clinical investigations in critical care

Transcription

1 clinical investigations in critical care Extended Evaluation of Recombinant Human Activated Protein C United States Trial (ENHANCE US)* A Single-Arm, Phase 3B, Multicenter Study of Drotrecogin Alfa (Activated) in Severe Sepsis Gordon R. Bernard, MD, FCCP; Benjamin D. Margolis, MD, FCCP; Harvey M. Shanies, MD, PhD, FCCP; E. Wesley Ely, MD, MPH, FCCP; Arthur P. Wheeler, MD, FCCP; Howard Levy, MB,BCh, PhD, FCCP; Kar Wong, PhD; and Theressa J. Wright, MD; for the Extended Evaluation of Recombinant Human Activated Protein C United States Investigators Study objective: To gather additional 28-day all-cause mortality and safety data among adult patients with severe sepsis who were treated with drotrecogin alfa (activated). Design: Prospective, single-arm, multicenter clinical trial. Setting: Eighty-five study sites in the United States and two in Puerto Rico. Participants: Adult patients (273 patients) with a diagnosis of severe sepsis, which was defined as a systemic inflammatory response due to acute infection and one or more sepsis-induced organ dysfunctions present for < 48 h, as in the recombinant human activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) trial. Interventions: Drotrecogin alfa (activated) [Xigris; Eli Lilly and Company; Indianapolis, IN], 24 g/kg/h, as a continuous IV infusion for a duration of 96 1h. Measurements and results: The primary end point was all-cause mortality, which was assessed 28 days after the start of the infusion of drotrecogin alfa (activated). Serious bleeding was monitored to day 28. Comparisons of mortality were made to treatment groups from two double-blind, placebo-controlled clinical trials (PROWESS United States and the Secretory Phospholipase A2 Inhibitor [spla2i] in Severe Sepsis trial) that used similarly defined patient populations from the United States. For the 273 adult patients enrolled in this study, the 28-day all-cause mortality rate was 26.4%. This mortality rate was 6% lower than that observed in the placebo groups in the PROWESS US trial (32.9%) and the spla2i trial (33.2%), and was similar to that of the group treated with drotrecogin alfa (activated) in the PROWESS US trial (24.4%). One nonfatal intracranial hemorrhage was reported in the Extended Evaluation of Recombinant Human Activated Protein C United States trial (ENHANCE US) [0.35%]. Serious bleeding events during the infusion period occurred in 11 patients (4.0%) compared to 10 patients (2.8%) in the PROWESS US drotrecogin alfa (activated) treatment group. Conclusions: Despite the limitations associated with comparisons across trials, this study provides confirmatory evidence of the efficacy and safety of drotrecogin alfa (activated) documented in the PROWESS trial. (CHEST 2004; 125: ) Key words: activated protein C; acute physiology and chronic health evaluation II; drotrecogin alfa (activated); organ dysfunction; safety; severe sepsis; survival Abbreviations: APACHE acute physiology and chronic health evaluation; CI confidence interval; ENHANCE Extended Evaluation of Recombinant Human Activated Protein C; PROWESS Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis; SOFA sequential organ failure assessment; spla2i secretory phospholipase A2 inhibitor 2206 Clinical Investigations in Critical Care

2 Severe sepsis, a generalized inflammatory host response to infection that results in coagulopathy, diffuse endovascular injury, and death, occurs in approximately 750,000 patients in the United States each year. 1 Patients with severe sepsis often experience failures of multiple organ systems, and recent estimates suggest that between 36% and 60% of all cases are fatal. 2,3 Despite technical advances in critical care, including new dialysis techniques and methods of mechanical ventilation, sophisticated nutritional support, and antimicrobial therapies, mortality from severe sepsis and its sequelae had remained largely unchanged until the recombinant human activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) trial. 4 In November 2001, drotrecogin alfa (activated) [Xigris; Eli Lilly and Company; Indianapolis, IN] was approved in the United States for the treatment of adult patients with severe sepsis who have a high risk of death. This approval followed the completion of a phase 3 clinical trial (1,690 patients) 4 and a supporting US phase 2 trial (131 patients). 5 The PROWESS trial was a double-blind, multicenter study that randomized adult patients with severe sepsis to receive a 96-h infusion of drotrecogin alfa (activated), 24 g/kg/h, or placebo in addition to usual care. The primary end point was 28-day all-cause mortality, and treatment with drotrecogin alfa (activated) was associated with a 19.4% relative risk reduction in death compared with placebo (p 0.005). The Extended Evaluation of Recombinant Human Activated Protein C (ENHANCE) trial was a global, single-arm, phase-3b clinical trial of drotrecogin alfa (activated) that was designed to gather additional mortality and safety data among patients with severe sepsis in a monitored clinical trial setting. For administrative and data management purposes, this *From the Division of Allergy, Pulmonary, and Critical Care Medicine (Drs. Bernard, Ely, and Wheeler), Vanderbilt University School of Medicine, Nashville, TN; Section of Pulmonary and Critical Care Medicine (Dr. Margolis), West Suburban Hospital, Oak Park, IL; Division of Pulmonary and Critical Care Medicine (Dr. Shanies), Mount Sinai School of Medicine, New York, NY; and Lilly Research Laboratories (Drs. Levy, Wong, and Wright), Eli Lilly and Company, Indianapolis, IN. A list of the ENHANCE US investigators in located in the Appendix. This study was supported by Eli Lilly and Company, Indianapolis, IN. Manuscript received December 31, 2002; revision accepted December 12, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( permissions@chestnet.org). Correspondence to: Gordon R. Bernard, MD, FCCP, Division of Allergy, Pulmonary and Critical Care Medicine, Department of Medicine, T-1218 Medical Center North, Vanderbilt University School of Medicine, Nashville, TN 37232; Gordon.Bernard@ vanderbilt.edu study was conducted as three separate and independent studies, and the study conducted in the United States (including Puerto Rico) concluded in November While the global ENHANCE study enrolled adult and pediatric patients, this report contains data only from the adult patients enrolled in the ENHANCE US study. Efficacy and safety data from the ENHANCE US study were compared with those from two completed, double-blind, placebocontrolled clinical trials that used similarly defined patient cohorts from the United States. Patients Materials and Methods From March 2001 through November 2001, 273 eligible adult patients were enrolled in this trial, which was conducted in 85 study sites in the United States and two in Puerto Rico. Patient selection criteria were identical to that for the PROWESS trial and have been provided in more detail elsewhere. 4 Briefly, patients were eligible for entry if they had a suspected or proven infection, evidence of systemic inflammatory response, and one or more sepsis-induced organ dysfunctions that had lasted for 48 h. Patients were excluded from the study if they were pregnant or breastfeeding, weighed 135 kg, had a platelet count of 30,000 cells/ L, had a known hypercoagulable condition, were at an increased risk of bleeding, experienced their first sepsis-induced organ failure 48 h prior to infusion, or were at a high risk of death from a nonseptic cause within 28 days. Ethics The institutional review board at each center approved the protocol, and written informed consent was obtained from every patient or their authorized representatives. The clinical coordinating center was available 24 h a day throughout the study to answer investigators questions regarding safety and patient eligibility, and for the reporting of serious adverse events. Treatment Patients at each center received drotrecogin alfa (activated), 24 g/kg body weight per hour, as a continuous infusion for 96 h. Guidelines recommended that the infusion be interrupted 2 h prior to any surgical procedure, and be resumed within 1 h after any minor percutaneous and 12 h after any major surgical procedure. The study protocol allowed the concomitant use of antibiotics, fluids, vasopressors, or ventilatory support at the discretion of the investigator. Evaluation Patients were observed until the 28-day time point (ie, 672 h after the start of the infusion) or death. Baseline characteristics were assessed within 48 h prior to the start of infusion. Specifically, demographics, preexisting conditions, organ function based on sequential organ failure assessment (SOFA), 6 acute physiology and chronic health evaluation (APACHE) II score, 7 presence of septic shock, and laboratory values were assessed at study entry. The assessment of vital signs, infection status, transfusion status, and adverse events continued through day 28, and SOFA scores CHEST / 125 / 6/ JUNE,

3 were collected on day 0 (ie, after the start of infusion) through day 4 in all patients and throughout the duration of the study if available. Patients were defined as having septic shock if they met the criteria for cardiovascular organ failure at any time within 6 h before the start of the infusion. Protein C deficiency was defined as a plasma protein C activity below the lower limit of normal (ie, 81%) immediately prior to the start of the infusion. Vasopressor use was defined by a cardiovascular SOFA score of 3 or 4 at baseline. Central and local laboratories were used for all measurements, including hematology (with platelet counts), clinical chemistry, and coagulation in this study. In addition to collecting mortality data, survivor location and status on day 28 were assessed and classified as follows: homerequiring no professional support, home-requiring paid professional support, or at a hospital or nursing home. All adverse events, including serious bleeding, were recorded up to 28 days after the start of infusion. Bleeding events were considered to be serious if they were fatal or life-threatening, or were associated with intracranial hemorrhage or the transfusion of 3 U packed RBCs per day for 2 consecutive days. Other adverse events considered to be serious included those resulting in prolonged hospitalization or disability, or any other medically significant reason, as determined by the investigator. Serious bleeding events were considered to be procedure-related if they were associated with an invasive or surgical intervention. Statistical Analysis The primary efficacy end point was all-cause mortality, assessed 28 days (672 h) after the start of infusion. Ninety-five percent confidence intervals (CIs) were calculated for mortality data using the normal approximation for the binomial method. Stratified analyses were performed on selected baseline covariates that were reflective of disease severity and were considered to be prognostic indicators of death (eg, the number of organ failures, APACHE II score, and presence of septic shock). APACHE II quartiles and halftiles were based on the cut points from the global PROWESS trial. In addition, Kaplan-Meier curves were generated for 28-day all-cause mortality and were compared with historical comparator groups (described below). Historical Comparators Comparisons of 28-day all-cause mortality data were made with treatment groups from two double-blind, placebo-controlled clinical trials that used similarly defined patient populations from the United States. These studies consisted of the US subset of patients enrolled in the PROWESS trial and patients who had been randomly assigned to receive placebo as part of a phase 2 placebo-controlled trial of a secretory phospholipase A2 inhibitor (spla2i) 8 in the treatment of severe sepsis. (In the spla2i trial, which was sponsored by Eli Lilly and Company, the investigational drug LY315920Na/S-5920 failed to show a statistically significant survival benefit using prospectively defined primary analyses. 8 ) To minimize differences in patient characteristics, severity of illness, and treatment practice patterns across trials, treatment groups only from US trials of severe sepsis were compared. Because the PROWESS US and spla2i trials recruited populations using similar inclusion/exclusion criteria, these patient populations were appropriate comparators. The primary difference among the comparator trials was that the spla2i placebo population did not exclude patients who were at a high risk for bleeding, because bleeding was not considered to be associated with this investigational drug. However, an increased risk of bleeding is associated with drotrecogin alfa (activated), and therefore patients at a high risk for bleeding were excluded from the ENHANCE US trial and the PROWESS US trial. In addition, the spla2i trial used slightly different criteria to define sepsis-induced organ failure. In the spla2i trial, respiratory failure was defined as a Pao 2 of 300 mm Hg (compared to 250 mm Hg in the PROWESS US and EN- HANCE US trials), and hematologic failure was defined as a 50% decrease in platelet count or a platelet count of 100,000 cells/ L (compared to a 50% decrease in platelet count or a platelet count 80,000 cells/ L in the PROWESS US and ENHANCE US trials). Baseline measures of disease severity were compared among the trials, but the spla2i trial did not collect data on septic shock, prior patient location, and protein C deficiency at baseline. Mortality results of the US trials were compared with the mortality of the global PROWESS trial. Results A total of 273 ENHANCE US adult patients received treatment with drotrecogin alfa (activated). Informed consent was obtained for 288 patients. Of these, 15 patients (5.2%) did not receive the study drug (on review, 13 of the 15 patients did not meet the entry criteria for the study, 1 patient withdrew consent before starting the infusion, and 1 patient withdrew from the study because of a physician decision). Patients in the study were scheduled to receive 24 g/kg/h drotrecogin alfa (activated) for 96 h (scheduled total dose, 2,304 g/kg). Of the 273 patients who received the study drug, 195 (71.4%) received drotrecogin alfa (activated) for a duration of at least 95 h. The mean actual dose of drotrecogin alfa (activated) received by treated patients was g/kg/h (median dose, g/kg/h). The mean total exposure to drotrecogin alfa (activated) was 1, g/kg (median exposure, 2, g/ kg). The total exposure was less than that scheduled primarily due to patient deaths. The most common reason for the interruption of drotrecogin alfa (activated) infusion was that the patient had to undergo a procedure. Patient Characteristics at Baseline As shown by Table 1, the overall ENHANCE US patient population was comparable to those in historical comparator trials at baseline. As in both comparator studies, most of the patients in the ENHANCE US trial were white (72.9%). In addition, all three studies enrolled patients with a similar mean age, prior patient location (not available for spla2i trial), and recent surgical history. The ENHANCE US trial enrolled fewer patients who were at least 75 years of age than did comparator trials. Primarily due to lower chronic health problem scores, mean APACHE II scores were lower in the ENHANCE US trial when compared to those in the PROWESS US trial. Based on other measures, however, patients in the 2208 Clinical Investigations in Critical Care

4 Table 1 Patient Characteristics at Baseline for the ENHANCE US Trial and Comparator Trial Groups Variables ENHANCE US Drotrecogin Alfa (Activated) Group (n 273) Drotrecogin Alfa (Activated) Group (n 352) PROWESS US Placebo Group (n 353) spla2i US Placebo Group* (n 196) Age, yr Age 75 yr, % Male gender, % White race, % Prior patient location, % Home Acute care hospital Skilled nursing home Other Recent surgical history, % Elective surgery Emergency surgery No history of surgery APACHE II score Mechanical ventilation, % Septic shock, % Use of any vasopressor, % organ failures, % organ failures, % Protein C deficiency, % Yes No Unknown Site of infection, % Lung Abdomen Urinary tract *Results from the spla2i trial. 8 Values given as mean SD. The spla2i trial used a slightly modified definition of organ failure (see Materials and Methods section). ENHANCE US trial may have had greater disease severity at baseline when compared with patients from the PROWESS US trial. For example, more patients in the ENHANCE US trial (when compared to the PROWESS US trial drotrecogin alfa [activated] and PROWESS US trial placebo groups) were in shock (74.0% vs 65.1% and 70.5%, respectively), were receiving mechanical ventilation (74.4% vs 67.6% and 73.9%, respectively), and were receiving therapy with vasopressors (63.7% vs 50.6% and 55.2%, respectively). In addition, more patients in the ENHANCE US trial had two or more organ failures (72.9% vs 66.5% and 70.5%, respectively) and three or more organ failures (39.9% vs 29.5% and 34.6%, respectively). The majority of the patients in the ENHANCE US trial and the PROW- ESS US trial were protein C-deficient ( 75%). Fewer patients in the ENHANCE US trial had a lung infection (46.9%) than in either the drotrecogin alfa (activated) treatment group (62.5%) or the placebo treatment group (55.8%) in the PROWESS US trial, or in the placebo treatment group of the spla2i trial (51.3%). Severity of illness also was evaluated by APACHE II score at baseline. Notably, more patients in the fourth APACHE II quartile (APACHE II score, 30) of the ENHANCE US trial required vasopressors (84.9% vs 69.0%, respectively), needed mechanical ventilation (94.3% vs 89.7%, respectively), had two or more organ failures (94.3% vs 82.8%, respectively), had three or more organ failures (69.8% vs 49.4%, respectively), and met the criteria for septic shock (92.5% vs 82.8%, respectively) compared with the fourth APACHE II quartile of the PROWESS US trial drotrecogin alfa (activated) group. Efficacy For all 273 adult patients enrolled in this study, the 28-day all-cause mortality rate was 26.4% (95% CI, 21.1 to 31.6%). Treatment with drotrecogin alfa (activated) in this trial was associated with a higher survival rate (6% higher) compared with those in the placebo groups in the PROWESS US trial and the spla2i trial. The mortality rate was similar to that observed in the drotrecogin alfa (activated)-treated CHEST / 125 / 6/ JUNE,

5 group of the PROWESS US trial. By day 28, Kaplan- Meier survival curves for the two drotrecogin alfa (activated) groups were distinct from the two placebo comparator groups, as seen in Figure 1. Mortality point estimates for populations treated with drotrecogin alfa (activated) in the ENHANCE US trial, the PROWESS US trial, and the global PROW- ESS trial were less than those in the comparable placebo groups (Fig 2). Table 2 summarizes mortality by APACHE II score and other baseline covariates indicative of disease severity. The mortality rates determined by APACHE II halftiles and quartiles were consistently lower among treatment groups using drotrecogin alfa (activated) compared with historical groups using placebo groups, with the exception of the EN- HANCE US trial first APACHE II quartile when compared to that of the PROWESS US trial (Table 2). In addition, the mortality rate was lower among patients in the ENHANCE US drotrecogin alfa (activated) treatment group with multiple organ failures ( 5.7% lower), mechanical ventilation (3.6% lower), and thrombocytopenia ( 11% lower) when compared with patients in the historical placebo groups. Mortality was related to patient age. In the EN- HANCE US trial drotrecogin alfa (activated) treatment group, the 28-day mortality rate was 35.6% (21 of 59 patients) among patients who were at least 75 years of age, compared with 23.8% (51 of 214 patients) among patients who were 75 years. Safety Bleeding was the only serious adverse event that was related to the administration of drotrecogin alfa (activated). In the ENHANCE US trial, the incidence of serious bleeding was 4.0% (95% CI, 1.7 to Figure 1. Kaplan-Meier estimates of survival among trial groups of adult patients from the United States with severe sepsis who were treated with drotrecogin alfa (activated) or placebo. The thick, solid line represents the ENHANCE US trial drotrecogin alfa (activated) treatment group Clinical Investigations in Critical Care

6 Figure 2. The 28-day mortality rates in trials of severe sepsis. The mortality point estimate for each treatment group is indicated by a solid square, and the 95% CI is indicated by the horizontal lines. The size of the point estimate symbol is proportional to the number of patients in the particular treatment group. The vertical line represents the 28-day mortality rate observed in the drotrecogin alfa (activated) treatment group of the ENHANCE US trial. 6.4%) for the study drug infusion period (defined as the time the infusion was initiated to 24 h after the time infusion was permanently discontinued) and 5.5% (95% CI, 2.8 to 8.2%) during the 28-day study period (Fig 3). When considering only serious bleeding events that were not procedure-related (ie, spontaneous bleeding events), the incidence of serious bleeding among patients in the ENHANCE US trial who had been treated with drotrecogin alfa (activated) was 1.8%, which was similar to the incidence among PROWESS US trial patients who had been treated with drotrecogin alfa (activated) [2.3%]. The rate of intracranial hemorrhage in the ENHANCE US trial was 0.35% (1 of 273 patients), which was comparable to the 0.56% rate (2 of 352 patients) among PROWESS US trial patients who had been treated with drotrecogin alfa (activated). In the ENHANCE US trial, a single intracranial hemorrhage (ie, a nonfatal petechial hemorrhage) occurred after the start of infusion in a 63-year-old woman who had meningitis. The most common sites of bleeding in patients in the ENHANCE US trial were GI, skin/soft tissue, and bone/joint. Serious bleeding is summarized by site and relationship to procedures in Table 3. Functional Status Among Survivors Patient location and status at day 28 among survivors treated with drotrecogin alfa (activated) in the ENHANCE US trial was comparable to that for survivors treated with drotrecogin alfa (activated) in the PROWESS US trial (Table 4). Of the 201 surviving adult patients in the ENHANCE US trial, 104 (51.7%) were at home on day 28 (80.8% of whom required no professional support). For the drotrecogin alfa (activated)-treated patients in the PROWESS US trial, 125 of 266 surviving patients (47.0%) were at home on day 28 (85.6% of whom required no professional support). Discussion The ENHANCE US study was an open-label, nonrandomized trial that allowed investigators to collect additional data about drotrecogin alfa (activated) in US patients with severe sepsis within a controlled clinical trial setting. Recognizing the limitations associated with comparing results across trials, treatment with drotrecogin alfa (activated) in the ENHANCE US trial resulted in a 28-day mor- CHEST / 125 / 6/ JUNE,

7 Table 2 Summary of 28-Day All-Cause Mortality by Baseline Variables Indicative of Disease Severity ENHANCE US Drotrecogin Alfa (Activated) Group (n 273) PROWESS US Drotrecogin Alfa (Activated) Group (n 352) PROWESS US Placebo Group (n 353) spla2i US Placebo Group* (n 196) Variables No./Total No. % No./Total No. % No./Total No. % No./Total No. % All patients 72/ / / / APACHE II halftile 25 27/ / / / / / / / APACHE II quartile 19 14/ / / / / / / / / / / / / / / / Organ failure Single 9/ / / / Multiple 63/ / / / Vasopressor Yes 54/ / / / No 18/ / / / Ventilator use Yes 65/ / / / No 7/ / / / Septic shock Yes 61/ / / No 11/ / / Thrombocytopenia 100,000 cells/ L 16/ / / / ,000 cells/ L 55/ / / / *Results from the spla2i trial. 8 APACHE II quartiles and halftiles were based on cutpoints from the global PROWESS trial. 4 The spla2i trial used a slightly modified definition of organ failure (see Materials and Methods section). tality rate that was similar to that of the drotrecogin alfa (activated) group in the PROWESS US trial. Consistent with a previous analysis of data from the global PROWESS trial, 9 the treatment effects of drotrecogin alfa (activated) were similarly observed among almost all subgroups in the ENHANCE US trial (ie, patients were segregated according to the number of organ failures, the use of a mechanical ventilator or vasopressor, and thrombocytopenia status), regardless of disease severity. The results from the ENHANCE US trial and the PROWESS US trial suggest that approximately one half of surviving patients with severe sepsis will be at home 28 days after the start of treatment with drotrecogin alfa (activated). The overall 28-day mortality rate in the EN- HANCE US trial was similar to that in drotrecogin alfa (activated)-treated groups from previous trials and was less than that in historical placebo treatment groups. Although the ENHANCE US trial patient population was younger and had a slightly lower mean APACHE II score at baseline, other characteristics indicated that patients in this population had a higher disease severity and were therefore less likely to survive during the 28-day study period than patients in the drotrecogin alfa (activated) group in the PROWESS US trial. For example, the EN- HANCE US trial had a higher percentage of patients who were receiving mechanical ventilation (74.4% vs 67.6%, respectively), met the criteria for septic shock (74.0% vs 65.1%, respectively), or were receiving vasopressors (63.7% vs 50.6%, respectively) at baseline than did the drotrecogin alfa (activated) group in the PROWESS US trial. In addition, patients in the ENHANCE US trial had a greater percentage of patients with two or more and three or more organ failures at baseline (72.9% vs 66.5% and 39.9% vs 29.5%, respectively). The slight mortality differences observed between the groups treated with drotrecogin alfa (activated) in the ENHANCE US trial and the PROWESS US trial may be explained by the smaller sample sizes and differences in baseline characteristics between the two groups. 28-Day Mortality Rate by Subgroups In previous subgroup analyses of the PROWESS trial, the absolute magnitude of the treatmentrelated benefit observed with drotrecogin alfa (activated) appeared to be related to APACHE II score at 2212 Clinical Investigations in Critical Care

8 Figure 3. Summary of 28-day serious bleeding events among patients receiving drotrecogin alfa (activated). By comparison, 1.1% of patients in the PROWESS US trial placebo group (353 patietns) experienced one or more serious bleeding events. One patient in the drotrecogin alfa (activated) group of the ENHANCE US trial experienced a procedure-related and non procedure-related serious bleeding event. study entry. 9,10 While drotrecogin alfa (activated) provided larger absolute risk reductions in mortality as patients baseline disease severity increased, the relative risk reduction associated with drotrecogin alfa (activated) treatment was consistent, regardless of disease severity at baseline. 9 The ENHANCE US trial was not designed to evaluate specific subgroups. However, when compared with historical comparators, the ENHANCE US trial subgroups (including APACHE II halftiles and quartiles) experienced a consistent treatment benefit from drotrecogin alfa (activated) therapy. Because the variability of point estimates decreased as sample size increased, mortality results from the ENHANCE US trial subgroups that included the greatest number of patients, such as APACHE II halftiles, provided the most valid comparisons with the results of the PROWESS US trial (Table 2). While patients in the ENHANCE US trial subgroups experienced higher survival when compared Table 3 Site of 28-Day Serious Bleeding Among Patients Receiving Drotrecogin Alfa (Activated)* Procedure-Related Non Procedure-Related Variables ENHANCE US (n 273) PROWESS US (n 352) ENHANCE US (n 273) PROWESS US (n 352) Bone/joint 3 (1.0) GI 1 (0.4) 0 2 (0.7) 2 (0.6) Genitourinary 0 1 (0.3) 0 1 (0.3) Intra-abdominal 1 (0.4) Intracranial (0.4) 2 (0.6) Intrathoracic 1 (0.4) 1 (0.3) 0 3 (0.9) Retroperitoneal 1 (0.4) 1 (0.3) 0 0 Skin or soft tissue 3 (1.0) 0 1 (0.4) 0 Other 0 2 (0.6) 0 0 Source unidentified 1 (0.4) 0 1 (0.4) 0 All serious bleeding events 11 (4.0) 5 (1.4) 5 (1.8) 8 (2.3) *Values given as No. (%). CHEST / 125 / 6/ JUNE,

9 Table 4 Location and Status of Survivors at Day 28* Variables ENHANCE US Drotrecogin Alfa (Activated) Group (n 201) PROWESS US Drotrecogin Alfa (Activated) Group (n 266) PROWESS US Placebo Group (n 237) At home Not requiring paid professional support 84 (41.8) 107 (40.2) 94 (39.7) Requiring paid professional support 20 (10.0) 18 (6.8) 18 (7.5) At a hospital or nursing facility 96 (47.8) 141 (53.0) 125 (52.7) Location unknown 1 (0.05) 0 0 *Values given as No. (%). with historical placebo groups, some differences in mortality point estimates between the ENHANCE US trial and the PROWESS US trial were evident among specific subgroups of patients who had been treated with drotrecogin alfa (activated). For example, patients in the fourth APACHE II quartile in the ENHANCE US trial experienced higher mortality when compared with patients treated with drotrecogin alfa (activated) in the same subgroup of the PROWESS US trial, most likely due to small sample sizes and the higher underlying disease severity (ie, more patients required therapy with vasopressors and mechanical ventilation, had two or more and three or more organ failures, and met the criteria for septic shock) of these patients at baseline. Yet, patients in the ENHANCE US trial, the fourth APACHE II quartile experienced lower mortality than did the fourth APACHE II quartile of historical placebo groups. Patients in the first APACHE II quartile (ie, APACHE II score, 19) were similarly heterogeneous across trials. Although the mortality rate was approximately 4% higher in the ENHANCE US trial subgroup when compared to treatment groups from the PROWESS US trial, this mortality rate compares favorably with that of the placebo group of the spla2i study (ie, a 5% higher survival rate). The 9% mortality difference between placebo groups in the first APACHE II quartile was the only important difference between historical placebo subgroups from the spla2i and PROWESS US trials, raising the possibility that the 12.9% mortality rate observed in the first APACHE II quartile of the PROWESS US trial may have been aberrant. Further support for this comes from the phase 3 trial 11 of tissue factor pathway inhibitor, in which the mortality rate was shown to be 21.7% among placebo patients with APACHE II scores of 20 at baseline. Patients in the ENHANCE US trial who were at a lower risk of death, based on variables other than APACHE II score (eg, patients with single organ failure, no vasopressor use, no ventilator use, or no septic shock), experienced lower mortality than did patients in the PROWESS US trial placebo group or the PROWESS US drotrecogin alfa (activated) group (Table 2). The ongoing phase IV Administration of Drotrecogin alfa (activated) in Early Severe Sepsis trial 12 of patients with severe sepsis who have a lower risk of death as determined by APACHE II score, number of organ failures, or physician assessment will evaluate the efficacy of drotrecogin alfa (activated) in this patient population. Data from the global PROWESS trial showed that older patients with severe sepsis who were treated with drotrecogin alfa (activated) had greater survival rates than older patients treated who were with placebo. 13 Patients who were 75 years of age in the drotrecogin alfa (activated) treatment group of the ENHANCE US trial experienced a 28-day mortality rate of 35.6%, which is comparable to the 33.7% mortality rate among patients in the drotrecogin alfa (activated) group who were 75 years of age in the global PROWESS trial and was less than the 49.2% rate among patients who were 75 years of age who were receiving placebo. 13 Risk of Bleeding The only serious adverse event that was associated with drotrecogin alfa (activated) treatment in the ENHANCE US trial or the PROWESS US trial was bleeding, which was related to its antithrombotic and profibrinolytic properties. 4,14 While any agent with similar properties can be expected to increase the risk of bleeding, this risk may be balanced by the prevention of thrombosis, which frequently occurs in patients with severe sepsis. For example, the results from combined clinical trials 9,14,15 showed that when compared with placebo, patients treated with drotrecogin alfa (activated) experienced numerically fewer thrombotic events and deaths from thrombotic events such as stroke and myocardial infarction. Furthermore, as shown by the global PROWESS trial, 4 the survival benefit from drotrecogin alfa (activated) therapy far outweighs the risk of bleeding. In the present study (ie, the ENHANCE US trial), more patients treated with drotrecogin alfa (activated) experienced at least one serious bleeding 2214 Clinical Investigations in Critical Care

10 event during the 28-day period compared with those in the PROWESS US trial (5.5% vs 3.7%, respectively). In the ENHANCE US trial, 69% of serious bleeding events were considered to be procedurerelated. In contrast, this procedure-related rate was only 38% among drotrecogin alfa (activated)-treated patients in the PROWESS US trial. When adjusted for these events, the incidence of spontaneous serious bleeding during the 28-day study period was similar in the ENHANCE US compared to that in the PROWESS US trial (1.8% vs 2.3%, respectively). Differences in observed bleeding rates between the two trials may be related to differences in the procedures performed, the timing of drug interruption and continuation, and patient characteristics (eg, degree of coagulopathy). The infusion of drotrecogin alfa (activated) should be discontinued 2 h before the patient undergoes invasive procedures, and should be resumed 12 h following major procedures and after ensuring adequate hemostasis. 16 Following uncomplicated invasive procedures, infusion may be resumed immediately. Because bleeding risk is increased during procedures that involve the instrumentation of large blood vessels or highly vascular organs, 14 physicians should show special caution before, during, and after such procedures. Meningitis associated with coagulopathy or severe thrombocytopenia (ie, platelet count, 30,000 cells/ L) may also potentially increase the risk of bleeding associated with drotrecogin alfa (activated). Limitations While the results were consistent with previously available data, the limitations of this study must be considered. The ENHANCE US trial was an openlabel trial. No randomization occurred, and no individual was blinded to the use of drotrecogin alfa (activated). The advantages of large, randomized clinical trials vs observational studies and the limitations of subgroup analyses have been welldocumented. Small sample sizes in this study and among analyzed patient subgroups resulted in large 95% CIs. Because the ENHANCE US trial was not powered to show a statistically significant mortality difference with the placebo comparator group in the PROWESS US trial, 95% CIs overlapped between the two groups (Fig 2). While mortality point estimates were consistent among the drotrecogin alfa (activated) and placebo groups reviewed, conclusions regarding the statistical significance of the mortality results of the drotrecogin alfa (activated) group in the ENHANCE US trial cannot be made. Mortality differences between comparable APACHE II quartiles are likely due to small sample sizes and patient heterogeneity among subgroups. In addition, APACHE II score may not be an appropriate predictor of mortality for individual patients, because certain parameters that are used to calculate APACHE II scores (eg, Glasgow coma score) may be unobtainable or unreliable in patients with severe sepsis. 24 The APACHE II scoring system has been validated as a predictor of mortality in populations, but not in individuals, and was not intended to be used as a tool for individual risk assessments. Other baseline covariates (eg, number of organ failures, the use of a ventilator or vasopressor, and thrombocytopenia) may be more reflective of the clinical status of septic patients, and therefore more predictive of mortality. 25,26 Conclusions In summary, this study provides evidence that the efficacy and safety of drotrecogin alfa (activated) is consistent with findings documented in the PROW- ESS trial. No clinically meaningful difference existed between all-cause mortality and bleeding among patients treated with drotrecogin alfa (activated) in the ENHANCE US trial and the PROWESS US trial. Treatment with drotrecogin alfa (activated) is associated with higher survival in patients with severe sepsis when compared to historical placebo control groups. The results from the ENHANCE US trial confirm the favorable benefit/risk profile of drotrecogin alfa (activated) as seen in the PROWESS trial. ACKNOWLEDGMENT: The authors thank all investigators who enrolled patients in this study, without whom this study would not have been possible, and Matthew Monberg for writing and editorial support. Appendix: The ENHANCE US Investigators M. Ahmed, Savannah, GA; T. Albertson, Sacramento, CA; N. Anas, Orange, CA; M. Astiz, New York, NY; A. Bacon III, Newark, DE; V. Bandi, Houston, TX; F. Booth, Buffalo, NY; W. Bray, Marietta, GA; R. Brilli, Cincinnati, OH; R. Carlson, Phoenix, AZ; M. Cheatham, Orlando, FL; S. Conrad, Shreveport, LA; R. Cooney, Hershey, PA; M. Cox, St. Louis, MO; M. Dolich, Tucson, AZ; D. Dries, St. Paul, MN; R. Fang, Lackland AFB, TX; R. Forse, Boston, MA; R. Fricker, Akron, OH; C. Givens, Newport News, VA; G. Gomez, Indianapolis, IN; J. Gottlieb, Philadelphia, PA; D. Graham, Springfield, IL; M. Gropper, San Francisco, CA; T. Gross, Iowa City, IA; M. Harrison, Grand Rapids, MI; R. Kearl, Phoenix, AZ; P. Kearney, Lexington, KY; J. Kennedy, Birmingham, AL; H. Kesselman, Boston, MA; G. Kinasewitz, Oklahoma City, OK; J. Kruse, Detroit, MI; J. Lamberti, Falls Church, VA; M. Levy, Providence, RI; R. Light, Nashville, TN; L. Lin, San Francisco, CA; S. Lisco, Boston, MA; T. Lo, Loma Linda, CA; R. Lodato, Houston, TX; L. Malave, Bayamon, PR; A. Martinez, Baltimore, MD; G. Mastuschak, St. Louis, MO; L. McDermott, Philadelphia, PA; J. McFeely, Berkewww.chestjournal.org CHEST / 125 / 6/ JUNE,

11 ley, CA; P. Morris, Winston-Salem, NC; D. Mueller, Lackland AFB, TX; C. Naum, Indianapolis, IN; M. Neff, Seattle, WA; M. Niederman, Mineola, NY; D. Ost, Manhasset, NY; J. Otoshi, Escondido, CA; R. Pearl, Stanford, CA; S. Pingleton, Kansas City, KS; R. Player, Birmingham, AL; R. Riker, Portland, ME; W. Rodriguez, San Juan, PR; G. Rubeiz, Indianapolis, IN; M. Rumbak, Tampa, FL; J. Samuel, Chicago, IL; E. Santomauro, Hackensack, NJ; R. St. John, Columbus, OH; M. Seneff, Washington, DC; M. Siegel, New Haven, CT; J. Siever, Phoenix, AZ; P. Sioson, Jackson, TN; J. Stasek, Salt Lake City, UT; J. Steingrub, Springfield, MA; J. Taylor, Tacoma, WA; R. Taylor, St. Louis, MO; J. Turner, Eldorado Hills, CA; T. Verville, Charlotte, NC; K. Waxman, Santa Barbara, CA; S. Weiss, Denver, CO; R. Wunderink, Memphis, TN; and P. Wright, Greensboro, NC. References 1 Angus DC, Birmingham MC, Balk RA, et al. E5 murine monoclonal antiendotoxin antibody in gram-negative sepsis: a randomized controlled trial. JAMA 2000; 283: Brun-Buisson C, Doyon F, Carlet J, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults: a multicenter prospective study in intensive care units; French ICU Group for Severe Sepsis. JAMA 1995; 274: Martin GS, Mannino DM, Eaton S, et al. The epidemiology of sepsis in the United States from 1979 through N Engl J Med 2003; 348: Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 2001; 344: Bernard GR, Ely EW, Wright TJ, et al. Safety and dose relationship of recombinant human activated protein C for coagulopathy in severe sepsis. Crit Care Med 2001; 29: Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsisrelated Organ Failure Assessment) score to describe organ dysfunction/failure: on behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22: Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985; 13: Abraham E, Naum C, Bandi V, et al. Efficacy and safety of LY315920Na/S-5920, a selective inhibitor of 14-kDa group IIA PLA2, in patients with suspected sepsis and organ failure. Crit Care Med 2003; 31: Ely EW, Laterre PF, Angus DC, et al. Drotrecogin alfa (activated) administration across clinically important subgroups of patients with severe sepsis. Crit Care Med 2003; 31: Siegel JP. Assessing the use of activated protein C in the treatment of severe sepsis. N Engl J Med 2002; 347: Abraham E, Reinhart K, Opal S, et al. Efficacy and safety of tifacogin (recombinant tissue factor pathway inhibitor) in severe sepsis. JAMA 2003; 290: Chalfin DB, Teres D, Rapoport J. A price for cost effectiveness: implications for recombinant human activated protein C (rhapc). Crit Care Med 2003; 31: Ely EW, Angus DC, Williams M, et al. Drotrecogin alfa (activated) treatment of older patients with severe sepsis. Clin Infect Dis 2003; 37: Bernard G, Macias W, Joyce D, et al. Safety assessment of drotrecogin alfa (activated) in the treatment of adult patients with severe sepsis. Crit Care 2003; 7: Dhainaut JF, Laterre PF, Janes JM, et al. Drotrecogin alfa (activated) in the treatment of severe sepsis patients with multiple-organ dysfunction: data from the PROWESS trial. Intensive Care Med 2003; 29: Xigris [package insert]. Indianapolis, IN: Eli Lilly and Company, Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000; Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000; 342: Pocock SJ, Elbourne DR. Randomized trials or observational tribulations? N Engl J Med 2000; 342: Yusuf S, Wittes J, Probstfield J, et al. Analysis and interpretation of treatment effects in subgroups of patients in randomized clinical trials. JAMA 1991; 266: Assmann SF, Pocock SJ, Enos LE, et al. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet 2000; 355: Moreira ED, Susser E. Guidelines on how to assess the validity of results presented in subgroup analysis of clinical trials. Rev Hosp Clin Fac Med Sao Paulo 2002; 57: Pocock SJ, Assmann SE, Enos LE, et al. Subgroup analysis, covariate adjustment and baseline comparisons in clinical trial reporting: current practice and problems. Stat Med 2002; 21: Tugrul S. The effects of IgM-enriched immunoglobulin preparations in patients with severe sepsis: another point of view; authors response [letter]. Crit Care 2002; 6: Martin C, Saran R, Leavey S, et al. Predicting the outcome of renal replacement therapy in severe acute renal failure. ASAIO J 2002; 48: Mathiak G, Neville L, Grass G. Targeting the coagulation cascade in sepsis: did we find the magic bullet? Crit Care Med 2003; 31: Clinical Investigations in Critical Care