal sepsis more after case reports than as the result of formal scientific assessment. Until there is agreement on how to assess illness severity on presentation and what constitutes conventional treatment, studies will always suffer from having too few patients in non-comparable groups. A national data collection project is needed to make sure we are all talking about the same thing before we can look at any advances.
obviously difficult
to
*Paul Cockwell, Caroline O S
1
2 3
5
1 Gedde-Dahl TW, Bjark P, Arne Høiby E, Host JH, Brunn JN. Severity of meningococcal disease: assessment by factors and scores and implications for patient management. Rev Infect Dis 1990; 12:
SiR-Reports of six children with severe meningococcal sepsis suggest that mortality may be lowered by the use of veno-arterial extracorporeal membrane oxygenation or continuous haemofiltration.1,2 If these observations are confirmed by clinical trials they will extend available therapies for this devastating disease. Best and colleagues2 state that one possible mechanism for the success of their treatment is the removal of proinflammatory mediators by haemofiltration. The assumption that the immunomodulating effect of haemofiltration is downregulatory is speculative, however, because removal of soluble mediators may be counterbalanced by activation of circulating leucocytes on the filter membrane. An additional explanation is the use of heparin in these patients. Heparin competes for glycosaminoglycan (GAG) binding sites with inflammatory mediators that have a necessity for GAG engagement to exert their biological effects. These molecules include several chemotactic cytokines (chemokines) which, when immobilised on endothelial cell surface GAGs, direct transendothelial leucocyte migration and are thus pivotal in the development of the inflammatory cascaded Displacement of these molecules by heparin or other GAG-binding proteins may abrogate their effect. Indeed, in-vitro studies have shown that heparin downregulates the action of mediators that have a
GAG-binding requirement.4
Protamine infused into the post-filter line reversed the anticoagulant effect of heparin in four of these patients. This technique is not exact, and even with a normal partial thromboplastin time there may remain sufficient free circulating heparin to exert a substantial immunomodulating In effect. addition, little is known about the pharmacokinetics of the protamine-heparin salt. Breakdown products from this complex may themselves exert a biological effect. Although currently out of vogue, heparin has previously been used as part of treatment regimens for aggressive immune
or
inflammatory
disorders." We do
not
advocate its
reintroduction, but, as a clearer understanding of the role of GAGs in regulating inflammatory and immune processes
Savage
Champion MP, Murdoch IA, Sajjanhar T, Marsh MJ. Extracorporeal membrane oxygenation for refractory shock in fulminant meningococcal sepsis. Lancet 1996; 347: 201-02. Best C, Walsh J, Sinclair J, Beattie J. Early haemo-diafiltration in meningococcal septicaemia. Lancet 1995; 347: 202. Tanaka Y, Adams DH, Shaw S. Proteoglycans on endothelial cells present adhesion inducing cytokines to leukocytes. Immunol Today 1993; 14: 111-15.
Royal Hospital for Sick Children, Yorkhill, Glasgow G3 8SJ, UK
973-92. 2 Bellomo R, Tipping P, Boyce N. Continuous veno-venous haemofiltration with dialysis removes cytokines from the circulation of septic patients. Crit Care Med 1993; 21: 522-26. 3 Reeves JH, Butt WW. Blood filtration in children with severe sepsis: safe adjunctive therapy. Intens Care Med 1995; 21: 500-04. 4 Pollack M. Editorial response: blood exchange and plasmapheresis in sepsis and septic shock. Clin Infect Dis 1992; 15: 431—33. 5 Champion MP, Murdoch IA, Sajjanhar T, Marsh MJ. Extracorporeal membrane oxygenation for refractory shock in fulminant meningococcal sepsis. Lancet 1996; 347: 201.
and other anti-GAG
Centre for Clinical Research in Immunology and Signalling, Medical School, University of Birmingham, Birmingham B15 2TT, UK
4
*Crispin Best, Jo Walsh, John Sinclair, Jim Beattie Intensive Care Unit,
develops, further evaluation of heparin therapies might be appropriate
Daubener W, Nockemann S, Gutsche M, Hadding U. Heparin inhibits the antiparasitic and immune modulatory effects of human recombinant interferon-&ggr;. Eur J Immunol 1995; 25: 688—92. Fye KH, Hancock D, Moutsopoulos H, Dumes DH, Arieff AI. Low dosage heparin in rapidly progressive glomerulonephritis. Arch Intern Med 1976; 136: 995—99.
Vertical transmission of SIR-A chronic G virus
hepatitis G
agent that may be responsible for acute and hepatitis has been identified,I,2 and named hepatitis (HGV). At present, HGV infection can be detected only by reverse-transcription polymerase chain reaction (RT-PCR), with primers of the helicase-like region. The risk of HGV infection seems to be increased in people who are also infected with hepatitis B (HBV) or hepatitis C (HCV) virus and in individuals with a history of intravenous drug addiction (IVDA).’3 The risk of mother-to-infant transmission of HGV is unclear. We studied 61 pregnant women and their newborn children. From these, 47 women were with and 14 without risk factors for HGV infection. The 61 infants were followed up for over a year for vertical transmission of HGV, HCV, and HIV-1 by RT-PCR every 3 months. 30 (49-2%) women were infected with HCV and 17 (27-9%) with HIV-1. 14 (22-9%) women had no detectable infection with HBV, HCV, or HIV-1 and had no history of IVDA. The rate of vertical transmission for HIV-1 was 11-8% (2/17) and for HCV 6-7% (2/30), which correspond to reported HCV and HIV-1 vertical transmission rates. 4,5 Nine (14-8%) women had detectable HGV viremia by PCR. Six of these belonged to the 30 HCV-infected women (20%) and three to the 17 HIV-1-infected women (17-6%). Six (66-7%) of the nine HGV-positive women had a history. of IVDA. Vertical HGV transmission was shown in three of nine infants (33-3%). Two of the three mothers of the three HGV-positive infants were infected with HIV-1 and the third with HCV. Cotransmission of two viruses could not be demonstrated. During follow-up for a mean of 13 months none of the three infants with HGV infection and none of the two infants with HCV infection became icteric or showed other biochemical or clinical signs of hepatitis. In the high-risk population we studied, there was a higher rate of HGV mother-to-infant transmission compared with HIV-1or HCV. new
*Heinz-Hubert Feucht, Bernhard Zöllner, Susanne Polywka, Rainer Laufs *Institute of Medical Microbiology and Immunology, Universitätskrankenhaus Eppendorf, 20246 Hamburg, Germany
1
2
Linnen J, Wages J, Zhang-Keck Z-Y, et al. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science 1996; 271: 505-08. Simons JN, Leary TP, Dawson GJ, et al. Isolation of novel virus-like sequences associated with human hepatitis. Nature Med 1995; 1: 564-69.
615