Abstract
Objective
An increasingly recognized prognostic factor for out-of-hospital-cardiac-arrest (OHCA) patients is the ischemia-reperfusion injury after restored blood circulation. Endothelial injury is common in patients resuscitated from cardiac arrest and is associated with poor outcome. This study was designed to investigate if iloprost infusion, a prostacyclin analogue, reduces endothelial damage in OHCA patients.
Methods
50 patients were randomized in a placebo controlled double-blinded trial and allocated 1:2 to 48-hours iloprost infusion, (1 ng/kg/min) or placebo (saline infusion). Endothelial biomarkers (soluble thrombomodulin (sTM), sE-selectin, syndecan-1, soluble vascular endothelial growth factor (sVEGF), vascular endothelial cadherine (VEcad), nucleosomes) and sympathoadrenal activation (epinephrine/norepinephrine) from baseline to 48 and 96-hours were evaluated.
Results
Iloprost infusion did not influence endothelial biomarkers by the 48-hour endpoint. A rebound effect was observed with higher biomarker plasma values in the iloprost group (sTM p=0.02; Syndecan p=0.004; nucleosomes p<0.001; VEcad p<0.03) after 96-hours. There was a significant difference in 180-day mortality in favor of placebo. There was no difference regarding total adverse events between groups (p=0.73). Two patients were withdrawn in the iloprost group due to hypotension.
Conclusions
The administration of low-dose iloprost (1ng/kg/min) to OHCA patients did not significantly influence endothelial biomarkers as measured by the 48- hour endpoint. A rebound effect was however observed in the 96-hour statistical model, with increasing endothelial biomarker levels after cessation of the iloprost-infusion.
Considerable research has been conducted in pre-hospital settings and in specialized cardiac arrest centers to improve survival and neurological outcome after out-of-hospital-cardiac-arrest (OHCA). During the latest decades, the survival rates of OHCA to hospital discharge have improved markedly, even so, survival rates are still considered low and extend from 2, 6, 9 and 11 percent for Asia, North America, Europe and Australia respectively. Another central assessment of OHCA survivors is the functional outcome after cardiac arrest; approximately half of the unconscious OHCA survivors suffer cognitive impairment due to hypoxic-ischemic brain injury. The high mortality rate among patients with initial return of spontaneous circulation (ROSC) is closely related to the development of the post-cardiac-arrest-syndrome (PCAS). PCAS involves ischemia-reperfusion injury and often multiple organ dysfunctions as a response to the periods of no flow, low flow and reperfusion during and following resuscitation.
During cardiac arrest the vascular endothelium, one of our largest organ structures, is exposed to universal ischemia. The consequential ischemia-reperfusion injury has in several studies been related to increased circulating levels of endothelial biomarkers reflecting endothelial activation and damage. Levels of endothelial derived biomarkers reflecting shedding of the glycocalyx (syndecan-1), endothelial activation (sE-selectin), endothelial permeability (soluble vascular endothelial growth factor (sVEGF); vascular endothelial cadherine (VEcad)) and endothelial damage (soluble thrombomodulin (sTM); nucleosomes) have recently been related to patient outcome in OHCA patients from post-hoc observational data. Microcirculatory dysfunction due to glycocalyx degradation, endothelial activation and cell damage, promotes vascular permeability, coagulopathy, and endothelial mediated inflammation. , , Clinically, the endothelial injury manifests as hypovolemia, impaired vaso-regulation and compromised oxygen delivery and utilization, which is increasingly recognized as a fundamental factor contributing to the development of multiple organ failure and poor outcome.
We hypothesize that therapeutic interventions directed towards the endothelium will improve endothelial integrity and outcome for patients with PCAS. Prostacyclin is released by endothelial cells and has become of interest because of its cytoprotective properties in addition to its antiplatelet and vasodilatory effects. Several studies with prostacyclin analogues administration have reported results indicating preserved endothelial integrity. The rationale being that prostacyclin analogues such as iloprost, may be beneficial as an endothelial rescue treatment in patients suffering from endothelial breakdown. Iloprost administered to critically ill OHCA patients are expected to limit endothelial activation and damage, including the glycocalyx shedding reflected by the primary outcome: endothelial biomarker change (syndecan-1, sE-selectin, VE-cad, VEGF, sTM and nucleosomes). Prostacyclin analogues are effective vasodilators and a randomized controlled trial of epoprostenol therapy for severe congestive heart failure (FIRST), showed that the systemic vasodilation acquired was detrimental in terms of clinical outcome. Prostacyclin analogue-induced vasodilation is dose dependent and doses of 0.5-2.0 ng/kg/min has previously been utilized in other trials without reported significant adverse side effects on blood pressure, central venous pressure, or platelet aggregation and is considered a safe intervention. , ,
The OHCA pathophysiologic profile, demonstrates excessive sympathoadrenal activation and endotheliopathy. This is consistent with other critically ill patient populations, where serum catecholamines display harmful effects on the endothelium. Systemic blood pressure targets are recommended as part of goal directed care in most post cardiac arrest care protocols and requires the use of catecholamines to achieve blood pressure targets. Blood pressure targets and exogeneous catecholamines may consequently influence the effect of iloprost on the endothelium.
The aim of this study was to assess safety and efficacy of iloprost administration on endothelial damage in OHCA patients. To evaluate a possible interaction between exogenously administered catecholamines and iloprost efficacy on the endothelium, participants were further randomized into two different blood pressure target groups (65 mmHg vs. 65 mmHg + 10%). ,
Patients & Methods
Trial Design
This study was a randomized, double-blinded, placebo-controlled, single center, 2×2 factorial prospective pilot trial. Patients were randomized to receive either 48 hours iloprost infusion (Bayer, Leverkusen, Germany) at a dose of 1ng/kg/min or placebo (0,9% saline) infusion at an equal volume. Both iloprost and placebo infusions were prepared for every patient by nurses at a separate department, ensuring blinding of both physicians and nurses caring for included patients. Patients were furthermore randomized in a 2×2 factorial design to a target blood pressure of 65mmHG or 65mmHg + 10% mmHg. Hence, patients were allocated in four different permutations: i) iloprost/65 mmHg, ii) iloprost/65 mmHg + 10%, iii) placebo/65 mmHg, and iv) placebo/65 mmHg + 10%. The design and rationale of the trial, along with the statistical analysis plan, have been published previously. In accordance with national guidelines and the Declaration of Helsinki, written informed consent was waived, delayed, or obtained from a legal surrogate, depending on the circumstances, and was obtained from each patient who was legally capable. The trial commenced in February 2016 and included participants during a period of 6 months. The end of trial defined by last patient’s last visit was completed in February 2017.
Monitoring and approvals
The trial protocol, amendments, written information and consent forms were approved by the Danish Health and Medicines Authority (Case no: 2014092629; EudraCT no: 2014-002998-11), Committee on Health Research Ethics (H-1-2014-097) and Danish Data Protection Agency (Journal no: 30-1330). The protocol is available at www.clinicaltrials.gov (identifier NCT02685618) and has been published in Trials Journal (manuscript ID: 1488709955200795). This trial was conducted and monitored according to guidelines for Good Clinical Practice.
Patients
Eligible patients were over 18 years of age with OHCA of presumed cardiac cause, Glasgow coma score (GCS) <8 and had sustained (>20min) ROSC. Patients with ROSC > 240 minutes before screening, un-witnessed arrest with asystole as the initial rhythm, suspected or known acute intracranial hemorrhage or stroke were excluded from the trial. The complete list of exclusion criteria is listed in Appendix 1 .
Standardized therapy: Target Temperature, Sedation and Hemodynamic Management
All patients received standardized temperature management. The core body temperature was set as quickly as possible at the predefined target temperature 36°C by utilization of Thermoguard XP temperature management system ( Zoll ), an intravenous cooling catheter system. The target core temperature was then maintained for 24 h. After the maintenance period, core temperature was gradually raised to normothermia defined as 37°C with rewarming no more than 0.5°C/hour. Body temperature was then maintained at normothermia 37 ±0.5°C for as long the as patient was comatose until 72 hours from sustained ROSC in both treatment groups.
Patients were sedated with propofol and fentanyl accordingly. To achieve the guideline supported goal of mean arterial artery pressure of 65mmHg during the intensive unit care stay, nor-adrenalin (mkg/kg/min solution) was administered in increments of 0.02 mkg/kg/min until a dose of 0.2 mkg/kg/min was reached. Dopamin (mkg/kg/min solution) was up-titrated until a maximal dose of 10 mkg/kg/min. In circumstances with low cardiac output, addition of inotropic agents such as: Milrinone (0.375-0.75 mkg/kg/min), Dobutamin (2-10 mkg/kg/min) or Levosomendan (0.05-0.2 mkg/kg/min) was administered.
Intervention and Randomization Procedure
Randomization was performed by a web-based site (Zenodotus eCRF) and administered by a webmaster unrelated to the trial. Each patient was assigned a unique trial and randomization number. Patients were randomized to receive either iloprost (Bayer, Leverkusen, Germany) or placebo and later randomized to a target blood pressure of 65mmHG or 65 mmHg + 10%, giving four permutations. Patients have been randomized 1:2 for iloprost vs placebo and 1:1 for a target blood pressure of 65 mmHg or 65 mmHg + 10%.
Blood Sampling
Blood samples were collected upon arrival (baseline) to the intensive care unit before commencing study drug/placebo infusion and at 6, 24, 48, 72 and 96 hours after the administration of study drug/placebo. Blood samples were taken for standard arterial and venous blood gas analysis (ABG, Radiometer ABL 725/735, Copenhagen, Denmark) and routine biochemistry and research analyses (Citrate, Lithium-heparin, EDTA samples). Routine biochemistry samples were analyzed in a DS/EN ISO 15189 standardized laboratory by a Sysmex XE-2100 (hemoglobin, PLT) and ACL TOP (INR and fibrinogen). All samples were centrifuged immediately at 4°C. Plasma and serum aliquots were stored at -80°C until thawed for analysis.
Biomarkers
Biomarkers of endothelial damage and activation and cathecholamines were analyzed by Enzyme linked immunosorbent assay (ELISA). Syndecan-1 (Diaclone, Nordic Biosite, Copenhagen, Denmark; Lower limit of detection (LLD) 4.94 ng/ml); sE-selectin (R&D Systems Europe Ltd., Abingdon, UK; LLD 0.009 ng/ml); soluble thrombomoduline (sTM) (Diaclone, Nordic Biosite, Copenhagen, Denmark; LLD 0.31 ng/ml); vascular endothelial cadherine (sVE-cad) (R&D Systems Europe, Ltd., Abingdon, UK; LLD 0.113 ng/ml). Soluble vascular endothelial growth factor (sVEGF) (Quantikine, R&D Systems Europe, Ltd., Abingdon, UK; LLD 3.5 pg/ml); nucleosomes (Cell Death Detection ELISA PLUS, relative quantification; Roche, Hvidovre, Denmark); Plasma (p)-adrenaline and p-noradrenaline (2-CAT ELISAFAST TRACK, Labor Diagnostica Nord GmbH & Co. KG, Nordhorn, Germany; LLD 10 pg/ml (adrenaline) and 50 pg/ml (noradrenaline), respectively);
Viscoelastic Hemostatic Assays; Thromboelastography & Multiplate
Thromboelastography (TEG) measures the physical properties of whole blood formation via a pin suspended in the sample cup. Citrated (3.2% citrate) blood samples were analyzed at 37 °C by the manufacturers recommendations (TEG 5000; Haemoscope Corporation, Niles, IL, USA). TEG assays were analyzed in heparinase cups. Key TEG parameters for Citrated Kaolin (CK) and Functional Fibrinogen (FF) assays: Reaction time (R) (rate of fibrin formation), angle (α) (clot growth kinetics and thrombin burst), maximum amplitude (MA) (maximum clot strength), lysis after 30 min (Ly30) and lysis after 60 minutes (Ly60) (proportional reduction in amplitude after MA, fibrinolysis) were recorded.
Multiplate® (Roche) tests analyses platelet aggregation by measuring the impedance between two sensor-electrodes submerged in the sample. Results were recorded as the area under the aggregation curve (AUC) (AU*min or U) at 6 min for each test. Blood samples were collected in lithium-heparinized tubes and analyzed at 37°C within 1 hour of sampling. The platelet response to iloprost/placebo intervention was analyzed regardless of antithrombotic or anticoagulation therapy. Aggregation was determined in response to the following platelet tests: ASPI-test (arachidonic acid) normal range 79-141 U; TRAP-test (thrombin-receptor activating peptide (TRAP)-6 normal range 92-151; RISTO-test (Ristocetin) normal range 65-116; ADP-test (adenosin diphosphat ) normal range 55-117 U.
Bleeding & Blood Transfusions
The total bleeding volume measured by drainage systems (ml) and administration of any blood products (red blood cells (RBC), fresh frozen plasma (FFP) and platelet concentrates) were recorded at 24, 48, 72 & 96 hours.
Sample Size
The sample size of 40 patients was defined by a power calculation based on a randomized study investigating the effect of iloprost infusion (1 ng/kg/min) on hemostasis and the endothelium in patients undergoing Whipple surgery. The primary biomarker of interest, sTM, was used for the power calculation: the increase in circulating sTM differed significantly among the active ((mean ±SD) 0.53 ±0.52 ng/ml) and placebo (2.14 ±2.19 ng/ml) group post-surgery (p=0.046). To detect a difference in Thrombomodulin plasma levels, 2-sided with a power of 0.85 (1-β) and α of 0.05, 19 patients were required in each group (STATA IC 14.1). Inclusion of patients continued until 40 patients had completed the 96-hour endpoint. The primary objective of the trial was the evaluation of the safety and efficacy of iloprost and thus no sample size adjustments with regards to blood pressure intervention was performed.
During preparations for the un-blinding of allocated treatments in the ENDO-RCA trial, a configuration error in the electronic randomization study was discovered. This caused the system to randomize the patients in a 1:2 fashion (iloprost vs. placebo) instead of the 1:1 randomization scheme set out in the study protocol. This has in no way affected neither the blinding of investigators and study personnel, nor the randomness of how the patients have been allocated throughout the trial. However, the intended sample size per group set out in the protocol was not achieved. In total, 40 patients were set out to be included per-protocol, due to patients being replaced if dropped out prior to the 96-hour endpoint, 50 patients were included increasing total sample size. In the intention to treat population 16 patients were allocated to the iloprost arm and 34 to the placebo arm. The inadequate randomization and imbalanced sample size were evaluated in the modified intention to treat population iloprost:placebo/13:33, in a two-sample mean t-test with unequal variance and unbalanced groups according to original sample power calculation giving a slightly underpowered trial (0.84 (1-β) with α of 0.05) (supplemental data Figure 1 .) The blood pressure intervention was randomized as planned in a 1:1 fashion and a feasibility study has been published. A letter of clarification has been sent to the Danish Health and Medicines Authority (Case no: 2014092629; EudraCT no: 2014-002998-11) and Trials Journal.
Statistics
This study was a pilot trial and thus all statistical tests were considered exploratory. Statistical analysis on primary endpoints, patient characteristics and mortality were performed on the modified intention to treat (mITT) population defined as all randomized patients except for patients withdrawing consent, not fulfilling inclusion criteria or/and never receiving the intervention. Statistical analyses were performed using SAS Enterprise Guide 7.1 (SAS-Institute Inc., Cary, NC) and IBM SPSS statistics 22 (IBM, Armonk, NY). Endothelial and other organ biomarker values were log10 transformed and analyzed by baseline corrected repeated measurement mixed models with Group, Time and Group*Time as fixed effects at the 48 and 96-hour endpoints. Variance components were unstructured and Tukey post hoc tests applied. Patient data were compared by Mann–Whitney U-tests and χ2/Fischer exact tests as appropriately and presented as medians with interquartile ranges (IQR) or as number (%). Mortality between groups was compared using χ2/Fischer exact tests, survival statistics and Kaplan Meier. Differences with a p<0.05 were considered statistically significant.
Funding
This Trial was funded by The Danish Council for Independent Research (Medical Sciences), The Capital Region of Denmark Fund , The Capital Hospital of Denmark Heart Center Fund , The Danish Heart Foundation , The Augustinus Fund and The Gangsteds Fund. The authors are solely responsible for the design and conduct of this study, all study analyses and drafting and editing of the paper.
Results
Patient Flow
In total, 50 patients were randomized. In the placebo arm: 1 patient was excluded due to foreign nationality. In the iloprost arm: 1 patient did not receive the intervention; 1 patient withdrew consent; 1 patient did not fulfill the inclusion criteria. The mITT population consisted of 46 patients with an allocation of 1:2 resulting in 33 patients allocated to placebo and 13 patients allocated to iloprost therapy. 44 patients completed the 48-hour endpoint and 40 patients completed the 96-hour endpoint ( Figure 1 ).
Patient characteristics
The patient population was mainly comprised of males. However, the iloprost group had significantly more females. Allocation of blood-pressure targets were evenly spread between patients receiving placebo or iloprost. With respect to patient characteristics and therapy, no significant differences were found ( Table 1 ).
| Variable | Unit | Placebo Median (IQR) | Prostacyclin Median (IQR) | p-value |
|---|---|---|---|---|
| N=33 | N=13 | |||
| Demographics | ||||
| Age | yrs | 63 (55-69) | 57 (50-67) | 0.41 |
| Gender | male | 31 (93.9%) | 9 (69.2%) | 0.05 |
| BMI | kg/m | 26.1 (24.2-29.3) | 25.3 (24.5-27.8) | 0.90 |
| Co-morbidity | ||||
| Previous hypertension | n (%) | 13 (39,4%) | 6 (46.2%) | 0.68 |
| Ischemic heart disease | n (%) | 7 (21,2%) | 4 (30.8%) | 0.70 |
| Heart failure | n (%) | 3 (9,1%) | 1 (7.7%) | 1.00 |
| Pacemaker | n (%) | 0 (0%) | 0 (0%) | – |
| Implantable cardioverter defibrillator | n (%) | 0 (0%) | 0 (0%) | – |
| Cardiac Resynchronization Therapy (CRT) | n (%) | 0 (0%) | 0 (0%) | – |
| Percutaneous coronary intervention (PCI) | n (%) | 2 (6,1%) | 1 (7.7%) | 1.00 |
| Coronary artery bypass graft (CABG) | n (%) | 1 (3%) | 1 (7.7%) | 0.49 |
| Valve surgery | n (%) | 1 (3%) | 1 (7.7%) | 0.49 |
| Stroke | n (%) | 4 (12%) | 1 (7.7%) | 1.00 |
| Pulmonary disease | n (%) | 1 (3%) | 1 (7.7%) | 0.49 |
| Nephropathy | n (%) | 8 (24,2%) | 2 (15.4%) | 0.70 |
| Dialysis | n (%) | 0 (0%) | 0 (0%) | – |
| Alcohol abuse | n (%) | 2 (6,1%) | 1 (7.7%) | 1.00 |
| Drug abuse | n (%) | 1 (3%) | 1 (7.7%) | 0.49 |
| Intervention | ||||
| Blood pressure group | rdz | (n=18) vs. (n= 15) | (n=8) vs. (n= 5) | 0.67 |
| Time from OHCA to Intervention | hh:min | 3:22 (2:47-4:04) | 2:58 (2:24-4:00) | 0.24 |
| Pre-hospital Therapy | ||||
| Witnessed arrest | n (%) | 33 (100%) | 13 (100%) | – |
| Bystander cardiopulmonary resuscitation(CPR) | n (%) | 32 (97%) | 11 (84.6%) | 0.19 |
| Chokable Rythm | n (%) | 30 (94.7%) | 12 (92.3%) | 1.00 |
| Pre-hosp automated external defibrillator (AED) | n (%) | 9 (28.1%) | 1 (7.7%) | 0.24 |
| Pre-hosp Adrenalin | mg | 1.15 (0-3) | 0.75 (0-2.51) | 0.52 |
| Pre-hosp Amiodaron | mg | 0 (0-300) | 0 (0-150) | 0.61 |
| Pre-hosp Heparin U | n (%) | 7 (22%) | 4 (12.5%) | 0.70 |
| Pre-hosp Acetylsalicylic acid | n (%) | 9 (28%) | 4 (12.5%) | 1.00 |
| Time to ROSC | min | 19 (12-30) | 15 (14-16) | 0.42 |
| Shock Corrected | n (%) | 2 (6.3%) | 2 (15.4%) | 0.57 |
| Clinical Characteristics on ICU admission | ||||
| Temperature | Celcius | 36 (35.5-36.6) | 35.7 (35.0-36.2) | 0.55 |
| Serum pH | 7.23 (7.17-7.27) | 7.26 (7.14-7.31) | 0.69 | |
| Serum lactate | mmol/L | 4.1 (2.2-8.5) | 5.7 (3.5-8.4) | 0.33 |
| Serum Creatinin | μmol/L | 109 (101-133) | 105 (94-139) | 0.49 |
| In hospital Therapy | ||||
| Coronary arteriography (CAG) on admission | n (%) | 33 (100%) | 12 (92.3%) | 0.28 |
| PCI on admission | n (%) | 18 (58.1%) | 7 (53.8%) | 0.78 |
| PCI full revascularization | n (%) | 14 (82.4%) | 6 (85.7%) | 1.00 |
| CABG acute | n (%) | 3 (9.1%) | 0 (0%) | 0.55 |
| Pacemaker implant | n (%) | 5 (15.2%) | 0 (0%) | 0.30 |
| ICD implant | n (%) | 12 (37.5%) | 1 (7.7%) | 0.07 |
| ventricular assist device (VAD) | n (%) | 1 (3%) | 0 (0%) | 1.00 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
