A Pilot Study on Early Versus Delayed Hypertonic Saline Dextran Resuscitation in a Porcine Model of Near-Lethal Liver Injury: Early Hemodynamic Response and Short-Term Survival

We studied the effects of early versus delayed fluid resuscitation on hemodynamic response and short-term survival in a porcine model of severe hepatic injury associated with hemorrhagic shock. Eighteen anesthetized swine were randomized after
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A Pilot Study on Early  Versus  Delayed Hypertonic Saline DextranResuscitation in a Porcine Model of Near-Lethal Liver Injury: EarlyHemodynamic Response and Short-Term Survival Peep Talving, M.D., Ph.D., 1 and Louis Riddez, M.D., Ph.D.  Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden Submitted for publication March 22, 2006  Background.  We studied the effects of early  versus delayed fluid resuscitation on hemodynamic responseand short-term survival in a porcine model of severehepatic injury associated with hemorrhagic shock.  Materialsandmethods. Eighteenanesthetizedswinewere randomized after standardized liver injury intotwo groups: early resuscitation (ER,  n    9) and de-layed resuscitation (DR, n  9). The ER and DR groupswere resuscitated with hypertonic saline dextran(HSD) 20 min and 40 min after the injury, respectively.Mean arterial pressure (MAP), cardiac output (CO),and arterial blood gases were measured in addition tovascularbloodflowratesintheaorta,hepaticarteryandportal vein. The duration of follow-up was 100 min.  Results. MAPdecreasedfrom112  4to23  2mmHg (  P  <  0.05) during 20 min after the injury. Bolus infu-sion of HSD significantly elevated MAP, CO, and flowrates in the aorta, portal vein and common hepaticartery in both groups. Portal vein flow remained rela-tively high during the shock. Intra-abdominal bleed-ing (ER, 701    42 mL; DR 757    78 mL) and the mor-tality rate (ER 44%; DR 33%) did not differ between thegroups 100 min after injury (  P  >  0.05). Aortic flow,portal vein flow, common hepatic artery flow, MAP,CO, PaO 2 , PaCO 2 , base deficit, pH, hemoglobin mea-surements, and the volume of blood shed into the in-traperitoneal cavity did not affect survival in the Coxregression analysis. Conclusions.  Early  versus  delayed fluid infusionwith HSD resulted in a comparable hemodynamic re-sponse and survival 100 min after injury. No rebleed-ing was observed.  © 2006 Elsevier Inc. All rights reserved.  Key Words:  hemorrhagic shock; blunt hepatic in- jury; fluid resuscitation; hypertonic saline dextran;survival; hemodynamic response. INTRODUCTION The effect of early intravenous fluid resuscitation insevere hemorrhagic shock remains a critical issuewhen the source of bleeding is unknown or uncon-trolled. Many authors today advocate delayed or re-stricted intravenous fluid support until hemorrhagehas been controlled; the evolving attitude has gainedground because of the results of numerous laboratoryand clinical investigations [1–7].Most experimental studies on uncontrolled hemor-rhage have been fashioned after vascular injury mod-els. Bleeding has been induced through arterial cannu-las, arterial transverse cuts and, in some studies, using longitudinal abdominal aortic lesions produced by sur-gical steel-wire aortotomy [5, 8–11]. Despite the un- controlled hemorrhage  per se , these techniques havebeen associated with reproducible bleeds leading torelatively standardized states of shock.It is appropriate to note, however, that models of isolated arterial or venous injury may not be univer-sally applicable to blunt trauma victims. When itcomes to mimicking blunt trauma, only a few investi-gators have been successful in designing reproducibleand standardized solid organ injury models. In these,an aggressive fluid resuscitation mirrored those in the vascular models with increased bleeding and worsenedoutcomes [3, 4, 12]. Although aggressive fluid therapy may be associated with decreased survival, the criticalblood pressure to avoid total circulatory collapse inhemorrhagic shock still needs to be clarified. In addi- 1 To whom correspondence and reprint requests should be ad-dressed at Department of Surgery, Karolinska Trauma Center,Karolinska University Hospital, 171 76 Stockholm, Sweden. of Surgical Research  136,  273–279 (2006)doi:10.1016/j.jss.2006.07.016273  0022-4804/06 $32.00 © 2006 Elsevier Inc. All rights reserved.  tion, the proper time to initiate fluid therapy afterinjury also needs further research.The aim of the study was to investigate whether anearly  versus  delayed infusion of hypertonic saline dex-tran(HSD)afteranear-lethalhemorrhagebecauseofaliver injury would affect blood loss and short-term sur- vival. In addition, the portal vein and common hepaticartery hemodynamics were studied to evaluate differ-ences between venous and arterial responses to hem-orrhage in hepatic vascular inflow. MATERIALS AND METHODS The study was conducted at the Disaster and Emergency MedicalCenter, Stockholm, Sweden. The Ethics Committee for Animal Re-search, Umeå, Sweden, approved the study design. Experimental Animals Twenty-six Swedish landrace strain pigs with a mean weight of 23  0.7 kg, obtained from a single commercial animal breeder, werestudied. The animals were kept in an indoor holding unit for a daybefore they were used with free access to food and water until the dayof the experiment. Karolinska Institutet’s guidelines for the care anduse of laboratory animals were followed.  Anesthesia and Surgical Preparation Each animal received pre-medication consisting of an intramus-cular injection of 10 mL of ketamine hydrochloride 50 mg/mL, and0.5 mL of diazepam 5 mg/mL. An intravenous cannula was insertedinto a superficial ear vein, and anesthesia was induced with 0.5 mLof atropine 0.5 mg/mL and 3 to 4 mL of pentobarbital 60 mg/mLbefore tracheostomy was established through a midline neck inci-sion. The animal was mechanically ventilated by a Siemens Servo Ventilator 900B (Siemens-Elema, Solna, Sweden) to achieve normo-capnea using a min volume of 200 to 250 mL/kg. Complete anesthe-sia during the study was obtained by continuous infusion of theketamine hydrochloride 50 mg/mL at a rate of 14 to 16 mL/h. Ket-amine was chosen to reduce the depressive effect of general anesthe-sia on the hemodynamics [13]. A size 8-Fr catheter (Portex Ltd., Hythe, Kent, England) wasintroduced into the left external jugular vein through a left parame-dian neck incision to be used for the infusion of ketamine andresuscitation solutions. A second polyethylene catheter was intro-duced into the left common carotid artery for blood sampling andpressure monitoring. Through a right paramedian neck incision, aflow-directed thermodilution Swan-Ganz catheter (Baxter Health-care, Deerfield, IL) was inserted through the right external jugular vein and positioned with the distal port in the pulmonary artery tomeasure cardiac output (CO). The arterial pressure monitor and theSwan-Ganz catheter were connected to a Sirecust 1280 Patient Mon-itor and the pressures were printed using a Siredoc 220 Printer(Siemens, Boston, MA). A midline ventral celiotomy was performed and a Foley catheterwas placed in the bladder. Blood flow rates were measured continu-ously by three perivascular ultrasonic flow probes connected to twoTransonic T101 devices (Transonic System Inc., New York, NY). Wemeasured the flow in the portal vein, which was exposed proximallyto the liver (probe 165,65B). Probes were also applied to the commonhepatic artery (2SB99G) and to the abdominal aorta (4RB659), prox-imal to the aortic bifurcation. A tourniquet was applied around theinferior vena cava (IVC), below the diaphragm, and around theportal vein to obstruct backflow of the blood from the heart, IVC, andportal vein after the ECG had confirmed the asystolic state (Fig. 1). Model of Severe Hepatic Injury Associated WithUncontrolled Bleeding   After anesthesia and cannulations, the animals were allowed tostabilize for 10 min and the baseline hemodynamic features weremeasured. The liver injury was inflicted through 10-cm long inci-sions from the inferior margin of each liver lobe in the centraldirection according to a pre-measured ruler (Fig. 1). The cut edges of  the liver were allowed to bleed freely into the peritoneal cavity, andthe laparotomy incision was closed with a running suture. Study Groups The animals with a mean arterial blood pressure (MAP) of   35 mmHg before the initiation of fluid therapy ( n    18) were FIG. 1.  Liver injury model (IVC, inferior vena cava). 274  JOURNAL OF SURGICAL RESEARCH: VOL. 136, NO. 2, DECEMBER 2006  randomly allocated to two groups 1 min before the infusion wasintended to start by choosing one sealed piece of paper from thebucket containing 18 pieces bearing the treatment code. An earlyresuscitation group (ER,  n    9) was infused during 5 min with abolus dose of 0.7 mL/kg/min HSD (Pharmacia-Upjohn, Uppsala,Sweden) intravenously, starting 20 min after the injury. The secondgroup, the delayed resuscitation group (DR,  n  9), was infused withthe bolus dose of 0.7 mL/kg/min HSD intravenously, starting 40 minafter the injury. The allocation of the study groups was chosen tosimulate pre-hospital time intervals in the urban setting, where ittakes approximately 20 min for an ambulance to arrive at the sceneof injury and to initiate fluid therapy (ER) and another 20 min totransport the patient the hospital where fluid therapy is commenced(DR). To simulate the routine fluid resuscitation on arrival at hos-pital, all of the animals were infused with 0.4 mL/kg/min Ringer’sacetate (RA) solution intravenously (Ringer = s Acetate, Pharmacia,Uppsala, Sweden) over the remaining 55 min of the experiment,starting 45 min after the injury. The schematic study flowchart isdepicted in Fig. 2. After the experiment was discontinued, the abdominal incisionwas reopened and the tourniquets were pulled to obstruct the IVCand the portal vein. The free intraperitoneal blood was collected onweighed surgical swabs. The amount of blood loss was determined bythe difference in the wet and dry weights. Data Points Extracted and Statistical Analysis The data points accrued were aortic flow (mL/min), portal veinflow (mL/min), common hepatic artery flow (mL/min), MAP (mmHg),heart rate (HR) calculated as beats per min (bpm), cardiac output(CO, L/min), volume of blood in the intraperitoneal cavity after theexperiment (mL), and arterial blood gases, i.e., PaO 2  (kPA), PaCO 2 (kPA), pH (units), and O 2  saturation (kPA) obtained by means of ablood gas analyzer (GEM 3000; Instrumentation Laboratories, Lex-ington, MA). Baseline values (time  0) were obtained after surgicalinstrumentation. Post-injury values were obtained every min afterthe injury until resuscitation and then every 10 min until the end of the study at 100 min.The values are expressed as the mean    SEM. A repeated-measures ANOVA was used to investigate the mean difference be-tween the fluid resuscitation groups for the initial 20 post-injuryminutes with treatment groups and time as two factors. Anotherrepeated-measures ANOVA was used to analyze the early effects of fluid infusion. For this purpose, the variables were analyzed sepa-rately in the ER and DR groups, at 20 and 30 min and at 40 and50 min, respectively. A log-rank test was used to evaluate the differ-ence in mortality between the resuscitation groups. Because of thesmall sample size, a multivariate analysis, taking all of the indepen-dent variables into account at once, was not feasible. Therefore, aCox regression model, which included the treatment variable plusone of the other covariates, aortic flow, cardiac output etc., one at atime, was used.  P  0.05 was considered statistically significant. Thesurvival curve was plotted using the Kaplan-Meier analysis. It wasnot possible to establish the smallest hazard between the fluid re-suscitation groups that was of clinical importance and thus thehazard observed after a duration of 100 min had to serve as aguideline. To determine the required number of events required todetect a significant risk of mortality of 20% between the treatmentgroups, a power calculation for survival data were performed. Forthe sample size calculation, a power of 80% and a significance levelof 5% were assumed. The Statistica software, version 7.1 (StatSoft,Inc., Tulsa, OK) was used for all descriptive statistics and all of theanalyses presented in this paper. RESULTS Eight swine were excluded because of limited hypo-tension, i.e., MAP   35 mmHg before randomization.Eighteen animals were eventually included in thestudy and randomized. The results of the data pointsextracted during the experiment are shown in Figs.3–9. Repeated-measures ANOVA detected no signif- icant differences between treatment groups beforetreatment (MAP:  P    1.0; aortic flow:  P    0.18; com-mon hepatic artery flow:  P  0.9; portal vein flow:  P  0.7). The MAP decrease in all animals during the ini-tial 20-min after the injury is depicted in Fig. 3. A marked decrease in MAP from 110  5 mmHg to 20  2 mmHg (  P    0.05) and from 114    5 mmHg to 25   3 mmHg (  P    0.05) in the ER and DR group, respec-tively, was observed (Figs. 3–4). A significant early decrease was also observed in cardiac output (ER,  P  0.05; DR,  P  0.05), aortic flow (ER,  P  0.05; DR,  P  0.05), hepatic artery flow (ER,  P  0.05; DR,  P  0.05),and portal vein flow (ER,  P  0.05; DR,  P  0.05) (Figs.5–8). Heart rate (ER,  P    0.05; DR,  P    0.05), basedeficit (ER,  P    0.05; DR,  P    0.05), and PaCO 2  (ER, FIG. 3.  Mean arterial pressure measured during the initial20 min after massive liver injury in the ER and DR groups analyzedby repeated-measurements of ANOVA ( n  18,  P  0.05). FIG. 2.  Experimental flowchart (ER, early resuscitation; DR,delayed resuscitation; HSD, hypertonic saline dextran). 275 TALVING ET AL.: EARLY   VERSUS  DELAYED HYPERTONIC SALINE DEXTRAN RESUSCITATION   P    0.05; DR,  P    0.05) increased while no earlychanges were observed in hemoglobin concentration,PaO 2 , and pH.Resuscitation with HSD caused a significant in-crease in MAP (ER,  P    0.05; DR,  P    0.05), cardiacoutput (ER,  P    0.05; DR,  P    0.05), aortic flow (ER,  P  0.05; DR,  P  0.05), portal vein flow (ER,  P  0.05;DR,  P  0.05), and hepatic artery flow in the DR group,  P  0.05 (Figs. 4–8). Base deficit (Fig. 9) and the PaO 2 in DR group were not affected by resuscitation. Hepaticartery flow was not affected in the ER group. Heartrate and hemoglobin (Hb) levels were not affected byfluid infusion in any of the groups.The mean volume of the shed blood in the intraperi-toneal cavity was determined to be 701    42 mL and757    78 mL in the ER and DR groups, respectively(  P    0.05). During the experiment, death occurred infour animals in the ER group (44%) and in three ani-mals (33%) DR group. MAP at 20 min after the inflictedliver injury was measured and found to be 18.6    3mmHg and 25.2    2 mmHg in the succumbed andsurviving animals (  P    0.08). A total number of sixanimals in the DR group (66%) survived until intentionto treat at 40 min. Two surviving animals in the ERgroup showed a drop in MAP of   15 mmHg before thetreatment. The cumulative survival rate is demon-strated in a Kaplan-Meier chart (Fig. 10). The differ- ence in mortality rate between the study groups wasnot significant according to a log-rank test (  P  0.05).Covariates, such as aortic flow, portal vein flow, com-mon hepatic artery flow, MAP, HR, CO, PaO 2 , PaCO 2 ,base deficit (BD), pH, Hb, and volume of shed blood,showed no effect on mortality at the conclusion of theexperiment. Rebleeding was not observed in any of theanimals after the fluid infusion was started. FIG. 4.  MAP in uncontrolled hemorrhage after massive liverinjury in early ( n    9) and delayed ( n    9) resuscitation groups. Values are presented as the mean  SEM. FIG. 5.  Cardiac output in uncontrolled hemorrhage after mas-sive liver injury in early ( n    9) and delayed ( n    9) resuscitationgroups. Values are presented as the mean  SEM. FIG. 6.  Aortic flow after massive liver injury in early ( n  9) anddelayed ( n    9) resuscitation groups. Values are presented as themean  SEM. FIG.7.  Portal vein flow after massive liver injury in early ( n  9)and delayed ( n    9) resuscitation groups. Values are presented asthe mean  SEM. 276  JOURNAL OF SURGICAL RESEARCH: VOL. 136, NO. 2, DECEMBER 2006  DISCUSSION Hemorrhagic shock is one of the major causes of early death after severe trauma [14, 15]. During the last 4 decades, the standardized approach to the pa-tient in the state of hemorrhagic shock has focused onrestoring blood pressure with vigorous fluid resuscita-tion as early as possible after the traumatic event [16].However, there is a growing agreement that fluid ther-apy may improve survival only if initiated after thehemorrhage has been controlled [2, 5, 7, 17]. Hemosta-sis may be difficult to achieve in the prehospital setting [18, 19]. Further delays in achieving hemostasis arefrequently because of the time necessary for the diag-nosis and transfer to the operating room or to theangiography suite [20]. The delay may necessitate vig- orous intravenous shock treatment to buy time to avoidtotal circulatory collapse.In recent years, 250 mL of dextran 6% or polyethyl-ene hetastarch 6% mixed with hypertonic saline 7.5%or 7.2%, respectively, has been frequently infused inthe prehospital setting because of marked blood-pressure-elevating effects in patients with hemor-rhagic shock [21, 22]. The beneficial effects of HSD infusion have also been demonstrated in a meta-analysis of double-blinded prospective studies [23]. Be- ing suitable for rapid infusion and having a markedeffect on blood pressure, HSD was chosen as a resus-citation fluid in the present series.The present study attempted to mimic a major bluntliver injury of grade IV according to the Organ InjuryScale [24]. Complex hepatic injuries are relatively un- common, but frequently associated with a severe bloodloss [25–27]. Regardless of the cause of injury, the common denominator in complex liver lesions is a mas-sive blood loss mostly from large veins. Although thelow venous pressure allows hemostasis through coun-terpressure because of the expanding hematoma, thiscannot always be achieved in the shattered liver simi-lar to those in our model. Different treatment groupswere confirmed to be comparable before the random-ization by ANOVA in our series. We observed a pro-tracted hemorrhage lasting 6 to 7 min, compared toformer studies on uncontrolled hemorrhages from theabdominal aorta where the bleeding stopped at approx-imately 3 min after injury [11, 28]. The prolonged bleeding time may also be explained by the venousincapacity of vasoconstriction. Consequently, a morerapid decrease in blood flow was observed in the he-patic artery compared to the flow in the portal veinafter the liver injury. The vascular flow rates decreasedby 28% and 51% 1 min after injury in the portal veinand hepatic artery, respectively, and a difference wassustainable until the treatment. The difference can beexplained by one of many important effects of cathe-cholamine surge in the state of shock. The relativelyhigh blood flow in the portal vein despite the profoundshock confirms the value of early occlusion of hepatic vascular inflow using Pringle maneuver during explo-ration and repair of a liver injury. FIG. 8.  Common hepatic artery flow after massive liver injury inearly ( n    9) and delayed ( n    9) resuscitation groups. Values arepresented as the mean  SEM. FIG.9.  Base deficit after massive liver injury in early ( n  9) anddelayed ( n    9) resuscitation groups. Values are presented as themean  SEM. FIG. 10.  Kaplan-Meier cumulative survival function after mas-sive hepatic injury ( n  18) (  P  0.05). 277 TALVING ET AL.: EARLY   VERSUS  DELAYED HYPERTONIC SALINE DEXTRAN RESUSCITATION
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks