Critical care medicine.
`DUP--General Collection
`
`e—IVIet-Iiei1=1e,@«l
`
`www.ccmjoumalmg
`
` OFFICIAL JOURNAL OF THE SOCIETY OI’ CRITICAL CARE MEDICINE
`
`August 2002
`
`Volume 30, Number 8
`
`FEATURE ARTICLES
`Race and gender differences in acute respiratory distress syndrome deaths in the United
`States: An analysis of multiple—cause mortality data (1979—1996)
`Marc Moss et (11.
`
`Meta—analysis of hemodynamic optimization inhigh—risk patients
`Jack W. Kern et al.
`
`Comparison of multiple organ dysfunction scores in the prediction of hospital mortality
`in the critically ill
`
`Ville Pettzlc‘z' et al
`Solubletumor necrosis factor receptor p55 predicts cytokinemia and systemic
`inflammatory response after cardiopulmonary bypass
`
`Mahmoud ElBarbary et al.
`
`LABORATORY INVESTIGATIONS
`
`Improved resuscitation minimizes respiratory dysfunction and blunts interleukin-6 and
`nuclear factor— KB activation after traumatic hemorrhage
`
`Jeffrey A. Clarzdgeet al.
`Effect of pietreatment with interleultin 1B on inflammatory infiltrates and tissue
`damage after experimental endotoxic challenge
`Luis Te’llez-Gil et al.
`
`NEUROLOGIC CRITICAL CARE
`
`Management of severe head injury: Institutional variations in care and effect on
`outcome
`
`Eileen M. Bulger et (Il.
`
`PEDIATRIC CRITICAL CARE
`
`Pilot study of antibiotic cycling in a pediatric intensive care unit
`é”;
`William J. Moss et al.
`5/.
`V
`LIPPINCOTT
`‘
`WILL'AMS GW‘LK'NS Complete Table of Contents Inside
`7/IM“\
`‘
`‘
`”BL-.YOF
`Thisrmaterial wascopied
`atthe NLMaI‘Id maybe
`mi
`
`Su ijEtt U-S Cum-right Laws
`
`PROPERTY OF THE
`ONAL
`NAT]
`LIBRARY OF
`MEDICINE
`
`Page 1 of 9
`
`CSL EXHIBIT 1043
`
`Page 1 of 9
`
`CSL EXHIBIT 1043
`
`

`

`'
`
`- onchL CARE MEDICINE
`‘ Official Journal of the Sobiety of‘Critica'l Care Medicine j
`AUGUST 2002NOLUME 30/NUMBER 3
`
`y
`
`.
`
`FEATURE ARTICLES
`
`Race and gender differences in acute respiratory distress syndrome deaths in the United
`States: An analysis of multiple—cause mortality data (1979-1996) .................................................. 1679
`Marc Moss, David M. Mannino
`
`Meta-analysis of hemodynamic optimization in high-risk patients .................................................. 1686
`Jack W. Kern, William C. Shoemaker
`
`lnjurious ventilation induces widespread pulmonary epithelial expression of tumor necrosis
`factor-a and interleukin-6 messenger RNA ........................................................................................ 1693
`Lorraine N. Tremblay, Debra Miatto, Dutayba Hamid, Anand Govindarajan, Arthur S. Slutsky
`
`Acidification of formula with citric acid is equally effective and better tolerated than
`acidification with hydrochloric acid ..................................................................................
`John R. Mehall, Daniel A. Saltzman, Richard J. Jackson, Samuel D. Smith
`
`................ 1701
`
`CLINICAL INVESTIGATIONS
`
`Comparison of multiple organ dysfunction scores in the prediction of hospital mortality in the
`critically ill ......................................................
`.......................
`....................................... ........ .......... 1705
`Ville Patti/a, Markus Patti/a", Seppo Sarna, Petri Voutilainen, Oll/ Takkunen
`
`Soluble tumor necrosis factor receptor p55 predicts cytokinemia and systemic inflammatory
`response after cardiopulmonary bypass
`......................
`.............................................................
`Mahmoud ElBarbary, Khalid S. A. Khabar
`
`1712
`
`increased neutrophil migratory activity after major trauma: A factor in the etiology of acute
`respiratory distress syndrome? ................
`..................
`....................................................
`........ 1717
`lan Pallister, Colin Dent, Nicholas Top/ey
`
`I’OSTMASTISR
`
`© 2002 by Lippincott Williams & Wilkins.
`
`CRITICAL CARE MEDICINE (CCM/ISSN 0090-3493) is the official journal of the Society of Critical Care Medicine. and is published monthly
`(one volume a year beginning in January) by Linnincott Williams & Wilkins, 16522 Hunters Green Parkway,
`[liltieTStOWm MD 21740—2116.
`Business omees are located at 530 Walnut Street, Philadelphia, PA 19106—3621- Productlon offices are located at 351 West Camden Street.
`Baltimore, MD 212012436. Subscription Rates: SCCM Members: Annual dues include $65.00 for Journal subscription. Nonmembers: U.S.:
`I’ersonal $219.00; Institutional $341.00; Single copy $36.00. Outside the U.S.. excent Japan:
`Personal $302.00; Institutional $425.00; Single
`copy $36.00. Special in-training rate of $143.00 ($226.00 outside the U.S.). Foreign prices Exclude Japan. The CST number for Canadian subscribers
`is 893524239. The Canadian Publication Agreement Number is 40052291. Country Of origin USA. PRICES ARE SUBJECT TO CHANGE
`WITHOUT NOTICE. See Information for Subscribers for detailed instructions. Periodicals postage paid at IIagerstmm MD, and at additional
`mailing offices.
`I’OSTMASTER: Send address changes to CRITICAL CARE MEDICINE, PO. BOX 1550, Ilagcrstown. MI), 21741. Copyright
`
`Th is mate rial was taxes-pied
`at the NLM and may tea
`Subject; US C’prrighr.‘ Laws
`
`Page 2 of 9
`
`Page 2 of 9
`
`

`

`C1-inhibitor in patients with severe sepsis and septic shock:
`Beneficial effect on renal dysfunction
`
`Christoph Caliezi, MD; Sacha Zeerleder, MD; Maurice Redondo, MD; Bruno Regli, MD;
`Hans-Ulrich Rothen, MD, PhD; Regula Zürcher-Zenklusen, MD; Robert Rieben, PhD; Jan Devay, PhD;
`C. Erik Hack, MD; Bernhard Lämmle, MD; Walter A. Wuillemin, MD, PhD
`
`Objective: To investigate the efficacy and the safety of the
`parenteral administration of C1-inhibitor to patients with severe
`sepsis or septic shock.
`Design: Double blind, randomized, and placebo-controlled
`trial.
`Setting: Surgical and medical intensive care units of a tertiary
`care university hospital.
`Patients: Forty consecutive patients (20 C1-inhibitor/20 pla-
`cebo) who entered the intensive care unit with severe sepsis or
`septic shock.
`Intervention: C1-inhibitor intravenously in a 1-hr infusion,
`starting with 6000 IU, followed by 3000 IU, 2000 IU, and 1000 IU
`at 12-hr intervals, compared with placebo.
`Measurements and Main Results: C1-inhibitor administration
`
`significantly increased plasma C1-inhibitor antigen and activity
`levels during days 1– 4 (p < .007). Patients in the C1-inhibitor
`group had significantly lower serum creatinine concentrations on
`day 3 (p ⴝ .048) and 4 (p ⴝ .01) than placebo patients. Multiple
`organ dysfunction assessed by logistic organ dysfunction and
`sepsis-related organ failure assessment scores was less pro-
`nounced in patients treated with C1-inhibitor. Mortality rate was
`similar in both groups. There were no C1-inhhibitor-related side
`effects.
`Conclusions: C1-inhibitor administration attenuated renal im-
`pairment in patients with severe sepsis or septic shock. (Crit Care
`Med 2002; 30:1722–1728)
`KEY WORDS: severe sepsis; septic shock; C1-inhibitor; comple-
`ment system; renal failure; multiple organ dysfunction
`
`S epsis, severe sepsis, and septic
`
`shock are degrees of increasing
`severity of the inflammatory
`response to infection (1). Sep-
`sis often is induced by bacterial infections
`and is a leading cause of death in noncar-
`diologic intensive care units. Despite the
`use of potent antibiotics and improved
`intensive supportive care, the mortality
`rate of patients with severe sepsis and
`septic shock remains high, varying from
`20% to 50% (1). Sepsis is supposed to
`
`From the Central Laboratory of Hematology (CC,
`SZ, MR, RR, BL) and Department of Intensive Care (BR,
`H-UR, RZ-Z), University Hospital,
`Inselspital, Bern,
`Switzerland; Aventis Behring AG (JD), Zürich, Switzer-
`land; Central Laboratory of the Netherlands Red Cross
`Blood Transfusion Service and Department of Internal
`Medicine (CEH), Academic Hospital of the Free Univer-
`sity Amsterdam, Amsterdam, The Netherlands; and
`Division of Hematology (WAW), Department of Internal
`Medicine, Kantonsspital Luzern, Switzerland.
`Supported, in part, by grant 32-55312.98 from the
`Swiss National Foundation for Scientific Research and
`by an unrestricted grant
`from Aventis-Behring AG,
`Zürich, Switzerland.
`reprints to: Walter A.
`Address requests for
`Wuillemin, MD, PhD, Division of Haematology, Depart-
`ment of Internal Medicine, Kantonsspital, 6000 Luzern-
`16, Switzerland. E-mail: walter.wuillemin@KSL.ch
`Copyright © 2002 by Lippincott Williams & Wilkins
`
`1722
`
`result from the excessive activation of
`defense mechanisms, including the com-
`plement and contact activation systems,
`resulting in the release of endogenous
`inflammatory mediators (2). In patients
`with septic shock, significantly decreased
`plasma concentrations of FXII, FXI, and
`prekallikrein have been demonstrated
`(3). Activation of the contact system does
`not primarily lead to clotting activation
`but likely contributes to vasodilation by
`generation of kallikrein and subsequently
`bradykinin, a potent vasoactive peptide
`(4). Elevated concentrations of the acti-
`vated complement components C3a and
`C4a have been demonstrated in patients
`with sepsis and septic shock, and the
`plasma concentrations of C3a were sig-
`nificantly correlated with mortality rate
`(5).
`C1-inhibitor (C1-inh) is the only
`known inhibitor of the activated compo-
`nents C1s and C1r of the classic pathway
`of complement and a major inactivator of
`FXIIa, FXIa, and kallikrein (6, 7). C1-inh
`has a plasma half-life of about 28 hrs (8).
`It belongs to the superfamily of serine
`proteinase inhibitors (serpins) which also
`includes ␣1-antitrypsin, antithrombin,
`
`and plasminogen activator inhibitor type
`I. C1-inh is an acute phase protein, the
`plasma concentration of which may in-
`crease up to two-fold during infections
`(3). Like most other serpins, C1-inh can
`be inactivated by elastase released from
`activated neutrophils. Thus, inactivation
`of C1-inh may predominantly occur lo-
`cally in inflamed tissues (9). In patients
`with sepsis, an increase of inactivated,
`cleaved C1-inh compared with healthy
`volunteers has been demonstrated, and
`the concentrations of cleaved C1-inh
`were found to be positively correlated
`with mortality rate (8).
`C1-inh therapy was introduced into
`clinical medicine about 25 yrs ago as a
`replacement therapy for patients with he-
`reditary C1-inh deficiency, known as he-
`reditary angioedema (10). Preliminary
`evaluation of C1-inh therapy in a few
`patients with septic shock has shown an
`attenuation of complement and contact
`activation, a beneficial effect on hypoten-
`sion, and no toxic side effects during the
`study periods (11).
`We intended to investigate the efficacy
`and safety of the administration of C1-inh
`in patients with severe sepsis or septic
`
`Crit Care Med 2002 Vol. 30, No. 8
`
`Page 3 of 9
`
`

`

`shock and to look for clinical and labora-
`tory variables indicative of the effects of
`C1-Inh.
`
`MATERIAL AND METHODS
`
`The efficacy and safety of C1-inh were as-
`sessed in a randomized, double-blind, placebo-
`controlled study and were analyzed according
`to a prospectively developed protocol that used
`definitions adopted before the treatment allo-
`cation code was broken. The number of study
`participants was limited to 40 patients. Pri-
`mary end points were the effect of C1-inh on
`routine clinical and laboratory variables. Mor-
`tality analysis was a secondary end point only.
`A block randomization with patient blocks of n
`⫽ 10 for medical and surgical patients was
`used. The hospital pharmacy was responsible
`for randomizing and labeling the blinded vials
`and delivering the study medication to the
`responsible intensive care unit. Study audits
`were carried out by an independent center for
`study surveillance. The protocol was approved
`by the local ethics committee. The study was
`performed according to the guidelines of Good
`Clinical Practice and in accordance to the re-
`quests and prerequisites of the Declaration of
`Helsinki and its 1989 revision of Hong Kong.
`Informed consent was obtained from all pa-
`tients. In case of impaired consciousness, in-
`formed consent was obtained from a family
`member or the closest relative or partner of
`the patient, according to the national legal
`guidelines.
`Patients of the medical and surgical inten-
`sive care unit of the University Hospital Bern,
`Switzerland, were eligible if they met the in-
`clusion criteria for severe sepsis or septic
`shock according to the definitions of the
`American College of Chest Physicians consen-
`sus conference within a time period of 12 hrs
`(12). Patients were not eligible if they fulfilled
`one of the following exclusion criteria: age
`⬍18 yrs; pregnant or breast feeding; medical
`or surgical baseline condition with a suspected
`survival time of ⬍28 days; restriction to none
`or only limited cardiopulmonary resuscitation
`in case of deterioration; myocardial infarction
`in the last 7 days; cardiogenic shock; major
`bleeding with hemodynamic instability during
`the last 48 hrs; third-degree skin burn injury
`including ⬎20% of body surface; chronic re-
`nal
`insufficiency with need for dialysis;
`chronic liver insufficiency Child-Pugh B or C;
`known allergy against plasma proteins; known
`hereditary C1-inhibitor deficiency; treatment
`with other study medications in the previous
`30 days; or participation in another study.
`Patients enrolled in the trial were assigned
`randomly to receive either C1-inh or placebo
`(1% solution of human serum albumin). The
`study medication consisted of a purified and
`sterilized lyophilisate of human C1-inhibitor
`(Berinert HS, Aventis Behring AG, Zürich,
`
`Crit Care Med 2002 Vol. 30, No. 8
`
`Switzerland). The content of each vial (500 IU)
`had to be dissolved with 10 mL of physiologic
`0.9% sodium chloride solution before intrave-
`nous application. The patients were given
`6000 IU, followed by 3000 IU, 2000 IU, and
`1000 IU, respectively, at 12-hr intervals in a
`continuous infusion over 1 hr. The total dose
`administered was 12,000 IU or 2.88 g for C1-
`inh and 2.4 g for albumin, respectively. The
`additional fluid volume administered during
`36 hrs was 240 mL. A mixture with other
`infusions had to be avoided. The first dosage
`had to be given within a 12-hr period after
`study inclusion. Decisions regarding the use
`of antibiotics, intravenous fluid substitution,
`cardiovascular and respiratory support includ-
`ing vasopressor therapy, and surgical inter-
`ventions were left to the discretion of the
`attending physician and were not dictated by
`the study protocol. The vasopressor most of-
`ten used in these patients was norepinephrine.
`Patients were followed for 90 days or until
`death. The Simplified Acute Physiology Score
`II,
`logistic organ dysfunction (LOD) score
`(13), and sepsis-related organ failure assess-
`ment (SOFA) score (14) were calculated at
`study entry. Diagnostic investigations of sus-
`pected foci of infection (e.g., cultures of blood
`and tissue) were performed according to the
`decisions of the attending physician. Total
`fluid balance, vasopressor medication, respira-
`tory support, and LOD and SOFA scores were
`assessed daily from day 1 to day 4. LOD and
`SOFA score systems characterize the degree of
`organ dysfunction by a point score including
`the cardiovascular, pulmonary, neurologic, re-
`nal, hematologic, and hepatic systems. The
`range varies from 0 to 22 points in the LOD
`score and from 4 to 24 points in the SOFA
`score, the highest score of each system depict-
`ing severe dysfunction of all six organ systems.
`Vasopressor medication was registered as the
`24-hr amount of administered drug in milli-
`grams or micrograms per kilogram of body
`weight and was analyzed separately for differ-
`ent vasopressors like norepinephrine, epi-
`nephrine, dopamine, and dobutamine. The
`fluid balance was assessed as the 24-hr differ-
`ence between fluid administration and fluid
`loss per kilogram of body weight. Mortality
`was assessed on days 28 and 90. Unexpected
`adverse events and serious adverse events were
`recorded during the initial 28 days of the study
`period. Serious adverse events were defined
`according to the guidelines of Good Clinical
`Practice.
`Laboratory Analyses. Routine hemato-
`logic, chemistry, and hemostasis laboratory
`variables were obtained before the first infu-
`sion of the study medication and daily there-
`after until day 4. Plasma for the determination
`of C1-inh antigen and C1-inh activity was ob-
`tained before application of the study medica-
`tion, 1 hr after completion of the first infu-
`sion, and on days 1, 2, 3, and 4. C1-inh antigen
`
`was assessed by a nephelometric assay that
`used specific sheep antibodies directed against
`human C1-inh (normal range, 0.21– 0.39 g/L).
`C1-inh activity was measured by a chromo-
`genic assay that used the complement compo-
`nent C1s as a target protease (normal range,
`70 –130%). The complement component
`C4b/c, a characteristic activation product of
`the classic complement pathway, was mea-
`sured by an enzyme-linked immunosorbent
`assay by using a monoclonal antibody against
`a neoepitope on activated C4 as described else-
`where (15).
`Definitions and Criteria. The diagnosis of
`SIRS, sepsis, severe sepsis, and septic shock
`was established according to the definitions of
`the American College of Chest Physicians con-
`sensus conference (12). In addition, the fol-
`lowing definitions were used. Hypotension
`was defined as systolic blood pressure ⱕ90
`mm Hg or a decrease of systolic blood pres-
`sure ⱖ40 mm Hg for ⬎1 hr in the presence of
`adequate fluid challenge and in the absence of
`antihypertensive agents. Renal
`insufficiency
`was defined as one of the following criteria:
`urinary output ⬍0.5 mL/kg of body weight per
`hour or ⬍30 mL per 2 hrs; serum creatinine
`⬎180 ␮mol/L (2 mg/dL), or need for dialysis.
`Lactic acidosis was defined as a concentration
`of serum lactate ⬎2.2 mmol/L and an arterial
`pH of ⬍7.30 in the absence of respiratory
`acidosis. Central nervous system impairment
`was defined as a Glasgow Coma Scale score
`⬍11 points or a decrease of ⬎2 points com-
`pared with the baseline value. Liver insuffi-
`ciency was defined as the presence of two of
`the following criteria: total bilirubin concen-
`tration ⬎43 ␮mol/L; aspartate aminotransfer-
`ase or alanine aminotransferase concentra-
`tions more than twice the upper normal value;
`activated partial thromboplastin time ⬎1.5
`times the upper normal value (16). Dissemi-
`nated intravascular coagulation (DIC) was de-
`fined by all of the following criteria: platelet
`count ⬍100 ⫻ 109/L, activated partial throm-
`boplastin time ⬎1.5 times the upper normal
`value, and factor V clotting activity ⬍70% (17,
`18). Major bleeding was defined as bleeding
`with a decrease of hemoglobin concentration
`of ⬎20 g/L within 24 hrs, the need for trans-
`fusion of more than two units of packed red
`cells during 24 hrs, or any life-threatening or
`intracranial hemorrhage.
`Statistical Analysis. Data were expressed
`either as median and range or as mean ⫾ SEM.
`Differences between groups for quantitative
`data were analyzed by the nonparametric, un-
`paired, two-tailed Mann-Whitney U test (Wil-
`coxon rank sum test). Qualitative data were
`compared by nonparametric Fisher’s exact
`test. Survival rates were computed by using
`the method of Kaplan-Meier. Correlation anal-
`yses were done by Spearman rank correlation
`by using SAS Proc Mixed software (SAS pack-
`age, release 6.12, SAS Institute, Cary, NC; co-
`efficient of correlation ⫽ rs). The ␣ level, be-
`low which findings were
`considered
`statistically significant, was set at .05.
`
`1723
`
`Page 4 of 9
`
`

`

`RESULTS
`
`From June 1998 to June 1999, 96 con-
`secutive patients with severe sepsis or
`septic shock were evaluated. Forty pa-
`tients were enrolled into the study, 20
`each in the C1-inh and placebo groups.
`Fifty-six patients had to be excluded be-
`cause of positive exclusion criteria (n ⫽
`41), lack of informed consent (n ⫽ 11), or
`participation in another study (n ⫽ 4).
`Information was available on all of the
`participants through day 90. Inclusion
`criteria were severe sepsis in 16 and sep-
`tic shock in four patients of each group.
`The C1-inh and placebo groups were well
`balanced with respect to demographic
`characteristics and underlying diseases
`except for a statistically nonsignificant
`higher prevalence of coronary artery dis-
`ease in the C1-inh group (Table 1). No
`significant differences existed in baseline
`Simplified Acute Physiology Scores, LOD
`scores, and SOFA scores (Table 2). At
`study entry, the number of patients with
`liver insufficiency, renal
`insufficiency,
`DIC, and need for vasopressor therapy
`and mechanical ventilation and the in-
`tensive care unit length of stay before
`enrollment in both groups was compara-
`ble (Table 2). Significantly more patients
`in the C1-inh group had an intra-
`abdominal source of infection (nine vs.
`two, p ⫽ .03) Other sources of docu-
`mented infections were similar: respira-
`tory tract, six vs. 12 in the C1-inh and
`placebo groups, respectively; heart or
`mediastinum, three vs. two; urinary
`tract, one vs. one; sinus or meninges,
`zero vs. two; and skin, two vs. one. The
`main causative organisms in both
`groups were Staphylococci and Strep-
`tococci (Table 3). Positive blood cul-
`tures indicating bacteremia were docu-
`mented in seven patients of the C1-inh
`group and in two patients of the placebo
`group (p ⫽ .07). There were no signif-
`icant differences in the baseline labora-
`tory values at study entry with the ex-
`ception of alanine aminotransferase,
`the median concentration of which was
`significantly higher in the placebo
`group than in the C1-inh group (Table
`4). All patients received the full dosage
`of the study medication.
`Baseline plasma concentrations of C1-
`inh antigen and C1-inh activity were sim-
`ilar in both groups (Table 4). C1-inh ad-
`ministration significantly increased C1-
`inh antigen and C1-inh activity from days
`1 to 4 compared with the placebo pa-
`tients (p ⬍ .007; Fig. 1, A and B). As a
`
`result of the inhibition of the classic
`pathway of complement C4b/c, concen-
`trations in the C1-inh group decreased
`significantly on days 1–3 compared
`with the placebo group (p ⬍ .016; Fig.
`1C). In both treatment groups, C1-inh
`
`antigen and C1-inh activity were corre-
`lated negatively with LOD and SOFA
`scores. This correlation was significant
`for C1-inh activity with SOFA as well as
`with LOD score during days 1– 4 (rs ⬎
`.34, p ⬍ .029).
`
`Table 1. Demographic characteristics of patients
`
`Characteristics
`
`Age, yrs, median (range)
`Sex, male/female
`BMI, kg/m2, median (range)
`Medical/surgical, no.
`Underlying diseases, no.
`Neoplasm
`Diabetes mellitus
`Hypertension
`Coronary heart disease
`Alcoholism
`Immunosuppressiona
`Recent surgery
`
`Study Group
`
`Placebo
`(n ⫽ 20)
`
`61.2 (28–74)
`16/4
`27.9 (21–40)
`9/11
`
`C1 Inhibitor
`(n ⫽ 20)
`
`67.0 (41–74)
`17/3
`26.1 (15–43)
`9/11
`
`6
`1
`6
`2
`3
`4
`9
`
`8
`2
`6
`8
`2
`5
`9
`
`BMI, body mass index.
`aHuman immunodeficiency virus/acquired immunodeficiency syndrome,
`treatment (cytotoxic chemotherapy, steroids).
`
`immunosuppressive
`
`Table 2. Patients characteristics at study entry
`
`Study Group
`
`Characteristics
`
`Severe sepsis, no.
`Septic shock, no.
`Vasopressor therapy, no.
`Mechanical ventilation, no.
`SAPS II score, patients, median (range)
`LOD score, patients, median (range)
`SOFA score, patients, median (range)
`Days in ICU before enrollment, median (range)
`Liver insufficiency, no.
`Renal insufficiency, no.
`DIC, no.
`
`Placebo
`(n ⫽ 20)
`
`16
`4
`19
`19
`44.5 (28–59)
`5 (1–11)
`8 (6–18)
`2.0 (0–14)
`7
`8
`3
`
`C1 Inhibitor
`(n ⫽ 20)
`
`16
`4
`18
`17
`46.0 (28–68)
`4 (2–14)
`8 (4–16)
`1.0 (0–5)
`5
`8
`2
`
`SAPS, Simplified Acute Physiology Score; LOD, logistic organ dysfunction; SOFA, Sepsis-related
`organ failure assessment; ICU, intensive care unit; DIC, disseminated intravascular coagulation.
`
`Table 3. Isolates in culture-positive patients
`
`Isolate
`
`Staphylococcus aureus
`Streptococci
`Other Gram-positive
`Escherichia coli
`Pseudomonas
`Other Gram-negative
`Othera
`Culture positive
`Blood cultures positive
`
`Study Group
`
`Placebo
`(n ⫽ 20)
`
`C1 Inhibitor
`(n ⫽ 20)
`
`3
`3
`0
`2
`2
`5
`2
`15
`2
`
`6
`2
`2
`1
`0
`2
`3
`14
`7
`
`aParainfluenza virus, Mycoplasma pneumoniae, Mycobacterium tuberculosis, Candida albicans.
`
`1724
`
`Crit Care Med 2002 Vol. 30, No. 8
`
`Page 5 of 9
`
`

`

`Table 4. Baseline laboratory values, median (range)
`
`Study Group
`
`Data
`
`Hb, g/L
`Leukocytes, 10⫺9/L
`Platelets, 10⫺9/L
`CRP, mg/L
`Creatinine, ␮mol/L
`Albumin, g/L
`Fibrinogen, g/L
`ASAT, U/L
`ALAT, U/L
`pH
`aPTT, secs
`Factor V:C, %
`C1 inhibitor antigen, g/L
`C1 inhibitor activity, %
`
`Placebo
`(n ⫽ 20)
`
`92.0 (76–131)
`13.1 (1.5–24.0)
`173
`(28–484)
`214
`(4–417)
`126
`(77–717)
`17
`(11–27)
`5.5 (1.4–7.0)
`50
`(10–3340)
`35
`(13–1380)
`7.36 (7.09–7.42)
`49.9 (29.6–94.9)
`80
`(34–144)
`0.32 (0.16–0.60)
`90
`(40–209)
`
`C1 Inhibitor
`(n ⫽ 20)
`
`99.5 (78.0–145)
`9.4 (2.8–24.6)
`161
`(7–334)
`179
`(41–444)
`118
`(67–477)
`18
`(9–38)
`4.5 (0.9–7.7)
`41
`(6–447)
`(5–146)a
`20
`7.37 (7.23–7.50)
`51.1 (34.0–95.5)
`83
`(45–154)
`0.30 (0.16–0.53)
`82
`(41–171)
`
`ap ⫽ .02.
`Hb, hemoglobin; CRP, C-reactive protein; ASAT, aspartate transaminase; ALAT, alanine transam-
`inase; aPTT, activated partial thromboplastin time; V:C, clotting activity.
`
`The LOD and SOFA scores on days
`1– 4 were lower in the C1-inh group than
`in the placebo group, the difference being
`significant for LOD score on day 4 (p ⫽
`.043; Fig. 2, A and B). LOD and SOFA
`scores were positively correlated with
`each other (rs ⬎ .72, p ⬍ .0001). Fifteen
`patients in the C1-inh group and 16 in
`the placebo group were treated with nor-
`epinephrine. Dobutamine was given to 12
`and seven patients, dopamine to six and
`six patients, and epinephrine to two and
`three patients in the C1-inh and placebo
`groups, respectively. The patients in the
`C1-inh group required less daily amounts
`of norepinephrine per kilogram of body
`weight than the placebo patients: median,
`11 ␮g/kg per 24 hrs (range, 0 –3916) vs.
`57 ␮g/kg per 24 hrs (0 –922) (p ⫽ non-
`significant). There was no significant dif-
`ference in the daily fluid balance among
`the patients in the C1-inh and placebo
`groups. Serum creatinine decreased in
`the C1-inh group from days 1 to 4,
`whereas it slightly increased in the pla-
`cebo patients, the difference being signif-
`icant on days 3 and 4 (p ⫽ .048 and .017,
`respectively; Fig. 2C). Likewise, plasma
`concentrations of blood urea nitrogen
`were lower in the C1-inh compared with
`the placebo group, being significant on
`day 4: 11.0 mg/dL (6.4 –25.7) vs. 24.1
`mg/dL (3.1–55.9 mg/dL) (p ⫽ .046).
`Three patients in each group needed di-
`alysis before the first application of the
`study medication. During days 1– 4, five
`additional patients of the placebo group
`(30%) and three additional patients of the
`C1-inh group (18%) had to receive dial-
`
`Crit Care Med 2002 Vol. 30, No. 8
`
`ysis (p ⫽ nonsignificant). Three patients
`in each group developed liver insuffi-
`ciency and one patient in each group
`developed DIC during days 1– 4. There
`was no significant difference in the
`course of hemoglobin, hematocrit, leuko-
`cytes, platelets, sodium, potassium, as-
`partate aminotransferase, alanine amino-
`transferase, bilirubin, C-reactive protein,
`albumin, protein, prothrombin time, ac-
`tivated partial thromboplastin time, fi-
`brinogen, and factors II:C, V:C, VII:C, and
`X:C between the study groups.
`By day 28, six patients in the C1-inh
`group and four patients in the placebo
`group had died (p ⫽ nonsignificant). The
`90-day mortality rate was seven and six
`patients in the C1-inh and placebo
`groups, respectively (Fig. 3). All deaths
`were attributable to septic shock with the
`exception of one patient in the C1-inh
`group who died of intraoperative aortic
`rupture during cardiac surgery after sep-
`tic endocarditis. In both groups, patients
`with septic shock at study entry had a
`28-day mortality rate of 50% (two of
`four). There was no significant difference
`in proven bacteremia, Gram-negative
`stain of the causative organism, and
`whether a causative organism was de-
`tected among survivors and nonsurvi-
`vors. Nonsurvivors at 28 days had a sig-
`nificantly higher baseline Simplified
`Acute Physiology Score II score (median,
`52; range, 39 – 60 vs. median, 41; range,
`28 – 68; p ⫽ .028) and baseline SOFA
`score (median, 12.0; range, 7–14 vs. me-
`dian, 8.0; range, 4 –18, p ⫽ .002). Fatali-
`ties beyond the initial 28 days (C1-inh,
`
`Figure 1. Levels of C1-inhibitor antigen (C1-Inh-
`Ag; A), C1-inhibitor activity (C1-Inh-Act; B), and
`C4b/c (C) during the study period. Values in the
`C1-inh group (filled circles) and the placebo
`group (unfilled circles) are indicated as mean ⫾
`SEM. Normal values for C4b/c are ⬍8 nM. *p ⬍
`.05 for differences between the C1-inh and pla-
`cebo group.
`
`one patient; placebo, two patients) were
`all caused by sequelae of the initial septic
`process.
`There were six major bleedings in the
`C1-inh group (three patients on day 1,
`two patients on day 2, and one patient on
`day 4) and two in the placebo group (both
`on day 1; p ⫽ nonsignificant). All of them
`were judged to be related to the septic
`process, and none was surprising or in-
`explicable for the responsible physicians.
`There was no unexpected adverse event in
`the C1-inh group. Two patients in the
`placebo group experienced skin flushing
`
`1725
`
`Page 6 of 9
`
`

`

`absence of antibodies (3). The role of
`complement activation during sepsis
`seems to be dual: Some activation is nec-
`essary for an efficient clearance of bacte-
`ria or their products (23). On the other
`hand, inhibition of the biological effects
`of C5a in a baboon sepsis model attenu-
`ated lethal complications (24). C5a has
`been shown to induce hypotension and
`leukopenia and to enhance production of
`cytokines such as interleukin-1, tumor
`necrosis factor-␣, and interleukin-6 by
`monocytes (25). Elevated plasma concen-
`trations of C3a were reported to signifi-
`cantly correlate with mortality rates in
`patients with sepsis and septic shock 5).
`Administration of high doses of C1-
`inh in septic baboons has been demon-
`strated to prolong their survival time.
`Treated animals showed less organ dam-
`age (26). Particularly, the lung, kidneys,
`and adrenals seemed to be spared the
`microvascular thrombosis and infarction
`typically seen in untreated animals. C1-
`inh supplementation in these nonhuman
`primates challenged with lethal doses of
`Escherichia coli completely abrogated C4
`activation, indicating efficient inhibition
`of the classic complement pathway, and
`resulted in a less pronounced increase of
`creatinine,
`lactate dehydrogenase, and
`transaminases compared with untreated
`animals (26). Because the alternative
`pathway of complement is activated by
`many, if not all, bacteria, C1-inh likely
`will not affect the clearance of bacteria.
`Indeed, in the lethal E. coli model men-
`tioned previously, three-fold higher doses
`of C1-inh than those used in the present
`study did not affect the clearance of the
`microorganisms (26).
`In agreement with these findings, we
`were able to demonstrate marked but not
`complete attenuation of the generation of
`C4b/c in septic patients treated with high
`doses of C1-inh. However, because the
`generation of C4b/c was not suppressed
`completely and the inhibiting effect
`lasted only for 72 hrs, the dosage and
`duration of C1-inh administration in our
`study may not have been adequate. Base-
`line concentrations of C1-inh antigen and
`activity were within the normal range in
`our study patients. Because C1-inh is an
`acute phase protein and its plasma con-
`centrations would be expected to be ele-
`vated in septic patients, it has been pro-
`posed that there is a relative deficiency of
`functional C1-inh in patients with severe
`sepsis or septic shock attributable to in-
`creased enzymatic cleavage in inflamed
`or ischemic tissue. The administration of
`
`Crit Care Med 2002 Vol. 30, No. 8
`
`Figure 3. Comparison of the cumulative survival
`for patients who received C1-inhibitor (n ⫽ 20;
`dashed line) and those who received placebo (n ⫽
`20; solid line). The differences of the 28-day and
`90-day mortality rates between the C1-inhibitor
`and the placebo group were not statistically sig-
`nificant (p ⫽ .72 and 1.00, respectively).
`
`of high doses of C1-inh to patients with
`severe sepsis or septic shock. The addi-
`tion of C1-inh to conventional sepsis
`therapy had a beneficial effect on renal
`dysfunction. Mortality rate was equal in
`the C1-inh and the placebo group. There
`were no C1-inh-related side effects.
`Therapeutic application of C1-inh in
`human diseases other than hereditary an-
`gioedema already has been performed, for
`example, in patients with septic shock,
`vascular leak syndrome, or severe ther-
`mal injury (19). In a small study includ-
`ing five patients with septic shock, no
`patient died during the study period of 5
`days. Four of these patients needed less
`vasopressor therapy during C1-inh ad-
`ministration than before. No side effects
`were observed (20).
`Sepsis is supposed to result from the
`excessive activation of defense mecha-
`nisms,
`including the complement and
`contact systems, resulting in the release
`of endogenous inflammatory mediators
`(2). Activation of the contact system was
`suggested to occur because of endothelial
`injury or by direct activation of contact
`factors by bacterial
`lipopolysaccharides
`(21). Significantly decrease