Extent of the acute phase response in fulminant hepatic failure

Transcription

1 982 Institute of Liver Studies, King's College Hospital and Shool of Mediine and Dentistry, London S Izumi R D Hughes P G Langley J R B Pemambuo Roger Williams Correspondene to: Dr Roger Williams, Institute of Liver Studies, King's College Shool of Mediine and Dentistry, Bessemer Road, London SE5 9PJ. Aepted for publiation 10 November 1993 Gut 1994; 35: Extent of the aute phase response in fulminant hepati failure S Izumi, R D Hughes, P G Langley, J R B Pernambuo, Roger Williams Abstrat The extent of the aute phase response and the relation between aute phase proteins and ytokines in plasma was investigated in 50 patients with fulminant hepati failure. On admission, C reative protein was signifiantly higher in fulminant hepati failure (median: 12-4 ig/ml, range: jig/ml) than in 20 ontrols (median: 0.8,ug/ml, range: 0*3-2-9 jig/ml, p<0-001). Serial measurements showed that plasma C reative protein inreased daily after admission until day 5, the end of the study period. ot1-antitrypsin (AAT) (median: 69l1%, range: %) and fibrinogen (median: 1-10 g/l, range: g/l) were signifiantly lower in fulminant hepati failure on admission than in ontrols (AAT: median: 126%, range: %; fibrinogen: median 2-48 g/l, range: 1i g/l, p<0-001) and did not hange subsequently. Both AAT and fibrinogen were maintained at signifiantly higher onentrations in survivors than in those who did not. Baterial infetion ourred in 23 patients during the ourse of fulminant hepati failure, but did not influene the onentrations of these three proteins. Interleukin 6 was signifiantly higher in fulminant hepati failure (median: 21*2 pg/ml, range: pg/ml) than in ontrols (median: 2-4 pg/ml, range: 1i5-8-2 pg/ml, p<0001). There was a signifiant orrelation between interleukin 6 and the C reative protein onentrations in patients with viral hepatitis on admission and in all patients 48 hours later, onsistent with other evidene that interleukin 6 stimulates synthesis of this aute phase protein. (Gut 1994; 35: ) Inflammation, tissue injury or surgial trauma indue loal and systemi physiologial hanges known as the aute phase response to limit the progression of the injury and protet the organism against further damage. The kinetis and degree of inreased prodution of eah aute phase protein is harateristi. C reative protein is onsidered to be the most sensitive and speifi marker of the aute phase response,' with the plasma onentration rapidly rising several hundred fold. The physiologial role of C reative protein is omplex and remains unlear, but may inlude the opsonisation of bateria, immune omplexes, and foreign partiles.2 Other aute phase proteins with well defined funtion suh as ol-antitrypsin (AAT) and fibrinogen only inrease up to a maximum of five times their normal values. AAT ats as the most important inhibitor of neutrophil elastase,3 and fibrinogen plays a entral part in blood oagulation. In fulminant hepati failure, there is greatly redued synthesis of proteins, partiularly oagulation fators, by the damaged liver with the result that values of fator V and prolongation of the prothrombin time an be used as measures of hepati funtion. The oagulation inhibitors antithrombin-iii (AT-III),4 protein C,5 protein s,6 and heparin ofator II6 are also redued in fulminant hepati failure, as is the protease inhibitor AAT,7 although the latter is an aute phase protein. The liver is stimulated to produe the aute phase proteins by ytokines released mainly from ativated marophages suh as tumour nerosis fator,8 interleukin 1,9 10 and interleukin 6,11-13 although interleukin 6 seems to be the most important regulator of aute phase protein synthesis in humans The prodution of all three ytokines has been shown to be inreased in fulminant hepati failure.'7-19 It is unlear, however, if the liver an still produe an aute phase response in fulminant hepati failure, with the profound loss of liver syntheti funtion. In this study the aute phase proteins C reative protein, AAT, fibrinogen and, for omparison, antithrombin-iii (AT-III) (whih is not an aute phase protein) were determined in the plasma of 50 patients with fulminant hepati failure in relation to the ourse of the illness, as well as the prodution of ytokines interleukin 6 and interleukin 1. Patients and methods Fifty patients with fulminant hepati failure, as defined by Trey and Davidson,20 were studied (male: 19, female: 31, mean age: 31-7 years, range: 13-60), the auses being paraetamol overdose in 40 patients, viral hepatitis in eight (one hepatitis A, two hepatitis B, one hepatitis C, four hepatitis non-a non-b), and adverse drug reation in two patients (Table I). Twenty six patients were in grade 1 or 2 enephalopathy on admission and 10 of these deteriorated further to grade 3 or 4 enephalopathy. The remaining 24 ases were admitted in grade 3 or 4 enephalopathy. Of these 34 patients with grade 3 or 4 enephalopathy, seven patients had liver transplantation after seletion on the basis of poor prognosti riteria as used at King's College

2 Extent of the aute phase response in fulminant hepati failure 983 TABLE I Laboratory parameters in patients with fulminant hepati failure on admission Prothrombin AST Bilirubin Creatinine WBC Platelets time (INR) (IU/I) (p.mol/7) (pmolw/) (X 109/l) (x 1JO/) Paraetamol overdose (n = 40) median * 73* 157** range Viral and other auses (n= 10) median range Normal range AST= serum aspartate aminotransferase; WBC =white blood ell ounts; INR= intemational nornalised ratio. *p<0-001, **p<o-01. Hospital.2' Of the 43 patients with fulminant hepati failure who did not reeive a liver transplantation, 23 patients survived and 20 patients died. Ten patients reeived haemodialysis for renal failure. Only one patient reeived transfusion of fresh frozen plasma beause of oagulation abnormality, plasma exhange was only performed for patients about to have transplantation. Infetion was suspeted when the axillary temperature exeeded 38 C on two suessive four hour readings and a peripheral blood leuoyte ount exeeded 15X 1i09/1, and diagnosis was onfirmed by mirobiologial examination of blood, urine, and sputum, or from radiologial signs of pulmonary onsolidation. Eighteen patients had some evidene of infetion and five other patients were suspeted of infetion during the ourse of admission. Control samples were obtained from 20 laboratory staff (male: 10, female: 10, mean age: 31-8 years, range: 24-42). Blood samples were taken on admission and then daily until prothrombin time had returned to normal, death or liver transplantation, to a maximum of five days. In the nonsurvivor group, eight patients were available for study on day 5. Citrated blood samples were olleted on ie and immediately entrifuged for 10 minutes at 2000 g at 4 C. The resulting plasma was stored in aliquots at - 70 C until assay. TABLE II Approval for the study was obtained from the loal researh ethis ommittee. ASSAYS C reative protein was assayed by an enzyme linked immunosorbent assay (ELISA) based on the method of Highton.22 Anti-serum, peroxidase onjugated anti-serum, and C reative protein standard were obtained from Dako Ltd (High Wyombe, UK). Miroelisa plates (Dynateh, Billingshurst, UK) were oated with antibody to C reative protein (diluted 1:5000 in barbitone buffer ph 8 8) by inubation overnight. Serum samples were diluted (1:1000 and 1:10000) in phosphate buffered saline ph 7*4 ontaining 0-05% Tween and 0* lo% bovine serum albumin. The peroxidase substrate was 3,5',5,5' tetramethyl benzidine in itrate/aetate buffer ph 6-0. A standard urve was run on eah plate and all samples were measured in dupliate. The final olour hange was measured at 450 nm using a MR-5000 miroplate reader (Dynateh, Billingshurst, UK). AAT ativity was determined using an indiret hromogeni substrate assay (Quadrateh, Epsom, UK). Plasma was initially diluted in buffer ontaining methylamine to degrade x2-maroglobulin. AT-III was measured by an indiret hromogeni substrate method (Behring, Hounslow, UK) in whih thrombin was added to diluted plasma. The thrombin ativity remaining after Values of aute phase proteins and related parameters in patients with fulminant hepati failure on admission C reative protein AAT Fibrinogen A T-III Interleukin 1 Interleukin 6 (jig/mi) (o) (g/l) (o) (pg/ml) (pg/ml) All patients (n = 50) median 12.4* 69-1* 1.10* 22-8* * range Paraetamol overdose (n = 40) median 8 0*t 71.6* 1.16* 22-4*t * range Viral and other auses (n = 10) median 18.3* 61.5* 0-88* 13-1* * range Survivors (n = 23) median 13-3* 76-3** 1.45* 25-9* ** range Non-survivors (n = 20) median 11-7* 61-9* 1.01* 23-6* * range Controls (n = 20) median range AAT = a,-antitrypsin, AT-III = antithrombin-iii.*p<0 001, **p<0.01 v ontrols, tp<0 05 v viral and other auses, :p<0 05, and p<0-001 v non-survivors.

3 984 Izumi, Hughes, Langley, Pernambuo, Williams E -i 0) I.. a) C 0 E en Cu 80r 60K 40K 20H o Figure 2: Serial plasma onentrations of,-antitrypsin in Days of admission patients with fulminant hepati failure. Open irles: Figure 1f: Serial plasma onentrations of C reative protein survivors, filled irles: non-survivors. (Mean (SEM)). in patiei its with fulminant hepati failure. Open irles: *p<0 05; signifiant differenes between survivors and survivor,ats. J,t&r&G filled tirlesh r-tmta. non-urvivors rturts M-a"v$>urA. (Mean kirmurt fsfm)) non-survivors, Mann- Whitney U test. V1J inubation gave the AT-III ativity in the plasma. The olour hange was measured at 405 nm using a Ultrospe III spetrophotometer (Pharmaia LKB, Milton Keynes, UK). Fibrinogen was measured by the method of Clauss23 using a Fibrintimer (Behring, Hounslow, UK). Interleukin 6 was measured by the quantitative 'sandwih' enzyme immunoassay, Quantikine Human interleukin 6 Immunoassay obtained from British Biotehnology Produts (Abingdon, UK). This uses a monolonal antibody speifi for interleukin 6 oated onto a mirotitre plate. Interleukin 1 was assayed by ACE enzyme immunoassay based on a double antibody 'sandwih' tehnique kindly supplied by SPI BIO (Salay, Frane). ANALYSIS Signifiane was assessed within groups with analysis of variane and between groups using the Mann-Whitney U test. Correlation between variables was assessed by the Spearman rank orrelation oeffiient. Results are reported as median and range. Mean (SEM) is shown in the Figures. Results Plasma C reative protein on admission was signifiantly inreased ompared with ontrol subjets (Table II). Conentrations were signifiantly higher in patients with fulminant hepati failure resulting from viral hepatitis and auses other than paraetamol overdose (Table II). There were no signifiant differenes in C reative protein onentrations between survivors and non-survivors (Table II) or between 23 patients with and 27 without baterial infetion. Plasma C reative protein during the ourse of the illness was in general higher in non-survivors than in survivors from day 3 to day 5 (the end of the study period), although this did not reah statistial signifiane (Fig 1). Patients who reeived a liver transplant are not inluded in any serial data. Plasma AAT on admission was signifiantly lower in fulminant hepati failure than in ontrols (Table II). Although plasma onentrations of AAT were signifiantly higher in survivors than in non-survivors, the onentrations even in survivors were signifiantly lower than in ontrols (Table II). Plasma AAT was ,_ C.) Cu) 0 E en Cu Days of admission less redued than another protease inhibitor AT-III, whih is not an aute phase protein, and AT-Ill was also signifiantly lower in fulminant hepati failure than in ontrols (Table II). There were no signifiant differenes in plasma AAT between patients with and without baterial infetion, or between patients with fulminant hepati failure resulting from paraetamol overdose and viral hepatitis and other auses (Table II). The serial plasma onentrations of AAT did not hange after admission. AAT remained at signifiantly (p<005) higher onentrations in survivors than in non-survivors over the five days of study, and was maintained at more than 70% of normal (Fig 2). In ontrast with AAT, plasma AT-III in the overall fulminant hepati failure group ontinued to derease from the admission value to 10-3% of normal. Plasma fibrinogen on admission was also signifiantly lower in fulminant hepati failure than in ontrols (Table II). There were no signifiant differenes between patients with and without baterial infetion, or between patients with fulminant hepati failure resulting from paraetamol overdose and viral hepatitis and other auses or between survivors and non-survivors (Table II). Serial plasma onentrations of fibrinogen did not hange after admission. Survivors ontinued to have signifiantly higher fibrinogen onentrations than non-survivors from day 2 to day 5 and the values were maintained at more than 1-2 g/l. CYTOKINE CONCENTRATIONS AND CORRELATIONS WITH ACUTE PHASE PROTEINS On admission, plasma interleukin 6 was signifiantly higher in patients with fulminant hepati failure than in ontrols (Table II). Conentrations were signifiantly higher in patients who did not survive than in survivors (Table II), and no patient survived with an interleukin 6 onentration of more than 300 pg/ml. The pattern of interleukin 6 onentrations in fulminant hepati failure did not hange after admission. Interleukin 6 in nonsurvivors ontinued to be at signifiantly higher values than in survivors (day 2 and day 3; p<0-001 and p<0 05) (Fig 3). There was no signifiant differene in interleukin 6 onentrations on admission between patients with and without baterial infetion. There

4 Extent of the aute phase response in fulminant hepati failure 985 fulminant hepati failure ompared with other E, 1000 * aute linial situations. The response seen was similar irrespetive of whether the ause was viral hepatitis infetion or paraetamol X overdose. Viral infetion is not as strong a / stimulus for C reative protein prodution as baterial infetion.' 24 The seond is ommon in fulminant hepati failure,25 but in our study 10 there was no differene in C reative protein E 5 values between patients with ompliating X 1 2 ~ '~ baterial infetions and those without. In ful- Days of admission minant hepati failure, C reative protein Figure 3: Serial plasma onentrations of interleukin 6 in prodution may already be at a maximal rate, patients with fulminant hepati failure. Open irles: being limited by the severe loss of hepati survivors, filled irles: non-survivors. (Mean (SEM)). syntheti funtion part of the inflammatory **p<0o001, *p<005; signifiant differenes between survivors and non-survivors, Mann-Whitney U test. response to liver ell Further inrease invury. may not be possible even with the added stimulus of baterial infetion, as seen in our was also no signifiant differene between patients. Studies of patients with severe sepsis patients with fulminant hepati failure result- without liver failure have shown that C reative ing from paraetamol overdose and viral protein values are signifiantly higher in nonhepatitis and other auses on admission. survivors than in survivors,26 but in this study Plasma interleukin 1 onentrations on of patients with fulminant hepati failure there admission were not signifiantly different was no statistially signifiant differene between patients with fulminant hepati failure between plasma C reative protein values in and ontrols, between survivors and non- survivors and non-survivors. survivors, between patients with and without In the patients with fulminant hepati baterial infetion or between patients with failure, AAT and fibrinogen were dereased on fulminant hepati failure resulting from admission to 69 and 45% of normal values paraetamol overdose and viral hepatitis and respetively. The higher AAT and fibrinogen other auses (Table II). Both patients whose onentrations found in the survivors suggests interleukin 1 onentrations were more than that the prognosis is poor for those patients 20 pg/ml (24-4 and 144 pg/ml) had fulminant who annot maintain both AAT and fibrihepati failure resulting from hepatitis B virus nogen during the ourse of fulminant hepati infetion. failure, although there was some overlap The orrelation between interleukin 6 and between the two groups. Whether the derease C reative protein on admission was signifiant in values of the two proteins is also beause of in the patients with fulminant hepati failure their inreased onsumption is not lear. resulting from viral hepatitis (r=0-883, n=8, There is evidene for the latter in that p 0-004) but = not in the paraetamol overdose inreased values of elastase-aat omplexes patients (r= ). There was a orrelation are found in fulminant hepati failure7 and in all patients between interleukin 6 on admis- thus AAT, whih is the major inhibitor of sion and plasma C reative protein 48 hours elastase released from ativated neutrophils, is later (r= , n = 27, p = 0 034). onsumed in this proess. Similarly, dis- Interleukin 6 on admission orrelated seminated intravasular oagulation, whih inversely with AAT (r= , n = 38, onsumes fibrinogen, ours in patients with p 0010) and with fibrinogen (r= = , fulminant hepati failure In this study the n 38, = p = 0 022) onentrations 24 hours onentrations of AAT and fibrinogen were later. Interleukin 6 orrelated signifiantly greater than those of ATIII (23%), whih is with prothrombin time (r=0-291, n=47, also synthesised in the liver but is not an aute p 0-048) = on admission. Plasma interleukin 1 phase protein. This suggests that despite the orrelated signifiantly with fibrinogen on lower than normal onentrations of AAT and admission (r=0-56, n=28, p=0 002), but fibrinogen, there is a omparative inrease in not with the other aute phase proteins or synthesis of these proteins ompared with interleukin 6. ATIII; differenes in their rates of onsumption ould also be a fator in this. It would be diffiult to determine aurately the turnover Disussion of the individual proteins in an unstable ondi- These results show that an aute phase tion suh as fulminant hepati failure. response ours in fulminant hepati failure Interleukin 6 is known to be an important with an inrease in plasma C reative protein regulator of aute phase protein synthesis in and maintenane of higher than expeted humans Conentrations of interleukin 6 plasma onentrations of AAT and fibrinogen. were found to be inreased in patients with ful- In sepsis or trauma, plasma C reative protein minant hepati failure on admission to the rises rapidly to a maximum of 1000 times its intensive are unit. The ause of the inreased normal low onentration. In this study plasma interleukin 6 prodution is not lear, but was C reative protein onentrations were not related to the presene of baterial infeinreased by a fator of only 15 ompared with tion and thus probably does not result from ontrol values, showing that the extent of the endotoxin. It may be an inflammatory aute phase response is quite limited in response to massive ellular injury, involving

5 986 Izumi, Hughes, Langley, Pernambuo, Williams Kupffer and endothelial ells in the liver.29 In septi shok, interleukin 6 prodution peaked near the onset of sepsis and rapidly dereased to undetetable values within about 24 hours.30 The persistene of interleukin 6 in fulminant hepati failure ould point to dereased learane as an additional fator. The lak of orrelation between plasma interleukin 6 and serum reatinine would support the liver as being the main site of learane of interleukin 6, as previously reported.3' We found no evidene of inreased plasma interleukin 1 in this study, although in an earlier study inreased interleukin 1 and tumour nerosis fator prodution were found with fulminant hepati failure mononulear ells in vitro.19 In keeping with interleukin 6 being a stimulator of the aute phase response, a signifiant orrelation between plasma interleukin 6 and C reative protein has been desribed after surgery,32 33 in kidney rejetion,34 and in aute panreatitis.35 Peak values of interleukin 6 were found at two to four hours after an endotoxin hallenge in humans and this was assoiated with a rise in irulating C reative protein onentrations 24 hours after the endotoxin bolus.36 Similarly, after surgery maximum interleukin 6 values are attained 12 to 24 hours earlier than those of C reative protein.32 In this study, interleukin 6 orrelated with plasma C reative protein in the small group of patients with fulminant viral hepatitis on admission, and in all patients 48 hours later in the ourse of fulminant hepati failure, showing that C reative protein synthesis in fulminant hepati failure ould also be the result of the prior inrease in interleukin 6 onentrations. Thus, in fulminant hepati failure a strong stimulus is present for the aute phase response, but beause of the severely redued hepati funtion only limited protein synthesis ours. The ability of the liver to maintain a response suffiient to overome inreased onsumption of the aute phase proteins may be an important protetive mehanism. SI was a visiting linial researh fellow from the Department of Internal Mediine of Hyogo Prefetural Nishinomiya Hospital linked with The Seond Department of Internal Mediine, Osaka University Medial Shool, Japan. This study was presented in part as an abstrat at the British Soiety of Gastroenterology Meeting, University of Warwik, September Whiher JT, Evans SW. Aute phase proteins. Hospital Update 1990; 16: Kolb-Bahofen V. A review on the biologial properties of C-reative protein. Immunobiology 1991; 183: Ohlsson K, Olsson I. Neutral proteases of human granuloytes (III). Interation between human granuloyte elastase and plasma protease inhibitors. Sand 7 Clin Lab Invest 1974; 34: Dukert F. Behaviour of antithrombin III in liver disease. SandJ Gastroenterol 1973; 8 (suppl 19): Langley PG, Williams R. The effet of fulminant hepati failure on protein C antigen and ativity. Thromb Haemost 1985; 53: Langley PG, Williams R. Physiologial inhibitors of oagulation in fulminant hepati failure. Blood Coagul Fibrinolysis 1992; 3: Langley PG, Hughes RD, Rolando N, Williams R. Inreased elastase-,-antitrypsin omplex in fulminant hepati failure: relationship to baterial infetion and ativation of oagulation. Clin Chim Ata 1991;- 200: Delers F, Mangeney M, Raffa D, et al. Changes in rat liver mrna for s-1-aid-glyoprotein, apolipoprotein E, apolipoprotein B and beta-atin after mouse reombinant tumor nerosis fator injetion. Biohem Biophys Res Commun 1989; 161: Ramadori G, Sipe JD, Dinarello CA, Mizel SB, Colten HR. Pretranslational modulation of aute phase hepati protein synthesis by murine reombinant interleukin-1. JExp Med 1985; 162: Klapproth J, Castell JV, Geiger T, Andus T, Heinrih PC. Fate and biologial ation of human reombinant interleukin-1 in the rat in vivo. Eur J Immunol 1989; 19: Geiger T, Andus T, Klapproth J, Hirano T, Kishimoto T, Heinrih PC. Indution of rat aute-phase proteins by interleukin-6 in vivo. Eur J Immunol 1988; 18: Ramadori G, Van Damme J, Rieder H, Meyer zum Bushenfelde KH. Interleukin-6, the third mediator of aute-phase reation, modulates hepati protein synthesis in human and mouse: omparison with interleukin-1l and tumor nerosis fator a. Eur J Immunol 1988; 18: Marinkovi S, Jahreis GP, Wong GG, Baumann H. IL-6 modulates the synthesis of a speifi set of aute phase hepati proteins in vivo. J Immunol 1989; 142: Castell JV, Gomez-Lehon MJ, David M, Hirano T, Kishimoto T, Heinrih PC. Reombinant human interleukin-6 (IL-6/BSF-2/HSF) regulates the synthesis of aute phase proteins in human hepatoytes. FEBS Lett 1988; 232: Castell JV, Gomez-Lehon MJ, David M, et al. Interleukin- 6 is the major regulator of aute phase protein synthesis in adult human hepatoytes. FEBS Letu 1989; 242: Castell JV, Gomez-Lehon MJ, David M, Fabra R, Trullenque R, Heinrih PC. Aute-phase response of human hepatoytes: regulation of aute-phase protein synthesis by interleukin-6. Hepatology 1990; 12: Sheron N, Goka J, Wendon J, et al. Highly elevated plasma ytokines in fulminant hepati failure: orrelations with multiorgan failure and death. Hepatology 1990; 12: Shiratori Y, Moriwaki H, Kawashima Y, et al. Elevated interleukin-6 levels in sera of patients with fulminant hepatitis. GastroenterolJpn 1991; 26: Muto Y, Nouri-Aria K, Meager A, Alexander GJM, Eddleston ALWF, Williams R. Enhaned tumour nerosis fator and interleukin-1 in fulminant hepati failure. Lanet 1988; ii: Trey C, Davidson CS. The management of fulminant hepati failure. In: Popper H, Shaffner F, eds. Progress in liver disease. Vol 3. New York: Grune & Stratton, 1970: O'Grady JG, Alexander GJM, Hayllar KM, Williams R. Early indiators of prognosis in fulminant hepati failure. Gastroenterlogy 1989; 97: Highton J, Hessian P. A solid phase enzyme immunoassay for C-reative protein: Clinial value and the effet of rheumatoid fator. Jf Immunol Methods 1984; 68: Clauss A. Gerinnungsphysiologishes shnellmethode zur bestinnung des fibrinogens. Ata Haematol 1957; 17: Kushner I. The aute phase response: an overview. Methods Enzymol 1988; 163: Rolando N, Harvey F, Brahm J, et al. Baterial infetion in aute liver failure. Hepatology 1990; 11: Dofferhoff ASM, Bom VJJ, van Ingen J, et al. Baterial endotoxin: ytokine mediators and new therapies for sepsis. New York: Wiley-Liss, 1991: Hughes RD, Lane DA, Ireland H, Langley PG, Gimson AES, Williams R. Fibrinogen derivatives and platelet ativation produts in aute and hroni liver disease. Clin Si 1985; 68: Langley PG, Forbes AF, Hughes RD, Williams R. Thrombin-antithrombin III omplex in fulminant hepati failure: evidene for disseminated intravasular oagulation and relationship to outome. EurJ Clin Invest 1990; 20: Jirik FR, Podor TJ, Hirano T, et al. Baterial lipopolysaharide and inflammatory mediators augment IL-6 seretion by human endothelial ells. Jf Immunol 1989; 42: Calandra T, Gerain J, Heumann D, Baumgartner JP, Glauser MP, and the Swiss Duth J5 Immunoglobulin Study Group. High irulating levels of interleukin-6 in patients with septi shok: evolution during sepsis, prognosti value, and interplay with other ytokines. AmJMed 1991; 91: Castell JV, Geiger T, Gross V, et al. Plasma learane, organ distribution and target ells of interleukin-6/hepatoyte stimulating fator in the rat. Eur J Biohem 1988; 177: Oka Y, Murata A, Nishijima J, et al. Cirulating interleukin 6 as a useful marker for prediting postoperative ompliations. Cytokine 1992; 4: Ohzato H, Yoshizaki K, Nishimoto N, et al. Interleukin-6 as a new indiator of inflammatory status: detetion of serum levels of interleukin-6 and C-reative protein after surgery. Surgery 1992; 111: Van Qers MHJ, Van der Heyden AAPAM, Aarden LA. Interleukin-6 (IL6) in serum and urine of renal transplant reipients. Clin Exp Immunol 1988; 71: Heath DI, Cruikshank A, Gudgeon M, Jehanli A, Shenkin A, Imrie CW. Role of interleukin-6 in mediating the aute phase protein response and potential as an early means of severity assessment in aute panreatitis. Gut 1993; 34: Fong Y, Moldawer LI1, Marano M, et al. Endotoxemia eliits inreased irulating bl1-ifn/il-6 in man. Jlmmunol 1989; 142: