Omaha, NE. Offprint requests: William D. Haire, MD, 600 South 42 Street, Omaha, NE 68198-3330 .... concentrate (ATnativ, Baxter Healthcare, Glendale, CA) or.
Biology of Blood and Marrow Transplantation 4:142–150 (1998) © 1998 American Society for Blood and Marrow Transplantation
A prospective randomized double-blind trial of antithrombin III concentrate in the treatment of multiple-organ dysfunction syndrome during hematopoietic stem cell transplantation William D. Haire,1 Elizabeth I. Ruby,2 Lana C. Stephens,1 Elizabeth Reed,1 Stefano R. Tarantolo,1 Z. Steven Pavletic,1 Philip J. Bierman,1 Michael Bishop,1 Anne Kessinger,1 Julie Vose,1 James O. Armitage1 Departments of 1Internal Medicine and 2Preventive and Societal Medicine, University of Nebraska Medical Center, Omaha, NE Offprint requests: William D. Haire, MD, 600 South 42 Street, Omaha, NE 68198-3330 (Received 27 July 1998; accepted 7 October 1998)
ABSTRACT Many of the complications of high-dose therapy with hematopoietic stem cells are caused by or lead to the multipleorgan dysfunction syndrome (MODS). In hematopoietic stem cell transplantation (HSCT), acquired antithrombin III (ATIII) deficiency is independently associated with MODS to the exclusion of transplant type, preparative regimen, and bacteremia. In experimental settings, replacement of ATIII can ameliorate the severity of MODS that develops in response to a variety of pathologic stimuli, suggesting that ATIII supplementation might improve the clinical course of MODS in patients undergoing HSCT. We performed a study to determine if ATIII can improve the morbidity of MODS in HSCT. Forty-nine patients undergoing HSCT, who developed pulmonary dysfunction (oxygen saturation of 90%), central nervous system dysfunction (drop of 4 points in the mini-mental status exam), or hepatic dysfunction (bilirubin 34 µmol/L [2.0 mg%], weight gain of 5% over baseline, and abdominal pain, possibly of hepatic origin) with a concomitant ATIII activity of 84% were double-blind randomized to receive ATIII concentrate, 70 units/kg within 24 hours of recognition of initial organ dysfunction followed by 50 units/kg 8, 16, 48, and 72 hours later, or albumin placebo. The group randomized to ATIII had a lower severity-ofillness score (15.7 19.2 vs. 28.6 25.2, p 0.03), shorter duration of hospitalization (14.9 16.7 vs. 25.7 17.9 days, p 0.03), and lower hospital charges ($138,700 $23,500 vs. $206,400 $34,000). ATIII concentrate was associated with improved morbidity of MODS in patients undergoing HSCT when given early in the evolution of the syndrome.
KEY WORDS Hepatic veno-occlusive disease • Delerium • Idiopathic pneumonia syndrome • Systemic inflammatory response syndrome
INTRODUCTION The multiple-organ dysfunction syndrome (MODS) is a recently defined clinical entity intended to describe the progressive functional deterioration of a variety of organ systems, which occurs after exposure to a wide spectrum of pathologic stimuli including infection, burns, and trauma [1]. MODS is currently understood to result from the systemic inflammatory response to these stimuli, which has escaped physiologic control [2]. This inflammatory response is mediated by a complex and incompletely understood
This study was partially funded by Baxter Healthcare.
interplay of cytokines, hemostatic factors, the complement cascade, lymphocytes, phagocytes, vascular endothelial cells, and probably other elements currently unrecognized [2,3]. At some point in its evolution, mediators of the inflammatory response begin to generate systemic reactions that are detrimental to the function of many organs, or the regulators of the inflammatory response suppress these mediators to the point of immune deficiency with resultant predisposition to infection and other adverse consequences of “immunologic dissonance.” While the precise mechanism of the generation of these harmful reactions is not well delineated, functional derangements in vascular endothelial cells are probably involved. MODS is responsible for the majority of deaths in the intensive care unit (ICU) patient popula-
Antithrombin III Concentrate Treatment of MODS
tion [4]. Even modest degrees of MODS correlate closely with duration of hospitalization in patients who survive [4]. In survivors of illnesses or injuries requiring ICU care, development of MODS significantly prolongs hospitalization and increases costs of recovery to an average of 385,000 1996 dollars [4]. High-dose therapy and hematopoietic stem cell transplantation (HSCT) is associated with a variety of complications that can result in progressive functional deterioration of many organ systems and subsequent death. There are similarities between the complications of transplantation and MODS in other populations of critically ill patients. These similarities have been used to postulate that the complications seen in the two populations may be the result of similar, if not identical, processes. One of the similarities is an association with deficiencies of naturally occurring anticoagulants such as antithrombin III (ATIII) and protein C [5–7]. The prognosis of MODS is proportional to the degree of deficiency of these proteins [8–21]. In experimental models of MODS, generally using infection or endotoxin as a stimulus, altering the hemostatic system to minimize fibrin accumulation has prevented progressive organ dysfunction and mortality [22–36]. Administration of ATIII, either before or after the inciting stimulus, can prevent the fatal MODS that occurs in these models [29–35]. This model has been applied to human subjects with some success. When given to patients with bacteremia and disseminated intravascular coagulation, ATIII was associated with a trend to lower mortality [37]. When given to patients suffering multisystem trauma, ATIII use was associated with lower plasma levels of neutrophil elastase, interleukin (IL)6, IL-8, and significantly fewer days in the ICU and on ventilator support [38]. In none of the human or animal trials was any adverse effect of ATIII detected. With this background, a study was designed to test the hypothesis that ATIII supplementation would ameliorate the clinical manifestations of MODS from HSCT and that such improvement would lower the overall cost of the HSCT procedure.
MATERIALS AND METHODS Patients All patients with malignant disease who were admitted to the University of Nebraska Medical Center for HSCT and not committed to other confounding clinical trials were eligible to participate, except those for whom English was not their first language and who had not completed an eighth grade education (to allow reliable use of the minimental status examination; see below). Informed consent and high-dose therapy regimens The study was approved by the institutional review board. Informed consent for organ dysfunction monitoring and infusion of the study drug (if necessary) was obtained before beginning the high-dose therapy regimen for HSCT. The regimens used in this study have been previously described [5,39,40]. Organ dysfunction All patients were screened daily by one of two individuals for the development of central nervous system (CNS),
BB&MT
pulmonary, or hepatic dysfunction. These organ dysfunctions were chosen because of previous demonstrations of their contribution to MODS and subsequent mortality in HSCT as well as their close correlation with ATIII levels during HSCT [5]. For our analyses, organ dysfunction was considered as an all or none phenomenon rather than a spectrum of increasingly severe physiologic derangements, an approach supported by our experience in previous cohorts of patients undergoing HSCT [5]. Grading organ dysfunction severity at the time of presentation on a fourpoint scale of progressive severity added no value in predicting subsequent organ dysfunction or mortality compared with considering organ dysfunction and no organ dysfunction as dichotomous variables, similar to the experience using two types of organ dysfunction definitions to predict mortality in critically ill non–HSCT patients [41]. Organ dysfunctions were defined as follows: 1) CNS—A drop in the score of the standardized minimental status examination of 4 points from the pre-chemotherapy score. This score represents slightly more than two standard deviations from the mean of the differences in test/retest scores with this tool [42]. This definition has been shown to have a specificity of 100% for delirium in subjects with an education level of at least the eighth grade [43]. 2) Pulmonary—Finger oximetry reading of SA02 90% on two occasions on the same day separated by at least 2 hours. 3) Hepatic—A combination of bilirubin 2.0 mg%, a weight gain of 5% over prechemotherapy weight, and abdominal pain of possible hepatic origin. This definition has a high degree of concordance with histologically defined veno-occlusive disease from autopsy [44] and premortem [45] liver tissue. MODS Patients with single-organ dysfunction (either pulmonary, hepatic, or CNS) during HSCT who have a concomitant ATIII activity of 84% have been shown to have a 71.8% likelihood of progression to multiple-organ dysfunction in the following week [5]. Consequently, for the purposes of this study, patients with single-organ dysfunction with a concomitant ATIII activity of 84% of normal were defined as having MODS. Randomization Patients with MODS were randomized to receive ATIII concentrate (ATnativ, Baxter Healthcare, Glendale, CA) or albumin placebo (5% human albumin, Baxter Healthcare). The ATIII concentrate was given at a dose of 70 U/kg within 24 hours of organ dysfunction detection, followed by 50 U/kg 8, 16, 48, and 72 hours later. The total dose of ATIII concentrate was 270 U/kg, similar to the dose found effective in animal models [30,31,34] and human sepsis [37]. Albumin was given in a volume equal to the calculated volume of ATIII concentrate. Based on a recovery of 1.8%/Ukg and a half-disappearance time of 12 hours 14 days after beginning the high-dose therapy regimen for HSCT [46], the study dose of ATIII concentrate was estimated to provide an increment of 250% activity above pretreatment levels after the third dose, levels similar to those achieved in prior human studies found to be effective in shortening the
143
WD Haire et al.
Table 1. Characteristics of the population monitored for organ dysfunction Underlying disease Acute lymphocytic leukemia Acute myelogenous leukemia Breast carcinoma Chronic lymphocytic leukemia Chronic myeloid leukemia Hodgkin’s disease Multiple myeloma Lymphoma Miscellaneous Total Transplant type Allogeneic, related Allogeneic, unrelated Autologous bone marrow peripheral blood Total Preparative regimens Carmustine, cytarabine, etoposide, cyclophosphamide Cyclophosphamide, thiotepa Cyclophosphamide, thiotepa, hydroxyurea Carmustine, etoposide, cyclophosphamide Busulfan, cyclophosphamide Cyclophosphamide, total-body irradiation Cyclophosphamide, etoposide, total-body irradiation Other Total
6 14 51 3 19 11 12 68 2 186 47 7 33 99 186 50 18 35 19 5 34 24 1 186
duration of disseminated intravascular coagulation due to sepsis [37]. The duration of therapy was chosen based on previously published evidence from animal [29,30,31,33,34] and human [37] studies, suggesting efficacy with 72 hours of therapy or less. Previously performed multivariate analyses considering transplant type (allogeneic, autologous bone marrow, and autologous peripheral blood stem cell), type of preparative regimen, presence of positive blood culture, and AT III level have shown that only the AT III level was a consistent independent association of organ dysfunction during transplantation [5]. Because transplant type, treatment regimen (and, by inference, disease type), and bacteremia were not independent risk factors for organ dysfunction in this population, no attempt to stratify for these variables was made in this study. With no prior experimental experience available, the degree of change in the major outcomes of the study expected from ATIII supplementation was unknown. Consequently, accurate prediction of sample size necessary to establish a statistically significant difference between the two groups could not be precisely estimated. The factors ultimately governing the final number of subjects to be accrued were the number that would reasonably be required to show a trend to different outcomes between the two groups, the frequency with which organ dysfunction was seen in this single institution, and the amount of funding available to carry out the study. The number of 25 subjects in each arm was determined before beginning the study to meet these limitations.
144
Randomization was accomplished using a computer-generated list of random numbers. An unblinded research pharmacist reconstituted the requisite volume of study drug and provided it to the ward personnel in unmarked containers. Post-randomization monitoring After randomization, daily monitoring for organ dysfunction continued until hospital dismissal or death. Criteria for dismissal from the hospital were standardized before the beginning of this trial and included neutrophil recovery of 500/mm3, absence of fever or active infection, minimum oral intake of 1000 mL of fluid and 1000 kcal, controlled GVHD of grade less than II, ability to ambulate, and need for no more than one daily transfusion of red blood cells or platelets [47]. Decision for hospital dismissal was made without knowledge of study treatment allocation. The severityof-illness score was calculated after study completion by summing the number of organ dysfunctions (0–3) on each day of hospitalization from the point of randomization until dismissal or death. Thus, for example, a patient with CNS dysfunction on 6 days, pulmonary dysfunction on 3 days, both CNS and pulmonary dysfunction on 4 days, and no organ dysfunction for the remaining 7 days of hospitalization would have a severity-of-illness score of 17. The rationale of this approach is based on the observation that there is a direct relationship between the number of organ dysfunctions per unit time and mortality in critically ill non–HSCT patients [48], suggesting that patients with more organ dysfunctions per unit time were at greater risk of dying than those with fewer organ dysfunctions. Consequently, this measurement of severity of illness encompasses more than simply the duration of hospitalization. If the MODS or its therapy predisposed to early death, this calculation would be expected to be falsely indicative of a good outcome. In the absence of a higher mortality rate, a lower severity-of-illness score would be indicative of less severe MODS. Statistical analysis Data are expressed as the mean SEM. Wilcoxon’s rank sum test assessed the difference in duration of hospitalization, hospital charges, and severity-of-illness scores between the two groups. The Fisher’s exact test was used to look at the differences in the mortality distribution of the study groups. A p value 0.05 was considered significant.
RESULTS A total of 197 patients were approached to give informed written consent. Eight patients declined to give such consent, one patient had not completed an eighth grade education and so was disqualified, and two patients had committed to participate in another clinical trial. The demographics, diseases, preparative regimens, and types of HSCT performed for the remaining 186 patients entered into the study are outlined in Table 1. All 186 patients were evaluable. Significance of the definitions of organ dysfunction and MODS Organ dysfunction was detected in 54 patients. Of this group, 25 died during hospitalization (mortality 46.3%). The mortality rate was 0% in the remaining 132 patients who did not meet our criteria for organ dysfunction ( p 0.0001).
Antithrombin III Concentrate Treatment of MODS
MODS was diagnosed by the presence of an ATIII activity of 84% in 49 of these 54 patients. That our definition of MODS (single-organ dysfunction with a low ATIII level) was truly indicative of a high risk of multiple-organ dysfunction is illustrated by the observation that progression to a second- or third-organ dysfunction occurred in 29 of these patients (16 of the 25 patients randomized to placebo, 13 of the 24 patients randomized to ATIII), compared with zero of the five patients with single-organ dysfunction but without a low ATIII level (Fisher’s exact test, p 0.017). The significance of progression from single- to multiple-organ dysfunction is evident in the observation that of the 20 MODS patients who did not progress to second-organ dysfunction, only three died; 10 of the 19 who progressed to second- and all 10 of those who progressed to third-organ dysfunctions died (Fisher’s exact test, p 0.001). This also confirms the observation in non–HSCT patients that mortality is related to the absolute number of organ dysfunctions that occur during HSCT. In addition to predicting mortality, MODS was a strong indicator of in-hospital morbidity. Patients with MODS, but dismissed from the hospital alive, had longer (37.5 19.5 days vs. 15.7 9.6 days, p 0.0001) hospitalizations than those who did not develop MODS. The severity-of-illness score in the 23 randomized patients who died was 35.3 24.4 compared with 10.7 14.5 in the patients who were dismissed alive (Wilcoxon’s rank sum test, p 0.0003), confirming that mortality was related to the duration and the absolute number of organ dysfunctions. Results of randomization ATIII levels were 84% in 49 (90.7%) of the 54 patients who developed organ dysfunction (mean ATIII level 65.1 3.1%); these patients were randomized and received a complete course of the study drug. One patient was randomized and received the first dose of the study drug based on an erroneous laboratory report. No further study drug was given and this patient was not included in our analysis. The mean weight of the randomized patients was 76 kg, making the mean total dose of ATIII concentrate 20,520 U/patient. The diagnoses and types of HSCT procedures performed on the patients who came to randomization are listed in Table 2. There were no significant differences between the treatment groups with respect to these parameters. Although all patients with MODS during unrelated allogeneic transplants who developed organ dysfunction were randomized to receive the placebo, analysis of outcomes of these patients showed no difference in their severity-of-illness score, hospital length of stay, or hospital charges compared with the related allogeneic transplant patients (severity-of-illness score, 28.2 12.8 vs. 28.7 27.1; length of hospital stay, 40.0 8.0 vs. 41.7 19.7 days; and hospital charges, $233,863 $159,448 vs. $201,140 $173,789). Consequently, any differences between the outcomes of the ATIII and placebo groups were not due to inordinately adverse outcomes of the unrelated allogeneic transplant patients in the placebo group. The pattern of clinical progression from single- through multiple-organ dysfunction after randomization is outlined in Table 3. Of the 24 patients randomized to ATIII, 13 (54%) progressed to other organ dysfunctions compared with 16 (64%) of 25 patients randomized to placebo (p NS). In the ATIII treatment group, 12 presented with CNS dysfunction, six of whom progressed to other organ dys-
BB&MT
Table 2. Characteristics of the patients randomized to treatment with antithrombin III or placebo
Demographics Mean age (years) Men Women Total Transplant type Allogeneic, related Allogeneic, unrelated Autologous Total Source of stem cells Bone marrow Peripheral blood Total Underlying disease Acute lymphocytic leukemia Acute myeloid leukemia Breast carcinoma Chronic lymphocytic leukemia Chronic myeloid leukemia Hodgkin’s disease Multiple myeloma Lymphoma Total Preparative regimens Carmustine, cytarabine, etoposide, cyclophosphamide Cyclophosphamide, thiotepa Cyclophosphamide, thiotepa, hydroxyurea Carmustine, etoposide, cyclophosphamide Busulfan, cyclophosphamide Cyclophosphamide, total-body irradiation Cyclophosphamide, etoposide, total-body irradiation Total
Antithrombin III
Placebo
46.1 10.3 10 14 24
46.0 10.5 10 15 25
10 0 14 24
11 4 10 25
8 16 24
10 15 25
0 5 8 0 1 0 3 7 24
1 4 4 2 4 1 3 6 25
4 3
4 1
4
3
2 0
3 1
5
5
6 24
8 25
functions; 10 presented with pulmonary dysfunction, six of whom progressed to other organ dysfunctions; and two presented with hepatic dysfunction, one of whom progressed to other organ dysfunctions. In the group randomized to placebo, 10 presented with CNS dysfunction, six of whom progressed to other organ dysfunctions; nine presented with pulmonary dysfunction, six of whom progressed to other organ dysfunctions; and six presented with hepatic dysfunction, four of whom progressed to other organ dysfunctions. The outcome of therapy is summarized in Table 4. The duration of first-, second-, and third-organ dysfunctions tended to be shorter in patients randomized to ATIII, but differences in durations of individual organ dysfunctions did not reach statistical significance. However, when the clinical effect of the aggregate of these changes was considered, patients receiving ATIII fared better than those receiving placebo. Patients randomized to ATIII had a lower mean severity-ofillness score (15.7 19.2 vs. 25.7 17.9, p 0.03) and a
145
WD Haire et al.
Table 3. Organ dysfunction outcomes of treatment in patients treated with antithrombin and placeboa Antithrombin III CNS only Pulmonary only Hepatic only CNS, pulmonary Pulmonary, CNS Hepatic, CNS, pulmonary CNS, pulmonary, hepatic Pulmonary, CNS, hepatic Total Placebo CNS only Pulmonary only Hepatic only CNS, pulmonary Pulmonary, CNS Hepatic, CNS CNS, pulmonary, hepatic Pulmonary, hepatic Hepatic, CNS, pulmonary Pulmonary, hepatic, CNS Total
Table 4. Outcomes of treatment in patients treated with antithrombin and placeboa
6 4 1 5 4 1 1 2 24 4 3 2 4 4 1 2 1 3 1 25
a The type of organ dysfunction listed first in each column is the initial organ dysfunction. The following organ dysfunctions listed are those that occurred subsequently. To simplify reporting, the second and third organ dysfunctions are not necessarily in chronological order.
shorter hospitalization (14.9 16.7 vs. 25.7 17.9 days, p 0.03) after randomization than did those randomized to placebo. This improvement translated to a tendency toward lower hospital charges for the ATIII group ($138,700 $23,500 vs. $206,400 $34,000, p 0.06). There was a lower mortality rate in the group randomized to ATIII (39% [9/23] vs. 56% [14/25]), but the difference did not reach statistical significance ( p 0.19).
DISCUSSION MODS has been recognized as the leading cause of death among critically ill or injured patients [4], including patients undergoing HSCT [5]. Our data confirmed the unfavorable clinical course of MODS in HSCT and supported the validity of the use of simple clinical measurements to define the disorder and estimate outcomes. Patients meeting the criteria for single-organ dysfunction had a mortality rate approaching 50%, suggesting the presence of significant physiologic derangement at the time of presentation, despite the often mild deviation from normal organ function parameters at that time. Patients who developed no organ dysfunction by these criteria did not die. Approximately one-half of the patients presenting with pulmonary, CNS, or hepatic dysfunction, by our definitions, progressed to second- or third-organ dysfunctions. An ATIII level 84% combined with the presence of single-organ dysfunction further improved the predictive value of progression to multipleorgan dysfunctions. In this prospective study, the rate of development of subsequent organ dysfunction in untreated
146
Death Duration (days) 1st Organ dysfunction 2nd Organ dysfunction 3rd Organ dysfunction Severity-of-illness score Hospital length of stay (days post-randomization) Hospital charges (thousands of dollars)
ATIII (n=24)
Placebo (n=25)
p
9
14
0.19
3.5 (1.0, 63.0) 5.0 (1.0, 56.0) 16.0 (1.0, 34.0) 5.5 (1.0, 77.0)
8 (0, 66.0) 18.5 (1.0, 50.0) 28.5 (8.0, 51.0) 27.0 (3.0, 66.0)
0.37 0.15 0.24 0.03
7.5 (1.0, 62.0)
21.0 (3.0, 66.0)
0.03
96.1 (35.0, 46.5)
118.0 (51.0, 630.3) 0.06
a
Results are reported as medians, with minimum and maximum values in parentheses.
patients with single-organ dysfunction and a low ATIII level was 64%, very similar to the 70.5% predictive value previously reported for this combination [5]. The development of second- and third-organ dysfunctions was an ominous occurrence, with mortality reaching 100% among those who developed CNS, hepatic, and pulmonary dysfunction. Not only was the development of organ dysfunction an important determinant of mortality, but the duration of organ dysfunction was also important. When the number of organ dysfunctions per day was combined with the number of days of hospitalization, the resultant severity-of-illness score associated strongly with mortality. In addition to its relationship to death, the severity of MODS had other clinically relevant features. Patients who developed nonfatal MODS had greater morbidity than did similar patients who did not develop MODS, which likely translated into increased suffering, longer hospitalization, and higher financial costs. This was observed in both HSCT patients and in populations with other severe illnesses [4,5]. Consequently, mortality is not the only clinically relevant goal of proposed therapies for this syndrome. An effective therapy should lower the duration and cost of hospitalization and lower the severityof-illness score without increasing mortality. Unfortunately, to date no specific therapy has been identified. The only therapy available for MODS is the elimination (where possible) of its inciting stimuli and the provision of supportive care until dysfunctional organs regain full activity. The results of this study suggest that the clinical course of MODS can be favorably altered with active therapy directed at the acquired ATIII deficiency. Patients given ATIII concentrate early in the evolution of MODS recovered faster and with a lower aggregate severity of illness than did patients given placebo. In addition to improved overall morbidity, ATIII therapy was associated with a trend of decreased recovery costs. In this study, the ATIII concentrate was provided without charge; thus, the reported hospital charges did not reflect the use of ATIII concentrate. However, if one factors in the cost of the ATIII concentrate (current average wholesale price of $.60/U), the mean cost of therapy in this study would have been $12,312 per patient. This compares favorably with the mean difference in hospital charges of $67,700 per patient between those receiving
Antithrombin III Concentrate Treatment of MODS
ATIII and those receiving placebo. Because ATIII therapy decreased morbidity of organ dysfunction and because organ dysfunction often leads to death, ATIII therapy might also improve the mortality of HSCT. The mortality rate was lower in the ATIII-treated group, but the difference was not statistically significant ( p 0.19), although the study was not designed to test this hypothesis. Data suggest that the evaluation of the effect of ATIII on MODS mortality requires a study randomizing 200–250 patients between active therapy and placebo. Because 25% of all transplant patients develop complications leading to MODS, such a study would require a population of 1000 transplant patients. And because 20,000 transplants are performed annually in the United States, this type of study seems feasible. In our study, treatment with ATIII resulted in a lower frequency of progression from single- to multiple-organ dysfunction (54% progression in the ATIII group vs. 64% in the placebo group) and a shorter duration of first-, second-, and third-organ dysfunctions. However, none of these differences reached statistical significance. The aggregate severity of all organ dysfunctions, as attested to by the severity-of-illness scores, was significantly improved with the use of ATIII, suggesting that individual organ dysfunctions were affected by this intervention as well. Perhaps, as with mortality considerations, the lack of statistically significant differences might be related to the small sample size used in this study. It is also reasonable to assume that organ dysfunctions detected, according to the definitions used in this study, were present several days before their diagnosis, but that our tools were too crude to detect them. Further research into methods of detecting clinically significant organ dysfunction may allow earlier intervention, possibly with better clinical outcomes, including improvement in frequency of progression from single- to multiple-organ dysfunction and duration of single-organ dysfunction. Additionally, higher doses of ATIII concentrate, dosing the concentrate to achieve a threshold level of plasma activity, or longer duration of ATIII supplementation may further improve outcomes. These issues should be addressed in subsequent clinical trials. The possibility of other explanations for the better outcomes observed in the ATIII-treated group should be carefully considered. All unrelated allogeneic transplant patients treated in this study were randomized to receive placebo. However, the outcomes of these patients were similar to those undergoing related allogeneic transplantation, suggesting that inordinately worse outcomes in the unrelated allogeneic transplant patients is unlikely to explain the differences in outcomes between the ATIII and placebo groups. The absolute numbers of autologous transplant patients was higher in the group randomized to ATIII, while the number of allogeneic transplant patients was higher in the group randomized to placebo. This imbalance may account for some of the differences between the outcomes of the two groups. Data are insufficient to prove or disprove this hypothesis. However, previous analyses have shown that while the frequency and type of organ dysfunction varies with type of transplant (allogeneic, autologous bone marrow, autologous peripheral stem cell), treatment regimen, and disease type, these are not independent risk factors for development of organ dysfunction during transplantation [5]. These analyses have shown the only consistent independent associate of
BB&MT
organ dysfunction is the ATIII level during the transplant course. This is consistent with the hypothesis that the inflammatory response is responsible for the generation of much of the organ dysfunction seen during transplantation, that ATIII is actively involved in the pathogenesis of this response, that the ATIII level is a marker of the severity of the inflammatory response, and that, while certain transplant-specific variables (type of transplant or preparative regimen) result in higher frequencies of complications, they do so by inducing a more robust inflammatory response. This suggests that differences between treatment and control arms in the areas of transplant type, treatment regimen, and bacteremia might not be relevant to differences in outcomes, except insofar as they induce a stronger inflammatory response as manifested by lower ATIII values. Potential contributions of these variable outcomes should be considered in any future trial of ATIII or other proposed therapy of MODS during HSCT by limiting studies to patients with transplant-specific variables associated with a high likelihood of inflammation-associated organ dysfunctions. The mechanism of action of ATIII concentrate is unknown. Serine proteases are activated during the administration of chemotherapy [49–51] and the genesis of the inflammatory reaction [6–7]. The serine protease thrombin can convert fibrinogen to fibrin monomer, levels of which predict development and fatal outcome of MODS in non–HSCT patients [52]. Additionally, thrombin has effects on vascular endothelial cells that are of major importance in the inflammatory process [53]. Thrombin causes these cells to secrete plasminogen activator inhibitor-1 [54], an inhibitor of fibrinolysis found in high concentrations in patients with hepatic dysfunction during HSCT [55]. Thrombin also induces the expression of P-selectin, facilitating polymorphonuclear leukocyte adhesion to endothelial cells and their subsequent degranulation [56,57]. Neutrophil degranulation releases elastase and proteinase 3, serine proteases that can mediate endothelial cell death by apoptosis [58] when they become procoagulant [59] and cause further activation of coagulation in a potentially self-propagating manner. Thrombin stimulates production of transforming growth factor beta [60], a predictor of pulmonary and hepatic dysfunction in HSCT patients [61]. When activated, another serine protease, factor XII, can generate bradykinin by activating prekallikrein and its subsequent cleavage of high-molecular-weight kininogen [62]. Bradykinin can produce many of the observed cellular and physiologic effects associated with inflammation [63], including increases in vascular permeability (part of the clinical syndrome of hepatic veno-occlusive disease) [43,44]. In addition to generation of bradykinin, plasma kallikrein can induce a variety of other inflammatory reactions that can lead to organ dysfunction. Specific inhibition of plasma kallikrein can prevent inflammation-induced pulmonary dysfunction in experimental settings [64]. Factor XII can also activate the complement cascade via the classical pathway, generating biologically active substances that are involved in the inflammatory reaction [62]. Inhibiting complement activation can prevent inflammation-induced pulmonary dysfunction in experimental inflammation [65]. ATIII as an inhibitor of serine proteases could be active at these and other interfaces of the coagula-
147
WD Haire et al.
tion and inflammatory systems, especially if present in supraphysiologic levels, and could favorably alter their contribution to the intensity of the systemic inflammatory response as one of its mechanisms of action. ATIII may also alter cytokine expression during inflammation through unknown mechanisms. When given to patients after multisystem trauma, ATIII resulted in lower levels of IL-6 and IL-8 [38]. The notion that modulators of the coagulation cascade can favorably affect the outcomes of inflammation is not limited to ATIII. In experimental settings, outcomes of inflammation due to sepsis or trauma can be improved by modulation of the hemostatic system by protein C [11,24], thrombomodulin [23], heparin [26], inhibitors of tissue factor [22], and recombinant tissue plasminogen activator [27,28]. Similar alterations in the hemostatic system using heparin or recombinant tissue plasminogen activator have been reported to be helpful in preventing and treating hepatic venoocclusive disease, one of the manifestations of MODS during HSCT [66–70]. The link between the coagulation and inflammatory systems may prove to be an important focus of therapy in HSCT and other disease processes. Perhaps further study will open the door to ways of providing patients with MODS more than nonspecific supportive therapy to prevent subsequent complications and death.
REFERENCES 1 Bone RC, Balk RA, Cerra FB: Definition for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis: the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference Committee. Chest 101:1644, 1992. 2 Bone RC: Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 125:680, 1996. 3 Bone RC, Grodzin CJ, Balk RA: Sepsis: a new hypothesis for pathogenesis of the disease process. Chest 112:235, 1997. 4 Barie PS, Hydo LJ: Influence of multiple organ dysfunction on duration of critical illness and hospitalization. Arch Surg 131:1318, 1996. 5 Haire WD, Ruby EI, Gordon BG, Patil KD, Stephens LC, Kotulak GD, Reed EC, Vose JM, Bierman PJ, Kesinger A, Armitage JO: Multiple organ dysfunction in bone marrow transplantation. JAMA 274:1289, 1995. 6 Bone RC: Modulators of coagulation. Arch Intern Med 152:1381, 1992. 7 Carvalho AC, Freeman NJ: How coagulation defects alter outcome in sepsis: survival may depend on reversing procoagulant conditions. J Crit Illn 9:51, 1994. 8 Brandtzaeg P, Sandset PM, Joo GB, Ovstebo R, Abilgaard U, Kierulf P: The quantitative association of plasma endotoxin, antithrombin, protein C, extrinsic pathway inhibitor and fibrinopeptide A in systemic meningococcal disease. Thromb Res 55:459, 1089. 9 Schipper HG, Roos J, Van der Muellen F, ten Cate JW: Antithrombin III deficiency in surgical intensive care patients. Thromb Res 21:73, 1981. 10 Lorente JA, Garcia-Frade LJ, Landin L, de Pablo R, Torrado C, Renes E, Garcia-Avello A: Time course of hemostatic abnormalities in sepsis and its relation to outcome. Chest 103:1536, 1993. 11 Fijnvandraat K, Derkx B, Peters M, Bijlmer R, Sturk A, Prins MH, van Deventer JH, ten Cate JW: Coagulation activation and tissue necro-
148
sis in meningococcal septic shock: severely reduced protein C levels predict a high mortality. Thromb Haemost 73:15, 1995. 12 Fourrier F, Chopin C, Goudemand J, Hendrycx S, Caron C, Rime A, Marey A, Lestavel P: Septic shock, multiple organ failure, and disseminated intravascular coagulation: compared patterns of antithrombin III, protein C and protein S deficiencies. Chest 101:816, 1992. 13 Owings JT, Bagley M, Gosselin R, Romac D, Disbrow E: Effect of critical injury on plasma antithrombin activity: low antithrombin levels are associated with thromboembolic complications. J Trauma 41:396, 1996. 14 Mesters RM, Mannucci PM, Coppola R, Keller T, Ostermann H, Kienast J: Factor VIIa and antithrombin activity during severe sepsis and septic shock in neutropenic patients. Blood 88:881, 1996. 15 Miller RS, Weatherford DA, Stein D, Crane MM, Stein M: Antithrombin III and trauma patients: factors that determine low levels. J Trauma 37:442, 1994. 16 Nedorn E, Wosegien F, Kemmler G: Antithrombin III and early prognosis in polytraumatized patients: a pilot study. Klin Wochenschr 69:817, 1991. 17 Jochum M, Gippner-Steppert C, Machleidt W, Fritz H: The role of phagocyte proteinases and proteinase inhibitors in multiple organ failure. Am J Respir Crit Care Med 150:S123, 1994. 18 Wilson RF, Farag A, Mammen EF, Fujii Y: Sepsis and antithrombin III, prekallikrein, and fibronectin levels in surgical patients. Am Surg 55:450, 1989. 19 Hesselvik JF, Blomback M, Brodin B, Maller R: Coagulation, fibrinolysis, and kallikrein systems in sepsis. Crit Care Med 17:724, 1989. 20 Nast-Kolb D, Waydhas C, Gippner-Steppert C, Schneider I, Trupka A, Rucholz S, Zettl R, Schweiberer L, Jochum M: Indicators of the post-traumatic inflammatory response correlate with organ failure in patients with multiple injuries. J Trauma 42:446, 1997. 21 Boveda JL, Ruiz JC, Monasterio J, Martin MC, Esteban F, Gargia JL, Nuvials AX, Angles A, Salgado A: Antithrombin III and protein C as indicators of multiple organ dysfunction syndrome in sepsis. Thromb Haemost 77:434, 1997. [abstr] 22 Dackiw APB, McGilvray ID, Woodside M, Nathens AB, Marshall JC, Rostein OD: Prevention of endotoxin-induced mortality by antitissue factor immunization. Arch Surg 131:1273, 1996. 23 Uchiba M, Okajima K, Murakami K, Nawa K, Okabe H, Takatsuki K: Recombinant human soluble thrombomodulin reduces endotoxininduced pulmonary vascular injury via protein C activation in rats. Thromb Haemost 74:1265, 1995. 24 Murakami K, Okajima K, Uchiba M, Johno M, Nakagaki T, Okabe H, Takatsuki K: Activated protein C attenuates endotoxin-induced pulmonary vascular injury by inhibiting activated leukocytes in rats. Blood 87:642, 1996. 25 Taylor FB, Chang A, Esmond CT, D’Angelo A, Vigano-D’Angelo S, Blick KE: Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest 79:918, 1987. 26 Meyer J, Cox CS, Herndon DH, Nakazawa H, Lentz CW, Traber LD, Traber DL: Heparin in experimental hyperdynamic sepsis. Crit Care Med 21:84, 1993. 27 Hardaway RM, Williams CH: Prevention of multiple organ failure with plasminogen activator. Curr Ther Res 49:711, 1991. 28 Palmoa MJ, Paramo JA, Rocha E: Endotoxin-induced intravascular coagulation in rabbits: effect of tissue plasminogen activator vs urokinase on PAI generation, fibrin deposits and mortality. Thromb Haemost 74:1578, 1995. 29 Redens TB, Emerson TE: Antithrombin-III treatment limits disseminated intravascular coagulation in endotoxemia. Circ Shock 28:49, 1989.
Antithrombin III Concentrate Treatment of MODS
30 Taylor FB, Emerson TE, Jordan R, Chang AK, Blick KE: Antithrombin III prevents the lethal effects of Escherechia coli infusion in baboons. Circ Shock 26:227, 1988. 31 Emerson TE, Fournel MA, Leach WJ, Redens TB: Protection against disseminated intravascular coagulation and death by antithrombin III in the Escherichia coli endotoxemic rat. Circ Shock 21:1, 1987. 32 Emerson TE, Fournel MA, Redens TB, Taylor FB: Efficacy of antithrombin III supplementation in animal models of fulminant Escherichia coli endotoxemia or bacteremia. Am J Med 87:3B-27S, 1989. 33 Dickniete G, Paques EP: Reduction of mortality with antithrombin III in septicemic rats: a study of Klebsiella pneumoniae-induced sepsis. Thromb Haemost 69:98, 1993. 34 Redens TB, Leach WJ, Bogdnoff DA, Emerson TE: Synergistic protection from lung damage by combining antithrombin III and alpha-1protease inhibitor in the E. coli endotoxic sheep model. Circ Shock 26:15, 1988. 35 Kessler CM, Tang ZC, Jacobs HM, Szymanski LM: The suprapharmacologic dosing of antithrombin III concentrate for Staphylococcus aureus-induced disseminated intravascular coagulation in guinea pigs: substantial reduction in mortality and morbidity. Blood 89:4393, 1997. 36 Uchiba M, Okajima K, Murakami K, Johno M, Okabe H, Takatsuki K: Effect of human urinary thrombomodulin on endotoxin-induced intravascular coagulation and pulmonary vascular injury in rats. Am J Hematol 54:118, 1997. 37 Fourrier F, Chopin C, Huart J, Runge I, Caron C, Goudemand J: Double-blind, placebo-controlled trial of antithrombin III concentrates in septic shock with disseminated intravascular coagulation. Chest 104:882, 1993. 38 Jochum M: Influence of high-dose antithrombin concentrated therapy on the release of cellular proteinases, cytokines, and soluble adhesion molecules in acute inflammation. Semin Hematol 32:19, 1995. 39 Bishop MR, Anderson JR, Jackson JD, Bierman PJ, Reed EC, Vose JM, Armitage JOI, Warkentin PI, Kessinger A: High-dose therapy and peripheral blood progenitor cell transplantation: effects of recombinant human granulocyte-monocyte colony-stimulating factor on the autograft. Blood 83:610, 1994. 40 Martin-Algarra S, Bishop MR, Tarantolo S, Cowles MK, Reed E, Anderson JR, Vose JM, Bierman, P, Armitage JO, Kessinger A: Hematopoietic growth factors after HLA-identical allogeneic bone marrow transplantation in patients treated with methotrexate-containing graftvs.-host disease prophylaxis. Exp Hematol 23:1503, 1995. 41 Barie PS, Hydo LJ, Fischer E: A prospective comparison of two multiple organ dysfunction/failure scoring systems for prediction of mortality in critical surgical illness. J Trauma 37:660, 1994. 42 Folstein MF, Folstein SE, McHugh PR: “Mini-mental state” — a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189, 1975. 43 Anthony JC, LeResche L, Niaz U, von Korf MR, Folstein MF: Limits of the “mini-mental state” as a screening test for dementia and delerium among hospital patients. Psychol Med 12:397, 1982. 44 McDonald GB, Sharma P, Matthews DE, Shulman HM, Thomas ED: Veno-occlusive disease of the liver after bone marrow transplantation: diagnosis, incidence, and predisposing factors. Hepatology 4:116, 1984. 45 Carreras E, Granena A, Navasa M, Bruguera M, Marco V, Sierra J, Tassies MD, Garcia-Pagan JC, Marti JM, Bosch J, Rodes J, Rozman C: On the reliability of the clinical criteria for the diagnosis of hepatic venoocclusive disease. Ann Hematol 66:77, 1993. 46 Haire WD, Stephens LC, Kotulak GD, Ruby EI, Gordon BG: Pharmacokinetics of antithrombin concentrate during autologous hemato-
BB&MT
poietic stem cell transplantation. Bone Marrow Transplant 15:505, 1995. 47 Pavletic ZS, Bishop MR, Tarantolo SR, Martin-Allegra S, Bierman PJ, Vose JM, Reed EC, Gross TG, Kollath J, Nasrati K, Jackson JD, Armitage JO, Kessinger A: Hematopoietic recovery after allogeneic blood stem-cell transplantation compared with bone marrow transplantation in patients with hematologic malignancies. J Clin Oncol 15:1608, 1997. 48 Knaus WA, Wagner DP: Multiple systems organ failure: epidemiology and prognosis. Crit Care Clin 5:221, 1989. 49 Zurborn KH, Bruhn HD: Hypercoagulability and cytostatic tumor therapy. Blut 53:179, 1986. 50 Kuzel T, Esparaz B, Green D, Kies M: Thrombogenicity of intravenous 5-fluorouracil alone or in combination with cisplatin. Cancer 65:885, 1990. 51 Edwards RL, Klaus M, Matthews E, McCullen C, Bona RD, Rickles FR: Heparin abolishes the chemotherapy-induced increase in plasma fibrinopeptide A levels. Am J Med 89:25, 1990. 52 Bredbacka S, Blomback M, Winman B: Soluble fibrin: a predictor for the development and outcome of multiple organ failure. Am J Hematol 46:289, 1994. 53 Renesto P, Si-Tahar M, Moniatte M, Balloy V, Van Dorsselaer A, Pidard D, Chignard L: Specific inhibition of thrombin-induced cell activation by the neutrophil proteinases elastase, cathepsin G, and proteinase 3: evidence for distinct cleavage sites within the aminoterminal domain of the thrombin receptor. Blood 89:1944, 1997. 54 Gelehrter TD, Sznycer-Laszyk R: Thrombin induction of plasminogen activator-inhibitor in cultured human endothelial cells. J Clin Invest 77:165, 1986. 55 Salat C, Holler E, Kolb H, Reinhardt B, Pihusch R, Wilmanns W, Hiller E: Plasminogen activator inhibitor-1 confirms the diagnosis of hepatic veno-occlusive disease in patients with hyperbilirubinemia after bone marrow transplantation. Blood 89:2184, 1997. 56 Lorant DE, Patel KD, McIntyre M, McEver RP, Prescott SM, Zimmerman GA: Coexpression of GMP-140 and PAF by endothelium stimulated by histamine or thrombin: a juxtacrine system for adhesion and activation of neutrophils. J Cell Biol 115:223, 1991. 57 Wright DG, Gallin JI: Secretory responses of human neutrophils: exocytosis of specific (secondary) granules by human neutrophils during adherence in vitro and during exudation in vivo. J Immunol 123:285, 1979. 58 Yang JJ, Kettritz R, Falk RJ, Jennette JC, Gaido ML: Apoptosis of endothelial cells induced by the neutrophil serine proteases proteinase 3 and elastase. Am J Pathol 149:1617, 1996. 59 Bombeli T, Karsan A, Tait JF, Harlan JM: Apoptotic vascular endothelial cells become procoagulant. Blood 89:2429, 1997. 60 Yamabe Hosawa H, Inuma H, Kaizuki M, Tamura N, Tsunoda S, Baba Y, Shirato K, Onodera K: Thrombin stimulates production of transforming growth factor-beta by cultured human mesangial cells. Nephrol Dial Transplant 12:438, 1997. 61 Anscher MS, Peters WP, Reisenbichler H, Petros WP, Jintle RL: Transforming growth factor-beta as a predictor of lung and liver fibrosis after autologous bone marrow transplantation for advanced breast cancer. N Engl J Med 328:1592, 1993. 62 DeLa Cadena RA, Wachtfogel YT, Colman RW: Contact activation pathway: inflammation and coagulation. In: Colman RW, Hirsh J, Marder VJ, Salzman EW (eds) Hemostasis and Thrombosis: Basic Principles and Clinical Practice. 3rd ed. Philadelphia, PA: JP Lippincott, 219, 1994. 63 Fein AM, Bernard GR, Criner GJ, Fletcher EC, Good JT, Knaus WA, Levy H, Matuschak GM, Shanies HM, Taylor RM, Rodell TC: Treatment
149
WD Haire et al.
of severe systemic inflammatory response syndrome and sepsis with a novel bradykinin antagonist, deltibant (CP-0127). JAMA 277:482, 1997. 64 Uchiba M, Okajima K, Murakami K, Okabe H, Okamato S, Okada Y: Effects of plasma kallikrein specific inhibitor and active-site blocked factor VIIa on the pulmonary vascular injury induced by endotoxin in rats. Thromb Haemost 78:1209, 1997. 65 Salvatierra A, Velasco F, Rodriguez M, Alvarez A, Lopez-Pedrera R, Ramirez R, Carracedo J, Lopez F, Lopez-Pujol J, Guerrero R: C1-Esterase inhibitor prevents early pulmonary dysfunction after lung transplantation in the dog. Am J Respir Crit Care Med 155:1147, 1997. 66 Attal M, Huguet H, Huynh A, Charlet JP, Payen JL, Voigt JL, Brousset P, Selves J, Muller C, Pris J, Laurent G: Prevention of hepatic venoocclusive disease after bone marrow transplantation by continuous infusion of low-dose heparin. Blood 79:2834, 1991. 67 Demuynck H, Vandenberghe P, Verhoef GEG, Zachee P: Prevention
150
of veno-occlusive disease (VOD) of the liver after marrow and blood progenitor cell transplantation: a prospective, randomized study of different prophylactic regimens. Blood 86:620a, 1985. [abstr] 68 Or R, Nagler A, Shpilberg O, Elad S, Naparstek E, Kapelushnik J, Gillis S, Chetrit A, Slavin S, Eldor A: Low molecular weight heparin (LMWH) may prevent veno-occlusive disease (VOD) and induce accelerated platelets recovery following bone marrow transplantation (BMT). Thromb Haemost 73:974, 1995. [abstr] 69 Bearman SI, Shuhart M, Hinds MS, McDonald GB: Recombinant human tissue plasminogen activator for treatment of established severe veno-occlusive disease of the liver after bone marrow transplantation. Blood 80:2458, 1992. 70 Yu LC, Malkani I, Regueira O, Ode DL, Warrier RP: Recombinant tissue plasminogen activator (rt-PA) for veno-occlusive liver disease in pediatric autologous bone marrow transplant patients. Am J Hematol 46:194, 1994.
